I I

PROCEEDINGS

OF TUE

AMERICAN ACADEMY

OP

AKTS AND SCIENCES.

YoL. XXXII.

FROM MAY, 1896, TO MAY, 1897.

BOSTON, MASS.:

JOHN WILSON AND SON.
Sanibcrsitg 5|rcss,

1897.

Sit

xS 1 I

CONTENTS.

I. 1. Revision of the Genus Tridax.

2. Synopsis of the Mexican and Central American Species of the

Genus Mikania.

3. Revision of the Genus Zinnia.

4. Revision of the Mexican and Central American Species of the

Genus Galea.

5. A Provisional Key to the Species of Porophyllum ranging

North of the Isthmus of Panama.

6. Descriptions of New and little known Phanerogams, chiefly

from Oaxaca. By B. L. Robinsox and J. M. Gkeenmax 1

II. A Revision of the Atomic, JYeight of Magnesium. By Theo-

dore William Richar:^s axd Harry George Parker 53

III. On the Group of real Linear Transformations whose Invariant is

a real Quadratic Form. By Henry Taber 75

IV. Studies in Morphogenesis, VI. A Contribution to the Quantitative

Study of Correlated Variation and the Comparative Variability

of the Sexes. By C. B. Davenport and C. Bullard . 85

V. On the Butanes and Octanes in American Petroleum. By

Charles F. Mabery and Edward J. Hudson ... 99

VI. The Constituents of Pennsylvania, Ohio, and Canadian Petroleum

between 150° and 220°. By Charles F. Mabery ... 119

VII. Refractive Power of the Hydrocarbons and Chlorine Derivatives
described in the Preceding Paper. By Charles F. INIa-

BERY AND EdWARD J. HuDSON 177

VIII. On the Composition of a South American Petroleum. By Charles

F. Mabery and Arthur S. Kittelberger . . . 183

VI CONTENTS.

IX. The Genera of North American Melanopli. By Samuel H.

S<!UDDER 193

X. Cycle in the Life of the Individual {Ontogeny') and in the Evolu-
tion of its own Group (^Phyhgeny). By Alpheus Hyatt . 2C7

XI. The Viscosity of Mercury Vapor. By A. A. Noyes and II. M.

Goodwin 225

XII. On the Action of Ammonia upon Cupriammonium Acetohromide.
By Theodore William Richards and Robert Jay
forsythe 237

XIII. On Obtaining ifeteorological Records in the Upper Air by Means

of Kites and Balloons. By A. Lawrence Rotch . . . 243

XIV. The Energy Conditions necessary to produce the Rontgen Rays.

By John Trowbridge 253

XV. A Measure of Variability, and the Relation of Individual Varia-
tions to Specif c Differences. By Edwin Tenney Brewster 267

XVI. 1. Revision of the Mexican and Central American Species of
Houstonia.

2. Key to the Mexican Species of Liabum.

3. Description of Neio or Little known Plants from Mexico. By

J. M. Greenman 281

Proceedings 315

Biographical Notices 331

Francis James Child 333

Thomas Tracy Bouve 340

Francis Amasa Walker 344

Benjamin Apthorp Gould 355

List of the Fellows and Foreign Honorary Members . . . 361

Statutes and Standing Votes 369

Index 381

Proceedings of the American Academy of Arts and Sciences.

VuL. XXXIl. No. I. — NovEiMBLii, 1896. •

CONTRIBUTIONS FROM TIIF: GRAY HERBARIUM OF
HARVARD UNIVERSITY.

New Series. — No. X.
By B. L. Robinson and J. M. Gkeenman.

I. Revision of the Genus Tridax.

II. Synopsis of the Mexican and Central American Species of the
Genus Mikania.

III. Revision of the Genus Zinnia.

IV. Revision of the Mexican and Central American Species of the

Genus C'alea.

V. A Provisional Ke}- to the Species of Poropliyllum ranging
North of the Isthmus of Panama.

VI. Descriptions of New and little known Phanerogams, chieflj^
from Oaxaca.

CONTRIRUTIONS FROM THE GRAY HERBARIUM OF
HARVARD UNIVERSITY, NEW SERIES, No. X.

By B. L. Robinson and J. M. Greenman.

Presented May 13, 1896.

I. — REVISION OF THE GENUS TRIDAX.

TRIDAX, L. (Name supposed to come from rpt's, thrice, and SaKctv,
to bite, referriug to the trifid rays.) — Heads terminal upon long peduncles,
heterogamous and radiate or rarely homogamous and discoid. Involucre
campanulate to sub-cylindric, 2 - several-seriate, very rarely sub-uni-
seriate ; scales all or at least the inner scarious and commonly purple-
margined. Receptacle mostly conical. Ray-flowers, when present,
fertile ; ligules yellow, white, or roseate, clearly or obsoletely bilabiate ;
external lip 3-toothed or deeply trifid, obovate or oblong in outline,
patulous; the inner of 1 or 2 short erect teeth or sometimes wanting;
disk-flowers usually (if not always?) yellow, regular, o-toothed. Anthers
short, sagittate at the base, appeudaged at the apex. Style-branches
terminating in short or long subulate appendages. Achenes turbinate,
hirsute or upwardly silky-villous, very rarely glabrous or nearly so.
Pappus of several to many narrow ciliate scales attenuate to plumose
awns (except in G. duhia). — Hort. Cliff. 418, & Gen. no. 979; DC.
Prodr. v. 679 ; Benth. & Hook. f. Gen. ii. 392 ; HofFm. in Engl. &
Prantl, Nat. Pflanzenf. iv. Ab. 5, 247. BartoUna, Adans. Fam. ii. 124.
Sogalgina, Cass. Bull. Philom. 1818, & Diet. Sci. Nat. xlix. 397. Balbisia,
•AVilld. Spec. iii. 2214. Galinsogea, HBK. Nov. Gen. & Spec. iv. 252,
t. 386. PtUostephium, HBK. 1. c. 253, t. 387, 388. Carphostelphium,
Cass. Diet. Sci. Nat. xliv. 62. Mandonia^VfeAdi. Bull. Soc. Bot. Fr. xi.
50, t. 1, not Schz. Bip. — Pubescent annuals or perennials not rarely lig-
nescent at the base, with leaves opposite, petiolate or subsessile, subentire
or more often dentate or irregularly cleft or pinnatifid. About 22 known
species, two of them of the S. American Andes, the others confined to
Mexico, except a single species which extends also to Mauritius and
E. India.

4 PROCEEDINGS OF THE AMERICAN ACADEMY.

«

Subgenus Eutridax. Scales of the involucre 2 - several-seriate :
achenes densely silky-villous or hirsute : pappus scales terminating in
plumose awns.

§ 1. Scales of the involucre very unequal, regularly imbricated in
several to many series, gradually decreasing in size ; the outermost very
short, mostly scarious and rounded at the summit, very rarely somewhat
herbaceous or (in T. angustlfoUa) subacute,

* Heads discoid : pappus shorter than the achenes.

1. T. brachylepis, Hemsl. "Annual, erect," witli " slender gla-
brescent brauchlets" and "thickish ovate-lanceolate dentate or sometimes
obscurely lobed leaves." — Biol, Cent.-Am. Bot, ii, 207. — Cordillera of
Oaxaca, altitude 7,000 feet, Galeotti. The only specimen, seen by the
writers, closely approximating in its characters Mr. Hemsley's descrip-
tion is no. 1423 of E. W. Nelson, collected in the Valley of Oaxaca,
altitude 5,500 to 7,500 feet, 20 September, 1894. But although only the
upper portion of the root is at hand it is not unlikely perennial.

2. T. tuberosa. Distinctly perennial ; the elongated woody root
at places tuberous-thickened : stem decumbent, subsimple, 3 feet in
height, leafy to the middle,. hirsute : leaves hirsute upon both surfaces,
3-nerved, 2 inches long, nearly half as broad, cuneate at the base and
3-cleft with sharply toothed acute lobes : heads about 6, nearly 7 lines
long and 6 lines in diameter : pappus only a third to half the length of
the achene. — Collected by C. G. Pringle, on the Sierra de San Felipe,
Oaxaca, altitude 7,000 to 8,500 feet, 17 November, 1894, no. 5644 a.

3. T. Pringlei. Perennial, decumbent, much branched from near
the base, 2 feet high, pubescent throughout but less hirsute than the
preceding: root woody, tuberous-thickened at intervals: leaves lanceo-
late, dentate or subentire, obtusish, 1 to H inches long, scarcely a third
as broad, 1 -nerved: heads 2 to 4, very similar in all respects to those of
the preceding. — Collected by C. G. Pringle, on the Sierra de San
Felipe, Oaxaca, altitude 7,500 feet, 7 September, 1894, no. 5644. In
technical characters very close to the preceding species, but with markedly .
different foliage,

* * Heads radiate : ligules evident,

■I- Mexican species.

++ Eays 3'ellow.

= Pappus very short or none.

4. T. trilobata, Hemsl. 1. c, 208, Erect much branched glandidar-
pubescent annual, a foot or two high, with lance-oblong obtusish and
coarsely few-toothed or laciniate leaves (cuneate at the base) and numer-

ROBINSON AND GREENMAN. — GENUS THIDAX. 5

ous showy heads 10 lines to au inch in diameter, with dark purple
involucre and broad bright orange-yellow rays. — Galinsoga trilobata,
Cav. Icon. iii. 42, t. 282; But. Mag. t. 1895; Sweet, Brit. Fl. Gard.
t. 56. Sogalgina (rilu hat a, Cass. Diet. Sci. Nat. xlix. 397. — Valley of
Mexico, Bouvgeaii, no. 846; near Chapultepec, Schaffner ; and on calca-
reous bluffs. Flor de Maria, Pringle, no. 3148 ; also in Michoacau in
fields near Patzcuaro, Pringle, no. 4271. The form without any trace
of pappus does not appear to differ in any other particular. A specimen
cultivated in the Botanic Garden of Harvard University has leaves
oblong, subentire.

= = Pappus about equalling or somewhat exceeding the achenes.

5. T. balbisioid.es. Gray. Annual, much branched, pubescent :
branches divaricate or the lowest decumbent : leaves from lanceolate and
irregularly toothed to deeply ternately cleft or pinnately parted : heads
rather numerous, nearly or quite an inch in diameter, with convex or
conical disk and spreading showy rays : ligules (exclusive of tube) 2\ to
4 lines long, as broad or broader. — Proc. Am, Acad. xv. 39. T. coro-
nopifolia. Gray, 1. c, not Hemsl. Galinsogea halbisioides, HBK. Nov.
Gen. & Spec. iv. 253, t. 386. Sogalgina bulbisioides, Cass. 1. c. xlix.
398. — Originally collected in Guanajuato between the Valley of San-
tiago and Lake Palangeo, at 5,500 feet altitude, by Humboldt ^ Bonpland.
It is described and figured by Kunth, 1. c, as having entire or repand
ligules of suborbicular contour. No plants with this character have

- since been observed, and we follow Dr. Gray in referring to the species
the following, which differ only in their more or less distinctly 3-toothed
rays : Schajfner s no. 238, and Parry ^ Palmer's no. 509, both from San
Luis Potosi. Nor does T. leptophylla, Gray (1. c. xxi. 391), from Chihua-
hua [Palmer^ s no. 425, and Pringle' s no. 769), ajjpear to differ by any
constant or satisfactory character. A form from San Luis Potosi, repre-
sented by Parry Sf Palmer's no. 508, has the ligules sometimes 3-toothed
and sometimes divided nearly to the base into 3 oblong lobes. This
plant was rather confidently referred by Dr. Gray (1. c. xv. 39) to T.
coronopifolia, but it differs from that species decidedly in its much imbri-
cated inolucre, with very unequal and rounded scales, and in its attenuate
chaff. Nor is the ligule strap-like. In stating it to be so Dr. Gray had
presumably observed only one of the long oblong lobes of a very deejily
trifid ray.

6. T. petrophila. Distinctly perennial from a lignescent base :
stems several, very slender, erect or nearly so : branches ascending : leaves
very narrow, linear, entire, toothed, or with 2 or 3 short linear lobes :

6 PROCEEDINGS OF THE AMERICAN ACADEMY.

ray-flowers with ligules much smaller than in the j^receding, about 1^-
lines in diameter : the tube being relatively very long and slender (4 to 5
lines in length) : chaff terminating in a long and slender awn. — T. bal-
bisioides. Gray, 1. c. xxii. 430, not xv. 39. T. balhisioides, var. tenui-
folia, Gray, (ined.) in distribs. Palmer and Pringle. — Jalisco, on the
Rio Blanco, Palmer, no. 569 ; on rocky hills near Guadalajara, Pringle,
nos, 2179, 255 G. The later collections of Mr. Pringle show this plant
quite distinct in its woody base and reduced rays from T. balhisioides.

++ t-f Rays white, purplish, or roseate.
= Pappus much longer than the achene.

7. T. rosea, Schz. Bip. in herb. Hirsutulous annual, leafy near the
base and often with spreading branches : leaves linear-oblong, irregularly
toothed or trifid with toothed lobes : peduncles becoming very long (6 to
10 inclies), quite simple, naked or bearing 1 or 2 minute alternate brapts :
heads large, with the spreading rays an inch or more in diameter: ligules
oblong, 3 to 5 lines in length, slightly 3-toothed at the apex. — A good
but apparently unedited species, founded by Schultz upon Schaffner's
no. 60, from Guadalupe, and collected in the same locality by Bilimek
in 1865, no. 488, and Bourgeau, no. 586; also in Valley of Mexico,
Schaffner, no. 265.

= = Pappus sliorter than or barely equalling the achene.

8. T. Palmeri, Gray. Pubescent or puberulent, 2 feet high, usu-
ally with a few ascending branches, naked above : leaves mostly much
cleft or deeply and laciniately parted into narrow acute segments : rays
broad, obovate, truncate, essentially entire, roseate : disk-flowers greenish
yellow. — Proc. Am. Acad. xv. 38. — San Luis Potosi, on rocky bluffs
at Alvarez, altitude 8,000 feet, Parry ^ Palmer, nos. 489, 490, 482^,
also Schaffner, no. 236.

Var. indivisa, Robinson & Seaton. Somewhat stouter and more
densely pubescent : leaves ovate, subhastate, abruptly contracted into a
petiole, repand-dentate, not lobed. — Proc. Am. Acad, xxviii. 109. —
Ledges of canons on mountains near Lake Chapala, Jalisco, Pringle., no.
4332. Mr. Pringle states that the base is perennial.

■4- -1- South American species, of the Andes of Bolivia and Ecuador: leaves
narrow : pappus longer than the achenes.

9. T. Mandonii, Schz. Bip. Branched, hirsute : leaves sinuate or
repand-dentate : involucral scales rounded at the apex, sparingly puberu-
lent near the ends or quite glabrous : rays very small. — Bull. Soc. Bot.
Fr. xii. 82, & Linnaea, xxxiv. 536. Mandonia Boliviensis, Wedd. Bull.

ROBINSON AND GREENMAN. — GENUS TRIDAX. 7

Soc. Bot. Fr. xi. 51, t. 1. — Near Sorata, Province of Larecaja, altitude
2,090 to 3,000 meters, Mandon, no. 289.

10. T. angustifolia, Benth. & Hook. f. Decumbent and rooting
at the nodes, pubescent: leaves lance-linear, acute, denticulate or entire:
iuvolucral scales acutish, pubescent or puberulent on the outer surface ;
the outer ones inclining to be squarrose : ligules oblong, 2 or 3 lines in
length. — Gen. ii. 392. — Collected in the Andes of Ecuador by Spruce,
no. 5582.

§ 2. Involucral scales 2-3-seriate, less unequal ; the outer often her-
baceous and acutish.

* Leaves ovate to linear, pubescent to densely hirsute but not canescent.
•t- Leaves relatively broad, subentire, toothed, or divided into a few rather

broad lobes.
•w. Pappus longer than or about equalling the achenes.

= Perennials with ligneous or lignescent base : rays short, nearly orbicular,

3-tootiied.

11. T. procumbens, L. More or less densely hirsute with white
hairs: stems several to many from a woody root or stock, decumbent or
procumbent : leaves lanceolate to ovate-lanceolate, acute, sharply repand-
or sinuate-toothed, cuneate at the base : scales of the involucre oblong,
acutish, usually villous-hirsute : chaff persistent: rays yellow or (?) white.
— Spec. ii. 900. Amellus pedunculatiis, Ovt. fide Willd. Enum. 916.
Balbisia elongata, Willd. Spec. iii. 2214. B. peduncidata, HofFmgg. Verz.
Pfl. 228. B. divancata, Cass. Ann. Sci. Nat. xxiii. (1831) 90. —The
most widely distributed species ; Monterey, North Mexico, Eaton 4* Ed-
wards, also Pringle, no. 1920; Wartenberg, near Tantoyuca, Ervendberg,
no. 61; San Luis Potosi, Parry ^ Palmer, no. 520; " Coahuila and
Nuevo Leon," Palmer, nos. 629, 20G1 ; Orizaba, Schaffner, Gray, Botteri;
Jalisco, Palmer, no. 297; Colima, Palmer, no. 1186; Oaxaca, L. O.
Smith, no. 423, E. W. Nelson, nos. 2582, 2773 ; Guatemala, nos. 2370,
& 4200 of Donnell Smith's sets. Also on Elliott's Key, Fla., Simpson,
no. 553; Cuba, Wright, no. 2861. Introduced and now abundant in E.
India and Mauritius.

Var. ovatifolia. Less hirsute, pubescence shorter, finer, and tending
to become fuscous : base of stem ligneous : leaves ovate, acutish, serrate
with small subequal obtusish teeth, finely pubescent upon both surfaces,
8 to 10 lines long, 6 to 8 lines broad, rather abruptly contracted at the
base into short petioles : involucral bracts very broad, mostly obovate. —
Collected by E. W. Nelson in the vicinity of Yalalag, Oaxaca, July, 1894,
no. 948. Rather well marked in its foliage and possibly a distinct

8 PROCEEDINGS OF THE AMERICAN ACADEMY.

species, but agreeing with the type in all floral characters except the
somewhat broader involucral scales.

= = Annuals.

12. T. obovata, Turcz. Low, 4 to 6 inches in height: leaves
obovate, subentire, obtuse: ligules short, yellowish. — Bull. Soc. Nat.
Mosc. xxiv. 187; Walp. Ann. 238. — S. Mexico, on .Sierra San Pedro,
Nolasco, Talea, &c., Jurgensen, no. 124. Here we would refer speci-
mens collected near Acapulco by Palmer, no. 165,

13. T. erecta, Gray. Slender, erect, hirsutulous, branching, a
foot or so in height : leaves ovate-lanceolate and serrate or broadly ovate
in outline and rather deeply 3-parted with acutish lobes : heads rather
small : involucre ovate ; one or two of the outer scales often narrow,
oblong, and spreading : ligules short, only a line or two in length, 3-
toothed, yellow. — Proc. Am. Acad. xxi. 390, 433. — Southwestern
Chihuahua, Palmer, no. 285.

14. T. tenuifolia, Rose. Erect, pubescent, 1 to 2 feet high, branch-
ing : leaves ovate, petiolate, dentate but not lobed : outer scales of the
involucre acuminate : rays conspicuous, 3 to 5 lines long, pale yellow or
" bright white." — Contrib. U. S. Nat. Herb. iii. 319. T. bicolor, Gray,
Proc. Am. Acad. xxi. 391, as to pi. Palmer. — Southwestern Chihuahua,
on mountain sides above Batopilas, Palmer, no. 130.

Var. microcephala. Rose, 1. c. Described as diflTering in its smaller
heads and '' slightly different disk corollas." — " Near Lodiego, Palmer,
no. 1611."

+H- ++ Pappus shorter than the achenes.

= Annual species of Northwestern Mexico : heads large, an inch in diameter.

15. T. bicolor, Gray. Erect annual, with leaves ovate to rather
narrowly oblong-lanceolate, serrate-dentate, not lobed, narrowed to slender
petioles: ligules reddish or purplish, cuneate-oblong, 3-tootlied. — PI.
Fendl. 104. — Llanos in the Sierra Madre, W. Chihuahua, Wish'ze7ius,
no. 214; also among rocks, Bachimba, Thurber, no. 833, and on rocky
hills near the town of Chihuahua, Pringle, no. 638.

= = Species of Southern Mexico, hgnescent at tlie base.

1 6. T. Galeottii, Klatt. " Stem terete, branched, pilose : leaves
petiolate, cuneate, irregularly lobed, pilose upon both surfaces, 2 inches
long, 11 inches broad, 3-nerved : heads solitary, terminal; involucral
scales biseriate, the outer shorter: chaff 3-5-dentate at the apex." —
Leopoldina, xxiii. 6. Plilostephium Galeottii, Schz. Bip. ^c?e Klatt, I.e.
— Between San Andres and San Miguel, Liehmann, nos. 588, 693. and

ROBINSON AND GREENMAN. — GENUS TRIDAX. 9

Galeotti, no. 2472. A species not seen by the writers. The description
is translated and condensed from the original characterization.

-1- -t- Leaves very narrow, linear or lance-linear, or deeply cleft into narrow di-
visions : heads rather small, 4 to 6 lines broad : rays small, bilabiate, the
external lip tritid.

1 7. T. coronopifolia, Hemsl. Low, much branched and spreading,
pubescent or hirsute ; leaves lance-linear, denticulate, lacinately toothed
or piunatifid with narrow linear unequal acute segments : peduncles
slender, toward the summit finely strigillose with appressed white hairs,
and lacking the glandular-tipped hairs so common in the genus : rays
yellow (or white) : pappus scales uuequal, some of them subulate and
ciliated, others plumose with scarcely dilated base. — Biol. Cent.- Am.
Bot. ii. 207. Ptilostephmm coronopifoUum, HBK. Nov. Gen. & Spec,
iv. 255, t. 387. We have little hesitation in including in this species
also Tridax trijida, Gray, Proc. Am. Acad. xv. 39 {Ptilostephium tri-
fidum, HBK. 1. c. 255, t. 388), notwithstanding the considerable differ-
ence in the pappus represented by Kunth. After the examination of a
number of specimens we doubt the specific significance of the difference in
length of the pappus, and as to the difference in the breadth of the scales
that is often considerable upon the same achene. — Mexico without lo-
cality, 7%. Coulter, nos. 348, 430, and Berlandier ; State of Mexico,
Bilimek, no. 491, also Bourgemi, nos. 164, 705 ; Tacubaya, Schaffner ;
Mt. Orizaba, altitude 9,000 feet, Seaton, no. 273 ; Oaxaca, Soledad de
Etla, altitude 5,300 feet, L. C. Smith, no. 361, and by same collector at
Telixtlahuaca, no. 866 ; also by E. W. Nelson, in Valley of Oaxaca,
altitude 5,100 to 5,800 feet, no. 1229. The formal variety alboradiata
{T. trijida, var. alboradiata, Gray, Proc. Am. Acad. xv. 39), with white
rays, but apparently without other significant or constant differences, has
a somewhat more northerly distribution, having been collected at San
Luis Potosi, Schaffner, no. 239, Parry 8^ Palmer, no. 511 ; in Jalisco,
Pringle, no. 2902 ; and in Guanajuato, Duges, no. 438.

18. T. lanceolata, Klatt. "Lower leaves broadly lanceolate and
upper lance-linear, entire : chaff obovate." — Leopoldina, xxiii. 6. —
'• Tehuacan, Liebmann, no. 205; Cuernavaca, Berlandier, no. 1063." A
species not seen by the writers and (as to description) distinguished from
the preceding species chiefly by the characters quoted.

19. T. imbricata, Schz. Bip. "Leaves linear-lanceolate, entire,
or 1-3-toothed, pilose, ciliated on the margins : external scales of the
involucre obtuse, striate, dorsally puberulent." — Schz. Bip. in Klatt,
Flora, 1885, 202. — " Real del IMonte, Ehrenberg, no. 355." We suspect
this to be merely a form of T. coronopifolia.

10 PROCEEDINGS OF THE AMERICAN ACADEMY.

* * Whole plant caneseent-tomentose : leaves linear, entire.

20. T. candidissima, Gray. Very white-woolly, 4 to 6 inclies
high, densely leafy below : leaves 1 to If inches long, sessile by a sheath-
ing base : peduncles solitary, terminal, naked : heads homogamous :
involucral scales lanceolate, acute : pappus long. — Proc. Am. Acad. xv.
39. — On ashy soil, Angostura, San Luis Potosi, Parry ^ Palmer, no.
510. Not since rediscovered.

Subgenus Pseudotridax. Scales of the involucre sub-uniseriate :
achenes naerely papillose-puberulent or glabrescent: scales of the pappus
obtuse or obtusish, laciniately bordered : anomalous species connecting
this genus with Galinsoga.

21. T. (?) dubia, Rose. Decumbent or procumbent pubescent

herb; stems about 2 feet long: leaves ovate, serrate, acutish, petiolate :

heads subracemose or subcorymbose, 6 lines in diameter : involucral

scales very few, about 5 : ray-flowers with short golden yellow 3-toothed

ligules. — Contrib. U. S. Nat. Herb. i. 337, t. 33. — Along river bottoms,

Colima, Palmer, no. 1173; also in lowlands near San Bias, Tepic,

Lamb^ no. 609.

Species (doubtless of Eutridax but otherwise) of doubtful affinities: not seen

by the writers.

22. T. Ehrenbergii, Schz. Bip. '' Stem herbaceous, elongated,
climbing, sulcate, sparingly pilose, trichotomous : leaves rhomboidal,
slender-petioled, acuminately serrate, 3-nerved, above sparsely pubescent,
below pilose upon the nerves: head solitary, terminal, many-flowered:
involucre campanulate : scales scarious, laciniate on the margins : chaflF
scarious, trifid, costate, mucronate : flowers of the disk campanulate,
pilose, ciliated on the margin: pappus-scales fimbriate, shorter than the
sericeous achene." — Leopoldina xxiii. 6. — " Chinantla, Liebmann, no.
598. Leaves 20 lines long, 9 lines broad."

11. — SYNOPSIS OF THE MEXICAN AND CENTRAL
AMERICAN SPECIES OF THE GENUS MIKANIA.

MIK ANI A, Willd. (Dedicated to Professor Joseph Gottfried Mikan,
of Prague, born 1743, died 1814.) — Cylindrical involucre of four erect
concave obtuse or acute equal or subequal scales, sometimes with a shorter
external fifth scale. Receptacle small, naked. Flowers 4 (in M. punctata
"7"), tubular, with campanulate throat and 5-toothed limb. Anthers
appendaged at the apex, obtuse or truncate at the base. Style-branches
Ion"-, filiform-clavellate. Achenes 5-angled, without intermediate ribs,

•

ROBINSON AND GREENMAN. — GENUS MIKANIA. 11

l)uberulent or glabrous. Pappus of many seta; in a single row, mostly
about equalling or somewhat exceeding the acheue, bright white or more
commonly sordid, fuscous, or rufous. — Spec. iii. 1742; DC. Prodr.
V. 187, vii. 270; Benth. & Hook. f. Gen. ii. 246; Hoffm. in Engl. &
Prantl, Nat. Pflanzenf. iv. Ab- 5, 140. WiUugbceija, Neck. Elem. i. 82.
Corymanthelium, Kuuze, •Linnoea, xx. ID. — Mostly slender shrubby or
herbaceous twiners or " rarely erect." Leaves opposite, usually ovate,
cordate or hastate, and petiolate. Inflorescence spicately, raceraosely,
or corymbosely paniculate. A genus of some 175 good species nearly
confined to Tropical or Subtropical America. M. scandens is also
widely distributed in the United States and in the warmer parts of the
Old iVorld.

Subgenus 1. Cylindrolepis. Truncate scales of the involucre not
at all imbricated but strongly involute, each completely surrounding a
flower.

1. M. globosa, Coulter. Glabrous or nearly so: leaves ovate,
thickish, undulate-dentate, acuminate, rounded at the base, 5-nerved from
near the base: heads in dense globose clusters; these in lateral or ter-
minal panicles: each of the four flowers completely enveloped in an
involucral scale : mature achenes not seen. — Bot. Gaz. xx. 46, where
also called WillugbcBya globosa. — Santa Rosa, Guatemala, altitude 4,000
feet, Heyde ^ Lux (no. 3430 of Dounell Smith's sets). A noteworthy
species differing much in its involucral scales from any other known to
us. It is said by Professor Coulter to resemble closely M. smilacina in
" habit and structure." The likeness, however, does not extend to the
heads, which are here completely divided into four compartments by the
intrusion of the involute edges of the scales ; while in M. smilacina the
involucre is normally imbricated.

Subgenus 2. Imbricate. Scales of the involucre imbricated.
§ 1. Heads spicately or racemosely arranged on the opposite spread-
ing branches of ample pyramidal panicles.

* Heads pedicelled.

2. M. Houstonis, Willd. Glabrous climbing shrub, with ovate
acuminate entire petiolate leaves. — Spec. iii. 1742 ; DC. Prodr. v. 190.
— S. Mexico, Vera Cruz, Houston, and (ace. to Hemsl.) " Linden, no.
1169; Yucatan and. Tabasco, Johnson, no. 25 ; Guatemala, Las Esca-
millas, Hartweg, no. 535 ;.Chojoja, near Mazatenango, Bernoidli, no. 100;
Panama, Barbacoas, S. Hayes." The only plants in Herb. Gray agreeing
with the character of this species are Ervendherg's nos. 87, 222, from
Wartenberg, Huasteca, Mexico.

12 PROCEEDINGS OP THE, AMERICAN ACADEMY.

* * Heads sessile.
■>- Heads very small : pappus bright white : branchlets winged.

3. M. pterocaula, Schz. Bip. Glabrous twiner : leaves ovate,

acuminate, dentate, thin, 5-nerved from the base : branchlets 6-winged. —

Schz. Bip. in Hemsl. Biol. Cent.- Am. Bot. ii. 103 (name onlyj ; first

described by Klatt, Leopoldina, xx. 4. — Mirador, Liebmann, no. 101.

•*- ■*- Heads larger : pappus tawny or rufescent : branchlets terete, striate but

not winged : leaves entire.

4. M. leiostachya, Benth. Leaves large, firm in texture, ovate
with acuminate apex and rounded base, pinuatelj 5-nerved from above
the base, glabrous above, finely fuscous-pubescent on the veins beneath,
at length wholly glabrate, the larger ones 3 or 4 inches in breadth :
spikes 1 to 2 inches long. — PI. Hartw. 201. — Columbia and Panama,
Seemann, no. 446 ; also at Gatun Station, on the Panama Railway,
Hayes. Said also to extend southward to Peru.

5. M. Hookeriana, DC. Leaves narrower, thinner, 3-nerved :
spikes less than an inch in length: pappus rufous. — Prodr. v. 195. —
Ascribed to Nicaragua and Panama by Hemsley. We have seen no
specimens from any part of Mexico or Central America. The description
is drawn from Schomburgk's no. 479 from Brit. Guiana.

§ 2. Heads not distinctly spicate nor racemose, disposed in ample
terminal panicles : branchlets densely tawny-hirsute or woolly.

6. M. pyramidata, Donnell Smith. A tall climber densely cov-
ered with ferrugineous hirsute pubescence : panicle loose : heads not
glomerate : leaves ovate, acuminate, rounded at the base, pinnately 5-
nerved from somewhat above the base. — Bot. Gaz. xiii. 188. — Coban,
Dept. Alta Verapaz. Guatemala, altitude 4,300 f eet, wn 2\ierckheim (no.
1106 of Donnell Smith's sets).

7. M. eriophora, Schz. Bip. Densely tawny-tomentose, the inflores-
cence woolly : heads somewhat glomerate : leaves ovate, cordate, sharply
acuminate, pubescent above, tomentose beneath, o-nerved from the base,
3 to 4 inches long, 2 inches broad. — Schz. Bip. in Hemsl. 1. c. (name
only). — Mirador, Liebmann, no. 94.

§ 3. Heads disposed in roundish or flattish cymose corymbs.
* Heads rather large, when mature 4 to 5 lines in length.
■I- Scales of the involucre obtuse : leaves ovate, distinctly cuneate at the base.

8. M. olivacea, Klatt. -Leaves entire." — Bull. Soc. Bot. Belg.
xxxi. 195. — Forests o^f Buenos Ayres, Costa Rica, Plttier, no. 4433.

9. M. Guaco, Humb. & Bonpl. Leaves {ex icone) denticulate. —
PI. iEquin. ii. 84, t. 105, Said by Index Kew, to equal M. amara^

ROBINSON AND GREENMAN. GENUS MIKANIA. 13

Willd. Spec. iii. 17-i-l, but apparently very different from Aublet's plate
of Eupatorium parvijiorum, PI. Guiau. ii. t. 315. — Nicaragua and
Panama, ucc. to Hemsley.

-I- -!- Scales of the involucre obtuse : leaves cordate-hastate at the base.

10. M. punctata, Klatt, 1. c. Heads described as 7-flowered. —
Clairieres du General, Costa Rica, Pittier, no. 3434, and banks of a
stream at Buenos Ayres by same collector, no. 4934.

•t- -4- •(- Scales of the involucre acute : leaves cordate or hastate, rarely

subtruncate at the base.

11. M. cordifolia, Willd. 1. c. 1746. M. suaveolens, HBK. Nov.

Gen. & Spec. iv. 135. M. gonoclada, DC. Prodr. v. 199. M. Fendleri,

Klatt, Abh. Naturf. Gesellsch. Halle, xv. 4 (of reprint). — W. Louisiana,

Hale ; Mexico, San Luis Potosi, Palmer^ nos. 1079, 1115, and on rocky

slopes of Tamasopa Canon, Pringle, no. 3928 ; Colima, Palmer, no.

1207; Acapulco, Palmer, no. 565; Cordova, Bourgeau, no. 1812;

Wartenberg, Huasteca, Ervendherg, no. 82 ; Mirador, Sartorius ; also

somewhat doubtful specimens from Guatemala, altitude 4,600 feet,

Donnell Smith, no. 2366, and Navarro, Costa Rica, by the same collector,

no. 4855.

* * Heads decidedly smaller, 2| to 3 lines in length.

12. M. scandens, Willd. 1. c. 1743. Leaves smoothish, acuminate ;
hastate basal lobes little spreading. — 31. Orinocensis, HBK. 1. c. 134.
Eupatorium scandens, L. Spec. ii. 836 ; Jacq. Ic. Rar. t. 169. — From
N. Eng. to W. Canada, Florida, and Texas ; Mexico, Victoria, Tamau-
lipas, Berlandier, nos. 852, 2272 ; San Bias, W. G. Wright, no. 1339 ;
Cordova, Bourgeau, nos. 1632, 2184; Orizaba, Botteri ; Oaxaca at Cui-
catlan, L. C. Smith, no. 247, and near Reyes, E. W. Nelson, no. 1861 ;
Guatemala, Esquintla, Hayes, Masagua, Donnell Smith, no. 2394, and
Ambelice, Heyde 3f Lux (no. 3434 of Donnell Smith's sets) ; Nicaragua,
G. Wright ; Panama, Seemann ; also W. India, Trop. Am., and warmer
parts of the Old World.

13. M. dentiCTilata, Willd. 1. c. 1744. Leaves very scabrous,
obtusish, hastate basal lobes widely spreading. — S. Mexico at Jalapa
ace. to Lessing ; Guiana, Schomburgk. *

Dubious species.

M. ANGULATA, M. REPANDA, and M. Tlalixcotan, La Llave, El
Mosaico Mexicano, ii. 299 ; Seemann in Hook. Jour. Bot. & Kew Misc.
V. 79, from Cordova and vicinity, are unrecognized species founded solely
upon foliar and very doubtful characters.

M. coriacea. La Llave, 1. c, Seemann, 1. c, is from its alternate
leaves positively to be thrown out of the genus.

14 PROCEEDINGS OF THE AMERICAN ACADEMY.

III. — A REVISION OF THE GENUS ZINNIA.

ZINNIA, L. (Dedicated to Professor Johann Gottfried Zinn, of
Gottiiigen, born 1727, died 1759). — Heads radiate: disk-fiowers her-
maphrodite, fertile; ray-flowers pistillate, fertile, with suborbicular oval
or oblong sessile persistent white, yellow, red, or purple ligules. Involucre
ovate-cylindric or campanulate, the scales 3 - many-seriate, broad, closely
imbricated, obtuse or rounded, often more or less colored and slightly
inflated or subsquarrose just beneath the summit. Disk conical to colum-
nar: chaff scarious, more or less cariuate, enveloping the flowers, often
erose at the mostly obtusish apex. Corollas of the disk-flowers tubular
with narrow scarcely ampliate throat and 5-toothed limb. Anthers ap-
pendaged at the apex, entire at the base. Style-branches obtuse, scarcely
or not at all appendaged. Achenes laterally compressed, glabrous or cilio-
late on the edges, 2-toothed at the summit and frequently 1-awned from
the inner angle or rarely 2-awned ; the achenes of the rays triquetrous,
3-toothed, with or without 1 to 3 short or long awns. — Gen. ed. 6, no.
974; Gray, PI. Wright, i. 105; Benth. & Hook. f. Gen. ii. 357;
Hoffm. in Engl. & Prantl, Nat. Pflanzenf. iv. Ab. 5, 225. Orassina,
Scepin. Diss, and Lejica, Hill, Exot. Bot. t. 22, fide Endl. Samritaliopsis,
Schz. Bip. ace. to Benth. & Hook, f., 1. c. — Annuals, perennial herbs,
or suffrutescent plants with opposite mostly entire leaves and showy
terminal pedunculate or subsessile heads. About a dozen species known
in nature, and three or four others somewhat doubtfully distinguished in
horticulture, beside obviously artificial varieties and hybrids. The range
of the genus is from the 8. United States to Chili and Brazil, but it
attains its greatest specific diversity in Mexico.

§ 1. Low cespitose perennials, shrubby at the base and many-
stemmed : stems (or perhaps better subsimple branches) erect, crowded,
or fastigiate : root stout, ligneous : leaves strictly linear to acerose, often
fascicled and rigidulous, mostly rather pale. — §§ Diplothrix Sf Heterogyne,
Gray, 1. c.

* Ligules showy, much exceeding the achenes, white or pale yellow.

■1- Leaves 1-nerved.

1. Z. acerosa. Gray. Leaves acerose, obscurely 1-nerved, much
crowded, rather sharp-pointed but scarcely pungent. 6 to 8 or 10 lines
long. — PI. AVright. i. 106. Diplothrix acerosa, DC. Prodr. v. 611. — -.
Hills of W. Texas, near Pecos, Wright, no. 324, Thurhcr, no. 125;
Mexico, San Luis Potosi, Berlandier, no. 1343, Parry ^ Palmer, no.
440^ ; Coahuila, Palmer, nos. 577, 578.

ROBINSON AND GREENMAN. — GENUS ZINNIA. 15

2. Z. pumila, Gray. Very similar to the preceding : leaves linear,
flat, mostly less than half inch long, prominently 1-uerved. — PI. Feudl.
81, PI. Wright, i. 105, & ii. 86. — Hills, W. Texas, Wright, nos. 323,
1215, to Arizona, Camp Grant, Palmer, no. 122, near Tucson, Greene,
no. 1106, Camp Lowell, Pringle, Lemmon, nos. 91, 92, near Sta. Cata-
Hna, Lemmon, no. 3033, Lowell Mts., W. F. Parish ; Mexico, on high
plains near San Juan de la Vequeria and at '' Castaniola " {^= Cas-
tanuela ?), Gregg, no. 279 ; on llanos of Sonora, Schott ; east of Guada-
lupe Cailon, E. K. Smith; near Carneros Pass, Coahuila, Pringle, no.
2390 (distrib. as Z. acerosa) ; San Luis Potosi, Schaffner, no. 336, and
Parry Sf Palmer, nos. 439, 440. The technical distinctions between this
and the preceding are unsatisfactory at best, although the specimens are
for the most part pretty readily distinguished upon the foliar differences.

1- 1- Leaves 3-nerved.

3. Z. juniper if olia. Gray. Leaves somewhat longer than in the
related species, the larger ones inch or more in length, usually whitish
beneath : rays oblong, mostly 2 or even 3 times as long as broad, of deep
orange color. — PI. Wright, i. 105. Diplothrix junijierifolia, DC. Prodr.
V. 612. — North Mexico, mountains near .San Juan de Vanegas, Berlan-
dier, 1359; without locality, Gregg, no. 68; Santillo, Parry, no. 40, and
near same locality, Palmer, no. 576 ; and on limestone hills, Carneros
Pass, Pringle, no. 2404.

4. Z. grandiflora, Nutt. Leaves less than an inch in length :
rays pale or sulphur yellow, very broad, suborbicular in outline. — Nutt.
in Torr. & Gray, Fl. ii. 298 ; Torr. in Emory, Report Reconn. Calif, t. 4.
— Colorado, on bluffs near Pueblo, Greene, to W. Texas, Wright, nos.
322, 1213, Pope, Ft. Davis, Girard ; New Mexico, Fendler, no. 400, near
Santa Fe, Wislizenus, no. 415 ; Arizona, near Ft. Whipple, Coues Sf
Palmer, no. 2821, Ft. Apache, Palmer, no. 583, Mustang Mts., Pringle,
Huerfano, Parry, no. 106, Upper Canadian River, Emory; Sonora,
Thurher, no. 312, Smith; San Cedro, Lloyd, no. 401.

* * Ligules almost obsolete, shorter or scarcely longer than the achenes.

5. Z. anomala, Gray. Scabrous-pubescent: leaves 9 lines to inch
in length, line to line and a half in breadth : heads 3 to 5 lines in di-
ameter, appearing discoid or with evident but short yellow rays: disk-
flowers apparently orange-red ; the limb velvety-margined. — PI. Wright,
i. 106. — Prairies of W. Texas, Wright, nos. 325, 1216; near Saltillo,
Coahuila, Mexico, Palmer, no. 581. First coll. (ace. to Gray, 1. c.) by
Berlandier in Northern Mexico.

16 PROCEEDINGS OF THE AMERICAN ACADEMY.

§ 2. Erect or procumbent herbs, sometimes a little woody at the base ;
stems and branches loosely spreading ; leaves linear, lance-linear, or ellip-
tic-oblong : rays rather short, suborbicular or quadrate to oblong, 2 to 6
lines in length, white or sulphur yellow.

* Achenes with interrupted callous margins and somewhat tufted ciliation :

slender erect annual with small heads and very pale or bright white rays.

6. Z. -icolor, Hemsl. Becoming a foot or so in height: leaves
linear to lance-oblong, an inch to inch and half long, 1 to 4 lines broad,
obtuse. —Biol. Cent.-Am. Bot. ii. 153, as to syn., but not as to specimens
cited, except that of Mendez. Z. maritima, Gray, Proc. Am. Acad. xxii.
423, in part. Mendezia bicolor, DC. Prodr. v. 533 ; Deless. Icon. iv. t. 29.

— West of Guanajuato, Mendez; San Luis Potosi, Scliaffner^ no. 337;
Jalisco, at Tequila, Palmer, no. 355, and on slopes of canons near Guada-
lajara, P/7«^^e, no. 2313. The last two distHbuted as Z. maritima, from
which this erect white-rayed plant of the inland is amply distinct.

* * Achenes evenly margined and regularly ciliated ; rays yellow or orange.

7. Z. Greggii. Slender pubescent herb, becoming scabrous, erect
or decumbent merely at the base : leaves linear or nearly so, 1-3-nerved,
sessile : heads slender-peduncled, terminating the spreading nearly naked
branches, these bearing mostly only a single pair of linear leaves : rays
varying from very short-oblong to half inch in length : disk-flowers
orange ; ray-flowers pale yellow ; ray-achenes about a line in length :
disk-achenes bearing a single slender awn. — Z. bicolor, Hemsl. 1. c. as
to plants of Coulter and of Seemann, but not as to syn. Z. maritima,
Gray, Proc. Am. Acad. xxii. 423, as to narrow-leaved form, not HBK.

— Mexico, without locality, Gregg, 1848-1849, no. 1082; also Baites ;
W. Mexico, Seemann ; none of these specimens show the base perfectly,
but a plant apparently identical, collected by F. H. Lamb on plains at
Zopelote, Tepic, 9 February, 1895, no. 555, has a thickish perennial root.

8. Z. littoralis. Procumbent spreading herb, probably of biennial
or perennial duration : stems leafy, branched, striate, puberulent : leaves
elliptic-oblong or elliptic-lanceolate, obtuse or obtusish at both ends, 3 (or
obscurely 5)-uerved and reticulated, green on both sides, 4 to 10 lines
long, a third to half as broad : heads scarcely peduncled, borne at the
ends of leafy branches: rays orbicular or nearly so, pale yellow, striate
and greenish toward the ends beneath : disk-flowers bright orange-
colored ; chaff oblong, very obtuse or truncate : achenes with a very
narrow cartilaginous margin, ciliated : pappus of a single awn with or
without a shorter second one : achenes of the ray-flowers about li lines
long, tuberculate. — Collected at Mazatlan by Th. Coulter, and redis-

ROBINSON AND GREENMAN. — GENUS ZINNIA. 17

covered on dry liiUs of the coast at the same point by W. G. Wright.,
January, 1889, no. 1201 (distributed as Z. rnaritima ?) ; also by F. H.
Lamb on dry rocky cliffs at same place, 26 December, 1891, no. 325
(distributed as Z. inaritima). In the presence of good material of all
three species we cannot doubt the complete distinctness of tliis species
from the preceding and from the following.

§ 3. Herbs with ovate or elliptic-oblong leaves : head' strongly
bicolorous ; disk dark purple-brown, nearly black ; rays oblong, bright
yellow.

9. Z. rnaritima, HBK. Prostrate, much branched from the base :
stems spreading: leaves elliptic or elliptic-oblong, obtuse at the apex,
rather abruptly contracted to short but distinct pubescent petioles : heads
8 to 10 lines in diameter (incl. rays), terminal, mostly borne on long
naked peduncles : chaff narrowed above although obtusish at the
dark-colored point ; achenes of the disk-flowers strongly callous-
margined : rays oblong, golden yellow ; ray-acheues ^\ to 2 lines long.
— Nov. Gen. & Spec. iv. 25 1 , not Gray, I.e. — Acapulco, Humboldt Sf
Bo upland ; rediscovered at the same point by Palmer, February, 1895,
uo. 523. From its peduncled heads, oblong rays, callous-margined
achenes, etc., we cannot doubt that Dr. Palmer's plant represents
the real Z. rnaritima which came from the same locality so long ago.
Yet Palmer's plant has leaves considerably larger than those described
by Kunth.

10. Z. Palmeri, Gray. Erect branched annual, a foot or so in
height : leaves ovate, or lanceolate from an ovate cordate or subcordate
closely sessile base, acute or acuminate at the apex. — Proc. Am.
Acad. xxii. 423. — Jalisco, at Tequila, Palmer, no. 386, Pringle, no.
4557 ; also at Manzanillo, Palmer, no. 893, and Colima, Palmer, no.
893 a.

§ 4. Erect perennials with spreading branches and narrow linear or
oblong leaves: rays oblong, 4 to 12 lines in length, deep orange-
colored as well as the disk-flowers.

11. Z. linearis, Benth. Much branched, 8 inches to a foot or
two in height: heads many and very showy, of intense orange color
(persisting evt-u in old dried specimens): leaves linear or nearly so. —
PI. Hartw. 17. — Aguas Calientes, Hartweg, no. 117; San Luis
Potosi, near Morales, Schaffner, no. 210, and in same state by Parry
^ Palmer, no. 441 ; Jalisco, on the Rio Blanco, Palmer, no. 54, and
on hills near Guadalajara, Pringle, no. 1778.

VOL. xxxiJ. — 2

18 PROCEEDINGS OP THE AMERICAN ACADEMY.

"Var. latifolia, Rose. Leaves somewhat broader, oblong. —
Contrib. U. S. Nat. Herb. i. 102. — Alamos, Palmer^ no. 352.

§ 5. Distinctly annual herb with showy flowers : disk and rays
concolorous, yellow, red, or purple, or if discolorous the disk-flowers
yellow or greenish and the rays red or purple : leaves ovate, lanceo-
late, or elliptic.

» Leaves sessile or nearly so, entire.

•t- Achenes of the disk-flowers sliort and broad, obovate, 2 to 2\ lines in
length : stems hirsute with spreading hairs.

12. Z. angustifolia, HBK. Branching, H to 2 feet high:
leaves lance-linear from a sessile ovate base : disk very convex,
orange-yellow from the abundant acute exserted chaff (orange-yellow
minutely tipped with purple at the very summit); disk-flowers at first
orange and becoming darker with age : rays paler yellow\ — Nov.
Gen. «fe Spec. iv. 251 ; DC. Prodr. v. 536. — Originally collected in
the neighborhood of Guanajuato by Humholdt Sf Bonpland, and de-
scribed as having leaves " scarcely 2 lines broad." Our only authentic
specimen of this species is one of Meridez sent by De Candolle to
Dr. Gray and mentioned in the Prodromus as of this species. From
other details of the description we cannot doubt that De Candolle was
right in referring this plant to Kunth's species, although in it and in
all the following specimens attributed to this species the leaves are con-
siderably broader than originally described, varying from 3 to 8 lines
in breadth. Evidently identical with Mendez' plant are specimens
collected at Morelia, Ghieshreglit ; no. 369 {"iZ. Ghieshreghtii , Verlot,
Rev. Hort. 1862, 368; Vilmorin, Fl. PI. Terre, 971,^.2', Mexicana,
Hort. fide Vilmorin, 1. c), and near La Barca, Jalisco, Pringle, no.
3866 ; also cultivated plants from the Harvard Botanic Garden, dated
1861 and 1865. From this species, notwithstanding its still broader
leaves, we cannot confidently separate the horticultural Z. Haageana,
Regel, Gartenfl. x. 355, ex char.

1 3. Z. elegans, Jacq. Erect : stems less branched : leaves
broader, ovate or elliptic, closely sessile and clasping, inch or so in
breadth : disk-flowers yellow or orange, scarcely or not at all exceeded
by the purplish tipped chaff; rays numerous, purple or lilac. — Coll. &
Icon. PI. Rar. iii. t. 589 ; Sims, Bot. Mag. t. 527. — South Mexico,
Ziticuaro, Hartweg ; mountains of Tixtla near Cuernavaca, Berlandier,
no. 975 ; also without locality, GIdesbreght, no. 306. Extensively culti-
vated in various countries.

ROBINSON AND GREENMAN. — fiENUS ZINNIA. 19

■t- t- Achenes longer, narrower, oblong, ■) or 4 lines in length.

•w Hirsute with spreading hairs : rays rather broad, patulous, yellow

as well as the disk.

14. Z. pauciflora, L. Erect roughish-pubescent annual, some-
what corymbosely branched above : peduncles at maturity enlarged
upwards and fistulous : the yellow heads about an inch in diameter. —
Spec. ed. 3, 1269; Lam. 111. t. 685, f. 1 ; DC. Prodr. v. 535. Z. lutea,
Giertn. Frucht. ii. 459, t. 172. — Chiapas, Mexico, Nelson, no. 3074;
Andes of Peru, Mathews, no. 456, and Bolivia, Mandon, no. 38 ; also St.
Thomas, W. I., Eggers, no. 400. Introduced in W. Africa at Cape Verd.

*+ ++ Pubescence of the stem much finer, appressed or very rarely spreading:
rays red or purple, mostly narrow and suberect or scarcely spreading.

15. Z. multiflora, L. 1. c. Erect annual with habit of the pre-
ceding, or subsimple : leaves ovate and acute or ovate-oblong and
obtusish : peduncles (frequently short or none) often thickened upwards.
— L. f. Dec. Ups. 23, t. 12 ; Curtis, Bot. Mag. t. 149. Z. tenui flora,
Jacq. Icon. PL Rar. iii. 590. Z. leptopoda, DC. Prodr. v. 535 (merely
weak form or state with more slender peduncles). Z. Floridana, Raf.
New Fl. iv. 70. Z. intermedia, Engelm. in Wisliz. Tour N. Mex. 107.
Z. Mendocino, Philippi, Sert. Mendoc. Alt. 21, fide Baker, Fl. Bras. vi.
pt. 3, 178. — The commonest and most widely distributed species ex-
tending from Florida, Chapman, Curtiss, no. 1417, to Texas, Drummond,
no. 115, Lindheimer, no. 93 (but in these localities probably introduced;
see Gray, Syn. Fl. i. pt. 2, 253) ; S. Arizona near Ft. Huachuca, Lem-
mon, no. 2761; throughout Mexico, Sonora, Wright, no. 1213, Thnrber,
no. 920, Schott, Hartman, no. 104; Chihuahua, Thurber, no. 832, Schott,

Wislizenus, Palmer, nos. 115, 156, Pringle, no. 316; Coahuila, Palmer,
nos. 574, 575; San Luis Potosi, Parry 8^ Palmer, no. 438; Orizaba,
^OM/'yef«<, no. 1682, Botteri, nos. 73, 516, 940, Seaton, no. 345; Cor-
dova, Bourgeau, no. 1633; Guadalupe, Bourgeau, no. 500; Oaxaca,
Andrieux, no. 314, L. C. Smith, nos. 813, 957 ; Chiapas, Ghiesbreght, no.
126; Guatemala, Heyde ^ Lux (no. 3808 of Donnell Smith's sets);
Venezuela, Fendler, no. 1974; Bolivian Andes, Mandon, nos. 39, 40,
and Bang, no. 207.

The yellow-rayed specimens, referred by various authors to this spe-
cies, have in most instances, if not always, the pubescence of Z. pauci-
flora, to which it seems best to refer them.

* * Leaves elliptic, petiolate, dentate.

Z. Liebmannii, Benth. & Hook. f. Leaves squamulose-hirsute
above, ferrugineous beneath : peduncles (|uadrangular-sulcate : chaff

20 PROCEEDINGS OF THE A3IERICAN ACADEMY.

aculeate or uncinate. — Benth, & Hook. f. ace. to Klatt, Leopoldina,
xxiii. 2, where first described. — Mexico on the Rio Taba, Liebmann^
no. 552.

Doubtful species known cliiefly or exclusively from cultivated specimens.

Z. RcEZLii, Hort. Gard. Chron. 1872, page 1392 ; Hook. f. &
Jackson, Ind. Kew. ii. 1251, is a mere horticultural name for a yellow-
flowered annual species said to come from Mexico, but never properly
described.

Z. VERTiciLLATA, Andrews, Bot. Rep. iii. t. 189, is apparently only
a robust cultivated form of Z. mulli flora., with verticillate leaves, and
double series of rays : said also to come from S. Mexico.

Z. HYBRiDA, Romer & Usteri, Mag. Bot. St. 1 (1787), 49; Curtis,
Bot. Mag. t. 2123 {Z. grandiflora, Hort. fide DC. Prodr. v. 536, not
Nutt), is an annual with deep red rays, greenish disk, and chaff not
fringed at the apex : apparently only a form of Z. multiflora, with i-ays
becoming broad and somewhat double by cultivation.

Z. AMBiGUA, Salm-Dyck, and Z. discolor, Hort., are names only,
and wholly obscure.

IV. — REVISION OF THE MEXICAN AND CENTRAL
AMERICAN SPECIES OF THE GENUS GALEA.

GALEA, L., R. Br. (Name of obscure origin. The derivation
from KaAds, beautiful, is unsatisfactory, and at best very doubtful.) —
Heads mostly small or of medium size, radiate or discoid. Involucre
ovoid, cylindrical, or campanulate; its scales pluriseriate, imbricated,
usually very unequal, the outer gradually shorter, all scarious or the
outer (rarely all) herbaceous or herbaceous-tipped. Receptacle small,
convex, or flattish, paleaceous : chaff scarious, concave, rigid or thin and
hyaline. Ray-flowers when present fertile ; ligules yellow, white, or
roseate, entire or denticulate at the apex. Disk-flowers fertile, yellow or
white ; the limb of the corolla regular, deeply 5-cleft. Anthers appen-
daged at the apex and shortly sagittate-lobed at the base. Style-branches
subtruncate or with a very short appendage. Achenes slender, subterete
or more or less distinctly 4-5-angled, usually pubescent: pappus of 4 to
20 subequal scales ; these mostly fringed or ciliolate, rarely wanting,
when numerous narrow and acuminate or when fewer usually short and
blunt. — About 85 species of shrubs and perennial herbs (rarely climb-
ers), extending from Mexico to Tropical S. America. Leaves opposite,
simple, mostly ovate, oblong, or lanceolate, sessile or petiolate, mostly

ROBINSON AND GREENMAN. — GENUS GALEA. 21

serrate or dentate, rarely entire (in one or more S. American species
piuuatifi(l). Tiie limits of the genus here taken are essentially those of
Beutham & Hooker (Gen. ii. 390), and Hoffmann (in P^ngl. & Prautl,
Nat. Pflanzenf. iv. Ab. 5, 246), whose generic synonymy need not here
be repeated. The Mexican and Central American species may be sub-
divided according to the following characters.

Subgenus 1. Leontophthalmum, Benth. & Hook. f. Heads very
large (inch or more in diameter), radiate, few or solitary, long-peduncled :
scales of the involucre few-seriate, broad, the outer often herbaceous :
both disk- and ray-flowers yellow : scales of the pappus numerous. —
Gen. ii. 391. Leontophthalmum, Willd. Gesellsch. Natur. Fr. Berl. Mag.
1807, 40 ; HBK. Nov. Gen. & Spec. iv. t. 409. — Mostly S. American
species, only the following known from Mexico.

1 . C. megacephala. Erect, herbaceous, 2 or 3 feet high : stem
striate-angulate, hirsute-pubescent, leafy to the middle and terminating
in one or (more rarely) three long naked 1-headed peduncles (often 18
inches in length) : leaves thin, ample, hirsute upon both sides, rhombic-
ovate or with deltoid acute or obtusish coarsely dentate blade (3 to 4
inches long, nearly as broad) abruptly contracted at the base, then grad-
ually attenuate into a winged entire petiole of nearly equal length ; the
lowermost leaves smaller, obovate, and with rounded apex: heads, exclu-
sive of raj^s, 9 to 12 lines in diameter: flowers deep orange; ray-flowers
15 to 20, with oblong spreading ligules half inch in length: disk conical.
— Collected by E. W. Nelson at Sta. Efigenia, Oaxaca, altitude 500 feet,
18 July, 1895, no. 2844, and on top of ridge back of Tonala, Chiapas,
altitude 1,200 to 2,500 feet, 10 August, 1895, no. 2884. A plant with-
out close affinities in the Mexican species of the genus, but related to
several of the S. American.

Subgenus 2. Oteiza, Llav. (as gen.). Heads few or solitary, large,
9 to 15 lines in diameter, loosely cymose or (in C. elegans) somewhat
densely grouped at the ends of the branches : rays long (nearly or quite
half inch in length) white or roseate : leaves sessile or nearly so.

* Leaves oblong, narrowed at tlie base.

2. C. Palmeri, Gray. Herbaceous, erect or slightly decumbent :
stem simple or divided almost from the base, pubescent, 18 inches to 2
feet high : leaves 2 to 3 inches long, 4 to 6 lines broad, denticulate and
ciliated, 3-nerved : heads 1 to 9, in terminal loose cymes ; slender pedun-
cles nearly naked ; the floral leaves short and linear : involucral bracts
green, few-seriate and more nearly equal than is usual in the genus. —
Proc. Am. Acad. xxii. 430. — On the Rio Blanco, Jalisco, Palmer^ no.

22 PROCEEDINGS OF THE AMERICAN ACADEMY.

147; on slopes of a barranca near Guadalajara, Prmgle, uo. 2904; and
Michoacan, on grassy hills near Patzcuaro, Pringle, no. 4125 ; fl. July.
* * Leaves ovate, abrupt or cordate at the sessile or subsessile base.

3. C. elegans, DC. Leaves thickish, quite glabrous above, finely
pubescent upon the veins beneath, ovate to ovate-lanceolate, few-toothed,
long-acumiuate, considerably paler beneath : heads loosely or somewhat
closely cymose at the ends of the branches : ra3s about 8 or 10. — Prodro
V. 674. Oteiza acuminata, Llav. Reg. Trim. Mex. 1832, 41. — Valley
of San Luis Potosi, Schaffner, no. 237, and Parry 8^ Palmer^ no. 491 ;
and in Tultenango Canon, State of Mexico, Pringle, no. 4297, with
heads more aggregated ; fl. August till last of October. An imperfect
specimen collected by Baites, 1846, in Mexico, without more precise
locality, is probably of this species.

4. C. multiradiata, Seaton. Herbaceous, erect or decumbent,
with habit of the preceding but with sessile leaves thinner, less attenuate,
appressed-pubescent upon both surfaces and scarcely paler beneath : heads
on very long peduncles, 3 or 4 in number : rays 15 to 20. — Proc. Am.
Acad, xxviii. 120. — Wooded slopes of Mt. Orizaba, altitude 10,000 feet,
Seaton, no. 167, in part. In flower in August. With the type material
of this species in the Gray Herbarium were associated specimens of a
Sabazia and Tridax of similar habit.

5. C. sabazioides, Hemsl. " Slender herb, 1 to 1^ feet high, pro-
cumbent, rooting at the base, in habit very similar to Sabazia sarmentosa :
leaves decussate, ovate, petiolate, acute, crenate, 3-nerved, hirsute : heads
solitary, radiate, as large as in Tridax : bractioles [chaff] scarious, uni-
nerviate, pale fuscous, oblong-ovate, 2-dentate beneath the gradually and
narrowly acuminate summit: rays purplish white." — Biol. Cent.- Am.
Bot. ii. 206. Allocarpus sabazioides, Less. Linnaea, ix. 590. — Near San
Miguel del Soldado and La Joya, S. Mexico. No specimens exactly
agreeing with this description have been seen by the writers, and the
characters are here translated from the Latin of Lessing.

Subgenus 3. Eucalea, Benth. & Hook. f. (extended). Heads
relatively small and numerous, in close corymbs or somewhat umbellate :
rays short or none : scales of the pappus 7 to 20 : leaves sessile or short-
petioled, serrate or dentate (in one species subentire). — Gen. ii. 391.

* Inflorescences terminal, close, on long naked peduncles : upper leaves reduced to
small oblong or linear bracts : beads homogamous : pappus often reduced or
wholly wanting. — Cali/dermos, Lag.
I- Scales of the involucre densel}' fringed with yellowish glandular-tipped hairs.

6. C. thysanolepis. Slender erect herb, 18 inches to 2 feet high:
stem simple, striate, densely pubescent, prolonged at the summit into a

ROBINSON AND GREENMAN. — GENUS GALEA. 23

nearly naked peduncle bearing a dense cymose cluster of heads : leaves
ovate, acute, subcordate at the base, sharply dentate, roughish-pubescent
upon both surfaces, 3-5-nerved and coarsely reticulate-veined, 1 to 3
inches long, a third or half as broad, scarcely paler beneath, closely ses-
sile : inflorescence and especially the margins of the 3-seriate involucral
scales densely glandular-pubescent: heads about 18-flowered: chaff ir-
regularly erose, often constricted below the summit: achenes calvous,
terete : corolla pubescent. — Collected by E. W. Nelson on the summit of
the Sierra Madre near Chilpancingo, Guerrero, altitude 9,000 to 10,200
feet, in llower 24 December, 1894.

-(- -1- Scales of the involucre essentially glabrous.

7. C. pediincnlaris, IIBK. Erect or decumbent, 1 to 2 feet high,
pubescent and usually more or less scabrous : leaves ovate, closely sessile
by a broad abrupt or subcordate base, dentate, H to 3 inches long, nearly
half as broad : scales of the involucre yellow : pappus present, about two
thirds the length of the achenes. — Nov. Gen. & Spec. iv. 295, t. 408.
Calydermos pedimcularis, DC. Prodr. v. 669. C. scaber. Lag. Nov.
Gen. 25. Calebrachys pedimcularis, Cass. Diet. Sci. Nat. Iv. 277. —
South Mexico, Mt. Jorullo, at 3,000 feet, Humboldt <§• Bonplcmd ; Mira-
dor, Sartorius ; Chiapas, Ghiesbreght, no. 783, and^. W. Nelson, no. 32G1 ;
mountains near Guapimalpatt, Mexico, Schaffncr, no. 35; Chinantla,
Liebmann, no. 413 ; Boerego near Orizaba, Bourgeau, no. 3149; Mt.
Orizaba, altitude 4,000 feet, Seaton, no. 126, and altitude 10,000 feet by
same collector, no. 168; also in Orizaba, Hahn, no. 2692, Bilimek, no.
544, and Botteri, nos. 620, 803 ; Jalapa, Th. Coidter, no. 334, in part.

Var. epapposa, HBK. Closely similar in foliage and habit : achenes
wholly destitute of pappus. — HBK. ^(/e DC. Prodr. v. 669. — Near
Santa Rosa, ace. to HBK. 1. c. ; also San Luis Potosi, Parry 8^- Palmer,
no. 497 ; Las Canoas in the same state, Pringle, no. 3672 ; Valley of
Mexico near the Santa Fe, Bourgeau, no. 718 ; Jalapa, Th. Coulter, no.
334, in part, and near Reyes, Oaxaca, altitude 6,700 to 10,000 feet, E.
W. Nelson ; also Mexico without locality, Th. Coulter, no. 250, in part
(the other part of the same number being a Eupaforium) ; fl. August to
October.

Var. longifolia, (Lag.) Gray. Leaves lanceolate or lance-oblong,
opposite or ternate, relatively much narrower and more elongated ; the
lower 3 to 6 inches long: scales of the involucre yellow : pappus none.
— Proc. Am. Acad. xxii. 430, as to syn. Calydermos longifoUus, Lag. and
PI. Hartw. Calydermos scaler, var. Benth. PI. Hartw. 346. — Bolanos,
Hartweg, no. 122, at least as to long-leaved specimens.

24 PROCEEDINGS OP THE AMERICAN ACADEMY.

Var. livida. Leaves narrow, lanceolate to lance-oblong, very sca-
brous-pubescent : iuvolucral scales dark purple : pajjpus present. — C.
peduncularis, var. hngifolia, Gray, as to pi. Palmer. — On the Rio
Blanco, Jalisco, Palmer^ no. 317 ; and in damp shady canons near
Guadalajara, Primjle^ no. 2326 ; fl. August to October.

8. C. Liebmannii, Schz. Bip. Branches terete : leaves petiolate,
ovate-elliptic, 2 inches long, 9 lines broad, coriaceous, 3-nerved, glabrous
above, ferrugineous and chrysopunctate beneath : heads short-pedicelled,
6-flowered, disposed at the summits of the peduncles in umbelliform
corymbs : involucres cylindraceous : scales 3-seriate, ovate, subreflexed at
the apex : chaff membranaceous, ovate, acute : achenes glabrous ; scales
of the pappus 10, short. — Leopoldina, xxiii. 6. — GnalulM, Li ehmann,
no. 411. Not seen by the writers, and doubtfully referred to this subdi-
vision. The characters are drawn from the original description of Klatt.

* * Inflorescences more numerous, short-peduncled or sessile, together

forming leafy panicles.
•4- Pappus-scales rather few, ovate to lanceolate, shorter than the achenes.

*•*■ Inflorescences lax, somewhat umbelliform : pedicels slender and relatively long,
most of them considerably exceeding the heads in length.

9. C. salmesefolia, Hemsl. 1. c. Shrub, 2 to 3 feet high with
rhombic-ovate leaves acute at both ends, smoothish and lucid above, 1 to
li inches long, half as broad: heads 8-14-flowered. — Cnlydermos
salmecefoliiis, DC. Prodr. v. 670. — Between Tula and Tampico, Ta-
maulipas, Berlandier, nos. 718, 2135.

*+ *+ Inflorescence denser : pedicels mostly very short, seldom equalling or exceed
ing the heads (except in C. Zacatechichi, var. macrophiiUa).

= Involucre of most or all of the heads closely subtended by a few broad obtuse

herbaceous bractlets.

10. C. albida. Gray. A branching pubescent shrub: leaves ovate,
shortly petiolate, serrate, acutish at the apex, abruptly contracted at the
broad obtuse base, scabrous above, somewhat paler and moderately pubes-
cent below, 1 to li- inches long, two thirds as broad: heads numerous,
cymosely clustered at the ends of the leafy branches, on pedicels 1 to 9
lines in length, homogamous or with one or more reduced ray-flowers ;
flowers whitish : achenes pubescent, scales of the pappus acute. — Proc.
Am. Acad. xv. 38. — San Luis Potosi, in mountains near Morales,
Schnffner, no. 269, and later in the same state. Parry S)- Palmer, no.
448 ; in fl. in August.

11. C. hypoleuca. Similar in habit : leaves somewhat larger,
broadly ovate or suborbicular, sessile or very shortly petioled, scabrous

ROBINSON AND GREENMAN. — GENUS GALEA. 25

above, densely tomentose and canescent beneath, 1^, to 2 inches in
diameter : heads very short-pedicelled or subsessile : scales of the pappus
obtuse or rounded at the summit. — Oaxaca, Sierra de San Felipe,
altitude 6,000 feet, 7 September, 1894, Prlngle, no. 5784; dry hills in
the Valley of Oaxaca, altitude o, 100 to 5,800 feet, 8 September, 1894,
E. W. N^elson, no. 1217, and in neighboring locality without date, Nelson,
no. 1192; also in Valley of Etla, September, 1895, Alvarez, (L. C.
Smith's) no. 7G6.

= = Involucre nearly or quite naked, without subtending herbaceous bractlets,
or these small and scattered, lanceolate or subulate.

a. Leaves wholly glabrous, quite smooth and free from resinous dots or globules.

12. C. Nelsonii. Glabrous, copiously branched and very leafy:
leaves rhombic-ovate, coarsely and rather bluntly toothed, 3-uerved,
short-petioled, paler beneath, quite smooth but scarcely or not at all lucid
upon both sides, l}j to 2 inches long, half as broad: heads very numerous
in many terminal or subterminal close cymes, together forming a con-
siderable pyramidal inflorescence : involucral bracts light colored, striated,
obtuse : disk-flowers about 7 ; ray-flowers 2 or 3, with very short obscure
ligules : pappus of about 1 2 scales. — Collected by £J. W. Nelson on the
top of ridge back of Tonala, Chiapas, altitude 1,200 to 2,500 feet, 10
August, 1895, no. 2887.

b. Leaves pubescent, or at least covered on the lower surface with resinous atoms.

13. C. Zacatechichi, Schlecht. Much branched shrub with harsh
scabrous foliage : leaves rhombic-ovate to ovate-obloncr, short- oetioled.
described as deeply crenate, but more often serrate-dentate, acute, cune-
ate at the base: heads about 12-flowered, very short-pedicelled or ses-
sile in numerous small terminal cymes : scales of the involucre with
scarious and undulate margins. — Linnaea, ix. 589 ; DC. Prodr. v.
672; HofFm. in Engl. & Prantl, Nat. Pflanzenf. iv. Ab. 5, 246, f. 120,
A-C. — On hills near Hacienda de la Laguna and Jalapa, Schiede ;
Mirador, Sartorius ; Orizaba, Botteri, nos. 481, 488; San Pedro Sula,
Dept. Sta. Barbara, Honduras, Tliieme (no. 5300 of Donnell Smith's
sets, a form approaching var. macrophylla), and Guanagaza, Dept. Sta.
Rosa, Guatemala, Heyde ^ Lux (no. 6159 of same sets). The leaves
are reputed a remedy for cholera. A form probably of the same spe-
cies, but having the involucral bracts less scarious-undulate and more
often ciliolate, has been collected on hills near Guadalajara, by Palmer,
no. 352, and Pringle, no. 2475. Another form collected by Pringle in
the Sierra Madre near Monterey, Nuevo Leon, no. 2224, differs only in

26 PROCEEDINGS OF THE AMERICAN ACADEMY.

having the involucre calyculate with 1 to 3 lanceolate herbaceous-tipped
bractlets.

Var. rugosa. Habit, foliage, and inflorescence of the preceding
species: heads smaller, 5-10-flowered. — C. rugosa, Hemsl. Biol. Cent.-
Am. Bot. ii. 206. Calydermos rugosus, DC. Prodr. v. 670. — Cuerna-
vaca, Morelos, Berlandier, no. 1061 ; also in Orizaba, Bourgeaii, no.
3095 ; near Acapulco, Palmer, no. 52.

Var. macrophylla. Leaves mostly much larger, 2\ to 3 inches
or more in length, half as broad, less rugose : pedicels often equalling or
considerably exceeding the 12-flowered heads: achenes nearly 1^ lines in
length, the pappus only a third as long. — Collected by H. von Tuerckheim
at Cohan, Dept. Alta Verapaz, altitude 4,300 feet, February, 1888, and
distributed as C. Zacatechichi in Dounell Smith's Guatemalan sets, no.
1345 ; also by Heyde 8^ Lux, at Laguna de Ayarza, Dept. Jalapa,
altitude 8,000 feet, distributed in same set as G. salmecefolia, no. 3782.

I- -1- Scales of the pappus narrower, more numerous, equalling or exceeding

the achenes.

++ Involucre cylindrical, or narrowly campaiiulate, rather few-flowered ; scales
broad, scarious, glabrous or only the outermost herbaceous and somewhat
pubescent : pedicels less than an inch in length : plants pubescent.

= Heads numerous, radiate, with ligules evident.

a. Inflorescences terminal, cymose-paniculate or somewliat corymbose : heads
very numerous, small : involucre ecalyculate.

14. C. integrifolia, Hemsl, Shrub, 4 to 8 feet high: leaves ovate-
lanceolate, attenuate at the apex, rounded or obtusish at the short-
petioled base, cuspidate-denticulate, varying from papillose-pubescent
and very scabrous to smoothish and lucid, 2 to A'k inches long, a third to
half as broad; the veins very prominent beneath : heads 15-20-flowered :
disk-flowers yellow ; ray-flowers 5, white, destitute of pappus. — Biol.
Centr.-Am. Bot. ii. 205. AUocarpiis integrifolius, DC. Prodr. v. 676.
— Originally collected in Mexico without more exact locality by Kar-
winski ; later in JMirador by Sart07-ius, by Liebmann, no. 418, and by
E. W. Nelson, no. 86 ; Chiapas, Ghiesbreght, no. 565 ; Cordova, Bourgeau,
no. 1751, and A. Gray ; near Orizaba, altitude 4,000 feet, Pringle, no.
5915, and E. W. Nelson, no. 2; Oaxaca, on the Sierra de San Felipe,
altitude 8,000 ieet, Pr ingle, no. 6111 ; Cieneguilla, L. C. Smith, no. 381 ;
between Panixtlahuaca and Jaquila, altitude 1,000 to 5,000 feet, E. W.
Nelson, no. 2389; City of Mexico, Mrs. D. H. Sheldon; Guatemala, at
Cohan, Dept. Alia Verapaz, altitude 4,300 feet, von Tuerckheim (no. 379
of Dounell Smith's sets) ; Guarda Viejo, Dept. of Guatemala, altitude

ROBINSON AND GREENMAN. — GENUS GALEA. 27

5,000 feet, Donnell Smith, no. 2345 ; Teocinte, Dept. Sta. Barbara,
altitude 2,500 feet, Heyde Sf Lax (no. 4199 of Donnell Smith's sets) ;
San Rafael, Dept. Zacatepequez, altitude 6,500 feet, Donnell Smith, no.
2332. One of the commonest species of S. Mexico and Central
America.

Var. dentata. Coulter. Leaves with more pronounced dentation,
long caudate-acuminate : the floral oblong-linear, attenuate. — Bot. Gaz.
XX. 51. — Nebaj, Dept. Quiche, Guatemala, altitude 7,000 feet, Heyde S^
Lux (no. 4506 of Donnell Smith's sets). Hither we would refer Nelson's
no. 2513, collected between Suchiotepec and Miahuatlan, Oaxaca.

b. Inriorescences cymose-umbellate in tlie upper axils, togetlier forming a leafy
elongated or tliyrsoid panicle : involucre commonly calyculate with one or
more herbaceous bractlets.

15. C. axillaris, DC. Shrub: leaves ovate-lanceolate or ovate-
oblong, attenuate-acuminate, sharply serrate. — Prodr. v. 673. Mocinna
serrata, Lag. Nov. Gen. 31. — Mexico, Hcenke ; Valley of Cordova,
Bourgeau, no. 1675; between San Luis Potosi and Tampico, Palmer,
no. 1111. Passes into

Var. urticsefolia. Leaves shorter and relatively broader, ovate,
acute or acutish to barely acuminate, crenate-serrate. — Caleacte urtici-
folia, R. Br. Trans. Linn. Soc. xii. 109. Galea urticcBfolia, DC. 1. c.
674. — The commoner form, Mexico without locality, Gregg^ nos. 1002,
1042.; Orizaba, Schaffner, and A. Gray ; Wartenberg near Tantoyuca,
Huasteca, Ervendherg, no. 96 (passing to typical form) ; Colima, Palmer,
no. 1215 ; Jalisco, on rocky slopes near Guadalajara, Pr/w^'/e, no. 1788;
and in neighboring locality on Rio Blanco, Palmer, no. 675 (robust
form with ternate leaves) ; Guatemala on the Rio Amatitlan, altitude
3,900 feet, Donnell Smith, no. 2337, also Jumaytepeque, Dept. Santa
Rosa, altitude 6,000 feet, Heyde S^ Lux (no. 3790 of Donnell Smith's
sets) ; Costa Rica at Navarro, altitude 3,500 feet, Donnell Smith, no.
4857 ; Nicaragua, Wright.

= = Heads subumbellate at the ends of the branches or from the upper axils,
discoid : Chiapas and southward.

16. C. prunifolia, HBK. Shrub: leaves broadly elliptic-ovate,
crenate, obtuse, abruptly contracted to a subcuneate base, slender-peti-
oled, scabrous and rugose above, somewhat paler and scabrous beneath,
2 to 3^^ inches long, two thirds as broad: heads about 18-flowered. —
Nov. Gen. & Spec. iv. 294, t. 406. — A South American species
reaching the Isthmus of Panama, where collected by Seemann and later
by Hayes.

28 PROCEEDINGS OF THE AMERICAN ACADEMY.

17. C. trichotoma, Donuell Smith. Densely fuscons-toraentose:
leaves ovate, acutish, crenate-serrate, scabrous above, somewhat canes-
cent-tomentose beneath, 1 to 2} inches long, one half to two thirds as
broad ; stems somewhat tortuous as though scaudeut. — Bot. Gaz. xiii.
299. — Rocky mountain sides near Cohan, Dept. Alta Verapaz, altitude
4,300 feet, voM Tuerckheim (no. 1353 of Donnell Smith's sets); Chiapas,
table land about Ocuilapa, altitude 3,400 to 3,800 feet, E. W. Nelson,
no. 3004.

*-*■ ++ Heads campanulate, many-flowered : scales of the involucre glabrous : leaves
very large (2^ inches or more in breadth) : pedicels less than an inch in
length : Costa Rican species.

18. C. pellncidinervia, Klatt. Leaves membranaceous, broadly
ovate, 4 inches long, two thirds as broad, long-acuminate, serrate-dentate,
rounded at the base ; petiole 5 lines long, densely pilose : ligules 6, 2 lines
in length. — Bull. Soc. Bot. Belg. xxxi. 207. — In woods at Terraba,
altitude 900 feet, Pittier, nos. 3707, 3726. A species not seen by the
writers ; the description is condensed from the original characterization.
++++++ Heads campanulate, many-flowered : scales of the involucre multlseriate,

sub-herbaceous, very pubescent : pedicles half inch or less in length : leaves
smaller : Mexican species.

19. C. scabrifolia, Benth. & Hook. f. Shrub: leaves ovate-
oblong or oblong-lanceolate, attenuate, serrate or serrulate, thickish
above papillose-pubescent and very scabrous, below sparingly pubescent,
slightly paler and with veins very prominent : pedicels and iuvolucral
bracts densely canescent-tomentose : disk-flowers 125 or more; ray-
flowers 20 to 25, the latter without pappus: plant drying green. — Benth.
& Hook, f. ace. to Hemsl. Biol. Centr.-Am. Bot. ii. 206. Allocarpus
scabnfolius, Hook. & Arn. Bot. Beech. 300. Ferdinanda oppositifoUa,
Schz. Bip. in Seeraann, Bot. Herald (by error cited as Zaluzania by
Benth. & Hook. f. Gen. ii. 391). Perymenium album, Wats. Proc. Am.
Acad. XXV. ■■ 54. — N. W. Mexico in Sierra Madre, Seemann ; Jalisco, on
mountains near Lake Chapala, Pringle, no. 2438, in fl. December ;
Alamos, Palmer, no. 283.

20. C. submembranacea, Fernald. Similar in habit: leaves

thinner, nearly smooth and somewhat lucid, somewhat nigrescent in

drying. — Bot. Gaz. xx. 535. — • On mountain sides, Zopelote, Tepic,

altitude 3,000 feet. Lamb, no. 554.

+++*++++ Heads few ; campanulate : scales of the involucre broad, few-seriate,
glabrous: pedicels very long (1| to 2 inches) : plant essentially glabrous.

21. C. longipedicellata. Shrub, glabrous throughout except the
puberulent summits of the long pedicels: leaves elliptical, acutish at both

ROBINSON AND GREENMAN. GENUS GALEA. 29

ends, callous-denticulate, short-petioled, paler beneath, 3-nerved, 1^- to 2
inches long, 9 or 10 Hues broad: heads 2 to 4 at the ends of the branches
(springing from the opposite axils of the last pair or two pairs of leaves),
half inch in diamt-ter, homogamous : outer involucral scales scarcely
shorter than the inner, broadly oblong, rounded at the apex : scales of
the pappus linear-attenuate, about 20 in number : achenes glabrous. —
Collected by E. W. Nelson, near Choapam, altitude 3,800 to 4,500 feet,
28, 29 July, 1894, no. 898.

^- 4- -(- Doubtful and poorly cliaracterized species, probably referable to this
subdivision and very likely synonymous with some of the foregoing species.

22. C. cacosmioides, Less. " Bracteoles [by which is meant ap-
parently the chojf] broad, at the apex laciniate, acuminate and aristate :
involucres cylindrical, about 12-flowered: differing from C. solidaginia,
Kunth, not only in the form of the bracteoles, but also in the radiate
heads, and in the leaves obtuse at the base or acute but not acuminate
into petioles, sharply serrate, subglabrous, lucid above." — Liuntea, v.
157. — In open places near Jalapa, October, Schiede ^ Dep-pe. The
characters translated from the original Latin description.

23. C. brachiata, DC. "Shrub: branches brachiate : leaves ovate,
obtusely mucronate : heads fasciculate and mostly terminal." — Prodr.
V. 673 (whence descr.). Mocinna brachiata, Lag. Nov. Gen. 31. Galin-
sogea brachiata, Spreng. Syst. iii. 579. — "In Panamaide" ace. to
Lagasca, 1. c.

Subgenus 4. Tetrachyron, Benth. & Hook. f. Heads corymbose,
radiate : leaves oblong or lanceolate, often with interpetiolar dilation :
both disk- and ray-flowers yellow: scales of the pappus only 4. — Gen.
ii. 391. Tetrachyron, Schlecht. Linncea, xix. 744.

24. C. manicata, Benth. & Hook, f., 1. c. A tall glabrous shrub,
8 to 12 feet in height: leaves lanceolate, gradually narrowed at both
ends, serrate, 4 or 5 inches long, 9 lines broad, nigrescent in drying:
involucres somewhat turbinate-campanulate, ecalyculate: corymbs flat, 2
to 6 inches or more in diameter: rays about 12 : scales of the pappus a
third to half as long as the achene. — Benth. & Hook. f. ace. to Hemsl.
Biol. Cent. -Am. Bot. ii. 206. Tetrachyron manicatum, Schlecht. 1. c.
— Cordillera near Vera Cruz, Galeotti, no. 2309 ; Mirador, Liebmann,
no. 392, Dr. Berendt, and E. W. Nelson, no. 85 ; Orizaba, Schaffner^
Botteri, no. 807 ; also Sierra del Borrego, above Orizaba, altitude 4,500
feet, Pringle, no. 6133.

25. C. Orizabaensis, Klatt. Branchlets pentagonal : lower leaves
petiolate, large, 5 inches long, 3 inches wide, oblong, sinuate-undulate,

30 PROCEEDINGS OF THE AMERICAN ACADEMY.

glabrous above, pilose on the nerves beneath : heads pedicellate in
lanceolate-bracted corymbs : scales of the involucre biseriate, ovate,
striate : rays 4 or 5, oblong ; disk-flowers externally pubescent : chaff
ovate, scarious, obscurely dentate at the apex : scales of the pappus
linear-lanceolate, laciniate at the apex, shorter than the tetragonal hirsute
achene. — Leopoldina, xxiii. 6. Tetrachyron Orizabaeiisis, Schz. Bip.^rfe
Klatt, 1. c. — Peak of Orizaba, altitude 8,000 to 10,000 feet, Liebmann,
no. 390. Not seen by the writers. Description translated and condensed
fi'om the original characterization.

Subgenus 5. Tephrocalea, Gray. Heads few or solitary, rather
large for the genus : scales of the pappus 4 to 5 : leaves ovate, entire,
j}.brupt at the base and slender-petioled, cauescent-tomentose or -tomentu-
lose beneath. — Proc. Am. Acad. xv. 38.

26. C. discolor, Gray, 1. c. Leaves ovate-oblong, obtusish, mu-
cronulate, glabrous above, very finely tomentulose beneath : heads about
5 at the summits of the branches, 9 lines in diameter including the
spreading rays : pedicels long, an inch or more in length, very finely
puberulent or nearly smooth. — Mexico without locality, Th. Coulter, no.
351. A very distinct species apparently never rediscovered.

27. C. tomentosa. Gray, 1. c. Leaves ovate, subcordate, rounded
at the apex, densely tomentose and canescent upon both surfaces when
young but glabrate above : heads solitary, terminal : peduncles very
tomentose, rather stout and somewhat thickened upward. — Between San
Luis Potosi and Tampico, Palmer, no. 1108. Not since collected.

Species of uncertain affinities.

28. C. sessiliflora, Less. Shrub: leaves very obtuse or subcor-
date at the base: heads discoid, about 10-flowered: involucre cylindrical:
chaff broad, obovate, muticous at the apex. — Linnaea, v. 158. — Mexico,
Hnmholdt. A species as yet wholly obscure ; the characters are drawn
from Lessing's scanty descri^Dtion.

ROBINSON AND GREENMAN. — GENUS POROPHYLLUM. 31

v. — A PROVISIONAL KEY TO THE SPECIES OF PORO-
PHYLLUM, RANGING NORTH OF THE ISTHMUS
OF PANAMA.

* Leaves with broad obtuse ovate or elliptic-oblong usually thin lamina, rather

abruptly contracted at the base to a slender petiole.
^- Heads very long ; involucral scales at maturity 8 to 12 lines in length : pe-
duncles conspicuously clavate : erect scarcely distinct annuals.

1. P. macrocephalum, DC. Prodr. v. 648.

2. P. ruderale, Cass. Diet. Sci. Nat. xliii. 56.

3. P. ellipticum, Cass. 1. c.

•(- ■*- Heads shorter ; involucral scales about half inch in length : peduncles

scarcely or not at all thickened toward the summit.

*+ Leaves with pellucid glands on the surface as well as at the margins.

4. P. Ervendbergii, Gray, Proc. Am. Acad. xix. 35.

5. P. nummularinm, DC. 1. c. 649.

•t-f t-f Glands of the leaves marginal or none,
= Leaves broad, thickish : petioles rather short : probably a single species.

6. P. viridiflomm, DC. 1. c. 648.

7. P. Lindenii, Schz. Bip. in Seem. Bot. Herald, 308.

= = Leaves thin and delicate.
a. Heads nodding or almost pendulous.

8. P. nutans. Leafy shrub with slender terete fuscous branches
(marked with small light colored lenticels) and almost filiform branch-
lets : leaves thin, elliptic, entire, or crenulate, obtuse but often mucronu-
late at the apex, obtuse or acutish at the base, 8 to 14 lines long, half as
broad ; marginal glands 4 to 6 on each side the leaf; petioles filiform, 3
or 4 lines long : heads numerous, about 9 lines long : peduncles clustered
at the ends of the branches in 3's and 4's : involucral scales 5, oblong,
obtuse, 6 lines long, somewhat carinate, green, with double row of linear
glands : flowers white or nearly so, 9 lines in length ; limb of the corolla
sub-bilabiate, shallowly toothed : achenes 5 lines in length. — Collected
by 0. G. Pr ingle, in mountains near Lake Chapala, 16 December, 1889,
no. 2976, and by the same collector on rocky hills, Querendaro, Michoa-
can, 25 October, 1892 ; later by E. W. Nelson, between Chilapa and
Tixtla, Oaxaca, altitude 5,200 to 7,000 feet, no. 2170. Mr. Pringle's
plants were at first determined as P. Ervendbergii and so distributed.
On subsequent comparison with that species they appear thoroughly
distinct, having only marginal glands on the leaves, nodding instead of
erect peduncles, less deeply cleft corolla, and longer achenes.

32 PROCEEDINGS OF THE AMERICAN ACADEMY.

b. Heads erect : slender annual.

9. P. Pringlei, Robinson, Proc. Am. Acad, xxvii. 178.

c. Heads erect : shrubs.

1 0. P. Jonillense, Cass. '' Leaves 1 inch long, eglandular : petioles
6 to 7 lines long : scales of the involucre green." Diet. Sci. Nat. xliii.
57. Kleinia Juridlensis, HBK. Nov. Gen. & Spec. iv. 156, t. 356, whence
description.

11. P. Nelsonii. Branching shrub, 1 to 2 feet high: stems terete,
striate, purple : leaves chiefly opposite, rather small and distant ; blade
elliptic or oblong, obtuse, 6 to 8 lines long, 2 to 3 lines broad, rather
gradually narrowed to a slender petiole (2 or 3 lines in length) ; glands
few on the margin or often only a single one at the end of the mid-nerve :
heads subcorymbose at the ends of the branches, 7 lines long, about 30-
flowered; scales of the involucre 5, linear-oblong, acute, purple especially
toward the summit, flat, marked with 2 rows of linear glands : corollas
dark purple, 3|- inches long : achenes strongly tapering, 3 to 3^ lines in
length. — Collected by E. W. Nelson^ in Oaxaca, between Panixtlahuaca
and Jaquila, altitude 5,000 feet, 26 February, 1895, no. 2399 ; also
between Nopala and Mixistepic, altitude 800 to 4,000 feet, March, 1895.

* * Leaves rather broadly lanceolate, acute at both ends, slender-petioled.

12. P. Palmeri, Rose, Contrib. U. S. Nat. Herb. i. 338, t. 34.

* * * Leaves narrowly elliptic- or linear-oblong, obtuse or obtusish, gradually
narrowed to a sessile or short-petioled base.

13. P. Seemannii, Schz. Bip. 1. c.

14. P. obtusifolium, DC. 1. c. 650.

15. P. decumbens, DC. 1. c.

* * * * Leaves lance-linear, sessile, and amplexicaul.

16. P. amplexicaule. Engelm. in Gray, PI. Wright, i. 120.

***** Annuals or more commonly perennials, often shrubby at the base :
leaves very narrow ; truly linear or subulate, or lance-linear and acute or
acutish, sessile or subsessile, but not amplexicaul.

4- Mexican species : corolla shallowly and regularly 5-toothed ; teeth deltoid :
achenes 2\ to o\ lines long : involucral scales green or glaucous, not dark
purple.

17. P. scoparium, Gray. Leaves truly linear or terete, half a
line in breadth : peduncles inch or more in length. — PI. Wright, i. 119.

18. P. pausodynum. Shrub: stems geniculate, branched, covered
with smooth reddish brown cortex : branchlets straight, erect, sulcate-
angulate, glaucescent : leaves scattered, linear-lanceolate, narrowed to an
obtusish apex, attenuate at the base to a short petiole, slightly succulent,

ROBINSON AND GREENMAN. — GENUS POROPHYLLUM. 33

3-nerved, slightly reticulated, 2 inches long, 2 to 2h lines broad, entire ;
marginal glands elliptical, usually 2 or 3 on each side and often one at
the end of each leaf : heads 6 to 8 lines long, 25-30-flovvered, rather
densely clustered in terminal flat-topped leafy cymose corymbs (2 or 3
inches in breadth); individual peduncles shorter than the heads: scales
of the cylindrical involucre 5, oblong, obtuse or rounded at the apex, 5
lines long, 1 to l.V lines broad, pale green, with a double line of dark
linear glands near the middle : flowers white. — P. Seemanni, Wats.
Proc. Am. Acad. xxiv. 57 ; Rraudegee, Zoe, i. 313 ; not Schz. Bip. —
Collected by Dr. Edward Palmer in crevices of exposed rocks in high
mountains near Guaymas, Sonora, 1887, no. 279 (distributed as P.
Seemannii., var.). According to Dr. Palmer's notes the plant is called
Maravilla by the Mexicans and used for headache.

H- +- Mexican species with corolla more deeply cleft and more or less distinctly
bilabiate ; teeth lanceolate ; achenes 2| to 3-| lines in length.

*+ Achenes long and slender, tapering at the ape.x, about 4 lines long.

19. P. gracile, Benth. Leaves fleshy, subterete. — Bot. Sulph. 29.

20. P. n. sp. ? aff. P. gracile. Leaves not fleshy, flat, very narrowly
lance-luiear, 1 to 2 inches long, a line or two broad, with regular margi-
nal glands : heads solitary, terminal on very short slender peduncles :
cylindrical involucre green, slender ; scales with purple border and glands.
— A tall slender, much branched leafy plant collected at Fronteras,
Sonora, by C. V. Hartman, altitude 4,550 feet, 25 September, 1890, no. 8.

*-*■ +H- Achenes shorter, 2 to 3 lines long : involucre very dark-purple, often

pruinose-glaucous.

= Spreading annuals : very likely a single species.

2 1 . P. coloratum, DC. Prodr. v. 650.

22. P. tagetoides, DC. 1. c.

= = Perennials, a little shrubby at the base.
a. Involucral scales obovate, very broad and obtuse.

23. P. filifolmm, Gray, Proc. Am. Acad, xviii. 107, xix. 35.

6. Involucral scales narrower, oblong.

24. P. Linaria, DC. 1. c. 649.

-!- -I- M- Lower Californian shrubs, with geniculate and much branched stems,
short subulate fleshy leaves, and short many-flowered heads : achenes only
\\ lines in length.

25. P. crassifolium, Wats. Proc. Am. Acad. xxiv. 57.

26. P. tridentatum, Benth. Bot. Sulph. 30.

VOL. XXXII. — .3

34 PROCEEDINGS OF THE AMERICAN ACADEMY.

VL — DESCRIPTIONS OF NEW OR LITTLE KNOWN
PHANEROGAMS, CHIEFLY FROM OAXACA.

Discorea composita, Hemsl. Mr. E. W. Nelson has rediscovered
this si^ecies at Santa Etigenia, Oaxaca, altitude 500 feet, 18 July, 1895,
no. 2828, and also between Topana, Oaxaca, and Tonala, Chiapas, alti-
tude 200 to 500 feet, 1 to 3 August, 1895, no. 2855. The specimens
from Santa Efigeuia represent the hitherto unknown staminate plant
from which the following supplementary characters are derived : foliage
and inflorescence of the ^ plant as in 9 : segments of perianth 6, oblong,
obtuse ; the inner slightly larger : stamens 6, all perfect, free from each
other aud borne upon the base of the periautli-segments, the three opposite
the outer segments a little shorter than the others : filaments not notice-
ably rigid nor flattened, considerably exceeding the very short anthers :
the interior of the flower containing a conical rudimentary ovary.

Habenaria subauricmata. Glabrous, 5 to 10 inches hiorh:
tuberiform root single, ovoid, an inch long : stem flexuous, leafy : leaves
ovate, acutish or acute, 3-ribbed, sheathing by the slightly narrowed
base, 1 to li^ inches long, a third or half as broad: spike 2 to 6^ inches
long, several- many-flowered : bracts ovate-lanceolate, acuminate, about
equalling the ovary : flowers green : upper sepals 3-nerved, about 3 lines
long, obtusish, the upper broadly ovate, galeate, obtusish, 3-nerved, 3
lines long; the lateral ones narrowly ovate, obtusish, slightly exceeding
the upper one : lateral petals linear-oblong, obtusish, slightly falcate,
subauriculate on the lower side at the base, otherwise entire, 2 to 2i
lines long ; labellum deeply 3-parted, the divisions linear, the lateral
divaricately spreading and slightly curved upwards, 2\ to 3 lines long,
about equalling the middle lobe : spur slender, scarcely clavate, green,
about 4 lines long. — Collected by C. G. Pringle, on grassy slopes, Las
Sedas, Oaxaca, altitude 6,000 feet, August, 1894, no. 4830. Habit of
H. Jlexuosa, Lindh, and H. clypeata. Lindl.. but with lateral petals
undivided.

Spiranthes Oaxacana. Glabrous (except flowers) : root a fas-
cicle of numerous thickened tuberiform fibres : stem erect, a foot to a
foot and a half high bearing 1 or 2 foliar leaves at the base and loosely
sheathed for the rest of its length by membranaceous striate-veiny ovate-
acuminate bracts : leaves oblong, 3 inches or more in length, three
fourths inch in breadth, sheathing at the narrowed base, scarcely per-
sisting until the perfecting of the flowers : flowers in a dense ovate
subcapitate spike, slightly exceeding the conspicuous ovate-acuminate

ROBINSON AND GREENMAN. MEXICAN PLANTS. 35

silvery diaphanous brown-veined bracts : sepals finely pubescent on the
outer surface ; the lateral sepals linear-oblong, 6 lines in length, obtuse,
o-uerved, not fleshy ; the dorsal one ovate-lanceolate, obtusish, about the
same length : petals linear-oblong, obtusish, 3-nerved, about equalling
the sepals ; labellum long-unguiculate ; lamina oblong, 5 lines in lengtli
(about equalling or slightly exceeding the claw), 3^ lines broad, with a
median lance-shaped callous thickening just above the thickish claw;
the latter involute on the margin and somewhat sulcate above. — Col-
lected by C. G. Priugle, on hills above San Felipe, Oaxaca, altitude
6,000 feet, 31 December, 1894, no. GlOl.

Cranichis thysanochila. Glabrous throughout, 6 to 8 inches
high : roots several, elongated, thick, tuberiform, obtuse, covered with
fine short rootlets: leaves basal, ovate, acute, thin, smooth, 1^ to 3
inches long, half to two thirds as broad, subsessile ; the contracted base
membranous, amplexicaul : stem slender, flexuous, bearing about 5 lan-
ceolate attenuate sheathing green bracts : spike 2 to 4 inches in length,
only moderately dense : floral bracts ovate-lanceolate, sharply acuminate,
green, 3 lines in length, about equalling the ovary : flowers spreading,
nearly half inch in length : floral envelopes white ; sepals oblong, obtuse,
1-nerved, subequal, about 2 lines in length ; lateral petals broadly spatu-
late or narrowly obovate, obtuse, 1-nerved, about equalling the sepals in
length ; lip superior, obovate, cuueate, short-clawed, biauriculate at the
base, terminated at the apex by an emarginate and fimbriated appendage ;
the inner surface longitudinally somewhat 2-crested. — Collected by C.
G. Pringle, on calcareous banks, Las Hoyas Caiion, Oaxaca, altitude
4,500 feet, 2 November, 1894, no. 6023. This species differs from
others of the genus in the fringed apex of the lip.

Mlcrostylis platyglossa. Erect, a foot high, glabrous, 1-leaved :
bulb three fourths inch in diameter : sheath subsolitary, short, obtuse :
leaf oval, amplexicaul at the subcordate base and rounded or very obtuse
at the apex. If to 2h inches long, 15 to 18 lines broad: peduncles about
2 inches long ; racemes 6 to 8 inches long, bracts ovate, acute, a third to
half as long as the slender pedicels, these nearly horizontal, 1^^ lines
long : flower purple (the labellum sometimes lighter colored), inverted
by the torsion of the pedicels, bringing the labellum downward : sepals
oblong, obtuse, 1-nerved, about a line long, the upper odd one erect, the
lower pair reflexed : lateral petals linear, reflexed ; labellum broadly
deltoid, one third broader than long, obtuse, about a line in length, dark
purple or yellow (both colors sometimes occurring on the same specimen).
— Collected by C. G. Pringle, on the Sierra de San Felipe, Oaxaca,

36 PROCEEDINGS OF THE AMERICAN ACADEMY.

altitude about 10,000 feet, August, 1894, nos. 5614, 5614 a, and by E.
W. Nelson, in the same localit}^ no. 1140.

Microstylis streptopetala. Slender, 6 to 18 inches high, 1-
leaved : bulb oval, half inch in diameter ; sheaths 2, closely surrounding
the base of the stem, obtuse : leaf elliptic-oblong, cuneate at the amplexi-
caul base, acutish or obtuse at the apex, 1^ to 3 inches long, 4 to 8 lines
broad : naked and angled peduncle about equalling the stem (2 to 4
inches in length) : spicate inflorescence strict, slender, 2 to 8 inches long,
2 to 3 lines in thickness, very densely flowered above but looser below :
bracts very short ovate-deltoid, obtusish : flowers sessile : sejials o-nerved,
green, oblong, obtuse, the upper somewhat falcately incurved, Ih lines
long, green; lower 2 lines long: lateral petals linear, spirally coiled or
twisted, greenish white ; labellum deltoid, strongly auriculate, minutely
3-toothed at the apex, half as long as the upper sepals, in dried state
nearly black ; margins somewhat thickened and slightly incurved ; auricles
oblong and obtuse. — Collected in flower by C. G. Pringle on dry pine
ridges, Sierra de San Felipe, Oaxaca, at 9,000 to 10,000 feet altitude,
30 July, 1894, no. 4808. Most nearly related to 31. montana, Roth, but
with very different lip.

Phoradendron Forestierse. Glabrous throughout : branches
terete, olive-green ; branchlets ancipital : leaves narrowly oblong, with
rounded apex and cuneately naiTowed base, sessile, 1-nerved, or very
obscurely 3-nerved, yellowish green, inch to inch and a half long, IJ- to
2i lines wide : inflorescences of 9 plant axillary, solitary, opposite,
moniliform, 5 to 9 lines in length, flowers in 1 to 4 globular 10-12-
flowered clusters, becoming deeply imbedded in the substance of the
nodular rhachis ; the clusters 2i^ lines in diameter, tawny in color ;
the intervening necks 1 to 1^ lines long, about half enveloped in a loose
sheath ; the margins of the sockets holding the flowers finely ciliolate :
divisions of the perianth 3 (rarely 4), deltoid, the free portion not a
third of a line in length. — Collected by C. G. Pringle on hills between
Tehuacan and Esperanza, Puebla, altitude 6,000 feet, 23 December,
1895, no. 6290. A species parasitic on Forestiera, and apparently most
nearly related to P. brachystachynm, Oliv., which, however, is tomen-
tulose on the branchlets, has simple not moniliform inflorescences, and
larger more distinctly veined leaves.

Euphorbia Lnciismithii. Tall branching tomentulose glaucous
shrub, 10 to 15 feet in height: branches subterete, striate: leaves verti-
cillate, 2-5-nate, elliptical, obtuse at both ends or subacute at the base,
glabrous or glabrate above, paler and soft grayish-tomentulose beneath,

ROBINSON AND GREExNMAN. — MEXICAN PLANTS. 37

10 to 16 lines long, a third to half inch broad: petioles 4 to 6 lines long,
tomentulose : cymes compound, terminal, flat-topped, leafy : floral leaves
oblanceolate, cuneate, mucronulate, 1 -nerved, white or rarely red, about
.'? lines in length, a line in breadth: involucres campanulate, puberulent,
nearly sessile ; lobes 5, fimbriated ; glands 5, oblong ; appendages oblong
or subrotund, undulate, white, three fourths line long : capsules 3-lobed,
nearly 3 lines long, glabcous seeds ashy, oblong, somewhat 4-angled,
faces rugulose and marked with fine irregular brown lines. — Collected
by Lucius C. Smith, at Rancho de Calderon, altitude 5,500 feet, 13
August, 1894, no. 181 ; also at Jaycatlan, altitude 4,300 feet, 10
September and 4 November, 1894, no. 182; also by C. G. Pringle, in
rocky gulches, Monte Alban, Oaxaca, altitude 5,800 feet, 14 September
and 27 November, 1894, no. 4903, and by E. W. Nelson, six miles above
Dominguillo, altitude 4,500 to 5,500 feet, 30 October, 1894, no. 1880.
Most nearly related to E. leucocephala, Lotsy, from which it differs in
pubescence and form of appendages.

Euphorbia Oaxacana. Stems subterete, 2 to 5 feet high, green,
rather densely pubescent near the summit, soon glabrate : leaves alter-
nate, ovate-elliptic, entire, thin, obtuse at each end, appressed-villous on
both sides and ciliated, 10 to 16 lines long, half as broad; slender soft-
pubescent petioles becoming half inch in length : inflorescence a long
narrow compound somewhat secund naked panicle: its leafless branches
alternate, 1 to 3 inches in length, again branched and rather densely
flowered, tomentose; buds roseate, tomentose: involucres in subcapitate
peduncled cymes, white-tomentose as well as the short (1 to 1^ lines)
linear-spatulate branchlets ; glands 5, equal, oblong, with white sub-
rotund or oblong 2-3-crenulate appendages (about a line long) ; involucral
lobes fimbriate, green: styles deeply 2-parted; capsule green, glabrous,
1 J lines in diameter : seeds oval, ashy, faveolate. — Collected by C. G.
Pringle, on ledges, Monte Alban, Oaxaca, altitude 5,800 feet, 23 No-
vember, 1894, no. 6070.

Euphorbia subcaerulea. Erect much branched annual, glabrous
throughout, 2 feet in height : stem and slender branches terete, striate,
livid: leaves elliptic-ovate, entire, thin, green above, a little paler and
glaucescent beneath, rounded or very obtuse at the base, rounded or
retuse at the apex, 3 to 8 lines long, nearly two thirds as broad : petioles
filiform, nearly equalling the leaves : inflorescences open cymose-panicu-
late: floral leaves very small, elliptic-ovate to subrotund, subsessile,
white or bluish: involucres (including appendages) 1^ lines in diameter;
glands 5, oblong, sessile, with suborbicular entire appendages, these at

38 PROCEEDINGS OF THE AMERICAN ACADEMY.

first light blue, then chaugiug to white, half line in diameter, spreading :
stigmas deeply 2-parted ; divisions clavellate : capsule smooth, about a
line in diameter: seeds oblong, light blue, three fourths line in length,
deeply pitted and with a conspicuous caruncle. — Collected by C. G.
Pringle, on dry banks, Tomellin Canon, Oaxaca, altitude 3,500 feet, 9
December, 1895, no. 6265.

Cardiospermum Galapageium. A slender climber with fur-
rowed tomentulose stem aud biternate leaves : leaflets oblone lanceo-
late, attenuate, entire, abrupt at base, finely tomentose upon both surfaces,
slightly paler beneath, prominently 1-nerved and pinnately veined, the
lateral leaflets of each division much smaller than the terminal: peduncles
slender, spreading, tendriliferous beneath the umbelliform inflorescence :
sepals 4, the outer pair a line, the inner 2 lines long : glands upon the
upper side of the disk short, rounded, not at all cornute. — G. Corindum,
Rob. & Greenm. Am. Jour. Sci. ser. 3, L. 145, in part, not L. — Col-
lected in the southern i^art of Albemarle Island, July, 1891, by Dr. G.
Baur, no. 61. Distinguished by its narrow entire leaflets. C. integerri-
mum, Radlk., the only other species with similar foliage, known to the
writers, has a 5-sepaled calyx and cornute glands.

Erythrsea retnsa. Erect glabrous annual, 6 or 8 inches in height:
stem leafy, 4-angled, bi'anched ; branches erect or nearly so, mostly rather
short and again branched: leaves oblong-elliptical, obtuse or mucronu-
late at the end, somewhat narrowed at the base, half inch long, 2 to 3
lines broad, the uppermost narrower and lanceolate : pedicels both termi=
nal and lateral on the branches, short, 1 to 2 lines in length : flowers
4-merous, calyx-lobes lance-linear, attenuate, slightly carinate, 2 lines in
length, green with narrow scarious diaphanous margins : corolla-lobes
oblong, retuse, a line long, pale j^ellow in a dried state : fruit dark brown,
2\ lines long. — Collected by C. G. Pringle, in springy meadows, Sierra
de San Felipe, Oaxaca, altitude 7,500 feet, 11 December, 1895, no. 6300.
This species differs from E. divaricata^ Schaffiier, in its much shorter
pedicels and erect not widely spreading branches ; from E. tctramera,
Schiede, in its retuse by no means acute corolla-lobes : from E. stricta,
Schiede, in its 4-raerous flowers. The flowers ai-e small and appear to
remain closed and to be cleistogamous.

Nama Pringlei. Low slender branching annual of similar habit to
the following and with similar pubescence : leaves oblong, entire, rounded
at the apex, rather gradually narrowed at the base to a short slender peti-
ole : calyx as in the last : corolla showy, 5 lines long and equally broad,
with short yellow tube and deep blue spreading rather deeply 5-lobed

ROBINSON AND GREENMAN. — MEXICAN PLANTS. 39

limb: styles a line long: capsule rugose, broadly ovate with depressed
summit. — Collected by C. G. Pringle, on iiills near Tehuacan, Puebla,
altitude 5,500 i'eet, 24 December, 1895, no. (5286. A very attractive
species, nearly related to N. latifolium, Gray, but with much longer calyx
and larger corolla as well as different pubescence.

Nama Pueblense. Slender weak dichotomously branched annual,
3 to 6 inches high, spreading-pubescent : leaves ovate, obtuse, entire,
thin, 10 to 12 lines long, half as broad, rather abruptly contracted to a
slender petiole and covered on both surfaces with fine sub-appressed
hairs with globular bases ; and below with short golden yellow pedicelled
glands : flowers short-pedicelled, grouped by 2's or 3's near the axils or
extra-axillary and solitary somewhat above the fork of the terete purplish
stem : calyx-lobes in fruit 4 lines in length, spatulate, densely hirsutu-
lous with spreading white hairs : corolla small, pale blue (in dried speci-
men), tubular, with scarcely ampliated shortly 5-lobed limb, 2 to 2\ lines
long : styles half line in length : capsule narrowly oblong, hispidulous
near the apex, rugose, about 2 lines in length, obtusish : seeds brownish,
rugose. — Collected by C. G. Pringle, on hills near Tehuacan, Puebla,
altitude 5,500 feet, 24 December, 1895, no. 6287. Habitally near to
iV. latifolium^ Gray, but differing in its much longer more hirsute calyx,
narrower capsule, and having leaves evenly pubesceut over the entire
surface instead of chiefly on the veins.

Berendtia levigata. Shrub 2 to 3 feet high, glabrous but some-
what vernicose especially upon some of the younger parts : stems and
branches covered with a gray bark ; branchlets very leafy especially at
the ends : leaves rather broadly lanceolate or oblanceolate, sharply few-
toothed above the middle and acute, cuneate to an entire short-petioled
base, glabrous or (under a lens) slightly pulverulent but green on both
surfaces and sometimes a little lucid, of firm texture, 1^ to 1|^ inches
long, half as broad: peduncles opposite in the upper axils, 1 -flowered,
bibracteolate at or below the middle, the bractioles bearing rudimentary
buds in their axils : calyx strongly prismatic, as in Mimuhis, 6 lines in
length, with 5 short broad subequal acuminate-mucronate teeth : corolla
showy, much exserted, 1 h inches in length ; throat rather gradually
ampliated, orange with crimson spots ; limb patulous, of broad rounded
lobes : stamens scarcely exserted : capsule 6 lines in length. — Collected
by C. G. Pringle, on calcareous hills near Tehuacan, Puebla, altitude
5,500 feet, 24 December, 1895, no. 6294.

Castilleia aurea. Slender annual of § Epichroma, a foot high,
with terete glabrous flexuous branched stems and delicate pectinate

40 PROCEEDINGS OF THE AMERICAN ACADEMY.

foliage : leaves inch to inch aud a half long, with 6 to 9 filiform -linear
segments and narrow rhachis : flowers 9 lines in length, in long rather
dense puberulent racemes ; lower bracts much like the leaves, the upper
gradually reduced to small lanceolate or subulate scales ; pedicels erect,
slender, 2 to 4 lines long : calyx fuuuel-form, gradually and considerably
anipliated, with broad orifice oblique, scarcely split ventrally and not at
all dorsally, about half inch in length, golden yellow : corolla concolorous,
about 9 lines in length, puberulent upon the considerably exserted galea :
capsule ovate-oblong, acutish, 3 lines in length. — Collected by C G.
Pringle, on wet bluffs of barrancas above Cuernavaca, State of Morelos,
altitude 7,000 feet, 19 November, 1895, no. 6204. Nearly related to C
tenuifolia, Mart. & Gal., and C. gracilis, Benth., but distinguished from
the former by its yellow flowers, more dense and continuous inflorescence,
narrower acuter capsule, and smaller seeds ; from the latter by the color
of the flowers and much more exserted galea, denser racemes, etc.

Carlowrightia glandulosa. Low much branched shrub, 1 to 3
feet in height, densely glandular-tomentose and viscid : cortex of the older
branches pale gray : leaves ovate, acute, entire, cordate, soft-pubescent
upon both surfaces, the larger ones 12 to 15 lines in length, three fourths
as broad, the floral ovate-lanceolate to lanceolate, half inch in length,
somewhat acute at the base : petioles nearly or quite half the length of
the leaf: flowers 1 to 4 in the opposite axils, closely sessile: calyx-tube
very short, divisions narrowly linear-attenuate, 3 lines in length, glandu-
lar-pubescent : corolla bluish, 4-parted ; the emarginate posterior lobe
purple-veined and marked with yellow at the ceutre ; tube 2 lines long ;
lobes 4 to 5 lines in length : capsule glabrous and lucid, tlu-ee fourths
inch long, with pungent tip, seeds orbicular, dark brown with lighter
colored margin, 2i lines in diameter. — Collected by C. G. Pringle at
Monte Alban, near Oaxaca, altitude 5,500 feet, 5 December, 1895, no.
6276. This species is nearest C. pubens, Gray, but differs from it most
obviously in its cordate and considerably larger lower leaves, as well as
larger flowers and seeds.

Carlowrightia (?) Pringlei. Shrub, 3 to 5 feet high, with slender
glabrous terete stems covered with smooth grayish brown cortex : leaves
ovate-lanceolate, acute, entire, rounded at the base, glabrous on both
surfaces or minutely strigillose on the veins beneath, 8 to 12 lines long, a
third as broad ; petioles a fourth inch long, glabrous but often with a tuft
of white hairs at the base within : floriferous branches slender, recurved :
flowers sessile, secund, solitary or 2 together in the same axil, the oppo-
site axil being empty : bracts small, subulate : calyx-tube very short ;

ROBINSON AND GHEENMAN. — MEXICAN PLANTS. 41

segments 5, liuear-attenuute, glandular-pubescent, U lines long: corolla
bluish white (in dried specimen) : tube slender, 4 to G lines long, about
equalling the 4 lobes : anther-cells equal and subcoutiguous, muticous :
capsule glabrous, half inch in length. — Collected by C. G. Pringle,-dry
slopes, Tomellin Caiion, Oaxaca, altitude a,.0OO feet, 30 November, 1895,
no. G261, A plant which with about equal propriety might be referred
to Diaiithera.

Jacobinia candicans, Benth, & Hook. f. "Gen. ii. 1115" accord-
ing to Ilook. f. & Jacks. Ind. Kew. i. 1240. Dianthera candicans,
Benth. & Hook. f. "Gen. ii. 1113" according to Hemsl. Biol. Ceut.-
Am. Bot. ii. 517. Adhatoda candicans, Nees in DC. Prodr. xi. 396.
Jacobinia Mexicana, Hemsl. 1. c. 521, as to plants of Galeotti and of
Liebniann. Excellent and copious material of this doubtful plant has
now come to hand from the following sources. Vicinity of Cuicatlan,
Oaxaca, altitude 1,800 to 2,500 feet, E. W. Nelson, no. 1698, and in
same locality V. Gonzalez, no. 48 ; also in Tomellin Caiiou, altitude
2,500 feet, C. G. Pringle, nos. 5638, 6260. The plant is a shrub 3 to 5
feet high with the whole habit and inflorescence as well as corolla of
Jacobinia, to which it would seem best to refer it notwithstanding its
slightly disjoined anther cells. Indeed the very close habital resemblance
between this species and Jacobinia Mexicana has led to a confusion of
the two for Mr. Hemsley in the places cited ascribes Galeotti's no. 911
both to Dianthera candicans and Jacobinia Mexicana. The plants of
recent collection mentioned above correspond in all points to Liebmann's
specimen from Tehuacan, and differ much in calyx and bracts from
Seemann's plant of N. W. Mexico. In the latter 'plant (J. Mexicana,
Seem.) the bracts are minute, much shorter than the 5-toothed calyx.
In the former {J. candicans) the bracts about equal the calyx and this is
much more deeply 5-cleft. The corolla of J. candicans is bright scarlet
rather than purple as described by Nees.

Oldenlandia xestosperma. Erect glabrous perennial, several-
stemmed from a slightly ligneous base : stems terete, moderately branched
above, 1 to 2 feet high : leaves linear, 1-nerved, an inch to an inch and
three fourths in length, less than a line in breadth : inflorescences termi-
nal few-flowered cymes; bracts small, subulate; pedicels filiform, 1 to 3
lines in length: flowers strongly heterogone-dimorphous : calyx-lobes in
anthesis but half line long, subulate, about equalling the tube : corolla
purplish, nigrescent in drying, 4 lines in length, gradually widened from
the base ; limb of 4 triangular erect teeth : capsule obovate, 2 lines in
length, entirely inferior, dehiscing to the base: seeds slightly angled,

42 PROCEEDINGS OP THE AMERICAN ACADEMY.

light colored, yellowish, highly polished and shining, half line or less in
diameter. — Collected by C G. Pringle in open glades, Sierra de San
Felipe, Oaxaca, altitude 8,000 feet, June, 1894, no. 4692. Habit of
Houstonia purpurea, var. teuuifolia.

Eupatorium eriocarpum. Stem tall, 5 to 8 feet in height;
branches glabrate, angled, lucid, yellowish brown ; branchlets puberulent
but soon glabrate : leaves opposite, ovate or rhombic ovate, crenate-
serrate, acuminate to an obtusish apex, cuneate or rather abruptly con-
tracted at the mostly unequal-sided base, glabrous on both surfaces except
on the veins and at their axils, 2 to 3 inches in length, a third to half as
bi'oad, neither coriaceous nor rugose, scarcely paler beneath, pinnately
veined; petioles 2 to 3 lines long: thyrsoid inflorescences terminal on
the upper branchlets and together forming large ovate leafy panicles ;
slender peduncles and short pedicels often nodding, covered with fine
and somewhat sordid pubescence: heads mostly 5-flowered, 5 to 6 lines
in length : scales of the involucre ovate to ovate-oblong, obtuse, finely
striate, ciliolate, imbricated in 3 or 4 series : corolla-tube slender, 2^
lines in length, larger at the base than at the summit, without any ap-
parent throat ; teeth very short : style branches dark colored, strongly
clavate : achenes turbinate, so densely covered with white woolly or silky
pubescence as to obscure their 5 angles, about 2 lines in length, narrowed
at the base to a glabrous callous and somewhat pungent lip; pappus-
bristles white, slightly exceeding the achene, about 50 in number. —
Collected by C. G. Pringle, in Tomellin Cafion, Oaxaca, altitude 3.000
feet, 22 December, 1894, no. 6112.

Eupatorium rilpicola. Shrub, 5 to 10 feet in height: branchlets
reddish brown, terete, finely and densely f uscous-puberulent : leaves op-
posite, petiolate, ovate, acuminate, crenate-serrate, 3-nerved, rounded or
obtusish at the base, green and sparingly puberulent or glabrate above,
paler and grayish-tomentulose beneath, 14 to 18 lines long, 7 to 12 lines
broad; petioles reddish, puberulent, 3 lines long: inflorescences 6-10-
headed flat-topped umbelliform corymbs terminating short opposite lateral
and terminal branchlets (1 to 2 inches in length), and together forming
more or less elongated leafy panicles ; pedicels densely pubescent, ascend-
ing, 2 to 4 lines long, subtended at the base by very reduced leaves and
often bearing one or more short subulate bractlets : heads 3i to 4 lines in
height ; involucre relatively short, imbricated in about 2 series ; scales 10
to 15, subequal, 1^ lines in length, oblong, acutish, puberulent; flowers
about 16, purplish white: corolla 2i|-3 lines, gradually enlarged upward:
achenes linear, a little over a line in length, rather densely white-pubes-

ROBINSON AND GRKENMAN. — MEXICAN PLANTS. 43

cent. — Collected b}- C. G. Priugle ou dry ledges of the Sierra de Sau
Felipe, Oaxaca, altitude 7,500 feet, 9 October, 1894, no. 4970.

Chrysopsis Brandegei. Low villous perennial, 3 to 5 inches
high, with spreading branched scaly rootstock and numerous densely
tufted stems : leaves spatulate, entire, gradually narrowed below into
slender petioles, rounded at the summit but mucrouate-acuminate at the
very apex, 1-nerved, finely grayish-pubescent and covered with very
short yellowish glandular or i-esinous-tipped hairs (so short as to appear
when viewed from above merely as sessile globules), and spreading'villous
on the edges, 7 or 8 lines long including the petioles; lowest leaves
shorter, densely crowded at the base, and often canescent with more
appressed pubescence : peduncles terminal, slender, erect, flexuous, terete,
H to 2 inches long, villous with spreading hairs and also closely covered
with the minute glandular hairs, entirely leatiess but sometimes bearing
one or two short filiform bracts : heads solitary, discoid, 6 to 8 lines in
diameter, half inch in height, 40-45-flovvered : scales of the involucre
ver\' unequal, imbricated in about 4 series ; the outer ones very small, a
line in length, very villous, the inner oblong, purplish, rather abruptly
but acutely pointed, flat and smoothish, ciliolate on the margin, 3^- lines
long, half to two thirds line broad : corollas 5-toothed, golden yellow, gla-
brous, 4 lines long: pappus manifestly doubh, the outer bristles very un-
equal: stigmas sometimes 3: achenes (immature) densely silky-villous,
about a line long. — Chrysopsis sp., Brandegee, Zoe, iv. 206. — Collected
by T. S. Brandegee at San Pedro Martir, Northern Lower California,
May, 1893.

Bigelowia pyramidata. Shrub, 2 to 3 feet high, much branched :
branches white-woolly; branchlets striate : leaves fascicled, linear-acicu-
lar, entire, somewhat pungent, white-woolly beneath, strongly revolute at
the margins, pale green and grooved above, 4 to 14 lines long, half line
in breadth: inflorescences terminal pyramidal panicles: heads small,
numerous, sessile or subsessile, spicately arranged along the spreading-
ascending branches, subtended by fascicles of short spreading pungent
leaves, discoid, few (o-7)-flowered : scales of the involucre linear-lanceo-
late, acute, scarious, diaphanous: flowers pale yellow, pappus scarcely
tawny, of numerous fine unequal bristles ; young achenes silky, the ma-
ture not seen. Collected by C. G. Pringle, on hills above Oaxaca,
altitude 5,500 feet, 16 November, 1894, no. 6048. A species anomalous
in its spicate-paniculate inflorescence but with the other characters of
Bigeloivia.

Lagascea tomentosa. Rather stout: stem subsimple, terete.

44 PROCEEDINGS OF THE AMERICAN ACADEMY.

tomentose, also somewhat glandular and villous, purplish : leaves ample,
ovate, acuminate at the apex, acute at the petiolate base, dentate, tomen-
tose on both surfaces, pale beneath, 3 to 5 inches long, nearly Ih to 2
inches broad, scabrous on the margins : petioles 4 or 5 lines long, densely
pubescent: iuHorescences enveloped in ovate-lanceolate acuminate pubes-
cent and ciliate bracts (6 to 8 lines long, 2 to 3 lines broad) : involucres
numerous, closely aggregated, villous, 3 lines in length, 1 -flowered, un-
equally dentate, glandular in lines : corolla long, 6 lines in length, exter-
nally pubescent, the ampliate cylindrical throat exceeding the limb and
the more slender proper tube : mature achenes not seen. — Collected by
E. W. Nelson between Ayusina and Petatlan, Guerrero, altitude 5,000
to 7,000 feet, 14 December, 1894, no. 2121.

Trigonospermum tomentosum. Stout branching pubescent
herb: stem terete, brownish or dark colored: leaves rhombic-ovate, 3-
nerved from above the abruptly contracted then cuueately narrowed
base, serrulate, acuminate, green and becoming scabrous above, paler
and densely tomentose beneath, including the narrowed petiole-like base,
4 to 8 inches long, half as broad : inflorescence a much branched corym-
bose panicle, densely covered with short dark glandular-tipped hairs ;
bracts subulate : heads half inch in diameter : involucral scales about
2-seriate; the outer oblong, acutish, 3-nerved, ciliated; the inner broadly
obovate, abruptly acuminate : ray-flowers 5 ; ligules broad, reversed-del-
toid, deeply 3-lobed, bright orange-yellow, Sh lines long ; disk-flowers
about 25, concolorous : chaff hyaline, obovate-cuneate to suborbicular,
ciliated. — Collected by E. W. Nelson on the western slope of Mt.
Zempoaltepec, Oaxaca, altitude 7,700 to 8,000 feet, 5 to 13 July, 1894,
nos. 610, 617, and later between Panixtlahuaca and Jaquila, altitude
1,000 feet, 26 February, 1895.

Montanoa macrolepis. Stem, 3 to 5 feet in height, terete, fuscous,
glabrous or somewhat pubescent : leaves opposite, sinuately 3-lobed ; the
lamina 3 to 5 inches long, 2 to 3 inches broad, very scabrous-pubescent
above, somewhat paler and sparsely pubescent or glabrate beneath, 3-
nerved above the base, and contracted below to a somewhat toothed wing,
which does not quite reach the stem but terminates gradually or abruptly
(usually with two rounded auricles), leaving a short naked petiole;
lobes undulate-denticulate ; the lateral short, broad, blunt or again sub-
bilobed ; the terminal ovate, obtuse, acutish or even acuminate : heads
rather few and large, cymose-corymbed, individually pedunculate or
grouped by 2's or 3's at the ends of the branches : involucral scales sub-
biseriate, oblong, obtuse or rounded at the apex, 3^ lines long: ray-

ROBINSON AND GREENMAN. — MEXICAN PLANTS. 45

flowers 10-12 ; ligules 6 lines long, 2 to 3 lines broad ; disk-flowers
numerous, with the tube about a third as long as the ampliated throat :
fruiting heads globose, 1| inches in diameter; achenes 2 lines long:
chaff lanceolate-attenuate, straight-pointed or nearly so, puberulent or
almost glabrous except the strong ciliation of the margins, 4 to 5 lines
long iu authesis, becoming 8 lines long in fruit. — Collected by C. G.
Priugle, in gulches of hills of Las Sedas, Oaxaca, altitude 6,000 feet,
29 September, 1894, no. 4932, also by L. C. Smith, at Nacaltepec
(Salome), Oaxaca, altitude 6,500 feet, 21 September, 1895, no. 818.

Montaiioa Rosei. Shrub 8 or 10 feet high : leaves opposite,
sleuder-petioled, rhombic-ovate, serrate, not lobed, acuminate at the
apes, cuneate at the base, rather harsh in texture, scabrous and some-
what rugose above, scarcely paler, finely pubescent and glandular-dotted
beneath, 3 to 4 inches long, half as broad : corymbs ample: bracts linear:
involucral scales lance-attenuate, sub-uniseriate, silky-villous, 2h to 3
lines in length : disk-flowers about 4, with tube slender, nearly equalling
the throat ; rays about 3, about 2 lines iu length : chaff densely fulvous-
woolly. — 3Tontanoa (Enocoma) sp., Rose, Contrib. U. S. Nat. Herb. i.
103. — Collected by Dr. E. Palmer, at Alamos, W. Mexico, 26 March
to 8 April, 1890, no. 394.

Viguiera Nelsonii. Stem terete, densely silky-villous with white
subappressed hairs : leaves attenuate at both ends, sessile, 3-nerved from
above the base and pinnately veined, appressed silk3--vil]ous upon both
surfaces, more densely so and paler beneath, 3 to 6 inches or more in
length, ^ to 1^ inches in breadth : heads 12 to 20 in number, 12 to 18
lines in diameter, borne in a terminal corymbose panicle ; the individual
peduncles an inch or two long: involucral bracts 2-3-seriate, narrowly
oblong-lanceolate, silky-villous especially near the margins : somewhat
thickened at the base ; the tips lax and spreading : rays about 10, orange-
yellow, oblong, slightly 2-3-toothed at the apex, 6 to 8 lines in length :
disk-flowers more than 50, concolorous : chaff carinate, with stronsr
midrib excurrent as a spreading tip : achenes somewhat compressed and
inconspicuously 4-augled, appressed-villous : pappus of two aristje some-
what broadened at the base, and intermediate squamellje two on each
side, ovate, ciliate-fringed. — Collected by E. W. Nelson, between
Chilapa and Tixtla, Guerrero, altitude 5,200 to 7,000 feet, 17 December,
1894, no. 2169, and by L. C. Smith in mountains of Huitzo, Oaxaca,
altitude 6,500 feet, 16 November, 1895, no. 899. As to character of
achenes a dubious intermediate between Virjuiera and Encelia, but in
habit approaching more closely species of the former genus.

46 PROCEEDINGS OF THE AMERICAN ACADEMY.

Verbesina Nelsonii. Stout, apparently herbaceous : stem striate-
angulate, puberulent under a lens, internodes entirely wingless or with a
broadish irregular deciduous corky wing near the summit : leaves green
and glabrous on both surfaces, oblong, acuminate, penuinerved, crenate-
serrate, 6 to 8 inches long, H to 2 inches broad, below the middle
slightly narrowed and somewhat crisped or undulate ; the base broadly
auriculate and amplexicaul, midrib prominent beneath, pale : heads
numerous, of medium size, 4 to 5 lines in diameter, in a very dense
terminal compound corymb ; bracts oblong, small, scarcely herbaceous :
pedicels tomeutulose : involucre campanulate ; scales about 10, subuni-
seriate, oblong, obtusish, 2 lines long ; ray-flowers 4 to 6 ; ligules very
small, yellow, 2 to 2^ lines long, nearly a line in breadth : chafE obovate,
mucronate, pubescent on the outer surface especially on the prominent
keel : achenes a line long, hispidulous, with conspicuous wing on each
edge, awns subequal. — Collected by E. W. Nelson, between Ayusinapa
and Petallan, altitude 5,000 to 7,000 feet, 14 December, 1894, no. 2118.

Verbesina Smithii. Branching shrub : branches glabrous, wing-
less, covered with a pale grayish cortex roughened with numerous lenti-
cels : branchlets gray-tomentose : leaves alternate, lanceolate, attenuate at
both ends but obtusish and mucronulate at the apex, rather sliallowly ser-
rate-dentate except at the cuneate short-petioled base, bright green and
scabrous-puberulent above, white-tomentose beneath, 3 to 4 inches long,
10 to 14 lines broad : heads small, 2 lines high, 3 lines broad, not very
numerous in small terminal corymbs ; these 2 inches in diameter not
exceeding the surrounding leaves: bracts small, grayish-tomentose as
well as the pedicels : involucral scales about 3-seriate, elliptic-oblong,
obtuse, green, pubescent on the outer surface, ciliated, a line long :
disk-flowers 30 to 35 ; ray-flowers about 8 ; ligules yellow, H to 2 lines
long ; chaff obovate, obtusish, yellow toward the summit, puberulent on
the outer surface : achenes (immature) distinctly winged on each side ;
awns 2, slightly unequal. — Collected by L. C. Smith, at Jayacatlan,
Oaxaca, altitude 4,500 feet, 10 September, 1894, no. 132.

Verbesina trilobata. Stems 5 to 10 feet in height, branched,
glabrous, lucid, wingless, covered with light colored cortex : leaves
opposite, decussate, rhombic in general outline, strongly 3-lobed, den-
ticulate, gradually narrowed to a subsessile base, scabrous-puberulent
above, grayish-tomentose beneath, subpalmately and rather obscurely
3-5-nerved from considerably above the base and reticulate-veined ;
terminal lobe elongated, ovate to oblong-lanceolate, acuminate ; the
lateral much shorter and rather blunt ; sinuses rounded : corymbs much

ROBINSON AND GREENMAN. — MEXICAN PLANTS. 47

branched, terminal, leafj-bracted at the base ; bracts oblong-lanceolate,
acute; pedicels finely grayish-tomentulose ; heads rather small, numerous,
radiate, few-flowered; involucral scales about o-seriate ; the outer consid-
erably shorter, linear-oblong, obtuse, 1 to H lines long with subherbaceous
tips ; inner scales subscarious, about 3 lines long, u line broad, acute,
glabrous, except the ciliolated margins: disk-flowers about 12 ; corollas
glabrous ; ray-flowers 2 to 3 ; ligules goldeu-yellow, 3^ to 4 lines long :
mature achenes 2 lines long, conspicuously winged on each margin,
hispidulons on the surfaces ; awns subequal. — Collected by C. G.
Priiigle, in rocky gulches, Monte Alban, Oaxaca, altitude 5, GOO feet,
15 August, 4 October, 1894, no. 4875.

Verbesina variabilis. Shrub : branches striate-angled, mostly
wingless and nearly glabrous : branchlets hoary-puberulent or glabrous,
usually bearing narrow irregular deciduous brown corky wings de-
current from the bases of the petioles : leaves alternate, short-petioled,
ovate or lance-oblong, sharply and finely serrate or mucronulate-den-
ticulate, acute to shortly acuminate at the apex, cuueately narrowed at
the base, 2 to 3 inches long, 9 to 12 lines broad, above green, strigillose-
puberuleut and very scabrous to nearly smooth ; below scarcely paler,
soft-pubescent on the pinnate and reticulated veins or quite glabrous :
heads of middle size in terminal subsimple or compound corymbs,
pedicels grayish-tomeutose to merely puberulent: involucral scales about
2-seriate, green, oblong, obtusish ; ray-flowers 10 to 12, yellow; ligules
6 lines long, about 2 lines broad, finely 2-3-dentate at the apex ; disk-
flowers about 60, pubescent : chaff cuneate, the apex truncate with
short recurved apiculus : achenes very narrowly winged on both margins,
about H lines long: awns 2, subequal. — Collected in three rather
different forms, which, however, appear to have no satisfactory specific
differences. The first, which may be regarded as the typical form, has
rather broad dull green very pubescent leaves. It was collected by
C. G. Priiigle, Sierra de San Felipe, altitude 9,500 feet, 24 September,
1894, no. 4918, by E. W. Nelson, 18 miles southwest of the city of
Oaxaca, altitude 7,500 to 9,500 feet, 10-20 September, 1894, no. 1393,
and by C. Conzatti on Sierra de San Felipe, altitude 9,000 feet, 29
November, 1895, no. 31. A second form collected by Mr. Nelson on
the top of the Sierra Madre near Chilpancingo, Guerrero, altitude 9,000
to 10,200 feet, 24 December, 1894, no. 2240, differs only in having
narrower oblong leaves, which are somewhat lucid above and somewhat
less pubescent. A third form, secured by Mr. Nelson at same place and
date, no. 2215, has narrow oblong leaves, which are quite glabrous

48 PROCEEDINGS OP THE AMERICAN ACADEMY.

beneath, and slightly lucid but finely scabrous-puberulent above. The
species, to judge from characters, must be near V. Seemannii, Schz,
Bip.

Dahlia tenuis. Root a cluster of 6 or 8 stout fibres, each enlarged
and tuberiform in the middle : stem single, erect, very slender, 1 to 2^
feet high, simple below, covered with a short and dense pubescence,
almost tomentulose : leaves small for the genus, pinnate to bipinnate,
somewhat deltoid in general outline, on slender divaricately spreading
petioles of nearly their own length ; leaflets lanceolate, acute or acumi-
nate at both ends, finely and sharply serrate or irregularly 2-3-lobed,
green and nearly or quite glabrous above, pale and finely pubescent
beneath, 8 to 12 lines in length, 3 to 4 lines in breadth: heads few and
subcorymbose, or even solitary, including the rays 1^ to 2 inches in
diameter : outer involucre of about 6 narrow thickish obtuse bracts,
reflexed during anthesis ; the inner scarious bracts lance-oblong, about 6
lines in length : rays about 8, pistiliferous. — Collected by E. W. Nelson,
18 miles southwest of city of Oaxaca, altitude 7,500-9,500 feet, 10 to
20 September, 1894, no. 1364; also by C. G. Pringle, Sierra de Clavel-
linas, altitude 9,000 feet, 27 October, 1894, no. 5807; and by L. C.
Smith, on mountains of Telixtlahuaca, altitude 7,500 feet, 27 July, 1895,
no. 481.

Flaveria vaginata. Perennial with stout lignescent root : stems
several, ascending from a decumbent or even prostrate somewhat branched
base, terete, striated, purplish, with bilineate short grayish woolly pubes-
cence, leafy above, naked below except for the persistent and sheathing
bases of the fallen leaves : internodes very short : leaves linear-subulate,
clasping at the base, very gradually attenuate, often fascicled in the
axils l(-3)-nerved, rather pale green, finely ciliated toward the base:
heads small, closely aggregated into terminal solitary or corymbose-
paniculate gloraerules ; these simulating the normal involucrate heads of
the order: glomerules 6 to 8 lines in breadth, subtended by a few short
recurved foliaceous bracts, and containing 30 or more heads : involucral
scales 3 to 4 in each head, hyaline : ray-flower solitary, conspicuous, 2i
lines long, with oblong slightly 2-3-toothed yellow ligule : disk-flowers
5 to 7, yellow: achenes black, lucid, about 10-nerved. — Collected by
E. W. Nelson between Coixtlahuaca and Tamazulapam, Oaxaca, altitude
7,000 to 7,700 feet, 12 November, 1894, no. 1933.

Florestina pedata, Cass. With this species, Schkuhria gJomerata,
Rob. & Seaton, based on Mr. Pringle's nos. 4289 and 5006, and published
in Proc. Am. Acad, xxviii. 109, is identical.

ROBINSON AND CxREENMAN. — MEXICAN PLANTS. 49

Florestina platyphylla. Mr. Piiugle's no. 4075, collected on
Moute Alban, Oaxaca, and described in the Am. Jour. Sci. ser. 3, L.
156, as Schukltria platyphylla, appears upon furtlier examination to be
better placed in Florestina.

Dtsodia serratifolia, DC. A specimen collected by L. C. Smith,
below Jayacatlan, at 3,500 feet altitude, 9 P'ebruary, 1895, corresponds
with Ghiesbreght's no. 519, and with a fragment of the type from the
De Candollean Herbarium, in all respects, exce[)t in having the leaves
mostly alternate, not opposite as hitherto described. The uppermost
only are subopposite.

Liabuji GLAiiKUM, Hemsl. (Biol. Cent.-Am. Bot. ii. 232). This
showy species, originally collected by Bourgeau, no. 1401, at Cuernavaca,
in 1865, has now been rediscovered by Mr. Priugle in the same locality.
His excellent specimens show the flowers to be bright orange-yellow and
the leaves normally attenuate. Mr. Pringle notes that the plant attains
a height of 15 feet.

Liabum Pringlei. Half shrub, 3 or 4 feet high: stems terete,
fuscous-tomentulose : leaves opposite, ovate, acute, mucronulate-denticu-
late, green and puberulent above, densely white-tomentose and reticulated
with brownish veins below, rounded and 3-nerved at the base, 3 inches
long, two thirds as broad ; margins revolute ; naked petioles but 2 lines
long: heads few, large, 1^ inches in diameter, discoid: involucre cam-
panulate, multiseriate ; scales pale brown (in dried state), lanceolate,
attenuate, slightly sericeous: flowers about 75 in a head, yellow, 10 lines
in length : achenes 2\ lines long, sericeous : pappus of elongated stramin-
eous bristles and very short outer scales. — Collected by C. G. Pringle,
on mountains near Lake Chapala, Jalisco, altitude 7,000 feet, 18 October,
1895, no. 6215.

Cacalia peltata, HBK., var. Conzattii. Stem 1 to 3 feet high :
inflorescence lax, 1-20-headed ; heads 25-40-flowered ; involucre 12-
phyllous ; calyculate, scales covered with a densely spreading pubescence.
— Collected by Professor C. Conzatti on the Sierra de San Felipe,
Oaxaca, altitude 9,000 feet, 29 November, 1895, no. 27 ; also by C. G.
Pringle in the same region, altitude 10,000 feet, 13 December, 1895,
no. 6238.

Senecio prionoptenis. Erect herb, 1 to 3 feet in height : stems
somewhat flexuous. green, striate, slightly flocculose, x-ather broadly and
interruptedly winged by the decurrent bases of the leaves, simple to the
corymbose-paniculate flat-topped inflorescence : leaves lance-oblong, at-
tenuate to a very sharp narrow apex, irregularly serrate-dentate, half inch
VOL. xxxii. — 4

50 PROCEEDINGS OF THE AMERICAN ACADEMY.

broad, 3 or 4 inches long, glabrate above, floccose-tomentulose beneath,
not narrowed at the base but decurrent upon the stem for nearly two
inches in two broad wings, these sharply toothed especially near the
lower end : bracts lance-linear : pedicels spreading, 4 to 15 lines in length,
bearing several subulate bractlets : heads about 50, erect, half inch long
and including the 8 or 10 spreading oblong golden-yellow rays, about 9
lines in diameter: involucre calyculate, about 20-phylloiis ; scales atten-
uate ; disk-flowers 40 to 45, shortly 5-dentate : ligules 3 to 3i- lines long,
1 to 1-J- lines broad, 4-nerved. — Collected by C. G. Pringle, on hills, at
Las Sedas, Oaxaca, altitude 6,000 feet, 3 December, 1895, uo. 6282.

G-ochnatia Smitliii. Shrub ? leaves clustered near the ends of
the branches, oblong, entire, obtuse, cuneate at the base, thickish, grayish
and covered with a very fine short toraentum above (perhaps later glabres-
cent), much paler, tomeutose and veiny beneath : heads in numerous
slender close terminal globose glomerules, 8 lines long, about 6-flowered;
involucres green, very slender, and gradually tui'binate ; the scales in
many (8 to 10) series, extending as it were down upon the pedicels, ovate
to lanceolate, obtuse, pubescent : style-branches short, flattened, rounded :
divisions of the corolla subequal, narrow: acheue subvillous, 1^ to 2 lines
in length. — Collected by L. C. Smith, on hills of Cuicatlan, 30 April,
1895, at 3,000 feet altitude; also by E. W. Nelson along road from
Totolapa to San Carlos, altitude 3,000 to 3,800 feet, April, 1895, no.
2546. Noteworthy for its very long slender closely imbricated involu-
cres and globose inflorescences, the latter about 2 inches in diameter.

Perezia Cnemavacana. Glabrous : stems clustered, 2 to 3 feet
high, striate-angulate, purplish, leafy : leaves oblong, obtuse and mucron-
ulate at the apex, scarcely narrowed to the rounded subsessile base,
sharply denticulate with pungent teeth, reticulated and lucidulous on
both surfaces, 2 to 2h inches long, 9 lines broad, ascending, imbricated,
the uppermost gradually reduced : heads few, very large, about 30-
flowered, 12 to 15 lines long and broad, terminal on long many-bracted
peduncles : involucre turbinate, multiseriate ; the scales greenish, silky-
ciliate, the inner ones oblong, obtusish, 1^^ to 2 lines broad near the
summit, the outer gradually smaller and extending as subulate spreading
bracts some distance down the peduncle : corollas lilac in dried state, 9
lines in length : pappus tawny ; achenes puberulent, 31- lines long. — A
handsome species collected by C. G. Pringle, at Cuernavaca, altitude
7,000 feet, no. 6196.

Perezia umbratilis. Habit of P. niidicaulis : root horizontal, send-
ing off numerous stout fibres ; caudex multicipital, tufted with sordid

ROBINSON AND GUKENMAN. — MEXICAN PLANTS. 51

white wool : leaves all radical, ruuciuately pinnatifid, glabrous or nearly
so, petioled, thin, acutish ; lobes 5 to 7, luucronulate-dentate : stems
about a foot high, slender, dark colored, quite smooth, bearing a lew
scattered minute oppressed subulate bracts and at the summit 2 or 3
heads on slender ascendinjr branches: heads 9 lines lonij, about 18-
flowered : scales of the involucre in 5 series, obtuse, flattish (not thick-
ened nor firm in texture), dark colored, ciliated, otherwise glabrous, the
inner ones 5 lines long: flowers lilac in dried state, 7 to 8 lines long (in-
cluding the achenes). — Collected by C. G. Pringle, in shade, at Tomellin
Canon, Oaxaca, altitude 3,000 feet, 1 December, 1895, no. 5966, Closely
related to P. nudicauUs, Gray, but having a more numerously seriate in-
volucre with thinner flatter scales, and flowers nearly twice as large.

SONCHUS ASPEK, Vill. Although so generally distributed as a weed
upon waste heaps, etc., this plant does not appear to have been noted
in Mexican lists. It has been collected at San Diego, Chihuahua, by
Hartman, and near the city of Oaxaca, by Nelson, no, 1353.

Proceedings of the American Academy of Arts and Sciences.
Vol.. XXXII. No. 2. — Dfxkmbku, 1S9G.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGP:.

A BE VISION OF THE ATOMIC WEIGHT OF

31 A GNESimi

By Theodore William Richards and Harry George Parker.

%

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.

A REVISION OF THE ATOMIC WEIGHT OF

MAGNESIUM.

By Theodore William Richards and Harry George Parker.

Presented May 13, 1896.

Although numerous determinations of the atomic weight of magne-
sium have been made, the results obtained show such very wide variations
among themselves that the value in use at present cannot be accepted
with any certainty. It will not be necessary to review in detail all the
work published, as most of it was done more than forty years ago, before
quantitative methods had attained their present exactness ; but the follow-
ing table of methods used and results obtained will assist in a clear com-
prehension of the situation.

Previous work on the Atomic Weight of Magnesium.*

Synthesis of sulphate by the action of sulphuric acid on the oxide.

Berzelius, 1826. Lehrbuch, 5th edition, IIL 1227 25.3

Svanberg and Nordenfeldt, 1848. Erdmann's Journ.

Prak. Chem., 1848, XLV. 473 24.7

Bahr, 1852. Erdmann's Journ. Prak. Chem., 1852,

LVL 310 24.8

Marignac, 1884. Annal. Chim. Phys., 1884, (5.), L

289, 321 24.37

Conversion of sulphate into oxide.

Jacquelain, 1851. Annal. Chim. Phys., (3.), XXXIL

1 95 24.5

Determination of sulphuric acid in sulphate.

Gay-Lussac, 1820. Annal. Chim. Phys., XIIL 308 24.6

* We are indebted to F. W. Clarke for most of tlie above references.

56 PROCEEDINGS OF THE AMERICAN ACADEMY.

Scheerer, 1846. Poggeud. Ann., 1846, LXIX. 535 24.5

Scheerer, 1847, Later Correction. Poggeud. Ann.,

1847, LXX. 407 24.5?

Jacquelain, 1851. Annal. Chim. Phys., 1851, (3.),

XXXII. 195 24.2

Conversion of oxalate into oxide.

Svanberg and Nordeufeldt, 1848. Erdmann's Journ.

Prak. Chera., 1848, XLV. 473 24.7

Determination of chlorine in maguesic chloride.

Dumas, 1859. Ann. Chim. Phys., 1859, (3), LV.

129, 187 ' 24.6

Conversion of carbonate into the oxide.

Marchand and Scheerer, 1850. Erdmann's Journ. Prak.

Chem., 1850, L. 385 24.0

Scheerer, 1859, Later Correction. Liebig's Ann., 1859,

ex. 236 24.0

Conversion of metal into oxide.

Burton and Vorce, 1890. Am. Chem. Journ,, 1890,

XIL 219 24.29

It will be seen that, with the exception of the results obtained by the
precipitation of the sulphuric acid with barium chloride and the precipita-
tion of the chlorine with argentic nitrate, all the methods employed
involve the use of magnesic oxide. The fact that all such results are
untrustworthy was shown by T. W. Richards and E. F. Rogers* in their
work upon the occlusion of gases by the oxides of certain metals when
obtained by the ignition of various salts. The error from this source is so
large that it seems hopeless to apply a correction to previous work upon
the atomic weight of magnesium, as the amount of gas occluded depends
in a large degree upon the method and thoroughness of ignition.

Concerning the results obtained by the precipitation of the sulphuric
acid in magnesic sulphate, it is only necessary to point out the error due
to the occlusion of various soluble substances present in the solution
from which the precipitation was made. This error was recognized by
Scheerer, after publishing his results, and an approximate correction was
made; but such a correction does not merit much confidence, as will be
seen.

In the work of Dumas it is evident that some magnesic oxychloride

* Tliese Proceedings, Vol. XXVIII. p. 200.

RICHARDS AND PARKER. — ATOMIC WEIGHT OF MAGNESIUM. 57

was formed, ami he does not appear at all confident of the accuracy of
his results. From the experience of the writers it does not seem likely
that the method which he used would give magnesic chloride free from
the oxide.

Prellminary Experiments.

Because considerable experience had been gained in a previous re-
search.* upon the occlusion by baric sulphate of salts present in a solution
from which this insoluble salt was precipitated, it was thought that Gay-
Lussac's and Scheerer's method of precipitating magnesic sulphate with
baric chloride might now be used with advantage, applying subsequently
the necessary corrections for occluded substances. It had previously
been found that the concentration of the solution and the method of
pouring had a great deal to do with the amount of occlusion ; and hence
it seemed likely that, by working in a very dilute solution and pouring
the magnesic sulphate into the baric chloride with extreme slowness, the
occlusion of baric chloride might be lai-ge, but that the precipitate might
be free from magnesium. Several experiments were made to ascertain
the correctness of this supposition, but in each case it was found that,
notwithstanding the precautions adopted, a very notable quantity of
magnesium was occluded in the baiic sulphate. It had been the custom
in working upon this precipitation to fuse the weighed baric sulphate with
sodic carbonate, to extract the sodic chloride thus formed, and to deter-
mine the chlorine with argentic nitrate and calculate as baric chloride,
subtracting this amount from tlie total weight of baric sulphate found.
This method gave very satisfactory results, but of course it could not be
applied when the baric sulphate was mixed with magnesic chloride and
sulphate as well as baric chloride, for then no one could discover the
proportion in which each salt was present, with sufficient accuracy for
work upon atomic weights.

The possibility of obtaining .satisfactory resi;lts by the determination
of the chlorine in magnesic chloride was now considered. The great
disadvantage of this method, as is well known, is the extreme difficulty
of obtaining pure anhydrous magnesic chloride. The usual method of
iiTiiiting the double chloride of ammonium and magnesium was tried a
number of times, but it was found that a quantity of the ox3'chloride was
always formed. As indicators do not give a sharp reaction in the pres-
ence of magnesic salts, the hydrochloric acid driven off cannot be added

* These Proceedings, Vol. XXXI. p. 67

58 PROCEEDINGS OF THE AMERICAN ACADEMY.

afterwards by titrating back to the neutral point with a weak acid solu-
tion, and it is therefore necessary to obtain in the first place maguesic
chloride containing its full complement of acid.

The method was then modified by conducting the ignition of the double
salt in a tightly covered platinum crucible in a stream of hydrochloric
acid, instead of air. That a considerable quantity of oxychloride was
usually formed, even under these conditions, was easily ascertained by dis-
solving the resulting product in water, when the oxychloride remained as
an insoluble residue. In two or three cases, however, the amount of oxy-
chloride formed was comparatively small ; hence it was hoped that, if the
right conditions could be found, the chloride might be obtained in a pure
state. Another series of experiments with a modified apparatus was
therefore undertaken. The expulsion of the ammonic chloride was con-
ducted in a combustion tube and the number of drying tubes was increased,
so that the hydrochloric acid gas might be as free as possible from water.
The heat was applied very gradually, in order that the double chloride
might be almost anhydrous before the sublimation of the ammonic chloride
began. This method gave better results. It was observed that in two
or three experiments, where the conditions had been unusually favorable,
the resulting chloride gave a clear solution ; and it seemed therefore
probable that, if an apparatus could be devised to deliver a rapid stream
of hydrochloric acid gas entirely free from aqueous vapor, the method
might be successful.

Assuming that these conditions might be fulfilled, another difficulty
remained to be overcome ; for even if the magnesic chloride could be
obtained in the combustion tube free from water and oxychloride, the
problem still remained to weigh the salt without foreign admixture. If
the boat were allowed to remain in the tube until cool, and then removed
to a weighing bottle, the salt must absorb a very notable quantity of
moisture from the air in the operation, however quickly this operation
might be performed. The boat cannot be transferred to another tube
and reheated, as the moisture present reacts upon the chloride, forming
some oxychloride and liberating h3Tlrochloric acid. If it is taken from
the combustion tube while hot and allowed to cool in a weighing bottle,
the same effect is produced. Dumas had met with the same difficulties
in his work with this method, and he endeavored to compromise matters
by lemoving the boat from the combustion tube when it had only partly
cooled. As his subsequent results proved, however, the moisture from
the air reacted upon the chloride, forming some oxychloride, which
interfered seriously with the accuracy of his work. To obviate this

lilCIIARDS AND PARKKU. — ATOMIC WKIGHT OF MAGNESIUM. 59

difficulty the form of apparatus used by one of us* in drying strontic bro-
niide was altered so that the boat could be transferred directly from the
ignition tube to the weighing bottle without an instant's exposure to the
outside air. In order to accomplish this result the hard glass tube was
ground with a long tapering joint directly into the wider desiccating or
cooling tube used to contain the weighing bottle. This desiccating tube
had a sort of bulb or " pocket" blown upon one side of it, to receive the
stopper of tiie weighing bottle, thus allowing the boat to be pushed past
the stopper directly from the ignition tube into the bottle. Afterwards
the stopper could be rolled into place with a rod provided for the pur-
pose. Tlie arrangement was used with great success in a recent deter-
mination of the atomic weight of zincf to which it was equally applicable.
A reference to the annexed sketch will make the apparatus more com-
prehensible.

Fig. 1. — Bottling Apparatus, Horizontal Section.

A — weighing bottle. B = stopper of bottle. C C = bard glass tube.
D= Platinum boat containing fused magnesic chloride.

The desiccating apparatus for the hydrochloric acid gas consisted of
two towers, composed of a number of glass bulbs filled with beads, upon
which strong sulphuric acid was allowed to trickle from small reservoirs
at the top into suitable receptacles at the bottom. This apparatus was
constructed wholly of glass, with glass gridirons for flexibility, and ground
or sealed glass connections. Joints were made tight with syrupy phos-
phoric acid (Morley). The hydrochloric acid, after being evolved by
allowing strong sulphuric acid to run into a flask containing a strong
solution of hydrochloric acid, was passed through a wash bottle contain-
ing sulphuric acid, thence through the towers just described, afterwards
through a tube containing phosphoric pentoxide, and finally into the com-
bustion tube. The apparatus was so ari-anged that the current of air from
an aspirator could be passed through another set of towers, a duplicate of
those used for drying the acid gas. By means of stopcocks either dry

* Richards, These Proceedings, XXX. 383.

t Richards and Rogers, These Proceedings, XXXI. 158, 174.

60 PROCEEDINGS OF THE AMERICAN ACADEMY.

hydrochloric acid gas or dry air could be passed through the tube con-
tainiiisj the weio-hiiia; tube aud boat.

With the help of this contrivance it was found possible to drive off the
amnionic chloride in a current of dry hydrochloric acid, to drive off the
excess of acid from the fused maguesic chloride by means of a current of
perfectly dry air, and to shut up the pure salt in a weighing bottle with-
out the least possible means of access of a trace of aqueous vapor. The
details of the method will be described later ; magnesic chloride prepared
after this fashiou gives a perfectly clear solution in water. Since this
problem was solved, attention was now turned to the preparation of
materials for the atomic weight determinations.

Preparation of Materials.

The sample of ammonic magnesic chloride which will be hereafter
referred to as sample No. 1, was prepared as follows. About five hun-
dred grams of ordinary " C. P." magnesic chloride was saturated with
hydrogen sulphide, a small amount of ammonia was added, and the whole
was allowed to stand in a warm place for several days. To the super-
natant liquid after decantation a small quantity of very pure ammonic
oxalate was added. The magnesic chloride thus almost wholly freed
from calcium was again decanted ; and after more ammonic oxalate
had been added, the whole was allowed to stand, and the clear liquid was
yet once more decanted. The solution was then evaporated to dryness,
and the resulting cake dried in an oven and ignited in a platinum dish.
The mixture of magnesic oxide and oxychloride thus formed was washed
with the aid of a filter pump for about sixty hours. At the end of this
time, although the wash water contained no sodium, the insoluble pre-
cipitate was not free from that metal. The precipitate was therefore
dissolved in hydrochloric acid, previously distilled in platinum for the
purpose, and the solution was filtered. In order to eliminate the sodium,
a portion of the magnesium was precipitated by passing into the solution
a current of ammonia gas. The precipitate formed b}' this very wasteful
process was washed for several days, at the end of which time it was found
to be free from any appreciable traces of sodium and potassium, when
tested with the spectroscope.

Ammonic chloride was now prepared by mixing streams of ammonia
and hydrochloric acid gas. This gave ammonic chloride mixed probably
to a certain extent with various amines, but free from inorganic salts.
As the amines must be driven off later, it was not thought worth while
to take the trouble of removingr them at this staire of the work.

RICHARDS AND PARKER. ATOMIC WEIGHT OP MAGNESIUM. 61

The solution of ammouic chloride thus prepared was added to the
solution of magnesic chloride obtained by dissolving the oxyciiloride in hy-
drochloric acitl in proportions corresponding to formula Mg.Cl.,(NIl4)CI,
and the mixture was carefully evaporated to dryness and gently heated
in an oven. It is of course unnecessary to say that all the latter part
of this purification was done as far as possible in platinum. The solid
cake was powdered in an agate mortar, and placed in a glass stop-
pered bottle which was kept in a closed jar. The doable chloride
thus prepared was then tested with the spectroscope, but no impurities
could be discovered ; and its solution in water was perfectly clear. Tests
were made with amnionic oxalate and baric chloride, but in neither case
was a precipitate formed on long standing.

The second sample of magnesic chloride was treated in a similar way
up to the point where it was necessary to get rid of sodium and potas-
sium. The solution was evaporated to dryness in a platinum dish with
the aid of an alcohol lamp, and the resulting cake was gently ignited and
then washed for a long time, nothing but platinum being allowed to
come in contact with the material from this time forth, and all the heat-
ing being done by means of alcohol lamps to avoid the danger of con-
tamination of sulphur from illuminating gas. The oxychloride thus
formed was then dissolved in pure hydrochloric acid and filtered. By
evaporating down again the magnesium was again rendered insoluble.
This process was repeated again and again, until there was no trace of
sodium or potassium remaining.

The ammonic chloride necessary for the preparation of the double salt
from this second sample of magnesic chloride was prepared by digesting
ammonic chloride with nitric acid to destroy the amines.* It was then
dried, sublimed several times, recrystallized five or six times from its
aqueous solution, and again sublimed in a current of air which had been
passed through wash bottles containing respectively a concentrated
solution of potash and sulphuric acid. After having been sublimed in
this manner about ten or twelve times, it was dissolved in redistilled
water and added to the sample of magnesic chloride. The whole was
then filtered, evaporated to dryness, partly dehydrated, broken up and
placed in a glass stoppered bottle. The usual tests were made as to its
purity, but no traces of foreign matter were discovered.

The third sample of magnesic chloride, which was used for the final
experiment in the last series, was at first treated in about the same way

* Kruss, Liebig's Annal., CCXXXVIII. 51.

62 PROCEEDINGS OF THE AMERICAN ACADEMY.

as the others. The precautions taken were somewhat greater, aud the
fractional precipitation with amnionic oxalate was continued long after
the last traces of calcium discoverable by the spectroscope had disap-
peared from the precipitates of magnesic oxalate. The ammonic mag-
nesic chloride, already very pure, prepared from this sample, was then
crystallized eight or ten times, the last six or eight recrystallizations
being conducted in platinum. From over a kilogram of magnesic
chloride used in the beginning, the portion finally separated out consisted
only of a few grams. This sample showed no traces of the sodium line
when tested with the spectroscope ; indeed, several other samples, ob-
tained from the mother liquors of the purest sample, gave equally satis-
factory negative spectroscopic results. Since the magnesic chloride had
contained in the first place a very noticeable amount of sodic chloride,
the fact of the complete elimination of the impurity seemed a satisfactory
indication of the elimination of other foreiijn materials. The double
chloride was dried over an alcohol lamp, and treated in the same manner
as the other samples.

Purification of Silver.

No very great labor was expended iipon the purification of the first
quantity of silver, as the chlorine in magnesic chloride was to be pre-
cipitated with an undetermined excess of silver nitrate. Residues were
therefore worked up by dissolving silver (obtained by reduction with
zinc) in nitric acid, precipitating the metal as chloride, and converting
tlie chloride into metallic silver by means of invert sugar. The reduced
silver, after having been fused into buttons, was thoroughly washed and
dissolved in nitric acid. The solution of argentic nitrate thus obtained
was diluted very much with water, allowed to stand, aud filtered just
previous to using.

With the second sample, on the other hand, much greater care was
taken, as it was designed in this case to ascertain the direct ratio between
silver and magnesic chloride. The material came partly from some
refined silver, purchased in the market, aud partly from some pure silver
residues remaining from previous work. The silver was jirecipitated
from a solution of the nitrate with pure hydrochloric acid, and reduced
by means of invert sugar and pure sodic hydrate, the sodic hydrate
having been previously freed from heavy metals by electrolysis. Both
the chloride and reduced silver were very thoroughly washed, the silver
was dissolved in pure nitric acid, and the process was repeated. After
this cycle of operations had been performed four or five times, the re-

RICHARDS AND PARKER. — ATOMIC WEIGHT OF MAGNESIUM. 63

duced silver was fused on a cupel of sugar charcoal before the blowpipe.
The resulting button was scrubbed witli sand, and made the anode of a
weak galvanic circuit in a solution of argentic nitrate prepared from the
same silver. The cathode was a piece of pure silver wire, upon which
the whole of the silver was deposited iu a crystalline mass. The silver
crystals were then removed from the solution and fused in a vacuum
upon a boat of pure lime,* which was contained in a porcelain tube.
Such a boat may be made by lining a porcelain boat with a mixture of
three parts of pure lime and one part of pure anhydrous calcic nitrate,
and igniting the mixture. The i)orcelaia boat is thus covered with a
firm, coherent layer of pure calcic oxide. In order to prevent the pos-
sibility of a trace of organic matter distilling off from the rubber stoppers
usually used to close sucli a tube, a set of hollow brass stoppers were
made, through which a current of cold water circulated. This latter
device is due to a suggestion of Professor Henipel. The construction of
this piece of apparatus is evident from the diagram.

<-N

°t.-5^

j^

es^2^^^ ^

Fig. 2. Apparatus for Fusing Silver, Vertical Section.

A is connected with Sprengel pump. B B = hollow brass stoppers in porcelain
tube. C = boat of lime containing silver. D =r " window " for observation.
E E E E = rubber packing of stopper. F = Fletcher furnace.

Of course the button after fusion showed no trace of spirting from con-
tained oxygen. It was scrubbed with distilled water and clean sand,
and divided into small pieces by means of a clean steel chisel. The
fragments were alternately boiled in strong hydrochloric acid and digested
in ammonia water, this process being repeated ten or fifteen times. The
silver was finally washed with distilled water and afterwards kept in a
desiccator, which was opened only when necessary to weigh out silver
for a determination.

A portion of the second sample was treated in the same way, except
that in the end it was fused on sugar charcoal before the blowpipe and

* These Proceedings, XXX. 379; XXXI. 173.

64 PROCEEDINGS OF THE AMERICAN ACADEMY.

cooled in the reducing flame. Particular 25ains were taken to prevent
the absorption of oxygen, and the button did not show the slightest trace
of having contained this gas. From this portion wire was prepared of
various thicknesses, by means of a draw plate ; and the weights of given
lengths of these wires were determined, so that small weights could be
made with considerable accuracy. Of course the wire was treated in the
same fashion as the rest of the silver, in order to remove any iron which
might be present on the surface.

The third and fourth samples of silver were prepared in the same man-
ner as the second, the starting point being the pure residues left from the
analyses made with previous samples. No qualitative nor quantitative
difference could be observed between any of these preparations of silver.
Fused upon sugar charcoal, they melted to a clear globule free from any
film, — a fact which in itself, according to Stas, is an excellent test of the
purity of silver, — and all gave practically the same results in later
determinations.

All water used was redistilled with potassic permangajiate, some of it
being condensed in a platinum condenser and some of it by means of a
tube of pure block tin, which was carefully tested in order to prove the
absence of an impurity of lead. Considerable quantities upon evaporation
in platinum left a scarcely appreciable residue, there being apparently no
difference between the water condensed in tin and that in platinum.*
The water was prepared as short a time as possible before being used,
and was carefully kept in a suitable bottle fitted with a siphon, air being
admitted to the bottle through a filter of cotton wool. It was carefully
tested for chlorine by means of the nephelometer from time to time.

The sulphuric acid used for the preliminary drying of the gases
was the usual " chemically pure " acid of the Laboratory, of a specific
gravity of about 1.83. For the final drying this acid was boiled down in
platinum.

Weighing.

The balance used was a long-armed Becker, sensitive to about one
thirtieth of a milligram with the largest load that it was required to carry
during the investigation, while the weights were a good set of gold plated
ones, which were kept in the balance case under a glass cover. These
weights were very carefully compared with one another, and all weigh-
ings were, of course, reduced to the vacuum standard. The specific

* See These Proceedings, XXVI. 249; XXX. 380.

RICHARDS AND I'AIIKEU. — ATOMIC WEIGHT OP MAGNESIUM. 65

gravity of magnesic chloride used for this computation was the value
2.177 deterruiued by Playfair and Joule. Weighing \va.s done by sub-
stitution, the object to be weighed being placed on the right-hand pan
and balanced by tare weights on the left. In general, the precautions
used in the recent work done in this Laboratoi-y upon co|)per, barium,
strontium, and zinc were adhered to with great care.* We are indebted
to the Cyrus M, Warren Fund of Harvard University for some of the
platinum ware used in the following work.

The atomic weights used in this investigation were as follows : —

O . . . 16.000 Ag. . . . 107.930

CI . . . 35. -156

Method of Work.

The method of operating may be inferred from the description of the
apparatus. The platinum boat, after having been weighed within its
wei"-hing bottle, was filled with the double chloride of ammonium and
magnesium and placed in position in the ignition tube, resting upon a
sort of carriage of platinum foil. The weighing bottle was placed with
its stopper in appropriate position in the " bottling tube," as previously
described. A current of dry hydrochloric acid gas was then passed
through the apparatus and the ignition tube was heated by a suitable
arrangement of burners. At first the I'esidual moisture was driven off
by the heat and carried away by the stream of gas. When as much
water as possible was expelled in this manner, the heat was slowly
increased so that the ammonic chloride commenced to vaporize. It was
found that the sublimation commenced before the salt was freed from the
last traces of moisture, but an effort was always made, by the very
gradual increase of heat, to make this proportion of water as small as
possible ; and it is probable that the salt was practically anhydrous some
time before the last of the ammonic chloride was sublimed. When no
further evolution of ammonic chloride could be observed, the heat was
increased until the tube and boat were heated to redness, and the mag-
nesic chloride had fused into a clear, colorless limpid liquid. It requires
a very excellent piece of combustion tubing to stand the heat necessary
to fuse magnesic chloride, and a number of tubes were spoiled during the
course of the work. In the first series of determinations the boat was
allowed to cool while the current of hydrochloric acid gas was still pass-

* Kichards, These Proceedings, XXVI. 240 ; XXVIII. 1 ; XXIX. 55 ; XXX.
369 ; XXXI. 158.

VOL. XXXII. — -5

66 PROCEEDINGS OP THE AMERICAN ACADEMY.

ing. The tubes containing the boat and weighing bottle were then
thoroughly washed out with a current of air dried in an apparatus simi-
lar to that used for drying the hydrochloric acid gas, as previously
described. After it was certain that all of the acid gas had been dis-
placed, and while the current of air was passing rapidly to prevent any
diffusion of moist air back into the apparatus, the bulbs were removed
fi'om the farther end of the ignition tube, and the boat was pushed into
the bottle in the manner already described. The boat itself remained
constant in weight during these operations, showing that the magnesic
chloride had not acted upon it.

After weighing, the boat and its contents were placed in a large glass-
stoppered Erlenmeyer flask, and the magnesic chloride was dissolved in
pure water. The chlorine was precipitated with a dilute solution of
argentic nitrate ; * and after a thorough shaking the whole was allowed
to stand in the dark over night. The argentic chloride was washed by
decantation a number of times, with vigorous shaking, and was finally
collected upon a Gooch crucible in the usual manner. The precipitate
was dried from five to ten hours in an oven, carefully protected from
dust and dirt, and weighed. After weighing, the cake of jirecipitate,
together with some adherent asbestos, was removed to a tared porcelain
crucible and heated uutil it bes;an to fuse. The crucible was ajjain
weighed, and the loss of weight, if any, noted, and subtracted from the
weiu;ht of the Gooch crucible and contents. The filtrate, containing a
little dissolved argentic chloride, was evaporated down to small bulk
and filtered through a very small filter ; and the weight of the precip-
itate was added to the weight of the first portion. In some cases the
small amount of argentic chloride present was determined with the
nephelometer. t

The wash water from the precipitate collected on the Gooch crucible
was also run through a small filter to make sure that no asbestos had
been carried away from the crucible in the process of washing ; and this
correction, when appreciable, was applied in the appropriate place.

The washing and filtration were both performed in dim orange light,
which had been suitably tested as to its non-actinic properties. Even
after fusing the argentic chloride was almost colorless, showing that only
unessential traces had been decomposed by the light.

The result of the first series of five experiments is given below. These

* This solution contained never more than one per cent of silver,
t See Tliese Proceedings, XXX. 385.

RICHARDS AND PARKEU. — ATOMIC WEIGHT OP MAGNESIUM. 67

determiuatious were consecutive, except that one determination met with
an accident and was not completed.

SERIES I.

No. of
Exp.

Sample of
MgOl,
used.

Sample of
Ag used.

Wei','ht of

MgCl,.

Weight of
AgCl.

Iljitio.

MgCU : 2 AgCl =:

lUO : n.

Atomic
Wei^lu
of Mg.

1

i.;»o50

4.01952

300.975

24.368

2

1 51601

4.56369

301.033

24.350

3

1.32413

3.98528

300.974

24.369

4

1.40GG4

4.23297

300.928

24384

5

1.25487

3.77070

300.963

24.373

Average

24.369

A careful consideration of the possible constant errors involved in the
foreirointr results leads to the belief that the fiorures found are too high
rather than too low, as the presence either of a small amount of water or
of oxjchloride in the magnesic chloride would tend in this direction.

Second Series of Determinations.

In order to drive all the subliming ammonic chloride to the farther
end of the combustfbn tube during the ignition, it had been found neces-
sary that the current of gas should be very considerable ; and hence it
was desirable to construct a piece of apparatus which should deliver the
various gases rapidly, but nevertheless as dry as it is possible to obtain
them. It was also desirable to work with larger quantities of materials
than could be handled in the former apparatus. For these reasons an-
other piece of apparatus was constructed to dry the hydrochloric acid
gas ; this apparatus contained several flasks of sulphuric acid, three very
efficient towers containing the same acid, which was constantly renewed,
and a long tube containing resublimed phosphoric pentoxidg. One of the
towers is shown on the following page. The whole apparatus was fused
or ground together, thus wholly avoiding rubber or cork connections.

In the following determinations the boat was allowed to cool in an at-
mosphere of dry nitrogen, as a further precaution against a possible
partial decomposition of the sensitive magnesic chloride. As soon as the

68

PROCEEDINGS OP THE AMERICAN ACADEMY.

salt had been fused, a current of dry nitrogen was passed into the com-
bustion tube and the hydrochloric acid generator was disconnected. The
nitrogen was prepared by passing mixed air and ammonia over rolls of

copper gauze heated to redness, the excess of
ammonia being removed by passing the gases
through wash bottles containing dilute sulphuric
acid ; and the nitrogen was dried in a set of
towers similar to those used for drying the cur-
rent of air. AVlien the tube was cool, the cur-
rent of dry air was turned on, and the tube and
its contents washed out as in previous experi-
ments.

As there were no especial objections against
the use of rubber connections and stoppers in
the part of the apparatus used for drying the
air, several large towers were employed, each
filled with crushed pumice stone and saturated
with sulphuric acid previous to using. Both air
and nitrogen were finally dried by resublimed
4/ ^Ipw I phosphoric pentoxide. The bottling and com-

'"' vv*<y// lIU bustion tubes were of the same construction as

in the former apparatus, except that they were
larger.

In the second series, the method of icrnitinof
the double salt to obtain the magiiesic chloride
was the same as in the first ; but the method
of estimating the amount of chlorine present was
different. From the approximate atomic weight
of magnesium already found, a calculation was
made as to the amount of silver necessary ex-
actly to precipitate the chlorine present in the
sample of magnesic chloride taken. This amount
of silver was weighed out as nearly as possible, dissolved in nitric acid in
an Erlenmeyer flask, provided with a set of bulbs to catch the spray from
the evolution of gas, and added to the solution of magnesic chloride con-
tained in a large flask. The flask was thoroughly agitated in the dark,
and allowed to stand over night. Fifty cubic centimeters were then
withdrawn by means of a pipette, and tested by means of a nephelometer,
or apparatus for determining the amount of precipitate from the intensity
of the opalescence produced by it. This piece of apparatus was cod-

FiG. 3. — One of the Towers

USED FOR BRYING IlTDnO-

CHLORic Acid.
Seventy centimeters high.

RICHARDS AND I'ARKKR. — ATOMIC WEIGHT OF MAGNESIUM. 69

stnu'ted for the purpose, and consisted of two rectangular glass cells, with
a mirror enclosed in a dark case, so arranged that the column of liquid
contained iu the lower part of the cells could be viewed horizontally with-
out disturbance from surface reflections. A dark scieen was placed at
the further end of the cells, and the whole so arranged tliat light could
come to the eye only by reflection from solid particles which might be
suspended in the oolunni of liquid inspected. If the liquid was perfectly
clear, the field of vision remained black, but an extremely small amount
of precipitate produced a very marked change, and the intensity of opales-
cence was approximately proportional to the amount of precipitate. It
was found perfectly easy and certain, by this method, to distinguish the
difference between .002 and .003 of a milligram of argentic chloride or
between .004 and .005 of a milligram, and larger amounts in proportion
Tliis instrument gave such satisfaction in this research tiiat the method
will be worked out for various other reactions, and published later.

The method of using this apparatus was as follows. Twenty-five
cubic centimeters of the clear supernatant liquid from the flask contain-
ing the well shaken argentic chloride and magnesic nitrate were placed
in eich cell, five cubic centiijieters of a very dilute solution of argentic
nitrate being added to one, and five cubic centimeters of a correspond-
ingly dilute solution of ammonic chloride to the other. The silver
solution contained one milligram of silver to the cubic centimeter. An
unequal depth of cloudiness indicated an excess of either silver or
chlorine iu the original solution, and accordingly the amount necessary
for neutralization was run into the large flask containing precipitate and
solution from a burette. The solution was again allowed to stand in the
dark with occasional shaking, and after the precipitate had entirely sub-
sided was again tested as-before, and this cycle of operations was repeated
until the opalescences matched one another.* It will be observed that,
if water is added to the cell giving the more dense opalescence until the
effect becomes equal on both sides, the amount of dilution will give a
means of ascertaining the amount of precipitate in each cell. The appro-
priate corrections were then applied to the amount of silver taken. Due
allowance was made for the slightly diminishing volume of the solution
in the flask. The addition of one tenth of a milligram of silver to a litre
of solution produced a distinct change in the depth of color observed.
After the matching was completed, repeated trials were made with fresh

* For details of this method, see Stas, Mem. Acad. Belg., XLIII. Part II., and
Richards, These Proceedings, XXIX. 86; XXX. -385.

70

PROCEEDINGS OF THE AMERICAN ACADEMY.

portions of the solution to confirm the result ; and as the depth of opales-
cence as seen in the nephelometer was perfectly fiat, without disturbing
eflections, the end point could be determined with great precision.
Several results obtained in this manner are given in Series II.

SERIES II.

No. of
Exp.

Sample of

MgCl^
used.

Sample of
Ag used.

Weight of

Mg CI2.

Vi'eight of

Ag.

Ratio.

MgCU :2Ag =

100 : n.

Atomic
\Veight
ofMg.

6

7
8

1
1

1

1

1

2

2.78284
2.29360
2.36579

6.30284
5.19560
5.35989

226.490
226.526
226.558

24.395

24.379
24.366

Averacre

24 380

These results, however, do not merit great confidence ; for the appara-
tus, which had become somewhat complicated, did not work smoothly
at first, on account of some minor imperfections which were remedied
later. Besides this, careful consideration led to the suspicion that the
towers used for drying the air and nitrogen were not efficient enough to
remove the last traces of water. Of necessity the towers had to be
charged with sulphuric acid an hour or two before their final use, and
during that time a large part of the acid drained out of the pumice stone.
This surmise was fully confirmed by later experiments ; and since this
was the case, the second series must be rejected in the final estimate of
the atomic weight.

Third Series of Determinations.

In order to remedy the most serious defect of the second series, the
arrangement for drying the air and nitrogen was much enlarged and im-
proved. By pouring sulphuric acid into the safety funnels at the top of
the many towers, from time to time, during the passage of the gas, the
glass beads were kept thoroughly saturated during the whole process.
Tlie sulphuric acid having reached the bottom of the column drained out
of the tubes provided for that purpose into beakers below. It will be
seen that by this means the efficiency of the apparatus was far greater
than in the previous form. As a test, a very rapid stream of wet air

RICHARDS AND PARKKR. — ATOMIC WlOKiHT OF M.VGNESIUM. 71

from a water blast was i)assed through the apparatus and then through a
weighed phosphorus pentoxide bulb fur ncaily two hours, without the
slighte.xt appi-eciable increase of weight of the pentoxide bulb. The same
test was applied to the ajjpai-atus for drying the hydrochloric acid gas,
with tlie same result.

Willi the help of this important addition to the apparatus, another
series of deiermiuatious was now made. The somewhat lower result of
this series is undoubtedly due to the more perfect desiccation of the
eases : the agreement of the individual results is still not quite perfect,
but the series is undoubtedly far more reliable than the second.

SERIKS 111.

No. of
Exp.

Sample of
MgCl.
used.

Sample of
Ag used.

Weight of

MgCU.

Weight of

Ag. •

Ratio.

MgCU : 2 Ag =

100 : 11.

Atomic
Weight
OfMg.

9
10
11
12
13
14

1
1
1
2
2
2

2
2
2
2
3
3

1.99276
1.78870
2.12832
2.51483
2.40672
1.95005

4.51554
4.05256
4.82174
5.69714
5.45294
4.41747

226.597
226.565
226.551
226.542
226.571
226.531

24.349
24.363
24.369
24.373
24.360
24.377

Average

24.365

Fourth axd Final Series of Determinations.

The apparatus was now put in the best possible order, and the phos-
phorous pentoxide tubes were recharged, in order to make ready for a
series of determinations in which the very highest exactness was to be
aimed at. The purest samples of material were used, and all other pre-
cautions, learned from previous work, were taken to insure accuracy.
The following determinations were consecutive, with the exception of one
between Nos. 15 and 16, which was spoiled by a slight accident.

These results agree with one another as well as could possibly be
expected, for the difference between the extremes in the last series
corresponds to a difference of only one tenth of a milligram in the \vei<^ht
of the magnesic chloride. Since two wholly distinct samples of this salt

72

PROCEEDINGS OP THE AMERICAN ACADEMY.

SERIES IV.

No. of
Kxp.

Sample of
MgClj
used.

Sample of
Ag used.

Weight of
MgCU.

M'eight of

Ag.

Ratio.

MgCl., : 2 Ag =
lUO : u.

Atomic
Weight
ofiMg.

15
16
17
18
19
20

2
2

2
2
2
3

3
3
3

2
2
4

2.03402
1.91048
2.09932
1.82041
1.92065
1.11172

4.00855
4.32841
4.75635
4.12447
4.35151
2.51876

226.573
220.562
226.566
226.508
226.565
226.564

24.360
24.364
24.362
24.362
24.363
24.363

Avprflff'^^

24.362

Extreme difference . .

0 004

and three wholly distinct samples of silver were used in this series, we
may conclude that all ordinary accidental errors had been eliminated ;
and in a critical discussion of the result wer may limit ourselves to the
consideration of the possible constant errors of the process.

The most serious objection to the method is, of course, the possible
retention of water, of magnesic oxycliloride, or of ammouic chloride by
the magnesic salt.

With regard to the first two difficulties, it need only be said that the
gases used for drying the magnesic chloride were as dry as present pos-
sibilities permit them to be made. Tlie phosphorus pentoxide in the
last drying tube showed no trace of liquefaction at the close of the
research, but seemed to be as light and powdery as at first, in spite of
the fact that several hundred litres of gas had been passed over it. Any
trace of oxygen, as well as of aqueous vapor, was excluded from the hot
salt; for the hydrochloric acid gas was replaced by nitrogen, and this
was driven out in its turn by dry air only after the tube had cooled. A
means of proving absolutely that no water remained does not exist ; but
it is extremely hard to see how water could have gained access to the
carefully guarded magnesic chloride.

The fact that every sample of magnesic chloride used in the last series
gave an absolutely clear an'd transparent solution in water is additional
evidence of much weight ; for a very small trace of oxychloride would

RICHARDS AND PARKER. ATOMIC WEIGHT OF MAGNESIUM. 73

have sliowu itself iu opalescence. As a proof of this it may be stated
that iu experiment No. 12 of Series III. there was a perceptible cloudiness
upon the solution of the magnesic chloride in water, owing to a known
access of a trace of aqueous vapor, caused by a momentary stoppage of
the current of nitrogeu. This result is, however, scarcely at all different
from the others.

"With legard to the possible retention of amnionic chloride by the
majrnesium salt, it may be said : first, that none could be detected by
means of a Nessler solution ; and, secondly, that even if a small amount
had been retained, it would have made but a very slight difference in
the final result.

Our result is essentially the same, no matter whether the chlorine is
weiidied as arijentic chloride (Series I.), or the amount of silver neces-
sary to precipitate it is found (Serifs III. and IV.). This fact is satis-
factory evidence that that silver and chlorine were both pure, as well
as that no magnesic chloride was occluded by the argentic chloride.
Thus : —

From the ratio 2 AgCl : MgCls (Series I.), Mg = 24.369.
" " 2 Ag : MgCla (Series III.), Mg = 24.365.
« « 2Ag : MgCla (Series IV.), Mg = 24.362.

Upon comparing these figures with the older ones, they are seen to
agree surprisingly with Marignac's value obtained from work upon
magnesic oxide and sulphate (Mg = 24.37). Burton and Vorce's syn-
theses of magnesic oxide gave a lower value for magnesium (24.28);
but if these were corrected for a probable amount of gases in the mag-
nesic oxide, the result would probably be close to the present one. The
analytical chemist should not forget that the value 24.36 is one and a
half per cent higher than the round number 24, which has been so com-
monly accepted.

For reasons which must be manifest to any careful reader of the fore-
going paper, we accept the value given by the fourth and last of our
series as representing the most probable atomic weight of magnesium.
It remains only to state this result in terms of the usual unfortunately
varying standards of reference used by the scientific world.

If O = 16.000, Mg = 24.362.
If 0 = 1.5.96, Mg = 24.301.

IfO = 15.88, Mg = 24.179.

Cambridge, May 1, 1896.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. No. 3. — January, 1897,

OX THE GROUP OF REAL LINEAR TRANSFORMA-
TIONS WHOSE INVARIANT IS A REAL
QUADRATIC FORM.

By Henry Taber.

ON THE GROUP OF REAL LINEAR TRANSFORMATIONS

WHOSE INVARIANT IS A REAL

QUADRATIC FORM.

By Henry Tabkr.

Presented May 13, 1896.

In what follows G will denote the group of real linear homogeneous
transformations of determinant +1 whose invariaut is the real quadratic
form of non-zero determinant,

where Q is a real symmetric matrix.*

Since the quadratic form J is real, the roots of its characteristic
equation are all real. In the Proceedings of the London Mathematical
Society, Vol. XXVL, page 376, I have shown that for n = 4 the group
of all proper linear homogeneous transformations, real and imaginary,
whose invariant is the real quadratic form J, provided the roots of the
characteristic equation of JF are not all of the same sign, contains a real
transformation that cannot be generated by the repetition of any infini-
tesimal transformation of this group; and that therefore, a fortiori^
cannot be generated by the repetition of an infinitesimal transforma-
tion of group G. It follows that for w > 4 not every transformation of
group G can be generated by the repetition of an infinitesimal transfor-
mation of group G if the roots of the characteristic equation of J are not
all of the same sign.

On the other hand, if the roots of the characteristic equation of jp
are all positive or all negative, every transformation of group G can be
generated by the repetition of an infinitesimal transformation of this
group, t

* I employ the notation of Cayley's " Memoir on the Automorpliic Linear
Transformation of a Bipartite Quadric Function," Philosophical Transactions, 1858,
with tliis exception : the transverse of a matrix or linear transformation <p will be
denoted by 0 .

t If the roots of the characteristic equation of Jf are all of the same sign, group
G is isoinorphic with the group of real proper orthogonal substitutions ; and every

78 PROCEEDINGS OP THE AMERICAN ACADEMY.

I shall therefore assume that w > 4, that the roots of the character-
istic equaiiou of jf are not all of the same sigu ; and I shall show that
a transformation of group G can be generated by the repetition of an
infinitesimal transformation of group G if it is an even power of a
transformation of this group.

If the matrix or linear homogeneous transformation ^ transforms Jf
automorphically, it satisfies the matrical equation

^ n «^ = O, (1)

in whch ^ denotes the transverse or conjugate of 0. Conversely, if
this equation is satisfied, ^ transforms jp automorphically. The deter-
minant of any transformation satisfying this equation is equal to either
-f 1 or —1. By definition the totality of real proper solutions of equa-
tion (1) constitutes group G.

If </) is any real solution of equation (1), we may put

cji = (f)o (f)i = cfii cf)o,

where <^o, <^i, are polynomials in ^, are both real, and are both solutions
of equation (1), that is

Moreover — 1 is not a root of the characteristic equation of ^i, therefore
the determinant of <^i is equal to +1 ;* and consequently ^i is a trans-
formation of group G. Finally,

and therefore

</>' = 4>' i>i' = <^i'. (2)

That is, the second power of any real solution of equation (1) is the
second power of a transformation of group G.f

transformation of this group can be generated by the repetition of a real infini-
tesimal orthogonal transformation.

* The roots, other than ±1, of the characteristic equation of any solution of
equation (1) occur in pairs, the product of two of the same pair being unity.
The determinant of a linear transformation is equal to the product of the roots
of its characteristic equation.

t If — 1 is a root of multiplicity vi of the characteristic equation of (p, and if
the roots of this equation other than —1 are gi, r?.., etc. of multiplicity, respectively.
p-,, pn, etc., then

where

[(./) + !)'»- (.71 + 1)'"?' [{<i> + 1)"* - {fj.2 + 1)'"]P'

[- (.'/I + 1)"'P' ' I- (.9. + 1)'"]^' " "•

l-7^r + f,x^-etc.

TABER. — REAL LINEAR TRANSFORMATIONS. 79

Let now e^ denote the infinite series

1 +x + ^x'+3iX' + etc.,
convergent for any Hnite matrix. We have

(>) = e\
and for any integer lu,

moreover, if ^ and -^ are commutative,

gX + X ^ e^ e^ = e^ e\

Corresponding to any finite matrix, ^, of non-zero determinant can be

found a polynomial x i" ^ ^"^^ ^^^^

<j> = e^.
The infinite series

4!
and

are also convergent for any finite matrix, and are equal respectively to

Therefore, if x and e^ ^~ are both real, the second power of the latter is
equal to the identical transformation. For if x and e are real,

gxi^_g-x1'^^0;
that is,

Since the determinant of (^i is not zero, by what precedes, a polynomial
X in <^i can be found such that

and since — 1 is not a root of the characteristic equation of <^i, x niay be
so chosen that, if

^ = x«"S

we shall have

^ = —&,

80 PROCEEDINGS OF THE AMERICAN ACADEMY,

that is, & is skew symmetric. If now

where 0 and ly are real, both 6 and rj are skew symmetric, that is,

e = -d, ^ = -7). (3)

Since O is real, 9 O and rj O V — 1 are the real and imaginary parts
respectively of i9- O = x-' -^^^ since the latter is a polynomial in the real
matrix <^i, its real and imaginary parts, $Q and rj H V — 1? are polyno-
mials in <^i, and are therefore commutative. Consequently, by virtue of
a theorem given above,

on

_ ea + r,n |/-i _ en ^n i- 1

Since <^i is real, and since 6Q and therefore e^" is real, it follows

that e' ^~ is real. Therefore, by what precedes, since r]Q, is real,

(y,a\'^2^ 1.
Whence we have

and therefore by (2)

,o ,2 2 9.0

If now we put
where m is any positive integer, \]/ is real, and

Moreover, siuce 12 == li, and since, by (3), ^ = — ^, we have

and therefore

* For any positive integer p,

Therefore

TAKER. — REAL LINEAR TRANSFORMATIONS. 81

Finally, we may show in precisely the same way that c"^" is a real

solntioQ of eqnation (1) ; and therefore, since (e™^")^ = c"^^", it follows
that if/ is tlie second power of a solution of equation (I), and is thus of
determinant +1, Wherefore, t/^ is a transformation of group G.*

2

By takincc m sufficiently ^reat, —6Ci may be made as nearly as we

please equal to zero, and therefore i// = e^^" may be made as nearly as
we please equal to the identical transformation. Wherefore, <fi- = ^i^
can be generated by the repetition of an infinitesimal transformation of
group G.

The converse is of course also true ; that is to say, every transforma-
tion generated by the repetition of an infinitesimal transformation of
group G is the joth power (for any positive integer p), and therefore the
second power, of a transformation of this group. This may be shown as
follows. The transformation (ft, if infinitesimal, may be put equal to
1 + 8« . X) where Bt is infinitesimal. If <^ is a transformation of group
G, 8t and ^ are both real, and we have

(I + 8t . x) ^ (I + Bi • X) — $ ^ "1^ = ^ '
That is, neglecting terms involving the second power of 8t,

xfi + Ox = 0.

If we put ^ = X ^~S ^ is real, and the preceding equation becomes

fi5 0 4-0^0 = 0;
that is, ]

The transformation resulting from m repetitions of the infinitesimal
transformation

</)r= 1 + 8<.x= 1 + s< . ^n

is the transformation (1 + Si! . ^ O)'", which, if m is infinite, is equal to
e'" ' •^", or e'^" if we put t = m8t. The repetition of an infinitesimal
transformation of group G results in a transformation of this group ; and
therefore, as we have in fact already seen, e'^" for any real quantity t is a

* The matrix e'" is also a transformation of group G. The totality of trans-

formations c'" for all possible real values of m constitutes a continuous one term
sub-group which contains the identical transformation.

VOL. XXXII. — 6

82 PROCEEDINGS OF THE AMERICAN ACADEMY.

-ea

transformation of group G. But then e^ , for any positive integer p, is

a transformation of group G ; and since (c^ y = e'^", any transforma-
tion generated by the repetition of an infinitesimal transformation of
gioup G is the pth power, for any integer p, of a transformation of this
group.

It may be shown that any transformation of group G that cannot be
generated by the repetition of an infinitesimal transformation of group
G is tlie (2 p +l)th power of a transformation of this group for any
integer j).

For any transformation of group G that can be generated by the
rejjetition of an infinitesimal transformation of group G, the numbers
belonging to each negative root of the characteristic equation are all

even.*

Postscript.

Let (3 denote the group of all linear homogeneous transformations,
real and imaginary, of determinant +1 whose invariant is the real quad-
ratic form of non-zero determinant,

J = (n^ Xi, X., . . . x")- ;

and, as above, let G denote the sub-group of real transformations of
group ®.

A transformation of group (3 can be generated by the repetition of
an infinitesimal transformation of this group if — 1 is not a root of the
characteristic equation of the transformation, or if — 1 is a root of
the characteristic equation, provided the numbers belonging to — 1 are
all even.f

As stated above, if a transformation of group G can be generated
by the repetition of an infinitesimal transformation of this group, the
numbers belonging to each negative root of the characteristic equation
of the transformation are all even. These conditions, though necessary,
are not always sufficient. Thus, for n = 2, the form jf is transformed
automorphically if

/>*' ' 'Y* 'vi' ■ — ^v» •

* For definition of the numbers belonging to a root of the characteristic equa-
tion of a transformation, see Tiiese Preceedings, Vol. XXXI. p. .336.
t Proceedings of the London Mathematical Society, Vol. XXVI. p. 374.

TABKR. — RKAL LINEAR TRANSFORMATIONS, 83

auil this transformation of group (7, if the roots of the characteristic
equation of jf are not both of the same sign, is not the second power
of any transformation of group G, and therefore cannot be generated
by the repetition of an infinitesimal transformation of this group. Since
the numbers belonging to —I are all even, the transformation defined
by the above equations can be generated by the repetition of an infini-
tesimal transformation of group (5,

Jl-ly 28. 1896.

1

i

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. No. 4. — Dfxembkr, ISOG.

COXTRIRUTIONS FROM THE ZOOLOGICAL LABORATORY OF
THE MUSEUM OF COMPARATIVE ZOOLOGY AT HAR-
VARD COLLEGE, E. L. MARK, DIRECTOR, No. LXXIII.

STUDIES IN MORPHOGENESIS, VI.

A CONTRIBUTION TO THE QUANTITATIVE STUDY OF CORRE-
LATED VARIATION AND THE COMPARATIVE
VARIABILITY OF THE SEXES.

By C. B. Davenport and C. Bullard.

fP.f'

CONTRIBUTIONS FKOM THE ZOOLOGICAL LABORATORY OF THE

MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD

COLLEGE, E. L. MARK, DIRECTOR, No. LXXIII.

STUDIES IN MORPIIOGEXESIS, VI.

A CONTRIBUTION TO THE QUANTITATIVE STUDY OF CORRE-
LATED VARIATION AND THE COMPARATIVE
VARIABILITY OF THE SEXES.

By C. B. Davenpokt and C. Bullard.

Presented October 14, 1896.

The following quantitative study of variation is based upon counts of
the Miillerian glands of the fore legs of 4,000 swine. Our attention was
directed to these glands as favorable objects of study by Gertrude Grotty
Davenport, who had already collected some data concerning their varia-
bility. These data, together with valuable suggestions derived from her
own experience, she generously placed at our disposal.

The positions of the Miillerian glands ai-e indicated upon the wrist by
large openings or pits, about 1 mm. in diameter, which are found only
upon the inner aspect of the fore legs. The number of pits is variable.
Where there are several they occur, for the most part, in a single row
trending somewhat obliquely to the long axis of the leg.

Of the 8,000 legs examined, the arrangement of the glands was studied
on only 2,000 legs, 1,000 male and 1,000 female. The total number of
glands on a single leg varies from 0 to 10. When the number is large,
some of the glands are frequently found outside the main row. In no
case have we found more than nine glands in one row. We may call
those lying outside the main row lateral glands. The lateral glands
usually (six exceptions) occur at the upper (proximal) end of the series.
Their number does not usually exceed two, but in a single case we have
found four. These four glands lay in a secondary row parallel to the
main row, which contained five glands. In one other case, where three
lateral glands were found, these lay parallel to the main row of five.
When there are two glands they may lie .either in a line parallel with
the main row, or make any angle up to 90° with it. Lateral glands
occur more rarely when the total number of glands on the leg is small,
but we have found one extreme case in which the only two glands on the
leg occurred side by side, i. e. in a transverse row.

88

PROCEEDINGS OF THE AMERICAN ACADEMY.

The reduction in the number of glands takes place from one end, —
the distal end of the series. Generally, where there are only two or
three glands these occur high up and the normal diatance apart. Rarely,
however, the reduction is brought about in part by the failure to de-
velop in the middle of the series while glands develop near the extremes,
so that there is a broad hiatus in the series.

Since the proximal end of the series is that at which glands are most
likely to be formed, and since they tend to be produced more abundantly
there, this end, which occupies the region of the upper wrist, is to be
considered as the source of the moi'phogenic impulses which give rise to
the glands. Sometimes the embryonic Anlage does not develop beyond
this point ; sometimes, on the other hand, it develops along the whole
extent of the wrist in one row, and even forms an accessory "lateral"
row.

The total number of swine examined was, as stated, 4,000 ; of which
2,000 were males and 2,000 females. The total number of fore legs
examined was, accordingly, 8,000 ; 4,000 left and 4,000 right. All of
the observations fall, consequently, into four grovxps of 2,000 cases each;
namely, male right, male left, female I'ight, female left. These four
groups will be considered, for the most part, separately.

We first determined how many legs in each of these classes had no
glands, one gland, two glands, and so on. The results are given in the

following table.

TABLE I.

No. of Glands.

0

15

14

29

1

2

3

4

5

6

7

8

16
30

9

12
10

10

1
3

Total.

d II
d" L

225
241

466

353
336

437
430

411

429

297
295

155
159

314

78
53

131

2,000

2,000

Total d

689

867

840

592

46

22

4

4,000

9 R
$ L

15
21

36

209
213

422

365
361

482
438

414
432

277
288

134
149

283

72
69

22
16

38

8
11

2
2

2,000
2,000

Total ?

726

920

846

565

141

19

4

4,000

Total d+ 9

65

888

1415

1787

1686

1157

597

272

84

41

8

8,000

DAVENPORT AND BULLARD. — CORRELATED VARIATION.

89

In the following table, which is based upon the preceding, the num-
bers are all reduced to per milles. The two lines of totals are here,
accordingl}', replaced by means. A glance at this table shows a close
parallelism between the distribution of glands in the four cases.

TABLE IL
Summary per Mille.

Xo. of Glands.

0

1

2

3

4

5

C

7

8

9

10

d" R
d- L

7.5
7.0

112.5
120.5

176.5
168.0

218.5
215.0

205.5
214.5

148.5
147.5

77.5
79.5

39.0
26.5

8.0
15.0

6.0
5.0

5.5

0.5
1.5

1.0

Mean d"

7.2

116.5

172.2

216.8

210.0

148.0

78.5

32.8

11.5

9 R
? L

7.5
10.5

104.5
106.5

182.5
180.5

241.0
219.0

207.0
216.0

138.5
144.0

67.0
74.5

36.0
34.5

11.0
8.0

4.0
5.5

4.7

1.0
1.0

1.0

Mean ?

9.0

105.5

181.5

230.0

211.5

141.3

70.7

35.2

9.5

Mean of d
and ?

8.1

111.0

176.8

223.4

210.7

144.7

74.6

34.0

10.5

5.1

1.0

Several interesting questions now arise :

(1) How closely similar is the average number of glands in the two
sexes, and in the right and left leg of the same sex ?

(2) "Which sex shows the greater variability, and to what extent is it
greater? Is the relation between the variability of the right and left legs
closer than that between the two sexes ?

(3) How closely correlated are the numbers of glands on the right
and left legs of individuals ? That is to say, what are the chances that
a swine which has 2, 4, or 7 glands on the right leg will have the same
number on the left leg also?

1. J7ie Relation between the Abundance of Glands and the Sex or the

Side of the Body.

The average number of glands on a leg of either sex is determined by
dividing the total number of glands counted in that leg by the number of
individuals of that sex, in this investigation 2,000. This gives us the
following result : —

90

PROCEEDINGS OF THE AMERICAN ACADEMY.

S50

225

200

175

150

125

100

75

50

25

•+::::::::: :::V:::::::::

p fi

-llllllllllllllnNNllml

i __ : _j_^ Lii

::__ i^^A-^t--

::::::::::: :.^:::r:-::::

j_, i--»

:__:_:_:__:_/ \-^---

tj- ":Vi:::

^4 {^---

:::::::::::n:: :::: :^::

^ §

, r---

________________

:::::::::- |::::::::?:^:

, .f^__ r

_ \^

::::::::: |::::::::::d:r

:-::::::+r:::::::::i::t3

_ Vj

Ih^- l^A

+ -+

._.._.__f __..____._____.

±::::::E::::::=::::::::?

j-

;f^

'»

If

ti ::: ::::::::::::::: ::

1

1

^^

__ _

H^ :

:tr :: :: : :"-

5a -

u

if - - — T--

_:^^___ :::

::::i:f:::::::::::::::::q

:±:::=::::::::::::::::::

J —

__c__

(►+

_ U

;t - z

__::;__ :

— :_ ir- — — ---

1

f

il

^O-

__ \l

====?'===================

^::::i^::::::::::::::::::i

::::E^:::::::::::::::::::

1

___.5,__.._.._...__.._.._

h'

7

___^

:::::^t::::::::::::::::::

t

::::::c::::::::::::::::::

1

-\

U

i__2 -

^

\

5

— I

_..

::::: :::::::::::::::::::

2JL

4 J

-5

;p : ::_: : _

:: ::: _^_s

<>

:::::::::1r:^::::::::::::

:::::::i:::r:S:::::::::::

ji:: ::::::::::::::::::

:::::: i:::::^::: ::::::

iiL,

_i _i

:__;_:_: __::^::ii:fl:2

8

10

DAVENPORT AND BULLARD. — CORRELATED VARIATION. 91

1. Average number of glands in (^ R . . . . 3.547

2. " " " ^ L . . . . 3.540

3. " " " 9 R . . . . 3.501

4. " " " 9 L . . . . 3.521

Comparing the average of (1) and (2) with the average of (3) and (4)
it appears that the average number of glands in the males (3.544) is
tolerably close to that in the females (3.511) but that a real dilfereuce
exists between the two. llie (jhtiids are slif/htly less abundant in the
female than in the male in the ratio, 100 : 100.94. The average number
of glands on the right side of the body is so close to tliat on the left side
(3.524:3.531) that we may conclude: The average numbers of the
glands on the right leg and on the left leg taken without regard to sex are
about equal.

2. Variability correlated with the Sex and with the Side of the Body.

In seeking to determine whether, in this matter of glands, male or
female swine are the more variable, it is necessary to employ a method
of stating variability quantitatively. Quetelet, Stieda, and Galton *
have employed such a method, based upon the fact that the organs of
an animal vary about their mean dimensions to an extent and with a
frequency indicated by the probability-of-error equation,!

y = k.e-^'-''"-.

Two of the principal features involved in such a distribution are that
deviations of a given size are equally apt to occur above and below the
mean, and that small deviations are more apt to occur than large ones.
These and other characters of the '•' probability " curve are indicated in
that shown in dotted line in the accompanying diagram. The diagram
also shows the curve of distribution of the various numbers of glands oc-
curring on a leg, from 1 to 10. This curve is drawn from the right female
leg only ; the curve for the other legs would be very similar. We shall
speak in a moment of the method of construction of these curves ; but we
want now to call attention to the fairly close similarity of the two curves,
— that gained by observation and the theoretical one, — a similarity so
close that we are justified in concluding that the law of distribution of the
variants in the lesr glands of swine is the same as that of accidental errors.

* Quetelet, Lettres sur la theorie des probabilites, Bruxelles, 1840. Stieda,
in Archiv fiir Anthropologie, Bd. XIV. pp. 167-182. Galton, Natural Inheritance,
New York and London, 1889.

t See any text-book on " Least Squares."

92 PROCEEDINGS OF THE AMERICAN ACADEMY.

This being granted, we can express quantitatively the degree of varia-
tion in the glands by determining the average deviation in the number of
glands of any set of legs from the mean number of that set. Thus in
the right leg of the female the mean number of glands is approximately
3.5. Since there is no individual with 3,5 glands on the leg, every in-
dividual shows in the number of its leg glands a departure of at least
0.5 from the mean. Adding together the departures of every individual
and dividing by the total number of individuals (2,000) we get the mean

departure, which is known to mathematicians as the mean error, and is

■S3,

indicated by the formula , in which 2x indicates the sum of the

n

individual departures, a:, and 71 the total number of individuals. Pro-
ceeding in this way, the average departure, as an Index of Variability,
was determined to be as follows for each set of legs : —

Average departure of J^ R 1.41089

« " " J^ L 1.41083

« « " 9 R 1.36457

" " " 9 L 1.38766

These determinations indicate that the variability of the right and the
left legs, of the male is exactly the same to four places of decimals ; that
the variability of the right leg of the female is slightly less than that of
the left lecf, and that the male shows a greater variability than the female
in the ratio of about 1.411 : 1.376, or 1.025 : 1.000. In other words,
the male is 2.5% more variable than the female.

As we have seen, the variabilities of the right and of the left sides of
the male are practically equal. In the female, the left side is more va-
riable. Disregarding sex, we find the variability of the left side is to
that of the right as 1.3993 : 1.3877, or as 1.0084 : 1. That is to say,
the glands are 0.8% more variable on the left side than on the right.

Let us now compare the relative variability of symmetrical legs with
that of the two sexes. "We find the variability to be greater between the
same leg (say the right) in opposite sexes, than between symmetrical legs.
The relation may be expressed by the ratio 1.025 : 1.0084, or 1.016 : 1.
These numbers indicate a closer morphogenetic kinship between the two
legs of a symmetrical pair than between the corresponding leg in differ-
ent sexes.

We may now briefly indicate the method of constructing the proba-
bility curve in the diagram. The abscissas represent the numbers of
glands from 0 to 10 on a leg, and the ordinates the corresponding number

DAVENPORT AND BULLARD. — CORRELATED VARIATION. 93

of individuals per niille. The mean number of glands is 3.50 and the
index of variability is 1.3G4G. With these data, we can draw a prob-
ability curve including about the same area as our observed curve. This
curve, the continuous line, is drawn from the equation

h is the so called index of precision, and is equal to the reciprocal
of the index of variability divided by the square root of tt, thus,

n

h =

'^ '\/t

e is the base of the Napierian system of logarithms; namely, 2.1718.

h
k is a constant determined by multiplying the quotient of —i^ by

Y TT

the interval {dx) between successive values of x, in this case, 1 ; thus,

hdx .

X indicates deviations from the mean value, and y the corresponding
ordinates.

When X = 0, 7/ =: k, which is thus the length of the ordinate at the
mean value of x. Its value gives the percentage of cases which should
theoretically occur at the mean; it is in this case 23.3%. Like h, k
might be taken as a measure of precision, since it increases as varia-
bility diminishes.

3. The Degree of Correlation betioeen the Number of Glands on the
Right and the Left Legs of Individuals.

To get quantitative results in this matter we must employ a method
devised by Galton.* This method depends upon the following procedure
and considerations. Separate the right legs into as many lots as there are
degrees of deviation from the mean number of glands. These lots may
be called the subjects. Find for each of the subjects the mean deviation
in the number of glands on the left legs of the corresponding individuals
(the relative). The deviation of any subject and the deviation of the
corresponding relative are to be compared. In order to make this com-
parison instructive, we must take into account the fact that left legs (for
example) are more variable than right legs. In order to eliminate this

* Galton 's method is explained in liis paper in the Proc. Roy. Soc, Vol. XLV.
p. 135, 1889.

94 PROCEEDINGS OF THE AMERICAN ACADEMY.

difference, divide the deviation of the subjects by their index of variabil-
ity and the deviation of the corresponding relatives by their index of
variability. If now, the correlation is perfect, those causes which have
produced a deviation from the mean in the right leg will act in precisely
the same degree on the left leg also, and thus the deviation of any rela-
tive will not differ from the deviation of the corresponding subject. If,
under these circumstances, we divide the mean deviation of the relatives
by that of the subjects, the quotients will average 1. This average
quotient is called the Index of Correlation. Thus, the index of perfect
correlation is 1.

Let us suppose, on the contrary, that there is no correlation whatever
between the number of glands in any subject and in the corresponding
relative, then, no matter what the number of glands in any subject, the
number in the corresponding relative is just as apt to be large as small,
and will be equal to the average number of glands in the whole group ;
in other words, no matter what the deviation of the subject is, that of the
relative will be 0. The average quotient obtained, under these cir-
cumstances, by dividing the deviation of relatives by the deviation of the
subjects, will consequently always be 0. Thus the index of entire lack
of correlation is 0.

An inverse correlation, in which a positive deviation of the subject
from the mean shall always be accompanied by a negative deviation of
the relative, will be represented by a minus quantity. Thus the correla-
tion of any two sets of compared organs will lie between +1 and — 1.
The size of the fractions lying between ± 1 and 0 will serve to indicate
the degree of correlation.

The quotient, r, obtained by dividing the deviation (always in units of
the average deviation) of the left legs by that of the corresponding right
will not be the same for all the lots of individuals. The true index of
correlation, R, will be found by taking the average of all the ratios,
r, /, /', /", etc. This process of finding R may be somewhat abbre-
viated from the following considerations. We have seen that

Deviat. of Rel. d^

also that

_ Avg. Dept. of Rel. A^ .

Deviat. of Subj. dg

Avg. Dept. of Subj. Ag

R^- (r + r' + r" + . . . . r") = - 2r ;

DAVENPORT AND BULL.\RD. — COHRKLATED VARIATION.

95

consequently,

li =

n

T "•" T, "^ ^. "^ ' • • ^".

n

A 1 ^ rfr

since J, and J,, are constant. It results from this that it is only
necessary to find the mean of the ratios of the untrausmuted deviations

and multi[ily this by the quotient - . This saves a great many divisions,

and it has been the method pursued in our work.

After this statement of the method of expressing correlation we now
pass to a consideration of the results obtained with the glands of the pig.
We shall consider first the correlation between the number of glands on
the right leg of the male with that on the left leg of the male.

TABLE III.

d R CORRELATED WITH c? L.

Relatives.
Subjects.

8
4
2

\l

5

151

65

14

5

1

21

2

58

154

88
27

7

336

Zl

9

96

173

119

24

8

1

430

U

bl

&l

11

8/

9Z

101

No. of

Iiuiivs.

Right leg.

O

I3

0.600
1.360
2.306
3.197
3.888
4.784
5.510
6.141
6.500
7.333
9.000

=3 -ii-^-
, — '

g.a 6

0/-
Ir
2r
Zr
ir
5r
Gr
Ir
8r
9/-
lOr

<i
O

28

128

153

92

16

8

1

429

7
28
77
101
58
20
3

1

295

1
6

26
52
48
18
5
3

159

8

11

16

17

3

3

53

1
9
7
9
2
2

30

0
5

2
2

1

10

2
1

15
225
353

437

411

297

155

78

16

12

1

-2.940
-2.180
-1.234
-0.343
+0.348
+1.244
+1.970
+2.601
+2.960
+3.793
+5.460

-3.547
-2.547
-1547
-0.547
+0.453
+1.453
+2.453
+3.453
+4.453
+5.453
+6.453

.829
.856
.798
.627
.768
.856
.803
.753
.665
.696
.846

No. of Indivs.
Left Leg.

14

241

3

2,000

Mean,

.772

96 PROCEEDINGS OF THE AMERICAN ACADEMY.

lu this table the first columu names the subjects, of which there are as
many as there are numbers of glands, viz. 0 to 10. In the succeeding
columns on a single line is exhibited the distribution of the number of
glands on the corresponding left legs. The column headed " Means of
Lefts " gives the average number of glands on the left legs of the indi-
viduals which make up the corresponding subject. The column headed
"Deviation of Rel. («f,. )" gives the deviations of the corresponding
" Means of Lefts " from the mean number of left male glands. The
column " Dev. of Subject (t^ ) " gives the deviation of the subjects from
their mean number. The last column is the quotient of d,. divided by c?,.
This gives r, r', r" , etc. The last number in the column is the mean of

all these values of r. This number multiplied bv -^ will give R^ the

1 1 T^ ^< ( 1.41083\ . , . , , T 1

value sought. iJut —r- = -t———7^\ is nearly unity, so that the Index
* A, \ 1.4108'jy ^ •"

of Correlation of the number of glands of the right and left legs is .772.

Galton has shown that the same ratio holds true when relative and

subject are interchanged.

By a process similar to the preceding we have found that the ratio of
correlation of right and left legs in the female is .783. This ratio is so
similar to that obtained for males as to justify the conclusion that the
index of correlation in variahility of the leg glands is approximately equal
in the two sexes, and is about .111.

The conclusions from this study may now be summed up. We have
in the leg glands of swine a serially arranged system of organs develop-
ing, for the most part, in one line, starting at one point, and extending
out a variable distance. On such a system of organs we investigate
quantitatively the question, How closely similar are the morphogenic
processes which determine the resemblance of these glands on the oppo-
site sides of the body and in the two sexes? First of ail, the average
number of glands is tolerably but not strikingly close on the two fore
legs and in the two sexes. The glands are nearly 1% more abundant in
the male than in the female. AVhen we come to study their variability
we find that the variants are distributed in accordance with the proba-
bility curve, very nearly. (See diagram.) A curious lack of symmetry
results from the fact that, since the mean lies at 3.5, variation is limited to
3.5 in one direction, but is unlimited (reaches as a matter of fact to 6.5)
in the other. The degree of variability in the right and left legs is,
especially in the case of the male, strikingly similar, being 1.41089 and
1.41083 in the two cases respectively, the difference being within the

DAVENPORT AND HULLARD. — CORRELATED VARIATION. 97

errors of the metliod. The males are about 2.o% more variable than the
females. The glands are 0.8 '<) more variable on the left side than on
the right. The relative variability of the same leg in the different sexes
is about 1.6% greater than that of the two legs in the same sex. The
degree of correlation in the variability of the right and left legs is about

,777.

Cambridge, M.\ss., July 25, 1896.

VOL. XXXII. — /

Proceedings of tlie American Academy of Arts and Sciences.
Vol. XXXir. No. 5. — January, 1897.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF THE CASE SCHOOL OF APPLIED SCIENCE.

INVESTrOATIONS ON AMERICAN PETROLEUM.

By Charles F. Mabery.

XXVI. — ON THE BUTANES AND OCTANES IN
AMERICAN PETROLEUM.

By Charles F. Mabery and Edward J. Hudson.

Aid cj the wobk described ix this Paper was given by the Academy from the C. M. Warren

Fund for Chejiical Research.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF THE CASE SCHOOL OF APPLIED SCIENCE.

XXVL — ON THE BUTANES AND OCTANES IN
AMERICAN PETROLEUM*

By Charles F. Mabkuy and Edward J. Hudson.

Presented October 14, 1896.

As one of us (C. F. M.) has shown f- the form of the butanes and oc-
tanes in American petroleum is not so well understood as other constituents
of the series C„H2n + 2' Concerning the butanes, except that an analysis
of A "•as condensed at 0° by Ronalds gave the composition required for
butane, very little has been published beyond the fact that distillates have
been collected at 0° and at 8°-9°. It, therefore, seemed advisable to
devote some time to the separation of the butanes for the purpose of ascer-
taining with greater precision what hydrocarbons with these boiling points
are actually contained in petroleum.

The Butanes.

The early investigations of Pelouze and Cahours | indicated the pres-
ence of butane in their most volatile distillates. A portion collected
between 5° and 10° gave with chlorine a '^chlorbutyl," boiling point
6o°-70°. In a higher distillate, whose boiling point was not given, amyl
hydride was suspected by the formation of a " chloramyl " boiling at
9&°-103°, but this hydrocarbon was not further identified. In the first
analysis of the gas from Pennsylvania petroleum, Ronalds § condensed an
oil at 0°-l°, specific gravity 0.6000 at 0°, that gave the composition

* For efficient aid in the work of this paper, I should acknowledge my ohliga-
tions to Mr. W. H. Wliitfield, who selected portions of the work on the butanes as
the subject of a thesis for the degree of Bachelor of Science, and to my assistants,
Messrs. C. A. Soch and E. Davidson. — C. F. M.

+ Proc. Amer. Acad., XXXI. 23.

t Compt. Rend. 1862, p. 1211.

§ London Chem. Soc, 1865, p. 54.

102 PROCEEDINGS OF THE AMERICAN ACADEMY.

required for butane. Ronalds also collected an oil between 6° and 8°,
specific gravity 0.6004, vapor density 2.178, which he regarded as a
mixture of butane with higher constituents. As will be shown in this
fjaper, nothing further has been done toward verifying the pentane from
which the peutyl chloride of Pelouze and Cahours was obtained, although
it was really derived from the pentane boiling at 30°, as we have ascer-
tained after many trials.

Probably on account of difficulty in procuring an adequate supply of
material, Warren* gave less attention to the butanes than to the higher
constituents. He collected a distillate at 0°, which, on the basis of Ronald's
determination, he assumed was a butane. After repeated and careful
distillation through his regulated condenser, Warren also separated a
liquid, boiling point 8°-9°. Although nothing further was done towards
determining its composition, Warren believed this body " to have been
sufficiently purified to justify the conclusion supported by analogy that
there is a constituent boiling at about 8°-9°." Since no other attempts
have been made to identify the butanes, further study of these bodies was
evidently necessary.

In returning to this subject, we assumed at first that it would only be
necessary to enforce the results already obtained, which seemed to be
supported by our first determinations. f But as the work progressed, it
soon became evident that this assumption would not be justified. Taking
advantage of the extremely cold weather in January, 1895, through the
kindness of Messrs. Schofield, Shirmer, and Teagle, we procured 45
litres of a second distillate of the naphtha from Pennsylvania and Ohio
petroleum, the most volatile portion that could be condensed with the
atmospheric temperature below 0°F. Since the naphthas from Pennsyl-
vania and Ohio petroleum are re-distilled together, it is difficult to pro-
cure this portion from either crude oil, unless, indeed, a run of the crude
oil were made in the refinery with this object in view. But this did not
seem necessary, since there can be little doubt that the constituents of the
most volatile portion of Ohio and Pennsylvania petroleums are identical.
During the continued cold weather of January, February, and March of
the same year, distillation of this product was continued through a Warren
condenser, filled with a freezing mixture, and the temperature of the dis-
tillates was kept very low to avoid loss so far as possible. In this manner
a long series of distillations was carried on with small loss, aside from
transference, except of the most volatile distillates. In these volatile

* Proc. Amer. Acad., XXVII. 56. t Ibid., XXXI. 23.

MABERY AND HUDSON. — AMERICAN PETROLEUM. 103

portions, —20° to —10°, the distillation proceeded from the heat of the
surrounding sitmosphere, and it was regulated by cooling the still. The
results obtained in the study of these distillates were so unexpected, the
entire process of distillation was repeated several times during the colder
months of the present year until in all 250 litres of 88° and 92°
•rasoline have been distilled, the portions below 12°, twenty times. The
interesting results in the distillations of the present year, especially in
excluding a butane at 8°-9°, are due to the patience and care of Mr.
Whitfield.

In studving the derivatives of these hydrocarbons, the chlorine substi-
tution products were first formed by bringing toi^ether chlorine and the
vapor of the hydrocarbon. The method we employed differed somewhat
in details from the method of Schorlemmer.* In the reaction between
chlorine and the hydrocarbon, the intensity of light must be carefully
rculated. In direct suuli£;ht so much heat is developed that a flame
appears wiih the separation of carbon. In a cloudy day the action pro-
ceeds slowly, and with less light it is suspended. We obtained the most
satisfactory results both as regards the progress of the reaction and the
yield of the monochlor derivative in clear sunshine with a sheet of news-
paper interposed. In all our distillates the action of chlorine has pro-
ceeded with great readiness in the cold. The distillates collected below
0° were well cooled by means of a freezing mixture. The condensation
of the chlorinated product was more complete in a capacious bottle, — a
two-litre bottle for 100-200 c.c. of the oil, with the delivery tube extend-
ing half way from the top of the bottle to the surface of the liquid. The
heavier chlorine comes into intimate contact with the vapor of the oil,
and the product condenses for the most part on the side of the bottle.
Some of it unavoidably escapes with the hydrochloric acid, and it may be
condensed in part by placing in front another bottle well cooled, or still
better by passing the escaping gas through water. We have not found it
essential to separate by distillation the chlorinated product from the hydro-
carbon until the action was finished. If it be stopped with a small amount
of the hydrocarbon unchanged, which may easily be determined by ex-
perience, the product is nearly all the monochlor-compound. The use of
iodine suggested by Schorlemmer has not seemed essential, at least in the
formation of the l)utane derivatives, since, without it, there is no difficulty
in obtaining 80 per cent of the monochlor-compound, the remainder con-
sisting for the most part of unchanged hydrocarbon, nor in our experience

* Philosophical Transactions, CLXII. Ill (1872).

104 PROCEEDINGS OF THE AMERICAN ACADEMY.

is the yield increased by the use of iodine. In the separation by fractional
distillation of the products of the ohlorination, our experience is not in
accordance with that of Schorlemmer, who stated that neither the chlorides
nor the acetates could be separated by distillation. With long continued
distillation, as our results will show, the crude product of the clilorination
may be separated into its constituents sufficiently to yield constant boiling
points and satisfactory analytical data.

Since Beilstein* found that isobatane boils at — 17°. 5, it seemed rea-
sonable that it should be contained in our product — 10°, especially since
a considerable portion of this distillate condensed below — 15°. In sub-
mitting 200 grams of this distillate in several portions to the action of
chlorine with all possible precautions, a considerable portion of the chlo-
rine product was lost, probably because it consisted to a large extent of
the more volatile chlorine derivatives of propane. Although much of the
unchanged hydrocarbon must have escaped with the gaseous hydrochloric
acid, 100 grams remained after the chlorination, which was washed, dried,
and distilled. Prolonged distillation, however, failed to collect any con-
siderable portion within sufficiently close limits to indicate the presence
in appreciable quantity of an individual hydrocarbon. A small quantity,
perhaps 5 grams, collected between 65° and 70°, doubtless isobutane,
which had not been completely removed into the higher fractions at 0°.
But the product distributed itself along in degree fractions from 30° to
150° at no point in quantities larger than half a gram or one gram.

Between —10° and —2° much smaller quantities of the hj'drocarbon
distillates were obtained, but within the limits —2° and 2°, 300 grams
came together, and the distillate — 0°-l° gave, as its specific gravity at
0°, 0.6029, and at— 12°, 0.6141. Ronalds found as its specific gravity
0.6000. A vapor density determination gave the following value : —

0.0945 gram of the oil gave 73.6 c.c. of vapor at 5°, and under
370.4 mm. of mercury.

Calculated for C^E^a- Found.

2.01 2.07

In quantities of 100 grams each, the fractions —2° to 2° were con-
verted into the chlorine derivatives. The chlorinition of such a quantity
is of necessit}' somewhat tedious ; it requires continuous operation during
ten or twelve days, and it cannot be hastened, since too rapid absorption
generates heat, and rapid escape of hydrochloric acid carries off the pro-

* Ann. Chem. Pliarra., CXLIV. 10.

MABERY AND HUDSON. — AMERICAN PETROLEUM. 105

duct. Witli care we were able to complete the absortiou of sufficient
chloriue without any considerable loss, lu subjecting the chlorine
product to fractional separation, after the fifth distillation it began to
collect between 65° and 70°, and after .the twentieth, 80 per cent col-
lected at 68°-69°. Of the remainder, 40 grams below 60° proved to
consist mainly of unchanged hydrocarbon with sufficient chlorine product
to raise the boiling point. Distillation of the higher portions was con-
tinued in 1° fractions to 160°, beyond which very Utile remained. To
prevent a slight decomposition above 150°, at first distillation was con-
ducted in vacuo ; but the boiling points soon fell below the point of decom-
position. In this entire series of 1° distillates, only at 120°-121° did
any appreciable quantity collect, and here only to the extent of 2 to 3
grams. The portions collected at other points were doubtless mixtures
in very small amounts of higlier chlorine products. The liquid 68°-69°
was shown by analysis and vapor density to have the composition required
for monochlorbutane.

I. 0.1288 gram of the oil gave 0.2440 gram COo, and 0.1166 gram HoO.
II. 0.2025 gram of the oil gave 0.3845 gram CO.2, and 0.1778 gram HoO.

III. 0.2537 gram of the oil gave by the Carius method 0.3950 gram

AgCl.

IV. 0.2645 gram of the oil gave 0.4078 gram AgCl.

Calculated for Found.

CiHaCl. I. II. III. IV.

C 51.89 51.65 51.77

H 9.73 10.06 9.76

CI 38.38 38.50 38.13

In a determination of its specific gravity at 24°, this butyl chloride
gave 0.8690 ; at 27°. 8, another determination gave 0.8648. The latter
value agrees closely with that of Perin and Puchot,* who found at 27°. 8,
0.8650.

In determining its vapor density by the method of Hofmann, 0.0975
gram of the oil gave 65 c.c. of vapor at 100° and under a tension of
36.44 mm.

Calculated for C^HsCl. Found.

3.20 3.30

In further evidence of its identity, this butyl chloride was converted
into the acetate by heating it twenty-four hours to 140° with fused

* Ann. Chem. and Pharm., CXXXVI. 1276.

106 PROCEEDINGS OF THE AMERICAN ACADEMY.

potassic acetate and glacial acetic acid, according to the method of Schor-
lemmer. Upon diluting the contents of the tube, the oil which separated
distilled for the most part at 116°-117°, corresponding to the boiling
point of isobutyl acetate, 116°.^. The acetate was converted into the
alcohol by heating with alcoholic potassic hydrate, and its boiling point,
107°-108°, showel it to be isobutyl alcohol, boiling point 108°-109°.
Upon analysis the alcohol gave the required percentages of carbon and
hydrogen : —

0.1737 gram of the substance gave 0.4090 gram CO2, and 0.2113
gram HgO.

Calculated for CiHgOH. Found.

C 64.86 64.22

H 13,51 13.52

Isobutyl sulphide was also prepared by boiling the chloride with an
alcoholic solution of potassic sulphide. It gave the required percentage
of sulphur : —

0.1640 gram of the oil gave by the method of Carius 0.2G31 gram

Calculated for (C4Hg)2S. Found.

S 21.92 22.03

As already explained, prolonged fractional distillation failed to show
the presence of any other chlorine derivative of this hydrocarbon except
one that dLstilled tolerably constant at 121°-122°. The quantity of this
product was too limited to admit of complete purification ; while the
analyses showed that it still contained monochlorbutane, the resulis are
sufficiently close to establish the presence of a dichlorbutane. Since no
dichlorhu'ane with this boiling point has hitherto been found, it is to be
regretted that our supply of material was so limited. But on account of
the vast amount of labor that would be necessary to obtain an adequate
quantity, it has not seemed of sufficient importance in the pressure of
other portions of this work.

This substance gave the following results on analysis : —

I. 0.1390 gram of tlie oil gave by combustion with plumbic chromate

0.1960 gram CO2, and 0.1887 gram H.,0.
II. 0.2360 gram of the oil gave 0.5245 gram AgCl.

Found.
I. II.

c

Calculated for
C.H.Clj.

37.79

H

6.30

CI

55.91

38.46
7.08

54.94

MABERY AND HUDSON. — AMERICAN PETROLEUM. 107

The quantity of this product was insufficient for a determination
of its specific gravity. The following value was obtained as its vapor
density : —

0.1 Goo gram of the oil gave 87.2 c.c. of vapor at 182°, under a tension
of 4-47 mm.

Calculated for C^IIsCIj. Found.

4.39 4.16

Our first attempts towards separating the butanes were made before
the necessity of long continued diatillution was fully appreciated. Not
until after several trials was it observed to what an extent a small pro-
portion of isobutaue depressed the boiling point of pentane. In the
earlier work, after six or eight distillations the product was assumed to
he sufficiently pure for chlorination. A study of the chlorine deriva-
tives show^ed beyond question that the hydrocarbon was pentane. We
were still further misled by the boiling point of the pentyl chloride, 96°,
which did not correspond to that of any hitherto published. Referring
to our former notes on the vapor density of the distillate 8°-9°, we found
the following determinations : —

I, 0.0717 gram of the oil gave 45.5 c.c. of vapor at 16°, under a ten-
sion of 48.1 cm. of mercury.
II. 0.0849 gram of the oil gave 46.8 c.c. of vapor at 18°.5, under
a tension of 47.75 cm. of mercury.

III. 0.0999 gram of the oil gave 51.5 c.c. of vapor at 18°. 5, under a

tension of 51.75 cm. of mercury.

IV. 0.0751 gram of the oil gave 43.5 c.c. of vapor at 18°. 5, under a

tension of 45.76 cm. of mercury.

Calculated for

Found.

C4Hio- C^Hij.

I.

II. III.

IV.

2.01 2.49

2.04

2.38 2.35

2.36

It will be seen that determination I. corresponds closely to the compo-
sition of butane, and determinations 11., III., and IV. to that of pentane.
When these determinations were made, we expected results that would
substantiate what had previously been done on the butanes, and there-
fore accepted only the first result, which was the last one obtained, as
supporting this view. But in the light of further knowledge of these
bodies, the determinations thousht to be erroneous correspond more
nearly to pentane, although showing the influence of the small quantity
of isobutane. The first result was published in a former paper,* when

* These Proceedings, XXXI. 24.

108 PROCEEDINGS OF THE AMERICAN ACADEMY.

it was assumed that we had in hand the butane which Warren had pre-
viously separated. Suspecting, therefore, the purity of our hydrocarbon
distillate 8°-9°, in resuming this work the distillation was conducted
with a sufficient quantity of material, and carried far enough to separate
completely all fractions between 5° and 20°. In the fraction 8°-9°,
after the tenth distillation, two vapor density determinations gave,
(I.) 2.17, (II.) 2.17. In a distillate 6°-8°, Ronalds* obtained as its
vapor density 2.178, evidently indicating as our values do a mixture of
butane and pentane.

But continuing this distillation in single degree fractions, scarcely any-
thing remained within these limits after the twentieth distillation. Hav-
ing obtained the same result in several quantities of forty-five litres
each, collet-ted at different dates, it became evident that neither Pennsyl-
vania nor Ohio petroleum contains a single body with a boiling point
between these limits. Furthermore, we have not been able to detect
normal butane in any of these volatile distillates. There is not the least
difficulty in separating chlorine derivatives of these hydrocarbons by
fractional distillation. All the chlorine products we have prepared have
collected readily within the limits of temperature of their characteristic
boiling points. But in no instance in the chlorination of hydrocarbon
distillates collected between — 10° and 20° has a distillate collected at
77°-78°, the boiling point of normal butyl chloride. Having excluded
a constituent of Pennsylvania and Ohio petroleum within the limits men-
tioned above, and having observed that certain properties of isopentane
derivatives are not in all respects identical with those previously published,
it seemed to us of sufficient interest to devote some attention to the deriva-
tives of this hydrocarbon. "We therefore collected several hundred grams
of a distillate at 29°-30°, and exposed the oil to the action of chlorine in
several separate quantities until about 200 c.c. of the substitution product
was obtained. In generating such large quantities of chlorine, without
discomfort from leaks, we have found the large Berlin porcelain stills
extremely convenient. The chlorinated product was washed, dried, and
carried through a long course of distillations, until after the eighteenth dis-
tillation it collected to the extent of 85 per cent between 95° and 96°, under
a constant tension of 730 mm., which for convenience was selected for all
these distillations. In a large series of distillations a constant tension is
simply and very conveniently obtained by means of the tension regulator
elsewhere described,! with the stopcock manipulated by a lever movable

* These Proceedings, XXXI. 10. t Journ. Ch. Soc, 1865, p. 54.

MABERY AND HUDSON. — AMERICAN PETROLEUM. 109

on a horizontal support. By a suitable adjustment of side tubes, stop-
cocks and adapters, any number of distillation flasks may be introduced
into the train, ami regulated by the single stopcock. As many as fifteen
distillations collecting in single degree fractions have been in simultane-
ous operation with no interruption incident to changes of flasks or collec-
tion of distillates. It is quite impossible, in Cleveland, to carry on a
course of distillations extending through several weeks without some
form of tension regulator, on account of sudden and extreme changes in
the barometric pressure, ocasionally equivalent to 25 mm. within a few
hours.

Under 760 mm. and with the thermometer wholly in the vapor, chlor-
pentane distilled completely between 1)6° and 97°, for the most part at
9 6°. 5. From a distillate below 30°, Pelouze and Caliours obtained a
butylchloride, as already mentioned, which boiled at 98°-103°.

The purity of our product was shown by analysis : —

I. 0.1205 gram of the oil gave 0.2479 gram CO2, and 0.1139 gram
ILO.
II. 0.2166 gram of the oil gave 0.2952 gram AgCI.
III. 0.2942 gram of the oil gave 0.4031 gram AgCl.

Calculated for Found.

C5H„ci. I. II. rn.

C 56.33 56.13

H 10.33 10.51

CI 33.33 33.69 33.88

A determination of the specific gravity of this chlorpentane at 20°
gave 0.8750. Its vapor density by the Hofmann method was found to
have the following value : —

0.1291 gram of the oil gave 72.6 c.c. of vapor at 100°, and under a
tension of 384 mm.

Calculated for CgHnCl. Found.

3.68 3.72

In forming isopentyl acetate, the purified chloride was heated 48
hours to 150°-160° with potassic acetate and glacial acetic acid. When
washed, dried, and well fractioned, a small quantity collected at 134°-
135° (760 mm.), which is somewhat lower than tlie boiling point of
isoamyl acetate, 138°. 6. The amount of our material was not sufficient
to raise its boiling point, but it gave the required percentages of carbon
and hydrogen.

110 PROCEEDINGS OF THE AMERICAN ACADEMY.

0.1556 gram of the oil gave 0.3651 gram CO2, and 0.1514 gram H2O.

Calculated for CjHuCoHaOj. Found.

C 64.60 64.00

H 10.80 10.81

When heated during several hours with alcoholic potash, the acetate
was converted into the alcohol which was separated from the solution
with salt. On account of the small quantity of the alcohol obtained, its
boiling point could not be raised above 117°-120° (bciling point of in-
active amyl alcohol 131°), although its composition corresponded to that
of amyl alcohol.

0.1866 gram of the oil gave 0.4620 gram COo, and 0.2293 gram H2O.
0.1854 gram of the oil gave 0.4610 gram CO2, and 0.2315 gram H2O.

Calculated for Found.

CfilliiOH. I. II.

C 68.18 67.51 67.81

H 13.63 13.66 13.88

At higher temperatures to 150°. the crude chlorine product could be
distilled without decomposition. Above this point the residue was dis-
tilled m vacuo, which reduced the boiling point to such an extent that
further distillation could be carried on under atmospheric pressure,
leaving only a very small residue above 150°, probably of substitution
products containing a larger number of chlorine atoms.

At no point between 96° and 160° could a distillate be held constant,
except at 144°-145°. Here 5 grams collected after the fifteenth distil-
lation that distilled at 144°-146°, mostly at 145°, unSer 760 mm. and
with the mercury column wholly in the vapor. This oil proved on
analysis to have the composition required for dichlorpentane.

I. 0.1309 gram of the oil gave 0.2055 gram CO.., and 0.0888 gram

HoO.
II. 0.1976 gram of the oil gave 0.3993 gram AgCl.

Found.
I II.

42.84
7.54

49.96

A determination of the vapor density of this body gave a value corre-
sponding to dichlorpentane.

Calculated for

C.^HioCla-

c

42.56

H

7.09

CI

50.36

MABERY AND HUDSON. — AMERICAN PETROLEUM. Ill

0.1680 gram of the oil gave 80.6 c.c. of vapor at 182°, and under a
teiisiou of 426 lutu.

Calculated for C5H10CI,. Found.

4.88 4.78

The boiling point of this diclilorpentane is the same as that of amylene
chloride, 145°, described by Bauer.* It is to be regretted that the quan-
tity of our [n-oduct was so limited ; but on account of the great amount
of labor involved in its preparation, it did not seem advisable to attempt
the preparation of a larger amount.

These results indicate that petroleum contains but one butane, and that
is isobutane, whii-h collects at 0°. That this body is not normal butane
is shown by tlie boiling point, 67°-68°, of the monochlor derivative, the
boiling point, 116°, of the acetate, that of the alcohol, 108°-109°, and
that of the sul[)hide, 172°, It has hitherto been assumed that the butane
in petroleum is the normal hydrocarbon, because the butane prepared by
Frankland from ethyl iodide and zinc boils at 0°. But isobutane pre-
pared by Butlerow from tertiary butyl alcohol should boil at — 17.°o.
The differences between the properties of butane and isobutane are suffi-
ciently marked to permit, without difficulty, of distinguishing between the
isomeric forms.

Normal butyl chloride boils at 77°. j6, and the alcohol at 117°. Since
our series of derivatives from petroleum butane was prepared several
times with the same results, and never with the formation of compounds
with boiling points corresponding to normal butane, there can be no ques-
tion that normal butane is not contained in Pennsylvania nor in Ohio
petroleum. The serious question concerns the boiling point of isobutane
which we found in petroleum. As has been fully explained, our 0°
fractions were separated as many as four different times from the lightest
gasoline we could procure from the refinery. While we never failed to
collect distillates below — 10°, they were much smaller in quantity than
those in the vicinity of 0°. It is, therefore, certain that the butane we
had in hand was isobutane, and it seems equally certain that this hydro-
carbon collected at 0° and not at —17°.

The Octanes in Ohio Petroleum.

It has already been pointed out by one of us (C. F. M., he. cit.)
that the published accounts of the octanes in petroleum are not fully

* Zeit. fiir Chem., 1866, p. 380.

112 PROCEEDINGS OF THE AMERICAN ACADEMY.

concordant. Pelouze and Cahours first announced in their series of
hydrocarbons an octane, boiling point 116°-118°, and a little later
Schorlemmer* found a body in coal tar distilling at 119°-122°, and an-
other at 124°, both with the composition required for octane. Possibly
led to believe by the results of Beilstein and Kurbatoff that the distillate
119°.5-122° is hexahydroisoxylol, in liis later summation of the octanes
known Schorlemmer seems to repudiate the one boiling at 119°-122°,
which he had previously described. Warren recognized two octanes, one
boiling at 119°. 5, and another boiling at 127°. 6. These bodies were
identified alone by vapor density determinations and boiling points ; no
analyses were given. In the separation of the hydrocarbons resulting
from the distillation under pressure of menhaden oil, Engler discovered
the octane, diisobutyl, boiling point 109°. In support of the work of
Pelouze and Cahours, Lemoine f submitted American ligroine to
four fractional distillations, collecting distillates within limits of jive
degrees, and separated an octane, boiling point 121° at 779 mm. In
our experience, the hydrocarbons in petroleum with boiling points not
far removed can only be separated by the use of the best possible means
for fractional separation, in a long course of distillations. The separation
of the octanes, for example, requires a refinery distillate boiling between
120° and 170°, and it is not until the twentieth distillation that the
octanes begin to accumulate with any degree of purity. In all our ex-
periments after the fifth distillation, the octanes are still contained for
the most part in the portions above 125°, and they are separated very
slowly into their respective fractions. In undertaking further study of
the octanes in Ohio petroleum, we procured from the refinery of Messrs.
Schofield, Shirmer, and Teagle 72 litres of a burning oil distillate that
distilled in our hands, within the limits of ten degrees, in the following
proportions : —

—75° 75°-100° lOO^-llO" 110°-120° 120=-130° 130=-140° 140°-150° 150°-160° +160°

Litres 10 12 10 8 8 8 5 5 5

The sulphur compounds were precipitated from these fractions by means
of alcoholic mercuric chloride, the oils washed and dried, and the distilla-
tion continued within limits of 5°, 2°, and finally within 1°, under a
constant tension of 730 mm. In further explanation of what was said
above concerning the slow separation of these constituents, the following
record of the tenth distillation is given ; when it is borne in mind that

* Jour. Chem. Soc, XV. 419. t Bull. Soc. Chem., XLI. 161.

MABERY AND HUDSON. — AMERICAN PETROLEUM. 113

these weights are from G5 litres, the small amounts of the constituents
will be appreciated : —

llo>^-116^ llG-"'-H7^ 1170-118° 118''-119^ 119M20^ 120''-121^ 12^-122°

Grams 50 60 80 100 145 130 IGO

122''-123^ 123"'-124'^ 124-^-125=' 125°-126=' 126°-127° 127°-128=' 128°-129°

Grams 1-20 170 160 150 120 90 40

If the distillation had been stopped at this point, since on account of the
diminished tension the true boiliug points should be at least one degree
higher than those given, it might be inferred that one of the principal
constituents was a hydrocarbon boiling at 126°-127^.

No doubt this inference would be supported by the vapor density of
this product, but further distillation showed its fallacy. At the end of the
thirty-third distillation, 190 grams collected at 118°. 5-119°. 5, which
distilled between 119°. 5 and 120°, for the most part at 119°. 5, under
760 mm. with the mercury column wholly in the vapor. Its specific
gravity at 20° was found to be 0.7243. This liquid was assumed to
have the same composition as the octane whose identity was shown
by analysis and a vapor density determination.*

The vapor density of this product was found to support the same
composition : —

0.1287 gram of the oil gave 75 c.c. of vapor at 182°, and under a ten-
sion of 429 mm.

Calculated for CjHu. Found.

3.95 391

The specific gravity of the crude distillate was found to be 0.7256.
The presence of this hydrocarbon is still further assured by the |)ro-
longed distillation. Our boiling point is practically the same as that
of Warren, although his product from Pennsylvania petroleum must
have been contaminated to some extent by hexahydroisoxylol, which
was not shown to be present in Pennsylvania oil until long after
the work of Warren was completed. Our boiling point was taken
in a portion of the Ohio oil which had been subjected to prolonged
treatment for the removal of hexahydroisoxylol. The specific gravity
of this distillate purified by fuming sulphuric acid with the aid of heat
was 0.7230. Another portion purified with a mixture of nitric and
sulphuric acids gave as its specific gravity 0.7190. In the oil with

* Mabery, loc. cit.

VOL. XXXII. — 8

114 PROCEEDINGS OF THE AMERICAN ACADEMY.

the same boiling point separated from coal tar, Schorlemmer found
the specific gravity 0.7190 at 17°.5. In our product purified with
fuming sulphuric acid, the required percentages of carbon and hydrogen
were obtained.

0.1475 gram of the oil gave 0.4562 gram COg, and 0.2073 gram HgO.

Required for Cjllig.

Found.

c

84.22

84.35

H

15.80

15.62

In the formation of chlorine derivatives from this octane, the same
method was employed without cooling as in the case of ihe more
volatile distillates. It was ascertained tliat the best yield of mono-
chloroctane was given when the quantity of chlorine absorbed was 50
per cent in excess of the amount theoretically required to form the
monochlor derivative. Even with this excess, still a small amount of
the hydrocarbon remained unchanged. Since it was found that the
chlorinated compound could not be distilled under atmospheric pres-
sure without serious decomposition, after washing and drying it was
fractioned in vacuo under a tension of 50 mm., within limits of 10°,
5°, 2°, and finally for some time within 1°. Under 50 mm. fractions
collected at all points between 65° and 150°, but a larger quantity at
83°-84°, which under atmospheric pressure distilled at 164°-J66°.
The composition of this substance was determined by analysis : —

I. 0.1 479 gram of the oil gave 0.3523 gram COj, and 0.1537 gram HoO.
II. 0.1961 gram of the oil gave 0.1888 gram AgCl.
III. 0.2483 gram of the oil gave 0.2408 gram AgCl.

Calculated for

Found,

CsH^Cl.

I.

II.

c

64.65

64,95

H

11.45

11.54

CI

23.90

23.8

III.

23.97

A determination of vapor density gave a value required for mono-
chloroctane : —

0.1682 gram of the oil gave 78 c.c. of vapor at 182°, and under a
tension of 400 mm.

Required for CgHijCl. Found.

5.14 5.28

MABERY AND HUDSON. — AMERICAN PETROLEUM. 115

The monochloroctiine obtained by Pelouze and Cahours,* and by
Scliorlemmer,t from petroleum octane, boiling point 1G8°-172°, was evi-
dently a mixture of the two chloroctanes from the hydrocarbons 119°. 5
and 1*24°. It could not be otherwise, on account of the imperfect sepa-
ration of the hydrocarbons. In our experience, nothing less than thirty
distillations is sufficient for the separation of these bodies with any
degree of purity.

In comparing the results in this paper with those of others, it should
be borne in mind that our products were separated from Trenton lime-
stone petroleum, which was unknown at the time of the earlier study of
Pennsylvania oil. But there can be little doubt that these portions of
Pennsylvania and Ohio oils are identical so far as the principal constitu-
ents are concerned. That this is true of constituents with boiling points
above 150^ will be shown in another paper.

In many instances, even after the most careful purification, the specific
gravity of the petroleum hydrocarbons is somewhat higher than that of
the same hydrocarbons synthetically prepared. Schorlemmer | thought
this was due to fine differences of isomerism. But this was before the
discovery of uaphtenes in petroleum. It now seems very probable that
the higlier specific gravity is due to the difficulty in removing the last
trace of these bodies, especially since a small excess of carbon and a
slight deficiency in hydrogen for the formula C„H2„ + 2 accompanies the
higher specific gravity. A notable quality of the naphtenes is their
inertness toward reagents, which is doubtless greatly increased by large
dilution in the principal petroleum hydrocarbons.

In Russian petroleum, Markownikoif and Putochin § discovered isocto-
naphtene, boiling point 122°. 5. In looking for this hydrocarbon in Ohio
petroleum the fractions 120°-124° were carefully distilled many times,
until so little remained within these limits no individual constituent
could be present in any appreciable quantity, or in such quantity that it
could be collected by fractional distillation and identified. The octane
found by Lemoine at 121° is, therefore, also excluded.

The octane, boiling point 125°. 46, separated by Schorlemmer, was
assumed to be identical with normal octane formed from normal butyl
iodide by the action of sodium. "Warren found a somewhat hisfher
boiling point in the octane from Pennsylvania petroleum. This body

* Jahr. 1863, p. 528.

t Ann. Chem. Pliarm., CXXV. 112.

J Pliil. Trans., CLXXI. 451 (1880).

§ Ber. der deutsch. chem. Gesellsch., 1885, p. 1860.

116 PROCEEDINGS OF THE AMERICAN ACADEMY.

distilled between 126°. 8 and 129°. 1, or in the mean at 127°. 6 ; its com-
position was based upon a determination of its vapor density, but it was
not supported by analysis. The observations of Warren were appar-
ently confirmed by similar distillates separated from Ohio and Canadian
petroleum.* Distillates collected at 126°-127° from these oils after the
eleventh fraction gave values in vapor density determinations correspond-
ing to that of octane. But evidently such determinations, unsupported
by other data, are less reliable, especially in products not fur removed in
boiling points from other isomers. Although our former results appar-
ently confirmed the presence of an octane at 12G°-r27°, those values
were accepted as provisional, to be supported or modified by more pro-
longed distillations which have now been made.

Under a constant tension of 730 mm., forty-two distillations were
made between 121° and 130° through Hempel bead columns. Of the
last distillates scarcely any remanied at 126°-127°, or between this
point and 130°, and very little at 125°-126°. The greater portion,
200 grams, collected at 124°-125°, normal conditions. There is, there-
fore, in Ohio petroleum, no octane with a boiling point higher than 125°.
After purification with a mixture of nitric and sulphuric acids and
sodium, the distillate 124°-125° was shown by analysis to have the
composition for octane.

0.1471 gram of the oil gave 0.4544 gram CO2, and 0.2077 gram H5O.

Calculated for CgH,,. Found.

C 84.22 84.26

H 15.79 15.70

After thorough treatment with fuming sulphuric acid, this octane gave
0.7183 as its specific gravity. Another portion, carefully purified with
a mixture of nitric and sulphuric acids gave 0.7134. The specific
crravity of the synthetic hydrocarbon was given by Thorpe as 0.7188.
The crude distillate, with no purification, gave as its specific gravity
0.7243. Its vapor density was found by the method of Hofmaun.

0.1578 gram of the oil gave 84.4 c.c. of vapor at 182°, and under a
tension of 462 mm.

Calculated for CgHis. Found.

3.95 3.96

In the formation of monochloroctane from this distillate, the hydro-
carbon was exposed to the action of chlorine in the cold until the

* Mabery, Proc. Amer. Acad., XXXI. 32, 57.

MABERY AND HUDSON. — AMERICAN PETROLEUM. Ill

increase in weight was fifty per cent in excess of the quantity tlieoret-
ically requiiocl to i'oini nioiiocliloroctane. The protUict, containing a
small amount of tiie iiychocarbon still uni-luuiged, was wa:^hed, dried,
and submitted to iVactional distilhitiou in vacuo under oO mm., since it
was found that it couUl not be distilled under atmospheric pressure with-
out decomposition. A considerable portion collected at 89°-91°, that
distilled under atmospheric pressure, normal conditions, at 173°-174°.
Schorlemmer stated that ohloriae converts normal octane into a mixture
of primary octyl chloride, boiling point 171)°-180°, and secondary octyl
chloriile boiling at 17J^. None of our product collected at the point
corresponding to the normal chloride, although the quantity was not
sufficient to determine the boiling point with absolute precision, and
not sufficient to form oiher compounds. Pelouze and Cahours * gave
168°-172° as the boiling point of the chloride which they formed from
petroleum octane, but their hydrocarbon was evidently not fully purified.
This substance gave upon analysis percentages of carbon, hydrogen,
and chlorine required for chloroctane : —

I. 0.1222 gram of the oil gave 0.2880 gram CO2, and 0.1264 gram
H2O.
II. 0.1821 gram of the oil "ave 0.1792 gram AgCl.
III. 0.2203 gram of the oil gave 0.2200 gram AgCl.

Calculated for

Found.

CgH^Cl.

I.

II.

c

64.65

64.27

H

11.45

11.49

CI

23.90

24.3^

III.

24.69

A determination of vapor density gave a value required for chlor-
octane : —

0.1675 of the oil gave 76 c.c. of vapor at 182°, and under a tension
of 418 mm.

Calculated for CeHijCL Found.

5.14 5.16

The small amount of distillates above 91° in vacuo showed the pres-
ence in minute proportions of higher chlorinated products, but it would
require much larger quantities than we could conveniently procure to
ascertain their composition.

* Jahr. Fittig, 1863, p. 528.

118 PROCEEDINGS OF THE AMERICAN ACADEMY.

Since Engler* discovered tetramethylbutane or diisobutyl, boiling
point 108.5°, as one of the products in the distillation of fats under
pressure, with the possibility that this body might be present in Ohio
petroleum, we submitted the portions of this petroleum distilling between
100° and 115° to prolonged distillation within single degree limits. But
after treatment with nitric acid, no distillate collected in this vicinity.
Having in hand a series of fractions near 135°-136°, the boiling point
of hexahydroraesitylene, or mononaphtene, they were carried through a
longr course of distillations, and the small amount remaining within these
limits was examined with the aid of fuming nitric acid and fuming sul-
phuric acid, but no sulphonic acid was formed, and the very small amount
of nitro compound was not sufficient for a melting point.

In view of the fact that the series of petroleum hydrocarbons boiling
approximately at 38°, 68°, and 98°, have received little attention since
they were first discovered by Warren, I shall soon undertake an exam-
ination of these bodies, and also of petroleum nonane to which Warren
assigned the boiling point 151°.

* Ber. der deutsch. chem. Gesellsch., 1889, p. 592.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXIL No. G. — January, 1897.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF THE CASE SCHOOL OF APPLIED SCIENCE.

INVESTIGATIONS ON A3IER1CAN PETROLEUM,
By Charles F. Mabery.

XXVII. — THE CONSTITUENTS OF PENNSYLVANIA,
OHIO, AND CANADIAN PETROLEUM BETWEEN
150° AND 220°.

By Charles F. Mabery.

Aid in the wore described is this Paper was given by the Academy from the C. M. Warren

Fond for Chejiical Research.

XXVII.— THE CONSTITUENTS OF PENNSYLVANIA, OHIO,

AND CANADIAN PETROLEUM BETWEEN

150° AND 220°.

By Charles F. Mabery.

Presented October 14, 1896.

The conflicting statements published long ago, and still uncorrected,
concerning the composition of the portions of Pennsylvania petroleum
distilling above 150°, the absence of any information relating to the con-
stituents of Ohio and Canadian petroleum, and the erroneous assumptions
as to the composition of American petroleum based on the results of
Markovruikoff and his assistants in the Russian oil, taken together, render
a study of these portions of American petroleum of extreme interest. One
cannot fail to be impressed with the marked differences of opinion that
have prevailed as to the composition of the portions of American petro-
leum with higher boiling points. Some authorities, influenced by the
profound investigations of the Russian chemists on the Caucasus oil,
have believed that the constituents of American oil above 150° are
naphtenes, with a reservation as to whether the series C„H2„ announced
by Warren should be accepted as naphtenes. Others, depending on the
results of Pelouze and Cahours, which form the basis of the statements in
chemical literature concerning the constituents of these petroleums, have
held that the series C„H2„+o does not find its last representative in nonane,
boiling point 151°. Similar differences of opinion have been expressed
by practical oil men, those who have witnessed the development of the
petroleum industry from the beginning, concerning the composition of
these oils. Some believe that Canadian and Ohio oils are essentially and
fundamentally different from the Pennsyvania oil. Others hold that the
chief constituents are identical, and that differences observed in refinina:
are due to variations in the proportions of the principal constituents, and
to the presence in some oils of small amounts of other bodies that are not
found in all. I must admit that my earlier impressions on this subject
have not been verified in the light of the results which will be presented.

122 PROCEEDINGS OF THE AMERICAN ACADEMY.

This work was uudertakeu with the expectation that Pennsylvania oil
would prove to consist in its higher portions of a series C„Ho„, and it is
only after a great amount of laborious study, and. a vast accumulation of
analytical data, presenting unquestionable evidence that, for the hydro-
carbons distilling within these limits of temperature, except those col-
lecting at 196° and 216° from Canadian petroleum, this series must be
discarded, that I accept this conclusion. Concerning the composition of
Ohio and Canadian petroleum, in the beginning of this work, I had no
preconceived ideas.

The separation of constituents with higher boiling points presents
greater difficulties than those in the portions distilling below 150°. In
the lower portions there is no danger of decompositions during distillation,
both on account of greater stability of the hydrocarbons and because the
unstable bodies, such as the nitrogen, oxygen, and sulphur compounds, as
well as the more complex hydrocarbons, distil for the most part at higher
temperatures. Then distillation in air has little if any effect on the lower
constituents, while the portions distilling at higher temperatures cannot
be volatilized in the presence of air without more or less decomposition.
As has been shown, distillation in vacuo prevents changes due to inherent
instability and it excludes air, but it increases very much the labor of the
separations. It has been found to be especially serviceable in separating
the hydrocarbons under consideration, since these bodies are contained
for the most part in portions of the first distillate boiling above the
limits of cracking, and are consequently contaminated by the products of
decomposition when distilled in air, and the decomposition products are
difficult to remove. Evidently any experiments on a laboratory scale
may fail to reveal the presence of bodies that are present in proportion-
ately minute quantities. It would be interesting, and from a commercial
point of view doubtless profitable, to establish a more extended investi-
gation involving the manipulation of at least a hundred barrels of crude
oil, continuing the fractional separations of all portions until the individual
hydrocarbons were as perfectly isolated as is done with smaller quantities
in the laboratory. Such an investigation could only be undertaken at
large expense, and a long time would be required to reach desirable
results.

In undertaking this subject, the course to be followed was plain.
There is but one method for the separation of these hydrocarbons, and it
yields satisfactory results only after long and tedious application. Con-
cerning the question as to whether fractional distillation can be relied on
for the separation of hydrocarbons with boiling points not far removed,

MABRRY. — CONSTITUENTS OF PETROLEUM. 123

it can, I think, be stated with confidence that distilhilions many times
repeated under constant conditions may be expected to yield products
wliose composition can be determined, after suitable purification, as accu-
rately as the methods of analysis will permit. But hydrocarbons sepaiated
from petroleum in this manner, before purification, may be contaminated
by other bodies whose boiling points are nearly the same. As an illus-
tration, to ascertain whether Ohio petroleum contains a hydrocarbon
boiling at 162°, a course of fractional separations under 730 mm. was
repeated fifty times, forty-five times within one degree. One portion
after treatment with fuming sulphuric acid had its specific gravity changed
from 0.7717 to 0.7535, but It still distilled within the same limits.
Another portion of the same distillate, after treatment with common
concentrated sulphuric acid, gave a product with the same specific gravity
and the same boiling point as the oil before treatment. The crude distillate
contained a certain amount of mesitylene, boiling point 163°, sufficient, as
will appear later, to affect seriously the percentages of carbon and hydrogen.
In the main, the hydrocarbons described in this paper as collecting be-
tween 160° and 216° have the same boiling points as those described by
Warren, except one which collected at 162°. It is peculiar that these
bodies have nearly the same boiling points as the naphtenes separated by
Markownikoff from the Russian oil. In previous investigations on these
hydrocarbons, except those of Warren, evidently the course of distillation
was not carried far enough to separate with any degree of purity the
individual constituents. In the light of Warren's distillations and those
here presented, it is evident that Pelouze and Cahours could not have
carried their separations far enough to obtain individual products. Con-
cerning the thoroughness of Warren's separations there can be no ques-
tion. But the deficiency in his work on the hydrocarbons now under
consideration was the result of the limited knowledge then prevailing con-
cerning the general composition of petroleum. The aromatic hydrocar-
bons, the oxygen compounds, and the nitrogen compounds, were not then
recognized, or were merely suspected as constituents of petroleum. In
the purification of his distillates, the single method adopted by Warren
consisted in boiling with sodium, as shown by the following statement :
" I must state, however, once for all, that, unless specially mentioned, no
one of the bodies operated upon had received any chemical treatment
except that of boiling with sodium.*' * But that Warren suspected the
presence in his distillates of other bodies is indicated by a quotation from

* Proc. Amer. Acad , XXVII. 66.

124 PROCEEDINGS OF THE AMERICAN ACADEMY.

one of his private papers, already mentioned in another paper : " The
samples analyzed may have contained traces of more highly hydrogeuized
substances, and that it would be worth while to treat with HOSOa, and
HONOo, and remove these."* It is evidently an error to consider the
hydrocarbons C„H2„4_2, especially the lower members, as unstable toward
reagents. It has been our experience, in removing aromatic hydrocar-
bons from the distillate 160°-216°, that the principal constituents are not
affected in the cold by a mixture of nitric and sulphuric acids, nor by
fuming sulphuric acid even by warming. After purification, the boiling
points do not change in a long course of distillations. No doubt the
stability diminishes with a rise in boiling points, but it is sufficient to
permit of purification of the members below 21G°. All analyses in this
paper point to the general composition C„H2„+2, except for the Cana-
dian hydrocarbons 196° and 216°. If continued distillation or treatment
with the means of purification we have adopted produces decomposition,
the products should scarcely consist at least entirely of lower members
of the same series.

In the comparative examination, the results of which will be described
in this paper, two principal objects were kept in view, one of chief im-
portance to determine the series of hydrocarbons which form the main
body of American petroleum, and the other naturally following, to ascer-
tain whether the composition of Pennsylvania, Ohio, and Canadian oils
as regards their principal constituents is the same.

In undertaking a study of the portions of petroleum within the limits
of temperature mentioned above, it was at first determined to prepare all
distillates from the crude oils, and this has been done in part in the Ohio
and Canadian oils. But when it was found that cracking in refinino; did
not begin in any considerable extent below 225°, and that distillates from
the refinery resembled in all respects, at least in their principal con-
stituents, those obtained in vacuum distillation, refinery distillates were
more freely employed, especially from Pennsylvania oil. But most of the
results on Ohio and Canadian oils were obtained in vacuum distillates,
the preparation of which was described in a former paper.f Further assur-
ance .igainst decomposition in the Pennsylvania product was gained by
selecting during the "run" in the refinery distillation just that por-
tion of the distillates that corresponds in gravity to the constituents
desired. Experience showed that heavier portions of the crude distil-
late should be selected than those corresponding to the constituents

* Proc. Amer. Acad.. XXXI. 31. t Ibid., XXXI. 1.

MABERY. — CONSTITUENTS OF PETROLEUM. ' 125

required, since the boiling points invariably decline many degrees in
subsequent distillations.

The refinery distillate employed was taken at 48° -50° Baume, having
a specific gravity 0.7892 at 20°. It was very nearly colorless and gave no
odor of decomposition. Forty-five litres of this distillate was fractioued
within limits of 10°, 5°, 2°, and for a long time within 1°. It gradually
collected in heaps, as described by Warren, but the prolonged distillation
was necessary to separate higher and lower constituents from the mixtures
between.

Concerning the impression that Pennsylvania petroleum has the same
composition within tiiese limits as the Russian oil, which is based in part
on the results of Warren suggesting the series C„H„, and in part on the
erroneous statements of Hofer, that Markownikofl: found the same series
in Pennsylvania that he had reported in Russian oil, it may be stated
once for all that this identity is clearly excluded by the great difference
in specific gravity of the corresponding distillates, without reference to the
ditferences in percentage composition. These differences in composition
are indicated at the outset by the great differences in specific gravity of
crude distillates at 16°, as shown by Markownikoff and Oglobine.*

Baku. American.

150°-200° 0.786 0.757

200°-250° 0.824 0.788

250°-320° 0.861 0.809

The same differences appear between the individual constituents : t —

Baku Naphtenes. B.P. Specific Gravity.

Dekanaphtene CioHoq 160^-162° 0.795 (0°)

Endekanaphtene C^^^^ 180°-18o° 0.8119 (0°)

Dodekanaphtene CioHj^ 196°-197° 0.8055 (14°)

And the series C„H2„ of Warren, purified only by distillation and boil-
ing with sodium : t —

B.P. Specific Gravity at 0°.

Kutylene CioHgo 174°.9 0.7703

Margarylene C11H22 19o°.8 0.7822

Laurylene C12H24 216°.2 0.7905

* Ber. der deutsch. chem. Gesellsch., 1883, p. 1873.

t Ibid., p. 1877.

t Proc. Amer. Acad., XXVII. 15.

126 PROCEEDINGS OF THE AMERICAN ACADEMY.

As will be shown later, the specific gravity assigned by Warren to
these constituents is very materially diminished by the removal of the
aromatic hydrocarbons, and perhaps of other heavier bodies, that can only
be separated by the application of more vigorous means of purification
than Warren applied.

Decane, C10H22, 163°-164°.

When Pelouze and Cahours announced C^qH^^ as a constituent of
Pennsylvania petroleum boiling at 160° ("sensiblement"), the presence
in petroleum of the aromatic hydrocarbons had not been demonstrated,
and since treatment with concentrated sulphuric acid and carbonate of
soda was the sole means of purification, it is evident that their product
must have still contained mesitylene, boiling ])oint 163°, which, accord-
ing to Engler, is contained in petroleum to the extent of 0.2 per cent.
Then the specific gravity, 0.757 at 15°, assigned by them, is somewhat
higher than that of this decane with mesitylene entirely removed. War-
ren did not observe the collection at this point of a distillate in any con-
siderable quantity. But if, as it seems probable, Pelouze and Cahours
investigated, not Pennsylvania, but Canadian petroleum, the close agree-
ment between their specific gravity of decane at 160° with ours in
Canadian petroleum is explained. In attempting to ascertain whether a
hydrocarbon with this boiling point is present in Pennsylvania petroleum
in any considerable amount, the distillates 150°-170° were carried
through a long series of distillations, until several hundred grams col-
lected between 158° and 162° under 730 mm., and finally more than
100 grams between 162° and 163° under 760mm. One portion of this
distillate was dried over sodium for analysis.

0.1453 gram of the oil gave 0.4560 gram CO2, and 0.1865 gram HjO.

Calculated for

CloH22' tJjQH^o.

Found.

C

84.51 85.71

85.58

H

14.49 14.29

14.27

A determination of the specific gravity of this oil at 20° gave 0.7684.
Its vapor density was determined by the Hofmann method in the vapor
of aniline.

I. 0.1327 gram of the oil gave 66.8 c.c. of vapor at 182°, under a
tension of 373.2 ram.
II. 0.1113 gram of the oil gave 59.5 c.c. of vapor at 182°, under a
tension of 354.7 mm.

MABERY.-— CONSTITUENTS OF PETROLEUM. 127

Calculated for Found.

C,oH,,. I. IT.

4.92 5.21 5.16

A portion of the same distillate was treated with fuming sul[)huric acid
with the aid of heat, and occasional agitation. When first added, a
slight rise in temperature was observed, doubtless caused by the forma-
tion of mesitylene sulphonic acid. The acid was diluted, neutralized
witli baric carbonate, and the filtered solution evaporated nearly to dry-
ness. A barium salt separated in clusters of needles, sparingly soluble
in cold, more so in hot water. The quantity of this salt was too small
for analysis. Another portion of the same distillate, with a mixture of
nitric and sulphuric acids in the cold, formed an oily layer above the
acids, which became solid on standing, and after crystallization from hot
alcohol the needles that separated melted at 83°-84° ; melting point of
dinitromesitylene, 86°. When the dinitro product was warmed with
fuming nitric acid, it formed prisms sparingly soluble in alcohol, and
melting at 225° ; melting point of triuitromesitylene, 230°. In the treat-
ment with acids, 16 grams of the crude distillate gave 8 grams of the
purified oil, with a loss of 50 per cent. When treated with fuming sul-
[)huric acid, 16 grams of the crude product gave 9.5 grams pure oil, with
a loss of 40 per cent. Mesitylene formed, therefore, a considerable pro-
portion of the crude distillate. The purified oil had the faint odor
characteristic of the pure petroleum hydrocarbons. After the removal
of the hygroscopic moisture, the oil had no effect on bright metallic
sodium. It gave upon analysis values required for decane : —

0.1644 gram of the oil gave 0.5084 gram COo, and 0.2284 gram HoO.

Calculated for C,oHj2.

Found.

c

84.51

84.34

H

15.49

15.44

The specific gravity of the purified oil at 20° was 0.7479, a value
somewhat lower, as mentioned above, than that found by Pelouze and
Cahours.*

Its vapor density was determined by the Hofmann method : —

0.1221 gram of the oil gave 68.5 c.c. of vapor at 182°, and under a
tension of 352.8 mm.

Calculated for CjoHoj. Found.

4.92 4.91

* Ann. Chim. Phys., (4.), I. 62.

128 PROCEEDINGS OP THE AMERICAN ACADEMY.

Although the composition as shown by analysis seemed to be suffi-
cient to demonstrate the presence of a hydrocarbon at this point, evi-
dently analytical data alone would be insufficient, unless supported by a
constant boiling point. The distillate used in the determinations de-
scribed above was obtained after a long fractional separation, but without
purification except the removal of sulphur compounds, until after the
distillations were made. To prove beyond question the constancy in
boiling point, another portion of the crude distillate after the fifteenth
distillation was thoroughly agitated and warmed with fuming sulphuric
acid, washed, dried, and the distillation continued fourteen times longer.
About 100 grams of the oil collected so that it distilled at 163°-164°,
mostly at 164°, under a tension of 760 mm., and with the mercury
column all in the vapor. There seems therefore to be no reasonable
doubt that Pennsylvania petroleum contains a decane with this boiling
point. That this hydrocarbon is a decane is shown also by a determina-
tion of its molecular weight by the Beckmann method, which gave 142 ;
and the formula C10H22 requires 142.

In forming the chlorine derivative of the hydrocarbon CioHoo, boiling
point 163°-164°, from Pennsylvania petroleum, 45 grams of the purified
hydrocarbon was subjected to the action of chlorine in sunlight with a
sheet of paper interposed, until it increased in weight 17 grams. The
chlorine was delivered rapidly above the oil, and it was absorbed as fast
as it could be added. Hydrochloric acid escaped with effervescence, and
the rapidity of the reaction generated sufficient heat to maintain the tem-
perature in the vicinity of 70°. With a slower stream of chlorine in
other experiments the heat was dissipated without heating the oil, al-
though the chlorine was as readily absorbed. Without further treatment,
the chlorine product was subjected to fractional distillation in vacuo
under tension of 80 mm. After ten distillations, approximately 6 grams
collected at r2.5°-130°. which determinations of chlorine, carbon, and
hydrogen showed to be mouochlor decane : —

I. 0.2181 gram of the oil gave 0.5460 gram COo, and 0.2265 gram

HoO.
II. 0.2024 gram of the oil gave 0.1660 gram AgCl.

Calculated for

Found.

c

c,„n2,ci.
68.00

I.
68.23

H

11.90

11.55

CI

20.11

n.

20.29

MABERY. — CONSTITUENTS OP PETROLEUM. 129

The specific gravity of this chlordecane at 20° was 0.8914. To the
same product, separated by Pelouze and Cahours, no speciHc gravity was
given. The ehlordecanaphteue prepared by MarkownikofF and Oglobiue,*
from decanaphtene boiling at 1 GO^-Kj-i", gave as its specific gravity
0/j;jOO at 0°. It boiled at 205°-200° (Cor.). Chlordecane obtained by
Pelouze and Cahours boiled under atmospheric pressure at 200°-204°.
Our chlordecane could not be distilled under atmospheric pressure with-
out decomposition. As nearly as its boiling point could be determined,
it distilled at 197°-203°.

On account of the small quantity of the chlorine derivatives described in
this paper, the specific gravity determinations may not be strictly accurate.
But the errors are doubtless small, and they do not affect the value of the
determinations in establishing the identity of these derivatives.

In continuing the distillation of the higher chlorinated products from
163°-164° decane. Pennsylvania petroleum, approximately 15 c.c. of an
oil collected after the eighth distillation at 160°-170° that was heavier
than water. Even these high distillates, which could not be heated to
their boiling points without decomposition, could be distilled in vacuo
with little if any decomposition. This product gave on analysis the values
required for dichlordecane : —

I. 0.2293 gram of the oil gave 0.4829 gram CO2, and 0.1884 gram
HoO.
II. 0.2557 gram of the oil gave 0.3420 gram AgCl.

Calculated for

Cj„H„oCU.

I.

c

56.87

57.42

H

9.48

9.13

CI

33.65

Found.

II.

33.09

This dichlordecane gave 1.0187 as its specific gravity at 20°.

Decane, CioH„o, 173°-174°.

Above 163° (730° mm.) the distillates were small in amounts to 168°,
where they began to increase, and large quantities collected within de-
gree limits to 173°, when they again fell off to small volumes. The
fraction 169°-170° was selected for examination, which included analysis
of the crude distillate, analyses after purification with acids, and the

* Ann. Chim. Phys., (6.), II. 453.

VOL. XXXII. 9.

Calculated for

C,dH22- C,oH;q.

c

84.51 85.71

H

15.49 14.29

130 PROCEEDINGS OF THE AMERICAN ACADEMY.

formation of the chlorine derivatives. It is interesting to note that, with-
out purification, uupurified distillates give percentages of carbon and
hydrogen that correspond to the series C„Ho,,, or to numbers between
this series and the series C„H2„+2- The unpurified distillate 169°-170°
(730 mm.) gave the following results : —

0.1521 gram of the oil gave 0.4758 gram COj, and 0.2018 gram U^O.

Found.

85.31
14.72

The specific gravity of this distillate at 20° was found to be 0.7502.
Its vapor density, determined by the Hofmann method, gave the follow-
ing value : —

0.1533 gram of the oil gave 74.6 c.c. of vapor at 182°, under a tension
of 414.6 mm.

Calculated for C]oH2j. Found.

4.92 4.86

Evidently no particular value can be attached to sucli determinations
of vapor density, since, as shown by analysis, the oil is not composed of a
single body. The same is true of boiling points unsupported by other
values. The contaminating body frequently has the same or nearly the
same boiling point as the principal constituent, and after the removal of
the impurity the boiling point is not materially changed. Another por-
tion of the crude distillate was shaken with ordinary strong sulphuric acid
washed with caustic soda water, and dried with sodium.

I. 0.1476 gram of this oil gave 0.4586 gram COo, and 0.1980 gram HoO.
II. 0.1674 gram of this oil gave 0.5190 gram COo, and 0.2263 gram HgO.

C

H

The specific gravity of this oil was 0.7486. Its vapor density was
determined ; 0.1454 gram of the oil gave 73 c.c. of vapor at 182°, under
a tension of 410 mm.

Calculated for CjqHj,. Found.

4.92 4.75

I.

n.

84.72

84.57

14.90

15.03

MABRRY. CONSTITUENTS OF PETROLEUM. 131

Auother portion of the crude distillate was agitated with a mixture of
nitric and sulphuric acids. An oily nitro product separated above the
acids, which became partially crystalline on standing. It was doubtless
a nitro derivative of cymol ; boiling point of metacymol 174°-176°, of
paracyraol 17l°-172°.

The quantity of the nitro product was small, and the proportion of
cyniol was still further shown to be very small by treatment with fuming
sulphuric acid. 185 grams of the distillates 167°-170^, after the eighteenth
distillation, first by agitation in the cold, and then by warming to 125°,
gave a loss in weight of only 15 grams, or about 8 per cent of the weight
taken. A barium salt was formed of the sulphonic acid, but it appeared
as a thick gummy mass, so uninviting that nothing further was done with
it. The purification with fuming sulphuric acid was intended more espe-
cially to show whether the boiling point of the distillates would be mate-
rially affected. But the only difference observed was that they came
together more readily and completely within one degree, 169°-170° under
730 mm., and at 173°-174°, chiefly at 174°, under a tension of 760 mm.,
with the mercury column all in the vapor. It is interesting to observe
how slightly the boiling point was affected by the removal of the contam-
inating bodies. Evidently, the principal influence of the latter was in
preventing the hydrocarbon from collecting closely at its true boiling
point.

The remaining oil was boiled with sodium as lon^ as the metal was
atifected, and until it remained unchanged on standing with the oil. Even
after this treatment was repeated several times, the percentage of carbon
was somewhat too high and hydrogen too low : —

0.1503 gram of the oil gave 0.4664 gram COo, and 0.2050 gram H2O.

Calculated for CioHjj.

Found.

c

84.51

84.65

H

15.49

15.14

After this purification the specific gravity obtained was 0.7475. For
still further certainty as to the purity and composition of this distillate
another portion was warmed during several hours with fuming sulphuric
acid and shaken thoroughly. After washing and standing over sodium,
it was analyzed : —

0.1754 gram of the oil gave 0.5431 gram CO.,, and 0.2442 gram HoO.

Calculated for CioHjj.

Found

c

84.51

84.45

H

15.49

15.47

132 PROCEEDINGS OF THE AMERICAN ACADEMY.

In the latter treatment the acid solution when neutralized with baric
carbonate gave a barium salt in clusters of needles, sparingly soluble in
water, doubtless the barium salt of cymol-sulphonic acid, although the
quantity obtained was too small for analysis.

After this prolonged treatment with the fuming acid, the specific gravity
was found to be 0.7467, substantially the same as in the previous deter-
minations, and practically the same as the specific gravity of the decane
boiling at 1G3°.

The molecular weight of this hydrocarbon, as determined by the Beck-
mann method, using benzol as a solvent, was found to be 144; since the
formula C1QH22 requires 142, this determination leaves no doubt that
this body is decane, and not a higher homologue.

The chlorine derivatives of the decane 173°-174° were formed in the
same manner as those of its isomer. 32 grams of the purified distillate
was allowed to absorb 9 grams of chlorine, and the product was fractioned
in vacuo. After the fourth distillation a small quantity collected at 130°
-140°, which gave percentages of carbon, hydrogen, and chlorine required
for monochlordecane : —

I. 0.2082 gram of the product gave 0.1699 gram AgCl.
II. 0.2003 of the product gave 0.5038 gram CO2, and 0.2150 gram HgO.

Found.
I. II.

68.61
11.93
20.19

The specific gravity of this chlordecane at 20° was 0.8874, a value
somewhat lower than the specific gravity of the chlordecane formed from
163° decane, 0.8914. After the study of its other properties not enough
of this chlordecane remained for a determination of its boiling point under
atmospheric pressure. As the specific gravity of the chlorine derivative of
petroleum decane, Lemoine * gave 0.908. This value must have been
obtained from decane boiling at 162°, since the French chemists have
never recognized a decane in petroleum boiling at 173°. It was probably
obtained in an impure product, since it is very considerably higher than
our determination, which was made, as already shown, in chlordecane from
well purified decane.

In continuing the vacuum distillation of the chlorine products from

* Bull. Soc. Chim., XLI. 165.

Calculated for

CioH2iCl-

c

68.00

H

11.90

CI

20.11

Calculated for

*^10"20^'2'

c

5G.87

H

9.48

CI

33.65

MABERY. — CONSTITUENTS OP PETROLEUM. 133

173° decane, a considerable quantity collected at 170°-171°, 80mm.,
which distilled at 235° 240°, 717 mm. Its composition was shown
by analysis : —

0.2008 gram of the oil gave 0.4121 gram CO., and 0.1641 gram HoO.
0.2540 gram of the oil gave 0.3532 gram AgCl.

Found.
T. II.

55.96

9.08

34.38

This substance is, therefore, a dichlordecaue. The quantity of the
dichlordecane was barely sufficient for a determination of its specific
gravity; it gave at 20°, 1.0126.

The presence of a decane at 174° seems therefore to be established
in Pennsylvania petroleum, confirming the observations of Warren, who
alone of all those who have examined American petroleum found a
body at this point. If Pelouze and Cahours had carried their course
of distillations sufficiently far, they could not have failed to discover
this body, since it forms such a large proportion of the higher boiling
petroleum distillate. Those chemists did, however, collect a distillate
at 180°-182° which gave analytical values and a vapor density very
closely supporting the formula C11H24. For some time I was in doubt
as to whether a distillate might not persist at this point, and it was
only after a prolonged series of separations that it was possible to
separate the distillates 180°-182° entirely into higher and lower limits,
thus supporting the conclusion of Warren, that Pennsylvania petro-
leum contains no hydrocarbon in appreciable quantity boiling in the
vicinity of 180°.

Hendecane, C11H04, 196°.

With increasing boiling points the distillates showed less stability, as
indicated by more color in the residue in a long course of distillations,
althounrh this was not sufficient to interfere with the collection of a
homogeneous body within narrow limits of temperature. Between 189°
and 192° distillates early began to accumulate, and the quantity gradually
increased until 150 c.c. was obtained which distilled entirely at 190°-
191° under 730 mm.

On standing over sodium, a portion of this distillate deposited a red-
dish flocculent precipitate in very appreciable quantity, which indeed was

134 PROCEEDINGS OP THE AMERICAN ACADEMY.

observed in all unpurified distillates. On this account very little impor-
tance has been attached to the composition of these oils as shown by
aualj'sis, except to demonstrate what slight weight can be given to those
results. Drying with sodium is indispensable, on account, as shown in
another connection, of the difficulty in removing water by other means.

Determinations of carbon and hydrogen in the crude distillate 190°-
191° dried over sodium gave the following results : —

0.1521 gram of the oil gave 0.4787 gram COo, and 0.1972 gram HoO.

Calculated for C^H^^.

Found.

c

85.71

85.82

H

14.29

14.41

The specific gravity at 20° was found to be 0.7673. After thorough
agitation with sulphuric acid and wa&liiug with sodic hydrate and water,
the specific gravity was scarcely affected, 0.7662.

A comparison of the composition and specific gravity of the crude
distillate 190°-191°, 730 ram., as shown above, with the results of War-
ren and of Pelouze and Cahours, in connection with what follows after
further purification, reveals the cause of the lack of uniformity in the
earlier work. The specific gravity of Warren's analyzed product, 0.7721
at 15°, which was only purified by boiling with sodium, is substantially
the same as that of Pelouze and Cahours, 0.7780 at 20°, which was puri-
fied with concentrated sulphuric acid and sodic carbonate, but appreciably
higher than the specific gravity of our crude distillate, only dried over
sodium. The percentage composition of these products purified in this
manner may be more clearly understood if the results of analysis are
brought together, as in the following table : —

Warren.
I. II.

Mabery.

Pelouze and Cahours.
I. II.

C 85.60

85.33

85.82

84.79 84.58

H 14.80

14.65

14.41

15.42 15.36

Required for
C

85.72

C]iIIo4.

84.62

CuHoj.

84.70

H

14.29

15.38

15.30

It is difficult to see how results so closely supporting the formula
C„H2«+2 ^s those of Pelouze and Cahours could be obtained in a
product still containing aromatic hydrocarbons, as shown by their method

MABERY. — CONSTITUENTS OP PETROLEUM. 135

of purification aud by the higher specific gravity. In the case of this
particular hydrocarbon, the method of purification is not especially men-
tioned, but their general method, which is mentioned under nearly every
other member of the long series of hydrocarbons which they examined,
consisted in agitating the oil in the cold with concentrated sulphuric acid
and washing with sodic carbonate. Then their method included no
special means of drying the oil for analysis. It is quite impossible to
remove moisture entirely from these hygroscopic oils without the aid
of the most vigorous desiccating agents. Nothing less than standing
continuously over sodium, in my experience, will insure a perfectly
dry condition.

It is evident from the following description of their experiments that
Pelouze and Cahours tried the action of reagents on their oil, although
they seem not to have suspected the presence of an aromatic hjdrocarbon
Q,H2«-6: "Le brome, I'acide azotique fumant, I'acide sulfurique au maxi-
mum de concentration, I'acide azotique de Nordhausen ne I'attaquent pas
a froid. Le melange des acides azotique et sulfurique agit sur le car-
bure lorsqu'on maintient ces corps pendant quelque temps en ebullition.
Prolonge Ton Faction, on voit se former une petite quantite d'un
produi:; solide et cristallisable. II se separe en meme temps une huile
jaunatre un peu plus dense que I'eau. De plus, on demele dans les
vapeurs nitreuses Todeur caracteristique des acides volatils homologues
de I'acide acetique."

Upon agitation with nitric and sulphuric acids, the distillate 190°-191°
deposited a heavy nitro compound.

Our observations concerning the action of a mixture of concentrated
nitric and sulphuric acids in the cold on this distillate do not coincide
with those of Pelouze aud Cahours. Immediately upon vigorous agita-
tion, a heavy nitro compound separated as a heavy oil, forming a crystal-
line product on standing. 20 grams of the oil lost 4.5 grams by this
treatment, equivalent to 22 per cent of its weight. The nitro compound
was very sparingly soluble in hot alcohol, from which it crystallized in
needles, melting at 169°-170°. As will be seen later, nitro compounds
with the same melting point were separated from Ohio and Canadian
petroleum, which do not correspond in melting point to that of either
dinitrodurol, 205°, nor of dinitroisodurol, 156°, although crystallization
was continued until the melting point was pretty constant. It would
have been interesting to obtain sufficient of this substance for more com-
plete examination ; but, with the great amount of work on hand in con-
nection with the principal hydrocarbons, it did not seem best to allow our

136 PROCEEDINGS OP THE AMERICAN ACADEMY.

attention to be diverted from the maiu object iu view. It should not be
understood that the diminution in weight mentioned above on treating
with acids was due entirely to the removal of aromatic hydrocarbons.
In all the petroleum hydrocarbons we have separated, on treating with
nitric and sulphuric acids, a large proportion of the products has remained
in solution, and it has been possible to reduce them only very slowly
by heating with sodium. From the distillates above 150°, heavy dark
brown precipitates have been formed at first, followed by light flocculent
deposits resembling aluminic hydrate. On distilling after treating with
sodium, usually a considerable residue is left. If heating with sodium be
continued until there is no further action, the metal will remain bright in
the oil, and we have looked on this as an indication of purity.

The oil separated from the acid mixture was boiled with sodium until
it produced no further decomposition, shaken with sulphuric acid, and
allowed to stand a long time in contact with sodium. Since the metal
remained unaffected, the oil was assumed to be completely purified ;
nevertheless, analysis showed still remaining a trace of the less hydro-
genized body : —

0.1547 gram of the oil gave 0.4816 gram COo, and 0.2104 gram HoO.

Calculated for CuHji.

Found.

c

84.63

84.87

H

15.38

15.12

A determination of the specific gravity of the purified oil gave 0.7585.

To prove with greater certainty the composition of this hydrocarbon,
the crude distillate obtained as described above was heated on the steam
bath with fuming sulphuric acid, which gave some decomposition, and the
oil remaining washed thoroughly with caustic soda and water. 20 grams
of the crude distillate gave 16.5 grams of the purified oil, with a loss of
17.5 per cent, which is probably an approximate measure of the propor-
tion of aromatic hydrocarbons in this distillate. It is probable that the
loss in weight with the mixture of nitric and sulphuric acids included
some decomposition of the principal hydrocarbon. The oil treated with
fuming sulphuric acid was carried through ten distillations, after which it
collected for the most part within one degree, and 100 grams distilled
at 196°-] 97°, under 760 mm., and with the mercury column all in the
vapor. This product was again warmed with fuming sulphuric acid,
washed, and dried over sodium.

In the latter ex23eriment the acid was only slightly discolored. The

Calculated for

C„H,,.

c

84.62

H

15.38

MABERY. — CONSTITUENTS OF PETROLEUM. 137

oil gave the faint characteristic odor of these hydrocarbons, which is
easily recognized when the impurities are removed. It then gave per-
centages of carbon and hydrogen required for a hydrocarbon of the series

I. 0.1502 gram of the oil gave 0.4665 gram CO2, and 0.2090 gram

HoO.
II. 0.1661 gram of the oil gave 0.5157 gram COo, and 0.2322 gram
H,0.

Found.

I. n.

84.70 84.67

15.46 15.53

The specific gravity of the oil 196°-197°, after purification in this
manner, was found to be 0.7581. The close agreement in specific gravity
of the products after treatment with the mixture of nitric and sulphuric
acids, and with fuming sulphuric acid, as well as the proportions of carbon
and hydrogen, indicate that the oil was quite thoroughly purified.

That the formula of the purified hydrocarbon boiling at 196° is CnH24,
and not the next higher horaologue, receives further support by a deter-
mination of its molecular weight, in which Mr. Hudson obtained, by the
Beckmann method, the value 157, required for CnHo^, 156.

In the preparation of the chlorine derivatives of the hydrocarbon CnH.24,
boiling point 196°, Pennsylvania petroleum, 40 grams of the oil was
allowed to absorb 14 grams of chlorine, and the chlorine product was
fractioned in vacuo. After five distillations, lOc.c. collected at 145°-
150° (80 mm.), that distilled with decomposition at 225°-230°, bar.
747 mm. This product gave on analysis percentages of carbon, hy-
drogen, and chlorine required for C11H.23CI : —

I. 0.2291 gram of the substance gave 0.1713 gram AgCl.
II. 0.1933 gram of the substance gave 0.4914 gram CO2, and 0.2071

Found.
I. II.

09.30
11.91

18.49

The specific gravity of monochlorhendecane at 20° was found to be
0.8721, Pelouze and Cahours found that the hydrocarbon which they

gram

HoO.

Calculated for

^tl"23^^*

c

69.29

H

12.07

a

18.13

138 PROCEEDINGS OP THE AMERICAN ACADEMY.

collected at 196°-200° absorbed chlorine at a gentle heat with the for-
mation of a raonochlor derivative that boiled at 242°-245° with a specific
gravity of 0.9330 at 22°. These values are much in excess of the deter-
minations given above. Our boiling point under atmospheric pressure
was only approximate, since the chlorine product was considerably decom-
posed at those temperatures. In the action of chlorine, the substitution
proceeded with the greatest readiness as soon as the chlorine came in
contact with the hyrlrocarbon. No heat was necessary, although, if the
action proceeded rapidly, much heat was developed.

The composition of the product obtained by Pelouze and Cahours
corresponds to the formula C12H25CI, as shown by the following record
of their analyses : —

Calculated for CjjHjsCl. Found

C 70.41 70.34

H 12.23 12.37

CI 17.36 17.53

In further support of the formula CnHogCl, Mr. Hudson made a
determination of the molecular weight of the chloride by the Beckmann
method, using benzol as a solvent, in which he obtained 191 ; the formula
C11H20CI requires 190.

Evidently a dichlorhendecane was also formed in this chlorination,
since in the vacuum distillation, about 5 c.c. collected at 190° -200°, that
gave a percentage of chlorine two per cent too low for the required value.
But the quantity was too small to purify it sufficiently to give acceptable
results.

DODECANE, C12H26, 21 4° -21 6°.

Above 193°, 730 mm., the fractions were very small to 208°, but
between 208° and 210° much larger quantities collected, for the most
part at 209° to 210°. Without further purification except drying with
sodium, the crude distillate was analyzed : —

0.1392 gram of the oil gave 0.4355 gram CO2, and 0.1845 gram HgO.

Calculated for C^H^n.

Found.

c

85.71

85.33

H

14.29

14.73

A determination of its specific gravity gave 0.7745. After treatment
with concentrated sulphuric acid, the specific gravity was not changed, —
0.7744. Another portion of the crude distillate was shaken with nitric

MABERY. — CONSTITUENTS OF PETROLEUM. 139

and sulphuric acids, and allowed to stand several hours. After separation
from the small quantity of the heavy nitro product and the acid, the oil
was boiled with sodium until there was no further action, and the metal
remained unaffected on standing several weeks. The oily nitro product
deposited a small quantity of crystals on standing, but not enough to
purify for a determination of its melting point.

Purified in this manner, this distillate gave the following results on
analysis : —

I. 0.1547 gram of the oil gave 0.4816 gram CO2, and 0.2104 gram

HoO. ,

II. 0.1474 gram of the oil gave 0.4562 gram COo, and 0.2017 gram
HoO.

Calculated for Found.

C 84.72 84.87 84.41

H 15.28 15.12 15.21

Its specific gravity at 20° was found to be 0.7684.

Another portion of the crude fraction 209°-210°, purified by thorough
agitation and warming with fuming sulphuric acid, twice repeated, gave
the following results on analysis : —

0.1471 gram of the oil gave 0.4537 gram CO2, and 0.2068 gram HgO.

Calculated for CijHog.

Found.

c

84.72

84.12

H

15.28

15.63

The specific gravity of the oil after treatment with the fuming acid
was 0.7729, somewhat higher than the portion purified with nitric and
sulphuric acids. It is quite possible that the purification with fuming sul-
phuric acid, in this instance, was not capable of removing the contaminat-
ing bodies as thoroughly as the other method. The purified hydrocarbon
was found to distil completely at 214°-216°, under a tension of 760 mm.,
and with the mercury column all in the vapor. A determination of
its molecular weight by the Beckmauu method gave Mr. Hudson 173 ;
required for the formula CigHge, 170.

In the formation of the chlorine derivatives from the hydrocarbon
C12H26, Pennsylvania petroleum, 95 grams of the purified hydrocarbon
was exposed to the action of chlorine until 30 grams was absorbed.
Apparently chlorine substitutes as readily in the hydrocarbons of higher;

Calculated for

C]2H25C1.

c

70.42

H

12.23

CI

17.36

140 PROCEEDINGS OP THE AMERICAN ACADEMY.

as in those of lower boiling points. In fractioning under 80 mm. after
six distillations, about 20 c.c. collected at 142°-153°, which distilled at
230°-235° atmospheric pressure. In none of the distillates of chlorine
derivatives was hydrochloric acid detected as a product of decomposition
when distilled in vacuo. Under atmospheric pressure it was impossible
to carry on these distillations without serious decomposition. The com-
position of this fraction corresponded to that of CiaHajCl.

I. 0.1910 gram of the oil gave 0.1385 gram AgCl.
II. 0.1339 gram of the oil gave 0.3434 gram CO.,, and 0.1423 gram
H2O.

• Found.

I. II.

69.93
11.81
17.93

In determining the specific gravity of this chlorine derivative at 20°, it
gave 0.8919. A determination of its molecular weight by the Beckmann
method, with benzol as a solvent, gave Mr. Hudson 200 ; required for
C12H25CI, 204. The chloride formed by Pelouze and Cahours from
their distillate 216°-218° boiled at 2o8°-260°, somewhat higher than
the boiling point of the chloride described in this paper. From the
results of their analyses, they deduced the formula Ci^H^iGi.

The chlorination of this product was not carried far enough fo form
sufficient of the dichlor derivative to separate it by distillation. The
greater part of the oil after chlorination distilled at 212°-214°, the
boiling point of the hydrocarbon.

A description of these experiments, which are intended to establish the
composition of the hydrocarbon distilling at 214°-216°, should not be
concluded without a "statement of the relation they sustain to those of
Warren and of Pelouze and Cahours, the only experimenters who have
hitherto attempted the separation of this constituent from American
petroleum. The two determinations most nearly concerned in this dis-
cussion are specific gravity and percentage composition. In the following
comparison of specific gravity determinations and percentage composition,
it should be borne in mind that Warren purified his distillate only by
boiling with metallic sodium, Pelouze and Cahours by agitation with
concentrated sulphuric acid and washing with sodic carbonate ; with
those resnlts are brouglit together the determinations described in this
paper in the crude distillate, after agitation with concentrated sulphuric
acid, and after more thorough purification with the acid mixture : —

MABERY. — CONSTITUENTS OF

PETROLEUM. 141

Sp.

Warren, 15°.

gr. 0.7804

Pelouze & Cahours, 20 \

0.7960

Crude
Distillate.

0.7745

Mabery, 20°.
Agitation
with II.SO4. Purified.

0.7741 0.7684

c

I. II.
84.66 85.74

I. 11.
84.58 84.51

85.33

T. II.

84.87 84.41

H

14.85 14.66

15.37 15.45

14.73

15.12 15.21

The percentages of carbon and hydrogen required for the formula
C12H26 are, C 84.72, H 15.30; for the formula CisHog, C 84.78,
H 15.22. Evidently the dilFerences in percentage composition are not
sufficient to distinguish by analysis alone which of the two formulas is
the correct one. But the differences between either of these formulas
and the formula C12H24, C 85.71, H 14.29, should be readily shown by
analysis.

Having at hand a portion of the purified distillate 214°-216° that had
been treated with chlorine and the part not chlorinated distilled, I con-
tinued the distillation until the chlorine derivatives were completely
removed, and the hydrocarbon was treated with fuming sulphuric acid
and distilled over sodium. It then gave as its specific gravity 0.7729,
substantially the same as before chlorination. Normal dihexyl prepared
by the action of zinc and hydrochloric acid on normal hexyl iodide boils
at 21 4°. 5 and apparently is identical with a hydrocarbon having the same
composition, obtained by electrolysis of potassium cenanthylate (Schor-
lemmer). The latter has the specific gravity 0.7738 at 17°. It is there-
fore probable that the hydrocarbon CioHoe, boiling at 214°-216°, from
American petroleum, has the same form. The purified hydrocarbon then
gave the following percentages of carbon and hydrogen : —

I. 0.1599 gram of the oil gave 0.4978 gram CO2, and 0.2177 gram
H2O.
II. 0.1 617 gram of the oil gave 0.5021 gram COo, and 0.2270 gram H.O.
III. 0.1559 gram of the oil gave 0.2193 gram H2O. The CO2 was lost.

Calculated for Found.

CjoH,,;. I. II. III.

C 84.72 84.89 84.69

H 15.28 15.10 15.31 15.63

These results indicate the absence of any impurity not readily acted
on by chlorine.

As one of the results of this examination it can, I think, be safely
asserted that the constituents of Pennsylvania petroleum with boiling
points at 163°-164°, 173°-174°, 196°-197°, and at 215°-216°, constitute

142 PROCEEDINGS OP THE AMERICAN ACADEMY.

the main body of this petroleum within these limits, and whatever other
bodies may be present, they are to be found only iu comparatively small
quantities. As to the proportion of aromatic hydrocarbons in the crude
oil, no direct estimation can be based on these observations.

Fuming sulphuric acid removed 40 per cent of the crude distillate 160°-
161°, leaving decane as the remaining GO per cent. Whether the portion
removed was mesitylene alone may be questionable, but no doubt it formed
the larger part of the body uniting with the fuming acid. Even in such
prolonged fractional separations as those described in this paper, it cannot
be assumed that nearly all the decane 163°-164° was collected in the
fractions which should contain it, neither is it probable that the crude
distillate from which was selected the specimen for the separation of these
constituents contained all the decane in the corresponding quantity of
crude oil. But it is evident that the aromatic hydrocarbons are present
in no inconsiderable amounts, and without doubt these bodies have much
influence on the illuminating qualities of the oil. As shown by the slight
changes in specific gravity after the action of ordinary concentrated sul-
phuric acid, the action of this acid in the usual method of refinino- does not
include to any considerable extent the removal of the aromatic hydrocar-
bons. Its beneficial action seems to concern more especially the decom-
position products of distillation and certain constituents present in minute
quantities, such as the unsaturated hydrocarbons, the oxygen, and nitroc^en
compounds. It scarcely need be mentioned that this applies to the action
of sulphuric acid in the cold. When heated, without doubt a part of the
aromatic hydrocarbons would be removed, to the detriment of the burnino-
qualities. The formation of barium salts from all the distillates treated
with fuming sulphuric acid proves the presence in appreciable amounts of
a wide range of aromatic hydrocarbons.

Without reference to the percentages of carbon and hydrogen, the low
specific gravity of the hydrocarbons described above is sufficient to show
that they are not naphtenes. The following comparison of the unpuri-
fied distillates and the purified hydrocarbons as regards their specific
gravity with the naphtenes separated by Markownikoff from the Russian
oil will perhaps make these differences more clearly understood.

Baku naphtenes : —

B. p.

Specific Gravity.

^loHoQ

160°-162°

0.795 (0°)

v^ii rigs

180°-182°

0.8119 (0°)

C12H24

196°-197°

0.8055 (14°)

MABERY.

CONSTITUENTS OF PETEOLEUM.

143

Pennsylvania C„IIo„+2 : —

•r

Specific Gravity at 20°.

B. P.

Purified.

Unpurified,

Cion22

163°-164°

0.7479

0.7684

C10H22

173°-174°

0.7467

0.7502

C11H24

196°-197°

0.7.^81

0.7673

^12 "-26

214°-216°

0.767G

0.7745

Another important result of this study is the evidence that the main
body of the hydrocarbons in Pennsylvania petroleum within these limits
of temperature are members of the series C„H2„+o. Satisfactory analytical
data in proof of this conclusion have cost a vast expenditure of time and
effort, not only in the routine labor of separation and purification, but in
obtaining suflficiently close percentages of carbon and hydrogen. In such
a larfre number of determinations the ordinary method of combustion be-
comes exceedingly tedious, since all details must be vpatched with the
oreatest care. In the beginning of analysis there was not the slightest
evidence as to whether the hydrocarbons were of the series C,jIl2„+2»
or of the series C„Ho„. The extreme hygroscopic nature of these
bodies was not then appreciated, nor the refinement in purification neces-
sary to yield acceptable results. Many analyses were made before these
details were fullv understood.

Constituents of Ohio Trenton Limestone Petroleum.

The constituents of Ohio petroleum between 160° and 216° were sought
for in vacuum distillates obtained from the crude oil, as described in a
former paper.*

Distillation in vacuo was continued until all portions had collected
that could be brought together within the desired limits. Further con-
centration was carried on under atmospheric pressure, since this did not
occasion serious decomposition. Probably no petroleum distillates with
such high boiling points can entirely escape decomposition in a long
course of distillations, but small amounts of decomposition products can
DO doubt be removed by sulphuric acid, leaving the main body of the oil
free from contamination. After the first distillations in vacuo Ohio
distillates seem to suffer no more change during distillation than those
from Pennsylvania oil. Before separation into two degree fractions, the
sulphur compounds were removed so far as possible by precipitation with
mercuric chloride. As the distillation proceeded, the characteristic

* Proc. Amer. Acarl., XXXT. 2-5.

c

Calculated for

85.71 84.51

H

14.29 15.49

144 PROCEEDINGS OF THE AMEEICAN ACADEMY.

"heaps" began to appear in the vicinity of 160°, 730 mm., falling off
below 158° and above 163°. These portions were therefore carried
through a long course of separations, fifty altogether, forty-five within
one degree, after which they collected mainly at 159°-162°. A com-
bustion of the crude distillate 159°-160°, purified as explained above
only by mercuric chloride, gave the following results :

I. 0.1591 gram of the oil gave 0.4983 gram COo, and 0,2076 gram
HoO.

Found.

85.41
14.50

On the basis of these values alone, there should be no hesitation in
assigning to this oil the formula CioHoq. The specific gravity of this dis-
tillate was found to to be 0.7717. Its vapor density was determined by
the Hofmann method : —

0.1272 gram of the oil gave 67 c.c. of vapor at 182°, and under a ten-
sion of 374.4 mm.

Calculated for CjoHj,. Found.

4.92 4.96

After treatment with ordinary concentrated sulphuric acid, another
portion of this distillate gave 0.7678 as its specific gravity, and the fol-
lowing percentages of carbon and hydrogen : —

0.1416 gram of the oil gave 0.4452 gram CO,, and 0,1855 gram HoO.

C 85.76

H 14.56

Agitation of the crude distillate in the cold with a mixture of concen-
trated nitric and suljihuric acids caused the separation of a heavy oil
that deposited crystals of a nitro product on standing. After crystalli-
zation from hot alcohol, in which it is sparingly soluble, this substance
melted at 229°, showing it to be trinitromesitylene, melting point 280°-
232°. Complete removal of the nitro compound from decane required
prolonged boiling with sodium followed by agitation with sulphuric acid.
The hydrocarbon then no longer attacked the metal, nor tarnished it on
long standing. Purified in this manner, this oil gave values required
for decane : —

MABERY. — CONSTITUENTS OP PETUOLEUM. 145

0.1512 gram of the oil gave 0.4677 gram COo, and 0.2088 gram HgO.

Calculated for CioHjj- Found.

C 84.51 84.35

H 15.49 15.35

For further assurance as to the composition of this hydrocarbon, an-
other portion of the crude distillate was first agitated in the cold with
fuming sulphuric acid, which produced no heat, and then heated with
the acid on the steam bath. As in most of the crude oils when treated
in this manner, the acid turned dark when heated, and evidently ex-
tracted a considerable proportion of the oil. After washing with sodic
hydrate and water, the remaining oil gave only the faint odor characteris-
tic of the petroleum hydrocarbons. In many of the analyses, it will be
seen that the percentage of hydrogen is somewhat lower than should be
expected in a pure substance. This deficiency is doubtless due to a very
small proportion of the hydrocarbon with less hydrogen, the last traces
of which it is somewhat difficult to remove. The fraction 160°-162'^,
after the last purification, contained the following percentages of carbon
and hydrogen : —

0.1532 gram of the oil gave 0.4740 gram CO2, and 0.2113 gram H2O.

Calculated for CjoHjj. Found.

C 84.51 84.38

H 15.49 15.33

A determination of its specific gravity at 20° gave 0.7535, a value
slightly higher than the specific gravity of the corresponding fraction of
the Pennsylvania oil, doubtless due to the slight trace of impurity still
remaining. In determining the molecular weight of this decane by the
method of Beckmann, Mr. Hudson obtained 145; required for C10H22,
142.

The boiling point of this hydrocarbon, so thoroughly purified that
neither fuming nitric acid nor fuming sulphuric acid produced further
change, was found to be nearly the same as that of 163° decane from
Pennsylvania petroleum. 50 c.c. distilled entirely between 162°. 5 and
163°. 5 under a barometric pressure of 757.5 mm.

Since even a larger proportion of the crude distillate was removed in
combination with fuming sulphuric acid than from the Pennsylvania
distillate 163°, an attempt was made to ascertain the composition of the
barium salt, formed by neutralizing with baric carbonate and crystalliz-
ing the filtered solution, after boiling with bone black. A considerable

VOL. XXXII. — 10

146 PROCEEDINGS OF THE AMERICAN ACADEMY.

quantity of the barium salt crystallizing in needles was obtained, and the
percentage of barium found, 25.95, agreed closely with the theoretical
percentage of barium in barium mesitylene sulphonate (C9HiiS03)2Ba,
25.61. But the salt proved to be anhydrous, while according to Jacob-
sou,* barium mesitylene sulphonate crystallizes with 9 HgO. On account
of the limited supply of the barium salt, it was not possible to verify this
determination by further study ; but there can be no doubt of the possi-
bility of forming a barium salt of mesitylene sulphonic acid in this man-
ner, since trinitromesitylene melting at 225°-230°, melting point of
trinitro mesitylene, 230°, was formed without difficulty by the action
of fuming nitric acid on this distillate.

As further evidence of the composition of the hydrocarbon 163° Ohio
petroleum, 46 grams well purified with fuming sulphuric acid was ex-
posed to the action of chlorine until the weight had increased 11 grams.
The chlorinated product was fractioned twelve times under 80 mm., when
15 c.c. collected at 130°-135°, distilling at 200°-208° atmospheric pres-
sure. It was shown by analysis to have the composition required for
CioH.2iCl.

I. 0.2340 gram of the substance gave 0.1934 gram AgCI.
II. 0.1935 gram of the substance gave 0.4785 gram CO2, and 0.2119
gram HgO.
III. 0.1715 gram of the oil gave 0.4246 gram COg, and 0. 1963 gram HjO.

Calculated for

C10H21CI.

c

68.00

H

11.90

CI

20.11

n.

m.

67.44

67.53

12.17

12.73

20.45

In a determination of its specific gravity at 20°, this chlordecane gave
0.8958. In continuing the distillation of the chlorine product after the
twelfth fraction, 10 c.c. collected at 160°-170°, specific gravity. 1.0627
at 20°, which gave the percentages of carbon, hydrogen, and chlorine re-
quired for CioHoqCL.

I. 0.2984 gram of the substance gave 0.4105 gram AgCl.
II. 0. 2753 gram of the substance gave 0.5746 gram COo, and 0.2356
gram 11,0.

Calculated for

CioHoqCIj.

c

56.87

H

9.47

CI

33.65

Found.

II.

56.94
9.51

34.03

* Ann. Cliem. Pharm., XLVI. 95.

Calculated for

Cl,oII„j. ClioHji).

c

84.51 85.71

H

15.49 14.29

MABERY. — CONSTITUENTS OF PETROLEUM. 147

Normal Decanl, B. P. 173°-174°.

Above 162° (730 mm.), the distillates were small to 16S°. After long
continued distillation, 300 grams collected at 169°-170° (730 mm.), which,
without further purification, except drying over sodium, gave the follow-
ing percentages of carbon and hydrogen : —

I. 0.1434 gram of the oil gave 0.4486 gram COo, and 0.1927 gram H2O.
II. 0.1577 gram ofthe oil gave 0.4948 gram C02,and 0.2070 gram HoO.

Found.

I. n.

85.33 85.55

14.94 14.58

A determination of the specific gravity of this distillate gave 0.7621, a
value considerably higher than that of the corresponding unpurified Penn-
sylvania distillate, 0.7502.

0.1670 gram of the oil gave 79 c.c. of vapor at 182", under a tension of
366 mm.

Calculated for CiqHjj. Found.

4.90 5.15

A portion of the crude distillate was shaken with concentrated sul-
phuric acid, washed, and dried for analysis ; —

0.1504 gram of the oil gave 0.4708 gram CO2, and 0.1984 gram H2O.

C 85.37

H 14.67

The specific gravity of this oil after treatment with sulphuric acid was
somewhat lower than that of the crude distillate, 0.7580. A determina-
tion of its vapor density gave 5.02. The fraction 169°-170°, Ohio oil,
was the first to be submitted to the action of the mixture of nitric and
sul phuric acids. When a small quantity of the oil was heated to about 1 25°
during 24 hours with the acid mixture, nitrous fumes were freely evolved
and the volume of the oil gradually diminished until very little remained.
Upon diluting the acid, a heavy, tarry mass was precipitated, evidently a
product of decomposition of the hydrocarbon. In another experiment,
25 grams of the crude distillate were agitated in the cold with the acid
mixture, separated from the acid and the oily nitro product, washed, and
dried. After boiling with sodium and shaking with sulphuric acid, there

148 PROCEEDINGS OF THE AMERICAN ACADEMY.

remained of the quantity taken 17 grams, with a loss of 32 per cent. A
portion of the nitro product crystallized on standing, and after several
crystallizations from hot alcohol, in which it was sparingly soluble, the
melting point could not be raised above l64°-16o°. In testing the action
of sulphuric acid alone on the distillate 171°-172°, 25 grams of the crude
oil was agitated with three succeijsive portions of concentrated sulphuric
acid, washed with sodic hydrate and water, and dried. In the first treat-
ment, in which the oil lost in weight 1.8 grams, the acid was badly dis-
colored; the subsequent portions were not affected. The oil was next
heated gently with fuming sulphuric acid with agitation, which caused con-
siderable blackening, washed first with concentrated sulphuric acid, then
with sodic hydrate and water, and dried. After the last treatment, the
oil weighed 19.8 grams, with a loss of 25 percent.

It is evident from these observations that some care is necessary in
attempting to purify these hydrocarbons with concentrated nitric acid. In
the cold or under a gentle heat not prolonged, the decomposition is
probably not serious. But at higher temperatures, as shown above, the
action may proceed too far. The purified oil gave satisfactory results on
analysis : —

I. 0.1373 gram of the oil gave 0.4247 gram COg, and 0.1900 gram
H2O.
II. 0.1459 gram of the oil gave 0.4520 gram C02,and 0.2090 gram HoO.

Calculated for

Found.

OioHjj.

I. II.

c

84.51

84.36 84.47

H

15.49

15.38 15.92

Two determinations of the specific gravity at 20° of this oil in different
preparations gave 0.7519 and 0.7513.

In a still more extended purification, some of the crude distillate was
agitated thoroughly with the mixture of nitric and sulphuric acids,
allowed to stand several hours with the acids, and, after washing, heated
for some time with tin and hydrochloric acid. The washed and dried
oil was then warmed with fuming sulphuric acid, washed, dried, and dis-
tilled. It then gave 0.7482 as its specific gravity at 20°, more nearly
approaching the specific gravity of the Pennsylvania decane 173°, 0.7467,
and analysis indicated that the remaining oil had the composition of the
general formula C„H.2„+o : —

0.1449 gram of the oil gave 0.4494 gram CO^, and 0.2051 gram HaO.

MABERY CONSTITUENTS OF PETROLEUM. 149

Calculated for CuHj,. Found.

C 84.51 ' 84.57

H 15.49 15.73

Another quantity of the crude distillate was heated on the steam bath
during several hours with fuming sulphuric acid, occasionally shaking the
acid with the oil, boiled with sodium, distilled, and dried for analysis : —

0.1519 gram of the oil gave 0.4715 gram COo, and 0.2085 gram H._jO.

C 84.67

H 15.26

The specific gravity of this product was the same as before, 0.7514.
In a determination of the molecular weight of the hydrocarbon purified
by the mixture of nitric and sulphuric acids, tin and hydrochloric acid,
fuming sulphuric acid and sodium, Mr. Hudson obtained 142 ; required
for CjqHoo, 142. The fuming sulphuric acid solution separated from the
oil was neutralized with baric carbonate filtered and evaporated nearly to
dryness. On cooling, a barium salt separated in needles, which were
very sparingly soluble in cold water. The air-dried salt was anhydrous,
and it gave, by ignition with sulphuric acid, 25.46 per cent of barium.
Hexahydrocymo^, requires 23.8 jjer cent of barium, and iso-cymol 24. 3.
The first hydrocarbon boils at 171°-172°, m-cymol at 174°-176°, p-cymol
at 172°-173°, and pseudo-cymol at 169°. By fuming sulphuric acid at
least the iso-cymol if present in petroleum should be extracted, and it is
probable that this barium salt was an impure compound of cymolsulphonic
acid. That the decane boiling at 174° is the principal constituent of
Ohio petroleum at this point, and that it contains a considerable propor-
tion of aromatic hydrocarbons, has received ample proof. As to the
precise form of the aromatic hydrocarbons, the quantity of material it has
been expedient to manipulate in purifying the decane was not sufficient
to demonstrate. It would evidently be impossible to separate these
bodies from decane in any course of distillation, and it would require a
large quantity of the crude distillate, although not so thoroughly distilled
as the one prepared in this examination.

In still further evidence as to the formula of this hydrocarbon, the
chlorine derivatives were prepared by the action of chlorine on a por-
tion purified with the mixture of nitric and sulphuric acids and sodium.
83 grams of the hydrocarbon was allowed to absorb 30 grams of
chlorine, the theoretical quantity for the mono derivative being 20
grams. As an illustration of the readiness with which substitution

150 PROCEEDINGS OF THE AMERICAN ACADEMY.

takes place, in this experiment the stream of chlorine happened to be
exceptionally vigorous, but it was completely absorbed from the beginning
to the extent of 5 grams during the first ten minutes. Hydrochloric
acid escaped from the oil with brisk effervescence, although the chlorine
was delivered at some distance above the surface of the oil. This rapid
substitution at low tem[)eratures is quite unlike the substitution of
chlorine in dekanaphtene observed by Markownikoff and Oglobine, as
shown by ihe following statement:* " Lorsqu'on fait agir du chlore sec
sur les vapeurs du decanaplitene en ebullition et sous I'influence de
I'insolation directe, la reaction se fait lentement avec decragement d'acide
chlorhydrique et demande un tres grand exces de chlore en comparaison
de ce qu'il en faut d'apres le calcul theorique." Yet in the residue of
the hydrocarbon not acted on by chlorine, Markownikoff and Oglobine
found a somewhat lower percentage of carbon and a higher percentage
of hydrogen, from which, together with a slightly lower specific gravity,
0.792, than decanaphtene, 0.795, they infer that " ces chiffres semblent
indiquer la presence d'une quantite notable d'un hjdrocarbure sature."
In accordance with my observations on all the hydrocarbons from Penn-
sylvania, Ohio, and Canadian oil which I have chlorinated, these bodies
should be saturated long before the naphtenes, especially if the latter
substitute chlorine slowly at a boiling temperature.

In fractiouing in vacuo under 80 mm. the chlorinated oil, after several
distillations about 8 cc. collected at 134°-136°, with a small propor-
tion of unaffected hydrocarbon which came over at a lower temperature.
Analyses of this product gave numbers corresponding to the composition
of chlordecane : —

I. 0.2773 gram of the oil gave 0.6977 gram COo, and 0.2886 gram

HoO.
II. 0.1623 gram of the oil gave 0.4085 gram CO2, and 0.1679 gram
HoO.
III. 0.2117 gram of the oil gave 0.1707 gram AgCl.

III.

c

Calculated for
C,oH,,CI.

68.00

I.
68.60

Found.
II.

68.64

H

11.90

11.56

11.50

CI

20.11

19.93

A determination of the specific gravity of this chlordecane at 20° gave
0.8895. Under atmospheric pressure, it distilled with some decomposi-

* Ann. Chim. Phys., (6.), II. 453.

MABERY, — CONSTITUENTS OP PETROLEUM.

151

tion at 205°-2l0°. Its molecular weight, determined by the Beckmann
method, was found by Mr. Hudson to be 174 ; required for the formula
CioHoiCl, 176.5.

In continuing the fractional distillation in vacuo of the chlorine pro-
duct from 173° Ohio decane, a small quantity, 3 c.c, collected at 170°-
180°, 80 mm., which distilled under atmospheric pressure at 240°-243°.
On account of decomposition, this portion was fractioned only five
times, but it gave on analysis a percentage of chlorine corresponding
to dichlordecaue : —

I. 0.2690 gram of the oil gave 0.5549 gram CO2, and 0.2191 gram
H,0.
II. 0.2753 gram of the oil gave 0.5746 gram CO,, and 0.2111 gram

H,0.
III. 0.3147 gram of the oil gave 0.4256 gram AgCl.

ra.

33.46

c

Calculated for

56.87

I.
56L.24

Found.

n.

56.94

H

9.48

9.04

8.51

CI

33.65

A determination of the specific gravity of this dichlordecane at 20°
gave 1.0300, which may not be strictly correct on account of the small
quantity of the material.

In determining the boiling point of this decane, 50 c.c. of the oil,
purified as thoroughly as possible, distilled at 173°. 6-1 74°. 6, mostly
below 174°, under a tension oi 760 mm. and with the mercury column
all within the vapor.

The boiling point of normal decane was given by Krafft * as 173° at
760 mm., but 0.7304, its specific gravity at 20°, is sensibly lower than
it has been possible to reduce the decane from petroleum. Whether
this be due to the presence of a small amount of decanaphtene, which
appears to be nearly inert toward the methods of purification that
have been applied to these petroleum hydrocarbons, can only be deter-
mined in more extended experiments. In view of the ease in chlorina-
tiou of the hydrocarbons CJ^2n+2i ^^^ t,he difficulty in chlorinating the
naphtenes according to the experience of Markownikoff and Oglobine,
perhaps fractional chlorination should permit of the removal of decane,
and the naphtene would reveal its presence by analysis. As already

* Ber. der deutsch. chem. Gesellsch., XV. 1695.

152 PROCEEDINGS OF THE AMERICAN ACADEMY.

shown in the case of Pennsylvania dodecaue, page 141, the portion
remaining after incomplete chloriuation is the purified hydrocarbon

Hendecane, CuH24, 195°~196°.

Above 173° (760 mm.), fractions collected irregularly and in small
amounts to 188°. There was no indication of an accumulation in the
vicinity of 180°, but after precipitation of the sulphur compounds with
HgClj considerable quantities collected at 188°-192° (730 mm.), which
for the most part came together at 189°-19r. An analysis of the crude
distillate gave the following percentages of carbon and hydrogen : —

0.1444 gram of the oil gave 0.2536 gram C0„, and 0.1888 gram H2O.

Required for CnHjn

C 85.71 85.67

H 14.29 14.53

A specific gravity determination at 20° gave 0.7789, and a determina-
tion of its vapor density by the Hofmann method the following value : —

0.1286 gram of the oil gave 67 c.c. of vapor at 182°, and under
357.5 mm.

Calculated for Cj,H22- Found.

5.33 5.26

As in the corresponding crude Pennsylvania distillate, the percentage
composition of this hendecane based on the above analysis supports the
formula C„H2„, and it is nearly the same as the specific gravity found by
Warren, which gave the formula C„U^„ ; it is even closer to the value
0.7780, found by Pelouze and Cahours in their distillate from Pennsylva-
nia, or rather from American petroleum. Allusion is made in their papers
to distillates prepared from Canadian petroleum, as well as from Pennsyl-
vania petroleum, and their high specific gravity, much higher than Penn-
sylvania distillates yield, would seem to indicate that they had in hand
Canadian distillates. Another portion of the crude distillate 189°-190°
was agitated with a mixture of nitric and sulphuric acids, and allowed to
stand some time after the lieat of the first reaction had moderated. A
heavy nitro compound separated, which crystallized on standing. The
reaction was far more energetic than in the same distillate from Pennsyl-
vania petrolevmi, probably on account of a lartrer proportion of aromatic
hydrocarbons. A heavy nitro compound immediately separated as an oil,
which crystallized on standing. It was purified by crystallization from

MABERY. — CONSTITUENTS OP PETROLEUM. 153

alcohol, in which it is very sparingly soluble, and gave as its melting
point 158°-1G0'', which was not clianged by several crystallizations.
There are several aromatic hydrocarbons whose boiling points are in this
vicinity, but only one forms a nitro derivative with this melting point,
isodurol, whose triuitro derivative melts at 156°. It would be inter-
esting to examine these nitro products more fully, but for want of time
and material nothing further was done with them.

The specific gravity of the oil was very sensibly diminished by the
action of the acid mixture. The first determination gave 0.7688, which
was not diminished by further treatment, although this value is somewhat
higher than the specific gravity of the Pennsylvania distillate, 0.7581. In
determining the boiling point of this product, 50 c.c. distilled entirely at
195°-196°, mostly at 196°, under a tension o£ 760 mm. After purifi-
cation of a portion of the crude distillate with fuming sulphuric acid,
different preparations of the washed and dried oil gave in the hands of
different analysts the following percentages of carbon and hydrogen : —

I. 0.1508 gram of the oil gave 0.4683 gram COj, and 0.2055 gram

HoO.
II. 0.1520 gram of the oil gave 0.4715 gram CO^, and 0.2109 gram
HoO.
III. 0.1523 gram of the oil gave 0.4733 gram CO2, and 0.2180 gram
H2O.

Calculated for

Found.

CjiHoj.

I.

II.

in.

c

84.63

84.70

84.61

84.75

H

15.38

15.15

15.42

15.90

The specific gravity of the purified oil was found to be 0.7737 at 20°,
and the vapor density as follows : —

0.1421 gram of the oil gave 68.6 c.c. of vapor at 182° under 373.8 mm.

Calculated for C11H22. Found.

5.33 5.43

A determination of its molecular weight by Mr. Hudson gave 158;
required for the formula CiiH24, 156.

In attempting to form the chlorine derivatives of this hydrocarbon, 16
grams of the fraction 196°, Ohio petroleum, the small amount that
remained for this experiment, was treated with chlorine until it gained in
weight 4.5 grams. After the chlorine product was fractioned four times

Calculated for

CiiHjsCl.

c

69.29

H

12.07

CI

18.63

154 PROCEEDINGS OP THE AMERICAN ACADEMY.

it collected in very small quantity at lo0°-15o'^. Analysis gave results
corresponding to the formula CutLsCl.

I. 0.2042 gram of the substance gave 0.1548 gram AgCl.
11. 0.1938 gram of the oil gave 0.4765 gram 00.7, and 0.2071 gram HoO.

Found.
I. II.

69.74
11.88
18.71

Evidently the chlorination had proceeded so far that the monochlor-
liendecane could not be separated completely from the dichlor derivative,
as shown by the high boiling point, and the large percentage of carbon.

The small amount of the dichlor derivative was not sufficient when
purified to give satisfactory analytical data in support of its composition.

DODECANE, C12H26, 214°,

Above 193° the distillation was continued longer in vacuo. At the
end of the eighteenth, between 120° and 130°, 1200 grams collected in
single degree fractions, for the most part at 122° -124° and 128° -130°.

The latter distilled under atmospheric pressure at 212°-214°, and was
therefore selected for the study of the hydrocarbon which according to
the corresponding distillate in Pennsylvania oil should contain Ci^tLe-
The crude distillate gave the following percentages of carbon and
hydrogen : —

Calculated for CijHji. Found

C 85.51 85.76

H 14.49 14.55

The specific gravity of this distillate at 20° was 0.7877. 25 grams
of the crude distillate was heated with fuming sulphuric acid to 50°, and
kept warm on the steam bath during several hours. Very little sul|)hur-
ous acid was set free, although the acid became thick and dark. The loss
in weight of the oil was 4.5 grams, or 18 per cent. The remaining oil was
washed with sodic hydrate and water, and dried over sodium for analysis.

0.1500 gram of the oil gave 0.4650 gram COj, and 0.2088 gram 11./).

Calculated for Ci^Hjo.

Found.

c

84.70

84.54

H

15.30

15.47

MABERY. — CONSTITUENTS OF PETROLEUM. 155

A determination of its specitic gravity at 20° gave 0.7867, with scarcely
any cliange by the action of the fuming acid. Another portion of the
crude distillate was shaken with a mixture of nitric and sulphuric acids.
The great heat of the reaction was controlled by cooling, and the heavy
nitro product allowed to separate by standing. The principal reaction
was soon over, the nitro cora[)ound separating in needles on standing.

In hot alcohol the nitro derivative was readily soluble, but quite in-
soluble in cold alcohol. It melted at 165°, and was probably a deriva-
tive, or perhaps a mixture of derivatives, of the aromatic hydrocarbons
with boiling points in this vicinity, such as isohexylbenzol, or isoamyl-
benzol, boiling points 212°-214°, and quite probably contained in petro-
leum. After treatment with the acid mixture, the oil was boiled with
tin and hydrochloric acid, washed, and boiled with sodium, then distilled.
Its boiling point was not appreciably changed, but the specific gravity
was reduced to 0.7727, a value practically the same as that of the Penn-
sylvania dodecane. A determination of its molecular weight by the
Beckmann method gave 172; required for CioHoe, 170.

In forming the chlorine derivates of this hydrocarbon, 35 grams ab-
sorbed 8.5 grams of chlorine, and the product was fractioned under
80 mm. After the tenth distillation, about 15 c.c. collected at 150°-160°.
In this instance, as in one or two others, the distillate taken for chlorina-
tion was not purified, and in the distillation the vapors affected the eyes
seriously, doubtless on account of substitution in the side chain in the
aromatic hydrocarbons contained in the crude distillate. Either on
account of decomposition, or want of sufficient material for complete
separation, it was not possible to purify this product sufficiently to give
satisfactoi'Y numbers on analysis, although the results show the formation
of a mono- and a dichlor dodecane. The percentage of chlorine in the
distillate 150°-160° came two per cent too low for monochlor dodecane,
and no combustion was made. In a higher distillate collected at 190°-
200° the following numbers were obtained : —

I. 0.2224 gram of the oil gave 0.2607 gram AgCl.
II. 0.1952 gram of the oil gave 0.4411 gram CO.2, and 0.1816 gram
H2O.

Found
I. II.

61.63
10.33
28.99

Calculated for

CijHj^Cl,.

c

60.26

H

10.04

CI

29.71

156 PROCEEDINGS OF THE AMERICAN ACADEMY.

"While, therefore, the formula of dodecane iu Ohio petroleum is uot
suj^ported by such reliable data as those of the lower constituents the
presence of this hydrocarbon is established, I think, beyond question.

The principal constituents of Ohio petroleum have been shown to be
identical with those of Pennsylvania petroleum. The higher specific
gravity and peculiar qualities of the Ohio distillates depend on the
larger proportions of aromatic hydrocarbons, and perhaps of other
heavy constituents.

Constituents of Canadian Petroleum from the Corniferous

Limestone.

In continuing the study of the higher portions of Canadian petroleum,
the vacuum distillates 150°-300°* were carried through fifteen addi-
tional distillations under 50 mm. As the separations proceeded, the
fractions fell 50° or more in boiling points, large portions collecting
below 220°, the point where the separations could be continued without
serious decomposition under atmospheric pressure. In single degree
fractions after the fifth, distillation w^as continued until the operation
had been repeated in all twenty-nine times. As in the other oils, the
distillates collected mainly at 159°-16r, 168°-170°, 188°-19r, and
208°-210° (730 mm.). There is even greater necessity in the Canadian
than in the Ohio oil that the earlier distillations be carried on in vacuo,
on account of the greater quantity o£ sulphur compounds, but more
especially, as will be seen in another paper, on account of the greater
proportion of unsaturated hydrocarbons and the smaller proportions of
the members in question that distill between 160° and 216°. It is, there-
fore, otherwise impossible to obtain a large proportion of these higher
constituents uncontaminated by impurities due to cracking. But the
vacuum distillates are free from the intensely disagreeable odors due
to cracking, which are far more pronounced than any to be obtained from
Ohio oil. Nevertheless, the natural odor of these compounds, however
carefully they have been protected from decomposition before purifica-
tion, is more pungent than those from Ohio oil. Colorless when first
distilled, all these distillates become colored on standing, probably on
account of polymerism of the unsaturated hydrocarbons alluded to above,
and of other unstable bodies resembling the terpenes, which we have
good evidence are contained in petroleum. That this is unquestionably
the cause of the coloration w^e have abundant evidence iu the polymeri-

* Proc Amer. Acad., XXXI. 52.

MABERY. — CONSTITUENTS OF PETROLEUM. 157

zation aud the formation of heavier oils in these unsaturated bodies after
they have been separated and allowed to stand as long as two years. As
fuither evidence that these unstable bodies are the cause of the color,
after tliey have been removed, the purified oils remain colorless perma-
nently. The higher specific gravity observed in the first vacuum distil-
late from Canadian crude petroleum is still retained in the single degree
fractions, however long the fractional distillation is continued. But
although these portions are heavier than the distillates from Pennsyl-
vania and Ohio oils, they are so much lighter than the naphtenes that
the latter bodies cannot be present in any considerable quantity. In
general the ready and abundant formation of nitro derivatives and sul-
phonic acids is sufficient evidence that the greater specific gravity is due,
for the most part at least, to the aromatic constituents. It is quite prob-
able that naphtenes are present in small amounts, and that they are
accountable, as has been suggested, for the extreme difficulty in removing
the last traces of the less hydrogenized constituents by fuming sulphuric
acid, even after vigorous treatment several times repeated. In some
instances, nitric aud sulphuric acids have given nitro products that have
required very long and vigorous boiling with tin and hydrochloric acid
for complete reduction, or an equivalent treatment with sodium. Occa-
sionally the reduction with tin and acid has caused the separation of a
heavy red oil insoluble in the acid, consequently not an amine.

Decane, C10H20, 163°-164°.

The composition of the principal distillates from Canadian petroleum
below 150° was found to be represented by the series C„H2„+2- Above
this point the distillates were small to 158°, but between this point and
162° larger amounts collected and remained persistently at 159°-160°
(730 mm.). After treatment with fuming sulphuric acid, a portion of this
distillate was still further fractioned until 75 grams distilled at 163°-164°
under 760 mm., and with the mercury column all in the vapor. In the
crude distillate before treatment with the acid, carbon and hydrogen were
determined, with no other purification than drying over sodium. No less
efficient means of desiccation removes the water sufficiently for analysis.
But since a reddish flocculent precipitate separates when any of these dis-
tillates stand with sodium, evidently the percentages of carbon and hydro-
gen in the crude distillates cannot be accurately expressed by the results
of combustion. The precipitation is doubtless caused in part by sulphur
compovTuds which are not wholly removed by alcoholic mercuric chloride,
although it is observed when the quantity of sulphur is very small.

158 PROCEEDINGS OF THE AMERICAN ACADEMY.

0.1610 gram of the oil gave 0.5057 gram CO2, and 0.1989 gram H2O.

C 85.67

H 13.72

Before treatment with the acid, a determination of specific gravity at
20° gave 0.7785. After purification with the acid the specific gravity
was diminished to 0.7572, nearly the same value as that found in the cor-
responding fraction of Ohio oil. The penetrating odor of the crude dis-
tillate, resembling that of the aromatic hydrocarbons, disappeared entirely
after treatment with the fuming acid, and the purified oil gave only the
faint characteristic odor of the petroleum hydrocarbons C„H2„+2- The
acid was blackened and gave ofE much sulphurous acid. 30 grams gave
by this treatment 23.4 grams of the purified oil with a loss of 22 per cent.
It is interesting to note the higher specific gravity of the unpurified dis-
tillate 163°-164° in all these oils — Pennsylvania, Canadian, and Ohio
— than that of the higher distillate, 174°. This clearly indicates a larger
percentage of mesitylene and perliaps of decanaphtene. The difficulty
in removing entirely the heavier body would seem to indicate a trace of
the latter.

After the last distillation, the oil was again treated with the fuming
acid and dried over sodium for analysis : —

0.1723 gram of the oil gave 0.5358 gram COo, and 0.2304 gram HoO.

Found.

84.82
14.85

The oil was again subjected to the action of the acid and analyzed: —

0.1561 gram of the oil gave 0.4847 gram CO., and 0.2127 gram HoO.

C 84.67

H 15.14

The slight change in the percentages of carbon and hydrogen evidently
shows the presence of a heavy body that is but slowly affected by the
acid. This impurity cannot be mesitylene, since this hydrocarbon dissolves
readily in the fuming acid. Another portion of the crude distillate was
next vigorously agitated with a mixture of nitric and sulphuric acids,
allowing the temperature to rise spontaneously nearly to 100°. As soon
as the principal action ceased, the mixture cooled and an oily nitro body

Calculated for

CioHoo C10H22

c

85.71 84.51

H

14.29 15.49

MABERY. — CONSTITUENTS OF PETROLEUM. 159

separated, more remaining in solution. After several crystallizations from
alcohol, in which it is quite insoluble, the nitro derivative melted at 225°,
near the melting point of trinitromesitylene. The hydrocarbon oil was
separated from the acid and boiled several hours with tin and hydrochloric
acid, which caused the yellow color to be absorbed by the acid solution,
and a small amount of a heavy oil separated, insoluble in the acid, evi-
dently a product of the vigorous reduction. The crude distillate lost 20
per cent in weight by this purification. The oil remaining after the re-
duction, when shaken with fuming sulphuric acid, washed, and dried, gave
only the characteristic odor of the petroleum hydrocarbons. Upon ana-
lysis it gave the following results : —

0.1455 gram of the oil gave 0.4517 gram CO2, and 0.1980 gram HgO.

Calculated for OioUu^

Found.

c

84.51

84.67

H

15.49

15.12

While these results point clearly to the composition C„Ho„^2 ^or the
principal hydrocarbon with this boiling point, they also suggest, as has
already been mentioned, a constituent of Canadian oil so inert toward re-
agents that it cannot be removed by ordinary means. No doubt the diffi-
cultv is increased by the great dilution of this impurity in the larger body
of the decane. In its behavior toward reagents, dekanaphtene is sug-
gested. Whatever may be the composition of this body, it is doubtless
present in largest quantity in Canadian petroleum, a smaller amount in
Ohio oil, and a still smaller proportion in Pennsylvania oil. But even in
the Canadian oil, the proportion is evidently very small.

The formula of the principal constituent at 1 63° found still further
support in a determination of its molecular weight by the method of
Beckmann, in which Mr. Hudson obtained 142; the formula C10H22
requires 142.

The purified distillate 160°-161°, Canadian petroleum, behaved toward
chlorine precisely like the corresponding distillates from Ohio and Penn-
sylvania oils. 17 grams absorbed 4 grams chlorine, and was then frac-
tioned in vacuo. About 1 c.c. was obtained after five distillations at 200°-
204° atmospheric pressure, without decomposition, and it gave by analysis
the composition required for C10H21CI : —

I. 0.1922 gram of the substance gave 0.1536 gram AgCL
II. 0.1873 gram of the substance gave 0.4710 gram CO2. and 0.207'
gram HjO.

160 PROCEEDINGS OF THE AMERICAN ACADEMY.

Calculated for

C,oH2,Cl.

c

68.00

H

11.90

CI

20.11

Found.
I. II

68.58
12.30

19.77

The quantity of this monochlor decane was too limited to permit of
further determinations, as were also the higher portions for the separation
of a dichlor derivative sufficiently pure for analysis.

Having at hand a small quantity of the distillate 162°, Berea .Grit
petroleum, it seemed of interest to ascertain whether it would form a
chlorine derivative similar or identical with those of the oils now under
examination. 30 grams of the oil absorbed 8 grams of chlorine in forty-
five minutes, and when the product was fractioned it gave a small quan-
tity at 120°-130° with the composition required for CioHoiCl : —

I. 0.1931 gram of the substance gave 0.1564 gram AgCl.

II. 0.1821 gram of the substance gave 0.4486 gram CO2, and 0.1871

gram HgO.

Calculated for Found

CioH,iCl. I. II.

C 68.00 67.19

H 11.90 11.41

CI 20.11 20.04

The formation of this chlordecane confirms the results published in a
former paper * of this series, showing a hydrocarbon C10H22 in Berea
Grit petroleum with the boiling point 162°.

Decane, 173°-174°.

Above 163° (730 mm.) the absence of single bodies was indicated by
the small amounts of distillates to 168°, where the single degree frac-
tions began to increase in quantity. Between this point and 173° nearly
500 grams collected, the greater portion at 169°-171°, which could not
be brought closer together without further purification. After drying
over sodium, analysis of this distillate gave the following percentages of
carbon and hydrogen : —

0.1640 gram of the oil gave 0.5151 gram CO.,, and 2083 gram HgO.

C 85.66

H 14.11

* Amer. Cliem. Journ., XVIII. 1.

MABERY.

CONSTITUENTS OP PETROLEUM.

161

A determinatiou of the specific gravity of this distillate at 20° gave
0.7770, a value somewhat higher than the specific gravity of the corre-
sponding Ohio distillate. For the removal of the aromatic hydrocarbons
to determine the boiling point of the principal constituent, all the frac-
tions 168°-173° were agitated and warmed with fuming sul[)liuric acid,
washed, dried, and the distillation continued until after a few repetitions
the fractions came together at 1G9°-170'^ (730 mm.), or under 7G0 mm.,
and with the mercury column all in the vapor at 173°-174°. By this
treatment the specific gravity was reduced to 0.7614, and the percent-
ages of carbon and hydrogen changed in a proportionate degree : —

I. 0.1545 gram of the oil gave 0.4810 gram CO2, and 0.2091 gram
H.,0.
II. 0.1446 gram of the oil gave 0.4485 gram CO2, and 0.1946 gram

H,0.
III. 0.1678 gram of the oil gave 0.5204 gram CO., and 0.2246 gram
H2O.

Calculated for

Found.

C10H20 CioHoz.

I.

II.

III.

c

85.71 84.51

84.90

84.59

84.57

H

14.29 15.49

15.04

14.96

14.89

Although these results indicate t'le removal of a large portion of the
aromatic hydrocarbon, and analyses II. and III. were made of different
specimens with the action of the acid continued several hours, it is evi-
dent that the less hydrogenized body was not even then entirely re-
moved. In further confirmation of the presence still of this constituent,
one of the oils treated with the fuming acid was shaken with a mixture
of nitric and sulphuric acids. The solution became warm and an oily
nitro product separated above the acid. Since MarkownikofF preferred
purification of the Russian oil with sulphuric acid to avoid the formation
of objectionable nitro products, it was inferred that this acid should
remove completely the aromatic bodies. But in our experience
with American oils, complete purification cannot be reached with the
fuming acid alone. At first we relied on decomposition of the nitro
product with sodium ; but this required long digestions several times
repeated, and in Canadian distillates the residual hydrocarbon which
contained the nitrogen could not always be entirely removed, as
shown by analysis of a portion of crude distillate purified in this
manner : —

VOL. XXXII. — IJ

162 PROCEEDINGS OF THE AMERICAN ACADEMY.

I. 0.1656 gram of the oil gave 0.5156 gram CO2, and 0.2241 gram
H2O.
II. 0.1736 gram of the oil gave 0.5419 gram CO2, and 0.2345 gram
H2O.

Calculated for Found.

C10H22. 1. II.

C 84.51 84.90 85.13

H 15.49 15.04 15.01

The specific gravity of this distillate was found to be 0.7618, the
same as that purified by fuming sulphuric acid. Another portion of
the same crude distillate vs^as treated with the mixture of concentrated
nitric and sulphuric acids. The great heat developed was controlled by
cooling, and finally the heavy nitro product collected above the acid.
25 grams of the crude distillate lost 9 grams in the formation of the
nitro product, equivalent to 36 per cent. In another experiment 25
grams of the crude distillate treated with fuming sulphuric acid lost 6
grams, or 24 per cent. The oil remaining after the first experiment was
again treated with the mixture of acids, which caused further separation
of the nitro product. For the removal of the nitro compound, the oil
was boiled during several hours with tin and hydrochloric acid. The
reduction was very slow and a red oil separated, leaving the upper layer
colorless. The latter was then agitated with concentrated sulphuric acid,
washed, dried, and analyzed : —

0.1449 gram of the oil gave 0.4494 gram CO2, and 0.2051 gram HjO.

Calculated for CjoHoj. Found.

C 84.51 84.57

H 15.49 15.73

The oily nitro derivative of the aromatic hydrocarbon deposited crys-
tals on standing, which after recrystallization from hot alcohol, in which
it was very sparingly soluble, melted at 169°. The quantity obtained
was insufficient for analysis.

The specific gravity of the oil purified as described above was 0.7601,
and, as shown by the vigorous means of purification employed, it cannot
easily be reduced, although it is materially larger than the specific grav-
ity of the decane 173° separated from Ohio oil, 0.7513, and of the
decane from Pennsylvania oil, 0.7467. But the analyses show that the
principal constituent is a hydrocarbon of the series C„H2„ + 2- The
molecular weight of this hydrocarbon, determined by the Beckmann
method, gave Mr. Hudson 144 ; required by the formula C10H22, 142.

MABERY. — CONSTITUENTS OP PETROLEUM. 163

In forming the chlorine derivative of this fraction, 35 grams was
exposed to the action of chlorine until it had increased in weight 12
grams. After ten distillations in vacuo, 5 c.c. were collected at 135°-
138°, which upon analysis gave values required for C10H21CI: —

I. 0.1950 of the oil gave 0.4835 gram COo, and 0.2143 gram HgO.
II. 0.2083 gram of the oil gave 0.1672 gram AgCl.

Calculated for Found.

CioHjiCI. I. n.

C 68.00 67.69

H 11.90 12.21

CI 20.11 19.86

The crude distillate having been used in the preparation of the chlo-
rine derivatives, the action on the eyes during distillation was severe,
doubtless due to substitution in the side chain of the aromatic hydro-
carbons contained in the unpurified oil.

In distilling the higher fractions, 5 c.c. collected at 170°-180° that
distilled under atmospheric pressure at 205°-210°. It was shown by
analysis to have the composition required for C10H20CI2 : —

I. 0.2405 gram of the substance gave 0.3214 gram AgCl.
II. 0.2057 gram of the substance gave 0.4352 gram COo, and 0.1772

Found.
I. U.

57.69
9.57
33.02

The specific gravity of this dichlordecane was found to be 1.0484. A
determination of its molecular weight by the Beckmann method gave
207; the formula C10H20CI2 requires 211.

Hydrocarbon, C11H22.

Above 173°, the absence of a definite product was shown by the small
amount of the distillates to 188°. Especial attention was given to the
fractions in the vicinity of 180° with reference to the possibility of a
naphtene, since the napbtene Ci2H24 boiling at 180°-18.')° was separated
by Markownikoff from Russian oil, and Pelouze and Cahours found a
hydrocarbon C„H2„ + 2 boiling at 180°-182° in American (Canadian?)
petroleum. But the very small amounts collected within these limits

gram Hfi.

Calculated for

CioHaoClj.

C

56.87

H

9.48

CI

33.65

164 PROCEEDINGS OP THE AMERICAN ACADEMY.

precluded the presence of either of these bodies in auy appreciable
quantities in Canadian petroleum. The portions distilling between 188°
and 200° were gradually brought together at 189°-l9l°. A portion of
the crude distillate was treated with fuming sulphuric acid and still
further fractioned until for the most part it came together at 196°-197°
under 7G0 mm., and with the mercury column all in the vapor. In
drying the crude distillate with sodium for analysis the usual reddish
flocculent precipitate separated, which doubtless changed somewhat the
composition of the oil : —

0.1565 gram of the oil gave 0.4921 gram COo. and 0.2041 gram HjO.

C 85.74

H 14.49

This oil gave 0.7889 as its specific gravity at 20°, and after treatment
with ordinary sulphuric acid 0.7856, the slight difference indicating that
this acid has very little action in the cold on this distillate, although the
acid was considerably blackened, and sulphurous acid was observed. An-
other portion of the crude distillate with fuming sulphuric acid developed
no heat, and when warmed on the steam bath the acid was only slightly
colored. After this treatment the specific gravity was 0.7832, nearly the
same as before. A third portion of the original distillate was agitated
with a mixture of concentrated nitric and suli^huric acids. Very little
heat was developed, and the mixture was then warmed on the steam
bath, A small amount of nitro product separated as an oil, which crys-
tallized on standing. After crystallization from alcohol, the nitro deriva-
tive melted at 150°-154°, near the melting point of dinitroisodurol, 156°.
After heating a long time with sodium, until there was no further action,
the specific gravity was found to be 0.7785. The purified oil was sub-
mitted again to the same treatment, when its specific gravity was 0.7758.
Still another treatment of the same oil with the mixture of acids and
boiling with sodium, also boiling Avith tin and hydrochloric acid, and
agitating with fuming sulphuric acid, reduced the specific gravity only to
0.7729. Analysis I. was made of the oil after the first treatment with
the mixture of acids and sodium, and analysis II. of the oil after the
third treatment : —

I. 0.1562 gram of the oil gave 0.4901 gram CO.., and 0.2051 gram

H.,0.
II. 0.1628 gram of the oil gave 0.5093 gram COg, and 0.2095 gram
H2O.

M.\BERY. — CONSTITUENTS OP PETROLEUM. 165

c

Calculated for

84.70 85.71

Found.
I. II.

85.57 85.33

H

15.30 14.29

14.59 14.30

A determination of the molecular weight by the method of Beckmann
gave 154; the formula CuHoo requires 154.

These results indicate the series C^Hj,!, and that the series C„H2„+2
has ceased to represent the principal compositiou of Canadian [letroleum
at the boiling point 196°. But with reference to the series of hydro-
carbons that are now recognized as constituting the main body of petro-
leum, it is not easy to classify this hydrocarbon. It is certainly not an
unsaturated member of the ethylene series because it lacks additive
power for the halogens, fuming sulphui'ic acid, etc. Its specific gravity
is much less than that of the naphtene that MarkownikofF and Oglobiue
separated at 196°-197° from the Russian oil, 0.8010 at 20°. But it is
interesting to observe that the specific gravity of this hydrocarbon is
practically the same as Pelouze and Cahours found in the hydrocarbon
separated by them at 196°-200° from American petroleum, and which
yielded them analytical values, as has already been explained, page 134,
corresponding closely to the formula CioHoe- Since the specific gravity
of Pelouze and Cahours is so much larger than that of the crude distil-
late from Pennsylvania petroleum, 0.7673, it is difficult to escape
the conviction that their distillates were prepared from Canadian petro-
leum, especially since they allude to an examination of oil from Can-
ada, although the source of the particular oil from which w^ere separated
the individual hydrocarbons which they described is not evident from
their statements.

In studying the constituents of Pennsylvania oil, it has already been
shown that chlorine acts less readily on the principal hydrocarbons than
on other constituents. Since this difference on the action of chlorine
seemed to afford a means of ascertaining whether Canadian distillates
contain any of the series C„H2„ + 2< a small quantity, 16 grams, of the dis-
tillate 189°-190° purified with fuming sulphuric acid remaining after the
examination already described, was exposed to the action of chlorine
until three grams was absorbed, and the product was fractioned in
vacuo until a small portion, six grams, distilled at 189°-190° atmospheric
pressure. This fraction was boiled with sodium to remove, so far as
possible, any chlorine derivative remaining, shaken with sulphuric acid,
and again distilled. There was finally obtained about three grams that
gave percentages of carbon and hydrogen agreeing fairly well for C11H.24.

166 PROCEEDINGS OF THE AMERICAN ACADEMY.

The low values are due to a small amount of clilorine that was not
entirely removed even by the treatment with sodium : —

0.1558 gram of the oil gave 0.4803 gram CO2, and 0.2115 gram HgO.

Calculated for CuHji.

Found.

c

84.70

84.09

H

15.30

15.09

In the formation of chlorine derivatives from the purified fraction 196°,
Canadian petroleum, 25 grams absorbed 7 grams chlorine. After five
distillations in vacuo, 3 c.c. collected at 145°-150°, which distilled at
220°-228° atmospheric pressure, and by analysis gave values required
for CnHaiCl : —

I. 0.2378 gram of the substance gave 0.1747 gram AgCl.
II. 0.2005 gram of the oil gave 0.5105 gram COo, and 0.2097 gram
H2O.

Calculated for Found.

Ci,[l23Cl. I. II.

C 70.03 69.44

H 11.14 11.62

CI 18.83 18.18

A determination of the specific gravity of the raonochloride at 20°
gave 0.8882. With the small quantity of the higher distillates, it was
not possible to separate a dichlor derivative in any degree of purity.

The small amount of the monochloride evidently precluded the possi-
bility of ascertaining with any precision the true boiling point, and the
analysis is chiefly of value in determining the number of carbon atoms
in the molecule. The number of carbon atoms received still further
confirmation in a determination of the molecular weight by the Beckman
method, which gave 187 ; the formula CuHojCl requires 188.5.

Hydrocarbon, C12H24.

Above 195° (730 mm.) the distillates were small in quantity to 208'',
but between 208° and 212° about 300 grams collected, for the most part
at 208°-210°. The specific gravity of the crude distillate was found to
be 0.7947, and analysis gave the following percentages of carbon and
hydrogen : —

0.1554 gram of the oil gave 0.4878 gram CO.., and 0.1982 gram HoO.

Required for CnHon.

Found.

c

85.71

85.57

H

14.29

14.18

MABERY. — CONSTITUENTS OF PETROLEUM. 167

111 order to collect this distillate more closely, a portion of the crude
oil was treated with fuming sulphuric acid, washed, dried, and the distil-
lation continued. 37 grams of the oil gave 27 grams after purification.
After several distillations it collected at 2r2°-214° under 745 mm., and
with the mercury all wiiliin the vapor. Two determinations of specific
gravity at 20° gave (I.) 0.7851, and (II.) 0.7857. The carbon and
hydrogen were also determined : —

0.1586 gram of the oil gave 0.4960 gram COo, and 0.2105 gram HoO.

c

Calculated for
C12H24.

85.71

Found.
I.

85.27

H

14.29

14.74

These values point to the composition C„Ho„ for this constituent of the
Canadian oil. In further evidence as to the correctness of this result,
another portion of the crude distillate was shaken with a mixture of con-
centrated nitric and sulphuric acids, the intense heat at first generated
controlled, and after the principal action had ceased the solution was kept
warm for some time on the steam bath. Au oily nitro product collected
above the acid in considerable quantity, but it was not further examined.
The hvdrocarbon remaining was then agitated with concentrated suli>huric
acid, washed with caustic soda, which removed mucli more of the nitro
product from the oil, then with water, and boiled for some time with tin
and hydrochloric acid.

The oil remaining was then washed, dried, and warmed during several
hours with fuming sulphuric acid, which produced little change, and
boiled with sodium. After this purification, analysis still gave values
required for a hydrocarbon C„H2„.

I. 0.1383 gram of the oil gave 0.4332 gram COo, and 0.1849 gram HoO.
II. 0.1406 gram of the oil gave 0.1856 gram HoO.

c

€1,11,4.

85.71

I.

85.41

II.

Lost.

H

14.29

14.86

14.68

In forming the chlorine derivatives of this hydrocarbon, 26.5 grams of
the fraction 214°-216° was allowed to absorb chlorine until the weight
had increased 9.5 grams, and the product was fractioned eight times in
vacuo; 5 c.c. collected at 160°-1 70°, which gave on analysis the per-
centage composition required for CioH.^sCl.

168 PROCEEDINGS OF THE AMERICAN ACADEMY.

I. 0.2059 gram of the substance gave 0.1515 gram AgCl.
II. 0.2064 gram of the substance gave 0.5357 gram CO.., and 0.2128
gram HoO.

Calculated for Found.

CiaHosCl. I. II.

C 71.06 70.79

H 11.30 11.46

CI 17.53 18.20

The specific gravity of this chlorine derivative determined at 20° was
found to be 0.8960. The small quantity of distillate collected within
higher limits did not permit of the separation of a dichlor derivative in a
pure condition. A chlorine determination gave 27.21 ; required for
Ci2H22Cl2, 29.95. It formed a thick viscous oil that could scarcely be
distilled even in vacuo without decomposition. At a temperature
slightly liigher than where this product was collected the distillate was
largely decomposed.

The series having been determined bv these numbers, the number of
carbon atoms was demonstrated by the molecular weight, which was found
by the Beckmann method to be 172 ; the formula C10H.24 requires 168.

What has been said as to the jn'obability that Pelouze and Cahours
operated on Canadian petroleum receives further support in comparing
their results on the hydrocarbon they separated boiling at 216°-218°
with those described above, although certain important differences appear
between their results and mine.

The specific gravity assigned by them to the hydrocarbon 21 6° -2 18°
was 0.796 at 20°, which is practically the same as the specific gravity of
my crude distillate, 0.7947, given above. But their description of the
properties of this distillate are not in accordance with my observations.
Referring to the action of reagents on their product, they state : " Le
brome, I'acide azotique fumant, I'acide sulfurique fumant, ainsi que le
melange de ces deux acides se comportent a son egard comme avec le
compose precedent." And after stating, in the description of the pre-
ceding compound, that it is not attacked in the cold by bromine, fuming
nitric acid, sulphuric acid at the maximum of concentration with nitric
acid, nor by fuming sulphuric acid, they state : " Le melange des acides
azotique et sulfurique agit sur le earbure lorsqu'on maintient ces corps
pendant quelque temps en ebullition." That this observation does not rep-
resent correctly the behavior of this distillate from Pennsylvania nor Ohio,
nor Canadian petroleum, has been clearly shown by experiments described
in this paper. It is especially inapplicable to the Canadian distillate,

MABERY. — CONSTITUENTS OF PETROLEUM. 169

siuce this oil with a mixture of ordinary concentrated nitric and sulphuric
acids, on shaking, immediately develops sufficient heat to raise the tem-
perature to vigorous ebullition of the acid mixture, and to destroy a large
portion of the hydrocarbon. In all these experiments the initial reaction
had to be controlled by cooling.

Concerning the number of carbon atoms in the hydrocarbon I have
separated at this point, it seems to be well established by the molecular
weight, C12H24) and by the composition of the monochlor derivative,
C12H23CI. Yet in a product with a specific gravity much higher than
this purified hydrocarbon, Pelouze and Cahours obtained numbers by
analysis, as shown above, page 141, closely supporting the formula
C13H26. and from the hydrocarbon a chlorine derivative, C13H27CI, also
supported by analytical values closely corresponding to the theoretical
composition required for this formula.

Summary of Results.

1. Pennsylvania petroleum is composed chiefly between 150° and 220°
of decane, boiling point 163°-164°; decane, boiling point 173°-174°,
probably normal decane ; hendecane, boiling point 196°-197° ; and dode-
cane, boiling point 214°-216°. It contains also in smaller proportions
the series of aromatic hydrocarbons boiling within these limits. Allusion
has already been made to mesitylene, cumol, pseudocumol, cymol, iso-
cymol, durol, isodurol, and no doubt others could be identified with suffi-
cient quantities of the petroleum distillates.

2. The composition of Ohio Trenton Limestone petroleum within the
same limits is represented by the same members of the series C„Fl2„+2»
and the higher specific gravity of these distillates is caused by a larger
proportion of aromatic hydrocarbons.

3. The constituents of Canadian Corniferous limestone petroleum from
Petrolia, within these limits of temperature, are the same at 163° and
173°. But the hydrocarbons collecting at 196° and 214° have the com-
position represented by the series C„H2„. Probably a better knowledge
of these higher distillates will be gained when the true composition of
American petroleum above 220° has been ascertained. The proportion
of aromatic hydrocarbons is greater in Canadian than in Ohio petroleum.
There are indications in all these petroleums that the heavier constitu-
ents include other bodies than the aromatic hydrocarbons, which will
require for their identification the manipulation of large quantities of
distillates.

170 proceedings of the american academy.

General Conclusions.

The results described in this paper make it clear that no conclusions
can be arrived at concerning the composition of the principal constituents
of American petroleum between 151° and 216° without separating from
these bodies the various impurities with which they are contaminated in
the crude distillates. However far fractional distillation may be carried,
it is impossible to effect a separation of those bodies whose boiling points
do not differ by more than a few degrees. But fortunately the principal
constituents of petroleum are not affected by reagents under conditions
which allow the removal of the contaminating bodies. While the state-
ment of Pelouze and Cahours that the portions of American petroleum
under consideration are not affected by nitric acid nor by fuming sul-
phuric acid is not supported by the behavior of distillates used by me in
this examination, it is true that the principal constituents are not acted
upon by those reagents under conditions that permit of purification. A
casual examination of the literature of Pennsylvania petroleum is suffi-
cient to reveal the uncertainty and confusion in statements concerning
the composition of the portions with higher boiling points. The principal
constituents have more commonly been referred to the series C„H2„+2)
as suggested by Pelouze and Cahours, who included in this series all
the petroleum hydrocarbons, even the least volatile oils and paraffine,
although on the basis of Warren's investigations allusions have been
made to the unsaturated define hydrocarbons C^H,,, as constituting the
main body of Pennsylvania petroleum above 150°, and the belief has
been expressed that the naphtene series C^Hj,, should best explain AVar-
ren's results. As already stated, above 150° Pelouze and Cahours
separated distillates at 162°, 182°, 196°-200°, and 216°-218°, which,
with no especial purification, gave analytical data corresponding closely
with the theoretical values for the series C„H2„+2. That there are marked
differences in the specific gravity of crude and refined distillates appears
in the purification of all distillates described in this paper, and it is no
easy task to purify the crude distillates so that they shall yield satisfac-
tory analytical data. As shown in the following table, the specific gravity
determinations by Pelouze and Cahours in the products they analyzed
are essentially different from those in Pennsylvania distillates herein
presented. The percentages of carbon and hydrogen required for the
series C„H2„+2 were obtained in fractions whose specific gravity was even
liigher than our crude distillates, which gave values closely agreeing with
the series C„H2„. This difference in specific gravity can only be explained

MABERY.

CONSTITUENTS OF PETROLEUM.

171

by assuming, which is evidently true, that Pelouze and Cahours over-
looked the aromatic i)ydrocarboiis. It does not appear in their reference
to American petroleum whether they really operated on distillates from
Pennsylvania oil ; but, on the other hand, their specific gravity determi-
nations are not widely different from those of distillates from Canadian
oil to which occasional reference is made in their publications.

Pennsylvania Petroleum.

Fraction.
C. F. M.
Unpurified. Purified.

163°-164°

163°-164°

p. & c.

1«2°

Specific Gravity.
C. F. M. P. & C.

Series.

196°-197°

196°-197°

0.7684 (20°)
0.7479 (20°)

0.7673 (20°)
0.7581 (20°)

196°-200°

214° -216° 0.7745(20°)

214° -216° 0.7684(20°)

216° -218°

0.7780

0.796

Canadian Petroleum.

Fraction
Unpurified.

163°-164°
196°-197°
214° -216°

Fraction
Purified.

163°-164°
196°-197°
214° -216°

C„H2„

0.757(15°) CnR2n+2
C„H2„

C„H2„+2
C„n2„+2

C„H2,j

^v n 2n+2

C. F. M.

0.7785
0.7582
0.7889
0.7729
0.7947
0.7851

If the results of Pelouze and Cahours were really obtained in distil-
lates from Canadian petroleum, as seems probable, especially since
Pennsylvania distillates do not yield such high values, the speoific gravity
determinations in their lower distillates agree fairly well with mine in
the purified Canadian hydrocarbons.

The results of Warren, on the other hand, are consistent in giving num-
bers that account both in specific gravity and in percentage composition
for the complex mixture that crude petroleum distillates are known to be.
"While Warren made no attempts to purify his distillates, they were ob-
tained in a course of fractional separations far exceeding in efficiency
those of other experimenters. It is interesting to observe in the following

172 PROCEEDINGS OF THE AMERICAN ACADEMY.

table how closely the distillates from Pennsylvania petroleum analyzed
by Warren resemble in specific gravity the unpurified distillates described
in this paper.

Fraction. Fraction. Specific Gravity.

Unpurified. Purified. Mabery. Warren Series. ■

163°-164'' 0.7674 (20°) C^H^,,

173°.5 0.7502 (20°) CJi^,,

173°.5 0.7445 (20") C„H2„+,

175° 0.7598 (15°) C„H2„

196° 0.7673 (20°) C^Hj^

196° 0.7581 (20°) C„H2„+2

196°-197° 0.7721 (15°) C„H2„

216° 0.7745 (20°) CJI2,,

216° 0.7684 (20°) C„H2„h.2

214°-216° 0.7804 (15°) C„H2„

Warren found no single body at 162° ; his other hydrocarbons correspond
in boiling pouits vrith those described in this paper. The series C„H2„ of
Wai'ren, with a series of the same numerical composition in Russian
oil, led to the suggestion that the naphtenes might form an essential part
of Pennsylvania oil. This belief was encouraged by statements that
found their way into works on petroleum, that MarkownikofF discovered
the naphtenes also in Pennsylvania oil. The statement that Mar-
kownikoff investigated Pennsylvania petroleum is indeed erroneous,* and
a closer study shows that the naphtenes cannot be contained in Pennsyl-
vania petroleum in any considerable quantity on account of their higher
specific gravity as already shown (page 143):

Fraction.

Fraction.

Specific Gravity.

Series.

Unpurified.

Purified.

Mabery.

Marljownikoff.

163°-164°

0.7684

C„H2„

163°-164°

0.7479

^n^2n-H

160°-162°

0.795 (0°)

C„H2„

180°-185°

0.8119 (0°)

C„H2„

196°-197°

0.7678

C„H2„

196°-197°

0.7581

^n"2n+2

196°

0.8055 (14°)

i^„n. 2i,

214°-216°

0.7745

C„H2„

214°-216°

0.7684

C„H2„+2

* In a private communication, I am informcrl by Professor Markownikoff that
lie has not included Pennsylvania petroleum in liis investigations.

MABERY. — CONSTITUENTS OF PETROLEUM. 178

Even the mixtures that the crude Pennsylvania distillates have proved
to be are much lower in specific gravity than those with the same boiling
points from Russian oil. After purification, the Pennsylvania distillates
unquestionably have the composition of the series C„H2„+2-

Allusion has been made to the great difficulty in removing entirely the
constituent with less hydrogen from the distillates prepared tor analysis,
and also to the fact that, even after the most thorough purification, the
specific gravity of the purified distillate is appreciably higher than that of
the hydrocarbon with the same boiling point prepared by synthetic methods.
These facts may indicate the presence in small quantity of naphtenes
which are acted on only very slowly by reagents, especially when largely
diluted in the main body of the principal constituent.

Structure of the Petrgleuji Hydrocarbons, 160°-216°,

lu comparing the petroleum hydrocarbons with the corresponding bodies
synthetically prepared, it is but fair to state that the literature of the latter
is not altogether satisfactory. The properties of normal decane as it
was prepared by Lachowicz,* by the action of sodium on a mixture of
normal octyl bromide and ethyl iodide, and also by Kraflft,! by heating
caprinic acid with a mixture of hydriodic acid and phosphoric penta-
chloride, seem to define the corresponding petroleum hydrocarbon as
the normal compound.

The boiling point of the hydrocarbon synthetically prepared is 173°
under 760 mm., and its specific gravity 0.74o6 at 0°. The boiling point
assigned by me to petroleum decane is 173°. 5, and its specific gravity at
20° is (from Pennsylvania petroleum) 0.7486. Evidently the decane
obtained by Thorp and Young by heating solid paraffine, boiling at 166°-
168°, specific gravity 0.7394 at 13°. 5, is an impure normal hydrocarbon.

A decane has been obtained by several methods, boiling at various
temperatures between 157° and 162°. Active diamyl boiling at 159°-
162°, specific gravity 0.7463 at 22°, with a high dextro rotatory power,
was obtained by Just, on treating active amyl iodide with sodium. But
the petroleum decane has not the same form, since it shows no influence
on polarized light. The latter has with greater probability the same form
as diisoamyl, obtained by Wurtz on heating isoamyl iodide with sodium.
The boiling point of this secondary decane is given by Wurtz as 158°,
with no mention of barometric tension, and specific gravity as 0,7413.

* Ann. Chem. Pharm., CCXX. 179.

t Ber. der deutsch. chem. Gesellsch., XV. 1695.

174 PROCEEDINGS OF THE AMERICAN ACADEMYo

Petroleum decane boils at 163°, and its specific gravity at 20° is 0.7479.
lu a similar reaction using amjl bromide, Grimshaw * obtained a decane
boiling at 168°, under 751 mm. Either the normal bromide must have
been used in this reaction giving normal decane, or the boiling point of
the product is too high.

Kormal hendecaue (or undecane) was prepared by KrafFt f from the
aldehyde raulenol by the same reaction that he used for decane. Its
boiling point is given as 194°.5 under 760 mm., and the specific gravity as
0.7411 at 20°. Since the boiling point, 196°-197°, and specific gravity,
0.7581, of petroleum hendecane and the formula of this body, C11H24,
determined by its molecular weight, correspond so closely to normal hen-
decane. it has doubtless the same form. As already explained, the higher
specific gravity may be due to a small proportion of a naphtene.

The formula C12H26, which represents the petroleum hydrocarbon
boiling at 214°-216°, is supported by that of normal dodecane which
Krafi't t obtained by the reduction of laurinic acid. Normal dodecane
boils at 214°. 5, atmospheric pressure, and its specific gravity at 20° is
0.7511. The specific gravity of petroleum dodecane was found to be
0.7729.

Relation between Specific Gravity and Chemical Compo-
sition OF Petroleum Distillates.

One interesting result of the examinations described in this paper, and
other papers of this series, is the relation between the chemical composition
of the individual hydrocarbons and the density of the crude oils from
which they were prepared. Pennsylvania petroleum with the lowest spe-
cific gravity, 0.80-0.82, is composed below 220° of the hydrocarbons
C„H2„+2- Ohio oil. next higher in the scale of specific gravity, 0.82-0.85,
still contains below 220°, as its principal constituents, hydrocarbons of
the same series. As the density increases, in the Canadian oil, specific
gravity 0.85-0.88, the series C„Ho„_^.2 represents the principal constitu-
ents up to and including decane, boiling point 173°, the higher members
having the composition C„H2„. But oils with higher specific gravity,
such as a South American petroleum, specific gravity 0.941^0, described in
another paper of this serie-, contain only hydrocarbons of the series C„H2„,
and in Caucasus petroleum, specific gravity 0.88, Markownikoff and Oglo-
bine found as the principal constituents the naphtenes, series C„H2„.

* Ber. der deutsch. chem. Gesellsch., X. 1602. t Loc. cit., 1698.

t Ibid., X. 1697.

mabery. — constituents of petroleum. 175

Action of Sulphuric Acid on Petroleum Distillates.

Since the beginning of the petroleum industry, the sole method of re-
fining has depended on agitation with sulphuric acid ; yet, notwithstanding
the large quantities of material that sulphuric acid removes, as shown by
the immense sludge lieaps iu the vicinity of the refineries, the refiner has
not the slightest notion as to what the acid accomplishes beyond his
principal object, which is to prepare acceptable products for the market.
Neither has any scientific study been made of this problem, at least in
American petroleum, so far as revealed by published statements, but that
this is an interesting as well as a difficult subject appears fi-om results
described in several papers of this series. That ordinary concentrated
sulphuric acid has some effect on the density of petroleum distillates is
evident from numerous experiments described in this paper. But without
independent evidence, it cannot be determined just what constituents are
affected by the acid. That some of these constituents, present in small
quantity, are unstable and easily acted on by reagents is evident. The
refiner must avoid an elevation of temperature during treatment with the
acid, otherwise a color appears in the oil that is difficult to remove without
redistillation. Evidently an increase in temperature permits of trouble-
some chemical changes between the acid and the oil, with the formation
of products that remain in solution. The skilful refiner is also careful
to remove the acid by washing before adding caustic soda to avoid an ob-
jectionable color.

In studying the action of sulphuric acid on different distillates from the
Russian oil, MarkownikofF and Oglobine * attributed the influence of the
acid mainly to its action on the unsaturated hydrocarbons and the oxygen
compounds. Having separated a series of unsaturated hydrocarbons
from Canadian petroleum, (Mabery and Quayle, unpublished results,) and
ascertained the presence of the same bodies in Ohio petroleum,! on account
of their unstable character and the ease with which they polymerize
alone, or more readily when in combination with sulphuric acid, it is
evident that one important office of the acid is the removal of these com-
pounds. With reference to the oxygen compounds in American petro-
leum, they seem to collect for the most part at least in the distillates above
225°. In the composition of their unpurified distillates, Markownikoff
and Oglobine found a considerable difference between the total percent-
ages of carbon and hydrogen and 100 per cent, which they assigned to
oxygen. In analysis of unpurified distillates described in this paper, the

* Ann. Chim. Phys., (6.), II. 404. t Proc. Amer. Acad., XVII. 218.

176 PROCEEDINGS OP THE AMERICAN ACADEMY.

total percentages of carbon and hydrogen in most instances have ap-
proached 100 per cent to within the limit of the error of analysis, althoagh
it is true that these distillates were subjected to close fractional distillation
wiihiu 1°, while those of MarkownikofF and Oglobine were distilled only
a few times within limits of 5°. In all the distillations of Pennsylvania,
Ohio, and Canadian petroleum, a slight coloration of the still residue
after a long series has been observed, which may be due to a small amount
of oxygen compounds.

On standing with metallic sodium, all the unpurified distillates described
in this paper deposit flocculent precipitates, more or less colored, which
are evidently products of decomposition. This cannot be due to the
action on the principal hydrocarbon, since when purified such action
by sodium is not observed. Neither should the aromatic hydrocarbons
behave in this manner toward sodium. Tiiese precipitates must be
formed from the oxygen or the nitrogen compounds in the oils, probably
from the former. They cannot be due to decomposition of sulphur com-
pounds, since Pennsylvania oil contains only a very small percentage of
sulphur, and the other distillates were all treated with alcoholic mercuric
chloride. It is quite probable that ordinary sulphuric acid combines to a
certain extent with some of the aromatic hydrocarbons. A more extended
study of the action of sulphuric acid on a larger scale as in refining, would
doubtless be interesting and profitable.

Concerning the action of fuming sulphuric acid, no further explanation
is necessary than has been given in connection with the experiments which
describe its behavior toward these petroleum distillates.

So far as possible, all details of this work have been carried on under
my immediate and constant personal supervision. For efficient aid I am
indebted to the following gentlemen : Messrs C. A. Soch, private assist-
ant, 1894-95 ; E. Davidson, private assistant, 1895-96 ; E. J. Hudson,
molecular weight determinations, and other aid mainly in connection with
the chlorine derivatives of the hydrocarbons; W. F. Priebe, who selected
a portion of the work on the Pennsylvania distillates as the subject of a
thesis for the degree of Bachelor of Science ; and Messrs Giessen, Wat-
son, Worstall, Piwonka, Shaw, and AValker, for faithful assistance in the
routine work of distillation and combustions.

The work now in progress includes a study of the pentanes, hexanes,
and heptanes in Pennsylvania petroleum, and the composition of the por-
tions of Pennsylvania, Ohio, Canadian, Berea Grit, and South American
petroleums between 216° and 350°.

MARINE BIOLOGICAL LABORATORY.

Received 1^1 i

Accession No. / l~r 7 >J ,

Given by Xj^-^r-^-^-^L^^ . \Jixi^a..r'CL^.,

Place,

*^:*rio book OP pampblet is to be pemoved fpom the Iiab"
opatopy uiithout tbe pepmission of the Tpustees.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. No. 7. — January, lbi)7.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF THE CASE SCHOOL OF APPLIED SCIENCE,

INVESTIGATIONS ON AMERICAN PETROLEUM.
By Charles F. Mabery.

XXVni. — REFRACTIVE POWER OF THE HYDROCAR-
BONS AND CHLORINE DERIVATIVES DESCRIBED
IN THE PRECEDING PAPER.

By Charles F. Mabery and Edward J. Hudson.

AlB IN THE WORK DESCRIBED IN THIS PAPER WAS GIVEN BY THE ACADEMY FROM THE C. M. WaEREN

Fund for Chemical Research.

XXVIII.— REFRACTIVE POWER OF THE HYDROCARBONS

AND CHLORINE DERIVATIVES DESCRIBED IN

THE PRECEDING PAPER.

Bt Charles F. Mabeky and Edward J. Hudson.

Presented October 14, 1896.

"Within recent years the refractive power of the hydrocarbons and
their derivatives has received considerable attention by Briihl, Thomsen,
and others, and attempts have been made to draw conclusions based on
the relation between the index of refraction and density, to define struc-
tural relations. The experimental data have been obtained for the most
part from bodies with a limited number of carbon atoms, in which the
differences in refractive power are large. Independent of determina-
tions bearing on theoretical considerations, the refractive power of oils
and solutions has been accepted on practical grounds as a valuable
property for recognition and for determining degrees of purity.

While it was not to be expected that the determination of refractive
power in hydrocarbons with such a large number of carbon atoms as
those contained in the higher portions of petroleum would be serviceable
in ascertaining the structure of individual hydrocarbons, the wide differ-
ences in specific gravity between unpnrified and purified distillates led
us to believe that similar differences should be observed in refractive
power. When subjected to experimental proof, these differences were
easily verified.

The determinations of the angle of refraction were made in the latest
form of Pulfrich refractoraeter, made by Carl Zeiss, Jena, and the calcu-
lations of the index of refraction by the formula 1 = -y/^^ — sin^ I, in which
1 represents the index of refraction, iV^the angle of the prism, and / the
observed angle. The observations in this refractometer are rapidly made,
and the calculations are much simplified by the use of a table arranged by

180 PROCEEDINGS OP THE AMERICAN ACADEMY.

Pulfrich * so that from the observed angle the index may read directly
from the table.

In the following table giving the refractive power (the excess of the
index of refraction over unity), after the numbers representing the
refractive power of the distillates purified by a mixture of nitiic and
sulphuric acids, and with fuming sulphuric acid, the differences between
the refractive power before purification and afterwards are given.

Pennsylvania. Ohio. Canada.

163°.

Ref. Power, Dif. Rcf. Power, Dif. Ref. Power, Dif.

Unpurified distillate .4241 .4248 .4276

After treatment with

IINO3 and H2SO4 .4083 .0158 ,4123 .0125 .4133 .0143

After treatment with

H2S2O7 .4146 .0095 .4113 .0135 .4137 .0139

173° .5
Uupurified distillate .4163 .4239 .4245

After treatment with

HNOsandHsSO^ .4118 .0045 .4134 .0105 .4149 .0096

After treatment with

H2S2O7 .4093 .0070 .4118 .0121 .4138 .0107

196°.
Unpurified distillate .4214 .4251 .4309

After treatment with

HNO3 and H2SO4 .4158 .0056 .4214 .0037 .4219 .0090

After treatment with

H0S2O7 .4163 .0051 .4209 .0042 .4231 .0078

216°.
Unpurified distillate .4249 .4280 .4289

After treatment with

HNOsandlLSOi .4209 .0040 .4244 .0036 .4219 .0070

After treatment with

H2S2O7 .4209 .0040 .4241 .0039 .4212 .0077

Inspection of the columns in this table headed Kefractive Power shows
a gradual increase with the rise in boiling points, less regular in the un-

* Zeit. fiir Instrunientenkunde, 1888, p. 47.

MABERY AND HUDSON. — REFRACTIVE POWER. 181

purified distillates, but more uniform after purification. A comparison
of refractive power in the same di:?tillates from all the oils reveals higher
values in the Ohio and Canadian distillates than in those from Pennsyl-
vania oil, analogous to the differences in specific giavity referred to in the
preceding paper. In the unpuiilied distillate 163° from all sources, the
influence of the large proportions of mesitylene is apparent.

In determining the refractive power of the monochlor and dichlor
derivatives of the hydrocarbons between 160° and 216°, portions of the
product were used whose composition was determined by analysis, as
shown in the preceding paper. On account of the limited quantities
purification was not carried as far as would have been desirable, although
the proportion of other bodies was probably not sufficient to affect seri-
ously the results. The differences in refractive power between the
hydrocarbons and the monochlor derivatives, as well as those between
the monochlor and dichlor derivatives are sufficient to demonstrate the
influence of the chlorine atoms.

Monochlor derivatives : — Pa. Ohio. Can.

Distillate

125-130 (C10H21CI)
130-140 (C10H21CI)

a ii

145-150 (CuH.sCI)
140-145 (C12H25CI)

Dichlor derivatives : —

Distillate.

160-170 (C10H20CI2)

rom

164°

.4424

C(

ee

.4470

a

174°

.4445

it

a

.4437

i(

196°

.4433

a

ii

.4457

u

ii

.4461

a

216°

.4456

((

u

.4521

170-180 (C10H20CI2)

<( u

190-200 (Ci.&hCL)

•om

164°

.4639 .

«

u

.4770

((

174°

.4604

((

(C

.4640

<i

((

.4676

<(

216°

.4650

((

t<

.4747

182 PROCEEDINGS OF THE AMERICAN ACADEMY.

It seems peculiar that the hydrocarbon 216° Canadian petroleum which
gives analytical values corresponding to the formula C10H24 should
have the same boiliug point as the corresponding constituent of Penn-
sylvania and Ohio oils, and that it should give a chlorine derivative with
substantially the same boiling point. Probably a better understanding
of this hydrocarbon will be gained in the study of the higher constituents,
which is now in progress.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. No 8. — Januapy, 1897.

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF THE CASE SCHOOL OF APPLIED SCIENCE.

INVESTIGATIONS ON AMERICAN PETROLEUM.
By Charles F. Mabery.

XXIX. —OA^ THE COMPOSmON OF A SOUTH AMERI-
CAN PETROLEUM.

By Charles F. Mabery and Arthur S. Kittelberger.

Aid in the work described in this Paper was given by the Academy from the C. M. Warren

Fund for Chemical Research.

XXIX. — ON THE COMPOSITION OF A SOUTH
AMERICAN PETROLEUM.

By Charles F. Mabery and Arthur S. Kittelberger.

Presented October 14, 1896.

It has long been known that large deposits of petroleum are to be
found in various districts in Soutli America, but beyond a limited use of
the crude oil in lubrication, until recently no demand has been created
for this petroleum. In the Argentine Republic are found heavy oils that
deposit paratfine on distillation. In an examination of these oils by
Engler and Ottin, hydrocarbons were found of the series C„H2„+2 and
QjH,^ ; they are said to yield a good illuminating oil. The oil fields of
Peru and Bolivia have long been known, and large quantities of petro-
leum products are here obtained. The most promising oil fields hitherto
in South America are those of Venezuela, and those deposits are receiv-
ing more attention in the preparation of commercial products.

The peculiar character of South American petroleum was brought to
the attention of one of us (Mabery) in an examination which he was
called upon to make by Mr. Horace M. Wilson of Cambridge, Mass.,
who collected a specimen of oil while prospecting along the Magdalena
River in the United States of Colombia. A few miles from the river, in
a rocky section of country, he discovered oil oozing up through a pool of
water from a fissure in the rocks below. Oil was also observed escap-
ing in bubbles through the water in a brook, where it formed a beauti-
ful green layer three yards square. These sources and another within a
range of two thousand feet were the only ones observed m a distance of
fifteen miles. The geological formations in this section consisted of
sandstones and shales tilted in nearly a horizontal position. At a dis-
tance of forty miles was a very large deposit of asphalt, hard and brit-
tle like coal. About eiglity gallons of this oil was collected at a rate
of five gallons in twenty-four hours, of which the larger portion mixed
with lard oil was used as a lubricator on car axles, for which it was
found to be well adapted. Mr. Wilson brought home fifteen gallons of
the oil, which was placed at our disposal for this examination.

186 PROCEEDINGS OP THE AMERICAN ACADEMY.

The crude oil gave as its specific gravity 0.9480 at 20°. It is a
high sulpliur oil, as shown by the followiug determinations : (I.), 0.70;
(II.), 0.66 per cent. It absorbed bromine equivalent to 12.09 per cent.
It also contains a large quantity of nitrogen, as shown by the following
determinations: (I.), 0.321 ; (II.), 0.315 per cent. A determination of
ash in the crude oil was made, in which 29.1360 grams of the oil was
burned, and the ash ignited until all carbon was consumed. The residue
weighed 0.313 gram, equivalent to 0.011 per cent, or the same amount
of ash as is obtained from Ohio crude oil.* The ash contains much iron,
as shown by its brown color.

Determinations of carbon and hydrogen in the crude oil gave the
following results : —

I. II.

C 85.80 85.45

H 12.02 11.79

The crude oil was dark in color, thick and viscous, flowing very
slowly at ordinary temperatures. In attempting to distil it under ordi-
nary atmospheric pressure, nothing came over below 260°, and between
this point and 345° it distilled in the following proportions, beginning
with 290 c.c. : —

—310° 310°-345°

70 c.c. 70 CO.

Specific gravity, 0.8749 0.8615

Bromine absorption, 8.46 34.96

As the great increase in bromine absorption shows, the distillate
310°-345° was badly cracked, and it had the very disagreeable odor of
the worst decomposition products of high petroleum distillates. The
lower fraction had only the natural odor of an uudecomposed distillate.
Nothing could be distilled above 345°, and the residue was completely
coked. In respect to the instability of its least volatile portions, this oil
diff'ers from any North American oil that has come under my obser-
vation (Mabery). It was therefore evident that another method of
distillation must be resorted to with any hope of separating without de-
composition the principal constituents.

In subjecting the crude oil to distillation in vacuo under 50 mm., the
first distillate came over at 100°, and below 250° the followinfr weights

* Proc. Amer. Acad., XXXI. 20.

MABERY AND KITTELBERGER. — S. A. PETROLEUM. 187

were collected from 14.5 kilos, after the fourth distillation. It cannot
be assumed that the absence of more volatile constituents was the result
of evaporation by exposure on the surface of the water where the oil
was found, since it was collected as soon as it appeared. No doubt such
volatile portions as constitute gasoline on long exposure under such cir-
cumstances would be lost, but not constituents such as higher members
of the series C„H.2„+2, which do not appear in this oil.

—130= 130^-135° 135°-140^ 140='-loO° 150°- 155^ 155°-160° 160°-165°

Grams 265 100 115 225 200 65 95

Sp. gr. 0.8706 0.8736 0.8747

165°-170° 170°-175° ITSO-ISO^ 180°-185° 185°-190° 190°-195°

Grams 130 130 70 125 115 155

Sp. gr. 0.8784 0.8806 0.8846 0.8858 0.8867 0.8884

195^-200' 200°-205° 205°-210° 210°-215° 215°-220° 220°-225°

Grams 140 165 95 110 100 145
Sp. gr. 0.8926 0.8947 0.8964 0.8989 0.9020 0.9045

22o°-230° 230^-235^ 235^-240° 240°-245° 245°-250=' +250

Grams 145 150 100 93 75 130

Sp. gr. 0.9078 0.9113 0.9137

Distillation of the fractions below 150° amounting to 700 grams was
continued under atmospheric pressure, collecting within the limits of
two degrees. After five distillations, ten in all from the beginning,
heaps collected at 170°, 190°, and 212°, although in small quantities.
The distillation could not be pushed too far on account of decomposition.
The fraction 170°-172° (730 mm.) gave the following percentages of
carbon and hydrogen : —

0.1456 grams of the oil gave 0.4556 gram CO2, and 0.1864 gram H2O.

Calculated for CioHoq- Found.

C 85.71 85.33

H 14.29 14.23

Unfortunately there was not enough of this fraction to determine its
specific gravity. A determination of sulphur gave 0.05 per cent. Its
bromine absorption was found to be 3.5 per cent. No other heap appeared
in the distillates below 190°. At 190°-192° more collected, evidently
corresponding to a hydrocarbon boiling at 196°. A determination of the
specific gravity of this distillate gave 0.8331. It absorbed bromine

188 PROCEEDINGS OP THE AMERICAN ACADEMY.

amounting to 3.56 per cent. A determination of sulphur gave 0.10 per
cent. A combustion gave the following percentages of carbon and
hydrogen : —

I. 0.1696 gram of the oil gave 0.5332 gram COo. The water was lost.
II. 0.1542 gram of the oil gave 0.4846 gram CO2, and 0.1912 gram H2O.

c

Calculated for
C11H22.
85.71

I.
85.73

Found.

II.

85.69

H

14.29

13.78

This distillate was treated with a mixture of nitric and sulphuric acids,
then with fuming sulphuric acid. In the first treatment, a very small
quantity of a nitro compound separated as on oil, showing a trace of an
aromatic hydrocarbon. The fuming acid produced no appreciable action.

A determination of carbon and hydrogen in this product gave the fol-
lowing results : —

0.1618 gram of the oil gave 0.5019 gram CO2, and 0.2087 gram 11,0.

Calculated for CnHjj. Found.

C 85.71 85.47

H 14.29 14.33

A determination of the specific gravity of the oil after this treatment
gave 0.8333. Besides the very small projiortion of aromatic hydrocarbon
C„H2„_6, evidently this distillate is composed of a single body, although
analysis alone is not sufficient to determine whether its composition is
represented by the formula CnH^o or CinH24.

Above 192°, no distillates collected in appreciable amounts below 210°.
At 212°-214° a larger quantity collected, evidently corresponding to a
hydrocarbon boiling at 216°, which has been found in other oils. A de-
termination of its specific gravity gave 0.8483. It absorbed bromine
equivalent to 4.29 per cent, and contained 0.04 per cent of sulphur. A
determination of carbon and hydrogen gave the following percentages : —

0.1555 gram of the oil gave 0.4887 gram CO., and 0.1992 gram IlgO.

Calculated for CjoH^^. Found.

C 85.71 85.72

H 14.29 14.23

A small quantity of nitro product was formed, when this distillate was
treated with nitric and sulphuric acids, and fuming sulphuric acid then

MABERY AND KITTELBERGER. — S. A. PETROLEUM. 189

removed more of the iiitro conipouiid from the oil. The oil shaken with
sodic hydrate imparted a yellow color to the alkaliue solution; it was
then redistilled for analysis : —

0.1590 gram of the oil gave 0.4964 gram CO2, and 0.2037 gram HgO.

Calculated for CjjUj,-

Found.

c

85.71

85.48

H

14.29

14.23

A determination of the specific gravity of the oil after treatment with
acids save 0.8484, the same value as was obtained for the crude distillate.
It is therefore evident that this petroleum consists mainly of a single series
of hydrocarbons, with a mere trace of aromatic hydrocarbons C„H2„_6.
What this series is does not appear from these results, at least so far as
its relation to the series of hydrocarbons hitherto discovered in petroleum.
In the proportions of carbon and hydrogen, the series C^Hon is indicated.
Neither the crude oil nor any of these distillates deposit paraffine, even at
low temperatures. The higher distillates are thick and viscous, light
yellow in color, and unquestionably are uudecomposed constituents of the
crude oil.

With nitric and sulphuric acids the constituents described in this paper
are as slightly affected as are the naphtenes in the Russian oil. This fact,
together with the results of analysis corresponding to the series C^Hj^.g,
and the high specific gravity, point to a similar composition for these
bodies, and the small proportion of the crude oil distilling below 220°
seems, therefore, to be composed almost exclusively of naphtenes, which
is the first instance of an American petroleum having been found contain-
ing these hydrocarbons in any considerable quantity. The higher distil-
lates from this petroleum will receive further attention in connection with
the corresponding portions of other American oils.

Having at hand a specimen of petroleum from Oregon, resembling the
heavier California products, with a specific gravity nearly the same as
that of the South American oil described above, it seemed- of sufficient
interest to submit it to an examination in connection with that of the
South American oil. The crude oil was very thick and dark, with a
specific gravity, 0.9597 at 20°. Like the California oils, as shown by
Peckham, it contains a large percentage of nitrogen compounds; A
Kjeldahl determination gave 0.868 per cent of nitrogen. This is also a
high sulphur oil, as shown by a combustion which gave 1.19 per cent of
sulphur. It is also a high carbon oil ; a combustion gave 8G.06 per cent
of carbon, and 11.87 per cent of hydrogen. A very small proportion

190 PROCEEDINGS OF THE AMERICAN ACADEMY.

distils under atmospheric pressure without decompositiou ; even under
diminished pressure, the amounts distilling below 250° are small.

In attempting to separate the constituents of this oil from the distil-
lates first collected in vacuo, the portions below 250°, about the same in
amount as the corresponding distillates from the South American oil,
were carried through ten distillations under atmospheric pressure, which
brought together larger quantities at 169°-170°, 190°-191°, and 212°-
214°. The portions collected at 85°-150° in the first distillation gave as
its specific gravity at 20°, 0.8755 ; the distillate 150°-225°, 0.9038 ; and
the distillate 225°-250°, 0.9271. After the tenth distillation, the fraction
170°-17r gave as its specific gravity, 0.S200; the fraction 189°-190°,
0.8330; and the fraction 212°-214°, 0.853. A combustion of the frac-
tion 189° -190° gave the following percentages of carbon and hydrogen : —

0.1510 gram of the oil gave 0.4757 gram COg, and 0.1835 gram H2O.

C 85.90

H 13.51

After purification with nitric and sulphuric acids, the composition of
this oil was not essentially changed : —

0.1484 gram of the oil gave 0.4643 gram CO2, and 0.1839 gram H2O.

Calculated for C)iH2n. Found.

C 85.71 85.33

H 14.29 13.77

The fraction 212°-214°, without purification, gave the following
results : —

0.1457 gram of the oil gave 0.4625 gram COo, and 0.1757 gram FLO.

C 86.46

H 13.40

By treatment with a mixture of nitric and sulphuric acids, and then
with fuming sulphuric acid, this distillate was only slightly affected : —

0.1343 gram of the oil gave 0.4231 gram COo, and 0.1677 gram H2O.

Calculated for CnHai. Found.

C 85.71 85.90

H 14.29 13.88

Unfortunately the quantities of these distillates were too limited to
permit of further examination. The principal object, however, — to ascer-

MABERY AND KITTELBERGER. — S. A. PETROLEUM. 191

tain whether any members of the series C„H2„+2 were present at all, or
the aromatic hydrocarbons C„H2„_6 iu more than minute proportions, —
was attained. Evidently the main body of this oil is composed of a
series with less hydrogen than the former and more than the latter.
The oil distills iu such small quantities, even in vacuo, below 220°, it
will be interesting to ascertain the composition of the distillates above
this point.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. Xo. 9. — January, 1897.

THE GENERA OF NORTH AMERICAN MELANOPLI.

By Samuel H. Scudder.

THE GENERA OF NORTH AMERICAN MELANOPLI.

By Samuel H. Scudder.

Presented January 13, 1897.

In an extended paper upon the Melanopli (Pezotettiges Brunn.) to be
published by the U. S. National Museum, I have treated this group
monographically, with full descriptions of each genus and species, accom-
panied by keys for their determination. As some time must elapse before
the final publication of this work, which was completed in December,
1895, and will be abundantly illustrated, I present herewith, by per-
mission of the Museum, the table therein given for the determination of
the genera of Melanopli, appending a few remarks to show how the
species already known are distributed among them. The fuller paper
will contain the complete revision, with full synonymy.

Table of the Genera of North American Melanopli.

A^. Lateral margins of subgenital plate (last ventral segment) of male,
as seen laterally, straight throughout or very slightly convex, never at
all abruptly ampliate at the base.

b^. Body exceptionally slender ; mesosternal lobes subattingent in both
sexes ; prozona three times as long as metazona . . Gymnoscirtetes.
b^. Body not exceptionally slender ; mesosternal lobes in both sexes so
widely separated that the interspace between them is at most twice as
long as broad ; prozona not more than twice as long as metazona.
c\ Interspace between mesosternal lobes of female decidedly trans-
verse, sometimes twice as broad as long ; of male sometimes trans-
verse, sometimes quadrate or subquadrate ; tegmina lobiform, linear,
or wanting.

d^. Interspace between mesosternal lobes of male decidedly trans-
verse, as broad as or broader than the lobes ; pronotum without
lateral carinas ; tegmina ovate or wanting .... Netrosoma.

196 PROCEEDINGS OF THE AMERICAN ACADEMY.

<P. Interspace between mesosternal lobes of male quadrate or
subquadrate, or, if feebly transverse (as in Paradichroplus), not
so broad as the lobes, and then the pronotum furnished with
lateral carin* ; tegraiua ovate or linear.

e\ Subgenital plate of male pyramidal, pointed, a sliglit tubercle
extending beyond its posterior margin, but the margin extend-
ing well beyond the apex of the supraanal plate.

Paradichroplus.
e^. Subgenital plate of male more or less conically protuberant
apically, but its interior apical margin not surpassing or barely
surpassing the apex of the supraanal plate.

/^ Apical tubercle of subgenital plate small, extending bui
a short distance beyond the supraanal plate ; cerci of male
abruptly narrowed before the middle by excision of the in-
ferior margin, the apical half narrow ; lateral carin^e of
pronotum wholly wanting Phcedrotettix.

f^. Nearly the whole subgenital plate forming a blunt coni-
cal tubercle projecting some distance beyond the supraanal
plate ; cerci of male forming broad, apically decurved, subfal-
cate laminae ; lateral carinse of pronotum more or less dis-
tinct Conalccea.

c^. Interspace between mesosternal lobes of female generally longer
than broad, sometimes quadrate, rarely feebly transverse ; * of male
never at all transverse (except feebly in Sinaloa and Cephalotettix) ;
tegmina variable.

d}. Tegmina never fully developed, rarely as long as the pro-
notum, lateral and ovate or linear, or wholly wanting ; hind
margin of pronotum distinctly truncate ; fore and middle femora
of male (except in Phaulotettix) distinctly more gibbous than in
the female.

e^ Furcula of male wanting, or forming a pair of brief lobes,
at most no longer than broad.

f^. Last dorsal segment of male furnished mesially with a
pair of slightly protuberant rounded lobes; cerci of male
compressed laminate, beyond the slightly narrowing basal
portion equal or subequal, the tip curved downward or
inforiorly angulate at apex.

* Cephalotettix, in vvliich the female is unknown, is placed in this division.

SCUDDER. — GENERA OF MELANOPLI,

197

g^. Prosterual spine erect ; interepace between mesosternal
lobes of male nearly twice as long as broad; fore and
middle femora of male noticeably gibbous; subgeuital
plate of male terminating in a large conical tubercle.

Barytettix.

g^. Prosternal spine retrorse ; interspace between meso-
sternal lobes of male only a little longer than broad ; fore
and middle femora of male only slightly gibbous ; subgeni-
tal plate of male with no apical tubercle. . Phaulotettix.
f. Last dorsal segment of male entirely without projecting
lobes or furcula in any form, unless as exceptionally broad
and short sessile plates ; cerci of male (except in Cepha-
lotettix) apically acuminate or curved upwards.

g^. Head large, and eyes, at least in male, large and very
prominent, the breadth of the head distinctly exceeding
the greatest width of the pronotum ; interspace between
mesosternal lobes of male slightly transverse.

Cefhalotettix.

g^. Head normal and eyes not very prominent even in the
male, so that the breadth of the head does not exceed
the greatest width of the pronotum ; interspace between
mesosternal lobes of male distinctly longer than broad.
h^. Tegmina lobiform ; subgenital plate of male pro-
truding beyond the tip of the supraanal plate by less
than half the length of the latter ; cerci of male com-
pressed, subequal, the tip broad . . • . Rhabdotettix.

K^. Tegmina linear ; subgeuital plate of male protruding
beyond the tip of the supraanal plate by much more
than half the length of the latter ; cerci of male taper-
ing from the base, the tip acuminate . . Cyclocercus.

e^, Furcula of male consisting of a pair of parallel, attingent,
cylindrical processes, generally at least twice as long as broad.

/\ Tegmina lobiform ; interspace between mesosternal lobes
of male slightly transverse ; cerci of male forming compressed
subequal laminae Sinaloa.

f^. Tegmina wanting ; interspace between mesosternal lobes
of male longer than broad ; cerci of male styliform, conical.

Paraidemona.

198 PROCEEDINGS OF THE AMERICAN ACADEMY.

dr. Tegmiua fully developed or abbreviate, never much if any
shorter than the pronotum ; hind margin of pronotum distinctly
angulate ; fore and middle femora scarcely more gibbous in the male
than in the female (except in some species of Campylacantha).

e\ Tegmiua fully developed ; disk of pronotum nearly flat, the
lateral lobes nearly at right angles to it, the posterior margin
rectangulate or subrectaugulate ; prosternal spine quadrate, ap-
pressed, broadly truncate Aidemona.

e^. Tegraina abbreviate ; disk of pronotum tectiform, the pos-
terior margin obtusangulate ; prosternal spine more or less
conical and acuminate.

f^. Head not jDrominent, the summit very slightly arched
longitudinally ; prosternal spine erect ; f urcula of male com-
posed of projecting cylindrical fingers ; surface of body very
feebly pilose Hypochlora.

f^. Head prominent, the summit strongly arched longitudi-
nally ; prosternal spine more or less retrorse ; furcula of
male reduced to slight scarcely projecting lobes ; surface of
the body rather densely pilose .... Campylacantha.

A^. Lateral margins of subgenital plate of male suddenly ampliate to a
considerable degree at the base ; or if not to a considerable degree, then
the entire margin rather strongly convex or sinuate.

b\ Subgenital plate of male furnished with a distinct subapical tubercle
(^. e., one in which the apical margin does not pass through and form
a part of the summit of the tubercle, but where it is distinctly separated
from the summit), but not otherwise tumescent (see note under A^ b^).

c^. Median carina of pronotum well developed and equally developed
throughout, accompanied on the front of the prozoua by distinct
lateral carinte ; prosternal spine sharply acuminate ; tubercle of
subgenital plate directed wholly backward, occupying the middle
of the terminal portion of the plate ; furcula distinctly developed.

£otettix.

c^. Median carina of pronotum feebly developed and generally much
more feebly on the prozona than on the metazona, accompanied by
no lateral carinse whatever ; prosternal spine bluntly acuminate ;
tubercle of subgenital plate directed upward, or upward and back-
ward, occupying the upper extremity of the terminal portion of
the plate.

SCUDDER. — GENERA OF MELANOPLI. 199

d\ Body relatively slender and compressed, not much enlarged
at the raetathorax, particularly in the male ; disk of the pronotum
tectiform,* the prozona not distinguished from the metazona either
by its plane or by any lack of a median carina, which latter is
generally marked in color ; pronotum fully half as long again as
broad; hind femora long and slender; apical tubercle of male
abdomen prominent; furcula present as distinctly projecting lobes;
terminal segments of female abdomen not abbreviated, the ovi-
positor fully exserted ■ . . . . Hesperotettix.

eP. Body- relatively short and stout, considerably enlarged at the
raetathorax even in the male ; disk of pronotum generally convex
transversely ; the prozona slightly and independently tumid with
DO median carina, thus distinguishing it from the metazona ;t ^"nd
femora relatively short and stout ; apical tubercle of male abdo-
men not very prominent; furcula scarcely or not apparent; termi-
nal segments of female abdomen abbreviated, the ovipositor only

partially exserted ^oloplus.

b\ Subgenital plate of male with no distinct subapical tubercle, but
often apically prolonged or tumescent. |

c\ Meso- and metastethiura together, in both sexes, no longer or
scarcely longer than broad ; metastethium narrowing but little pos-
teriorly, so that the portion behind the metasternal lobes is but
little narrower than the rest, rarely (in the male) less than three
fourths its width ; cerci of male very simple, subconical, straight ;

ovipositor half concealed Bradynotes.

c^. Meso- and metastethium together, at least in the male and
nearly always in both sexes, distinctly longer than the width of
the metastethium ; the latter rapidly narrowing posteriorly, so that
the portion behind the metasternal lobes is not, or is hardly more
than, one half the greatest width of the metastethium ; cerci of
male variable ; ovipositor generally fully exserted.

d}. Interspace between mesosternal lobes of male distinctly trans-
verse. § as broad or nearly as broad as the lobes themselves ; of

* This feature is not so apparent in some species as in others.

t Tills feature is less marked in some species than in others.

X Tliere is a minute subapical tubercle in some species of the flabellifer group of
Melanoplus, but in these the male cerci are exceptionally broad ; while in the
species of the alternate category (^1- b'^) they are very slender and tapering.

§ In many cases the interspace is truncate-cuneiform or broadly clepsydral, in
which cases the breadth is to be measured in the middle.

200 PROCEEDINGS OF THE AMERICAN ACADPJMY.

the female distinctly or strongly transverse, fully as broad as or
(and generally) broader than the lobes ; metasternal lobes of male
generally distinctly distant, occasionally approximate ; of the female
generally more distant, the interspace in the latter generally as wide
as or wider than the frontal costa ; tegmina typically abbreviate.
e^. Face almost vertical ; eyes small but prominent and widely
distant; pronotum constricted in the middle, with deeply im-
pressed transverse sulci, and the lateral lobes not obliquely
. truncate apically in front; distinct lateral carinas.

Dendrotettix.
e-. Face a little oblique ; eyes rather large, not very prominent
and not very distant ; pronotum not, or but feebly, constricted
in the middle, with generally feebly impressed transverse sulci,
and the lateral lobes obliquely truncate apically on the anterior
section Podisma.

LP- Pi'onotum of subequal width, the sides nearly parallel ;

subgenital plate of male normal Podisma s. s.

p. Pronotum enlarging posteriorly, conspicuously in the
female; subgenital plate of male exceptionally expanded,
laterally tumid, and elevated premarginally . Eupodisma.]
d^. Interspace between mesosternal lobes generally longer or much
longer than broad in the male, almost never (cf. Mel. montanus
Thom.) in the least broader than long even when the sides of
the interspace are posteriorly divergent; generally quadrate in
the female, but more variable than in the other sex, sometimes
as narrow as there, more often subtransverse, occasionally in some
brachypterous forms (as in Mel. artemisice Brun., Mel. militaris
Scudd., Mel. altitudinum Scudd., and Asemoplus) distinctly trans-
verse ; in both sexes always distinctly, generally much, narrower
than the lobes (except in the females of the cases just cited, where
they are barely narrower) ; metasternal lobes of male generally
attingent or subattingent, rarely only approximate ; of the female
less distant than in the alternate category (^- i- c^ (fi), generally
approximate or subapproximate, the interspace generally narrower
than the frontal costa ; typically the tegmina are completely
developed.

e^. Face almost vertical or a little oblique, its angle with the
fastigium rarely less than 75° ; eyes rounded oval, never more
generally less than half as long again as broad; portion of

SCUDDER. — GENERA OF MELANOPLI. 201

metasteruum lying behind the lobes transverse, more than
twice as broad as long ; legniina normally present.
f^. Tegmina always present ; sides of first abdominal segment
with a distinct tympanum.

g^. Fastigium of vertex plane or convex ; eyes separated
widely, the space between them twice as broad as the
frontal costa ; pronotum furnished with distinct percur-
rent lateral carince ; tegmina abbreviate ; cerci apically

acuminate Paratylotropidia.

g^. Fastigium of vertex more or less depressed or with
elevated lateral margins ; eyes separated narrowly, at
most but little further apart than the width of the
frontal costa ; pronotum with indistinct * or no lateral
carinae ; tegmina fully developed or abbreviate ; cerci
variable, rarely acuminate apically.

h}. Inferior genicular lobe of hind femora with at least
a darker basal spot or transverse band ; cerci of male
variable, often enlarging apically.

i^. Dorsum of pronotum rarely or never twice as long
as the average breadth, generally only half as long
again, even in the male, generally constricted more
or less in the middle ; antennae even when longest (as
in Mel. pachardii Scudd., Mel. nitidus Scudd.) no
longer than the hind femora, and only twice as long
as the pronotum alone ; face rarely as declivent as in
Paroxya ; prozona usually a half longer than the
metazona.

j^. Head not large in proportion to pronotum, nor
prominent, but little longer than the prozona, unless
(as in Mel. spretus Uhl. e. g.) the latter is distinctly
transverse ; pronotum in no way subsellate, nor flar-
ing in front; tegmina, when fully developed, nar-
row, rarely {Mel. fasciatus Barnst.,t Mel. dawsoni
Scudd.,) rat'her broad, but then very distinctly
tapering, more or less tapering in distal half, at

* In a few species they are tolerably distinct.

t In form of tegmina and sparseness of neuration, tlie macropterous form of this
species is the Melanoplus most nearly allied to Phoetaliotes, and like it, the species
is dimorphic as to tegmina.

202 PROCEEDINGS OF THE AMERICAN ACADEMY.

a distance from the apex equal to the breadth of
the tegmina distinctly narrower than the metazona,
the intercalaries and cross-veins of the discoidal
area relatively numerous, at least in the apical
fourth and usually throughout, the venation in
general sharp and clearly defined, the area inter-
calata distinctly defined by the adjustment of the
veins at its distal extremity, the humeral vein
straight and apically arcuate, nearly always ter-
minating either on the apical margin or but a
short distance before it, running for some distance
almost exactly parallel to the costal margin or
merging insensibly into it ; cerci of male very
variable, very rarely {Mel. flabellatm Scudd., Mel.
puer Scudd.) styliform, and then the subgenital
plate is either exceptionally broad or only moder-
ately narrow, and the apical margin elevated.

Melanoplus.
j^. Head large in proportion to pronotum, nearly
half as long again as the long prozona ; pronotum
faintly subsellate, feebly flaring in front to receive
the head ; tegmina, when fully developed, broad
and subequal, hardly tapering in the distal half,
at a distance from the apex equal to the breadth
of the tegmina as broad as the metazona, the in-
tercalaries and cross-veins of the discoidal area
everywhere ^ew, the venation in general loose
and ill defined, the area intercalata not distinctly
marked by the adjustment of the veins at its distal
extremity, the humeral vein broadly sinuous, ter-
minating on the costal margin at least as far be-
fore the apex as the breadth of the tegmina,
nowhere running closely parallel to that margin
nor merging into it ; cerci of male styliform, the
subgenital plate very narrow, the margin not api-
cally elevated Phcetaliotes.

i^. Dorsum of pronotum twice as long as average
breadth, at least in the male, with no median con-
striction; antennjE, at least in the male, generally
longer than the hind femora and much more than

SCUDDER. — GENERA OF MELANOPLI. 203

twice as long as the pronotum, generally twice as

long as head and pronotum together ; face more de-

cliveut than in Melauoplus ; prozoua only about a

third longer than the metazoiia . . . Paroxya.

¥. Inferior genicular lobe of hind femora wholly pallid,

with no dark basal spot or transverse band; cerci of

male conical or subconioal or basally bullate, always

apically pointed.

V: Subgenital plate of male terminating in a pro-
nounced tubercle ; prosternal spine slender.

Poecilotettix.

%^. Subgenital plate of male, even when apically angu-
late, not furnished with an apical tubercle ; prosternal
spine stout.
j^. Relatively heavy bodied ; dorsal disk of the
prozona tumid independently of the metazona ;
pronotum distinctly angulate or convex behind ;
the portion of the metasternum lying behind the
lobes laterally extended, reaching to the coxas ;
tegmiua fully developed or abbreviate but over-
lapping, with many longitudinal veins ; cerci of
male very stout and bullate on basal half or more ;
abdomen of female bluntly rounded apically, the
posterior segments much abbreviated ; ovipositor
but slightly exserted (Edaleonotus.

p. Relatively slender bodied ; dorsal disk of pro-
zona not tumid independently of the metazona ;
pronotum truncate posteriorly ; portion of meta-
sternum lying behind the lobes laterally abbre-
viated, much narrower than the width between
coxa3 ; tegmina linear, lateral, distant, with only a
few longitudinal veins ; abdomen of female taper-
ing regularly to a pointed tip ; ovipositor normally
exserted Asemophis.

/-. Tegmina wanting; sides of first abdominal segment with
no tympanum Philocleon.

e-. Face rather strongly oblique, the angle it makes with the
fastigium varying about from 55° to 67° ; eyes elongate, almost
or quite twice as long as broad ; portion of metasternum lying

204 PROCEEDINGS OF THE AMERICAN ACADEMY.

behind the lobes subtriangular, not greatly broader than long ;
tegmiiia linear and lateral or absent .... Aptenopedes.

Gymnoscirtetes Bruner MS. contains a single undescribed species from
Florida, of a somewhat anomalous character.

Netrosoma gen. nov. is based on two species from Mexico, both of
them new.

Paradichroplus Brunner is represented in North America by two
species from Mexico and Central America, mexicanus Brunn., and
imricolor Stal,

Pha^drotettix gen. nov. has but a single and undescribed species, which
comes from Mexico and southern Texas.

Conalcsea gen. nov. has three species, all of them new ; they occur in
Mexico and southwestern New Mexico.

Barytettix gen. nov. is known only from Lower California, where it is
represented by two undescribed species.

Phaulotettix gen. nov. contains a single and undescribed species from
Mexico.

Cephalotettix gen. nov. is also monotypic and the species is new and
from Mexico.

Rhabdotettix gen. nov. contains three species from Texas and Mexico,
pilosus Stal and two undescribed forms.

Cyclocercus gen. nov. also contains three species found in northern
Mexico and southern Texas, all of them new.

Sinaloa gen. nov. is founded upon a single, as yet unpublished Mexican
species.

Paraidemona Brunner is found in Texas and northern Mexico, and is
restricted to punctata Stal and a new species, the other species referred to
it by Brunner being placed elsewhere.

Aidemona Brunner is confined to azteca Sauss., found in Mexico.

Hypochlora Brunner was originally founded on three species, but is
here restricted to only one of them, alba Dodge, found on the eastern
margin of the Rocky Mts.

Campy lacantha gen. nov. is composed of four species found in the
western United States east of the Rocky Mts., from Nebraska to Texas,
and in Durango, Mexico. The species are acutipennis Scudd., olicacea
Scudd., vivax Scudd., and one undescribed species.

Eotettix gen. nov. is founded on a single undescribed species from
Florida.

SCUDDER. — GENERA OF MELANOPLI. 205

Hesperotettix Scudder is found across the coutiuent, but only a single
species, brenpennis Thom., is known east of the Great Plains, and that
has been found only on or near the Atlantic border. The other species
are vin'dis Thorn., pacificus Brun. (undescr.), speciosus Scudd., and four
new species, including two which have often been confounded with one
of the foresoinj;.

jEoloplus <ren. nov. is confined to the western half of the United States
from the Yellowstone to the Mexican border. I have seen no species
from further east than western Kansas and Nebraska, so that it does not
appear to reach the prairie region. There are ten species, of which
those already published are regalis Dodge, chenopodii Brun., turnbulli
Thom., and plagosus Scudd.

Bradynotes Scudder is confined to the extreme northwestern United
States, and contains seven species, of which the only ones already
described are hispida Brun., and obesa Thom. (optmus Scudd.).

Dendrotettix Riley has but a single and dimorphic species, quercus Ril.
(longipennis Ril.), ranging from Missouri to Texas.

Podisma Latreille is used in replacement of Pezotettix Burmeister for
reasons given in Psyche (vii, 195). It has a wider distribution than any
other genus of Melanopli, being found in Europe and temperate Asia, as
well as in America. The American species occur in two great districts,
one in the west from Alberta to northern New Mexico, the other in the
east from western Ontario and New York to Maine. There are eight
American species, of which those already described are the following;
glacialis Scudd., stupefacta Scudd., dodgei Thom., marshallii Thom., and
oregonensis Thom. Eupodisma is a subgeneric term for primnoa Fisch.
de "W. of Siberia.

Paratylotropidia Brunner is based on an undescribed species found
from Dakota to Texas.

Melanoplus Stal is the dominant genus, with one hundred and thirty-
one species, but it is confined to North America. These species have
been grouped into twenty-eight series, but as, on account of its pre-eminent
importance, this genus is treated separately in another paper in press, no
further account of the species is given here.

Phcetaliotes gen. nov. is based upon a single species, nehrascensis
Thom., found in the western part of the Mississippi basin, and beyond its
latitudinal limits from Alberta to Mexico.

Paroxya Scudder has three species, mostly confined to the Atlantic
and Gulf States ; they are atlantica Scudd., hoosieri BlatchL, and
jioridana Thom.

206 PROCEEDINGS OF THE AMERICAN ACADEMY.

PcEcilotettix gen. nov. is represented by three species: picticornis
Ttiom., and two new species, found only on the Pacific coast near our
southern borders.

(Edaleonotus gen. nov. is based on the polymorphic species, collaris
Scudd., found on the Pacific coast.

Asemoplus gen. nov. is based on a single species, montanus Brun.,
found in the extreme northwestern United States.

Philocleon gen. nov. has but a single species, nigroviUatus Stal, from
Mexico.

Aptenopedes Scudder is known only from the Gulf States, with the
three species, sphenarioides Scudd., rufovittata Scudd., and aptera
Scudd.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. Xo. 10. — Makcii, 18SJ7.

CYCLE IN THE LIFE OF THE INDIVIDUAL {ONTO-
GENY) AND IN THE EVOLUTION OF ITS
OWN GROUP {PHYLOGENY).

By Alpheus Hyatt.

CYCLE IN THE LIFE OF THE INDIVIDUAL (ONTO-
GENY) AND IN THE EVOLUTION OF ITS
OWN GROUP (PHYLOGENY).*

By Alpheus Hyatt.

Presented October 14, 1896.

The organic cycle, as generally understood both by laymen and
scientists, and as usually described in literature, is, as a rule, considered
from a physiological rather than a structural point of view. The develop-
ment of the young, and the attainment of the adult or comparatively
permanent stage, complete the progressive stages. Old age, accompanied
by losses of characteristics and functions and consequent weakening of
the body, is retrogressive, and brings on second childhood, thus complet-
ing the cycle in the ontogeny.

My purpose to-night is to show that the cycle is also represented in
the life history of the individual by definite structural changes, and that
these have direct correlations with the history of the changes in the
forms of the group while evolving in time.f

The fundamental discoveries that are more than any other directly
useful in the study of the phenomena of the cycle, both in ontogeny and
phylogeny, may be briefly noticed as follows.

The opinion that the higher animals are complex colonial aggregates
of cells, which in structure are equivalents to the lowest and minutest
adult forms of the animal kingdom, the unicellular bodies of Protozoa,

* This paper was in part read before the Academy as a general summary of tlie
phenomena of cycles, but does not assume to be an exhaustive or even complete
account of the literature or theoretical views treated of.

t These correlations have been more fully stated in a number of publications by
the author, especially " Genesis of the Arietidas," Smithsonian Contribution 673, and
Mem. Mus. Comp. Zoology, Vol. XVI. ; " Bioplastology and the Related Branches
of Scientific Research," Proc. Bost. Soc. Nat. Hist., Vol. XX VI. ; and "Phylogeny
of an Acquired Characteristic," Proc. Am. Phil. Soc. Vol. XXXII. No. 14.3.

VOL. XXXII. 14

210 PROCEEDINGS OP THE AMERICAN ACADEMY.

has been steadily gaining in probability since it was first announced by
Oken in 1 805, in " Die Zeugung," Frankfurt bei Wesche, 8vo.* This
work we have not yet seen, but in the first edition of the "Naturphilo-
sophie," Jena, 1809, II., XII. Buoh, Zoogeuie, he describes protoplasm
as ' Punctsubstauz ' and as giving rise to the ' Blasenform oder Zellform'
in both animals and plants. Oken considered the lower animals " Poly-
pen, Medusen, Beroen, kurz alle Gallertthiere," to be composed of
' Punctsubstanz.' The nerves, the cartilage, and bones of higher animals
were considered as modifications of this form of ' protoplasm,' but the
skin and fleshy parts, including the viscera, were described as cellular,
" dem Fleisch liegt die Blaschenform zur Gruude"; again, on p. 30, "die
Eingeweide, welche am meistens aus Zellengeweb bestehen." Oken, in
XII., VIII. Buch, treats of the subject we are more immediately in-
terested with, and writes as follows : " Pflanzen und Thiere konnen nur .
Metamorphosen von Infusorien sein, . . . im kleinsten sind sie nur infu-
soriale Blaschen die durch verschiedene Combinationen sich verschiedene
Gestalten und zu hoheren Organismen aufwachsen," and also adds, on
p. 29, in anticipation of one of the points advanced by the author in his
" Larval Theory of the Origin of Cellular Tissues," f " auch besteht der
Samen aller Thiere aus Infusorien."

This author directly compares his cystic or intestinal animals, In^
fusoria, with ova, and speaks of them as oozoa, and in the preface to the
English edition of his Physiophilosophy, London, 1847, Ray Society, he
writes that all organic beings originate from and consist 'of vesicles or
cells. " Their production is nothing else than a regular agglomeration
of Infusoria ; not, of course, of species previously elaborated or perfect, hit of
mucoiis vesicles or points in general which first form themselves hj their
nnio7i or combination into particular species." Oken's view was based on
the resemblances existing between the Protozoa and the cells in the tissues

* Through the kindness of Prof. George B.Tur I have received his copy of tliis
book, and find to my surprise tliat it contains quite different statements of Oken's
theory from that of the " Naturpliilosopliie." Tiiis autlior made great advances in
knowledge between 180-5 and 1809. In "Die Zeugung" he relies for the proof
tliat tlie more complex animals are descended from what we now call unicellular
organisms upon the constant presence of some of these in decaying flesh, assuming
that they are the disintegrated elements of the body itself and not independent
productions.

He makes also, on p. 128, distinct statements asserting the parallelism of the
development of the individual to the evolution of the organic kingdom, but gives
only fanciful analyses in support of this position.

t Proc. of the Bost. Soc. Nat. Hist., Vol. XXIII., March 5, 1884.

HYATT. — ORGANIC CYCLES. 211

of the ]Met:izoa ; and it is evident he is entitled to be considered the first
teacher of tlie unicellular doctrine, an honor now universally given to
Von Siebold.

However imperfect and imaginative the results as compared with the
more objective statements of later observers, the author who wrote such
sentences as these had as clear ideas as the knowledge of his time per-
mitted, and was the Ilaeckel of the early part of this century, and, like
him, a great and successful leader, making many errors but also many
discoveries and '• blazing out " some of the paths that we are still
following.

Meckel* seems to have been the first author who brought together and
stated in a clear way the scattered observations and ideas with regard to
tlie correlations existing between the transient stages of development of
the individual and the so-called permanent modifications represented by
the similar characters in the adult stages of similar forms.

Meckel says : " Es giebt keinen guten Physiologen, den nicht die
Bemerkung frappirt hatte, dass die urspriiugliche Form aller Organismen
eine und dieseJbe I'st, unddass aus dieser einen Form sich alle, die niedrig-
steu, wie die hochsten, so entwickeln, dass diese die permanenten Formen
der ersten nur als voriibergehende Perioden durchlaufen. Aristotles,
Haller, Harvey, Kielmeyer, Autenrieth, und mehi-ere andere haben diese
Bemerkung, eutweder im Vortibergehen gemacht, oder, besonders die
letzten, hervorgehoben und far die Physiologie ewig denkwlirdige Re-
sultiite daraus abgeleitet.f

" Von dieseu niedriorsten Wirbelthieren an bis zu den hochsten Ge-
schlechtern lasst sich die Vergleichung zwischen dem Embryo der hohern
Thiere und den niedern im volkommenen Zustande vollstandiger und
treffender durchfuhreu.

" In der That giebt es ja eine Periode wo der Embryo des hochsten
Thieres, wie schon Aristotles sagt, nur die Gestalt einer Made hat, wo er
ohne aussere und innere Organisation, bloss ein kaum geformtes Klump-
chen von Polypensubstanz ist. Ungeachtet des Hervortretens von
Organen bleibt es docli nach, wegen des ganzlichen IMangels eines
innern Knochengertistes, eine zeitlang Wurm und Mollusk, und tritt

* Meckel, "Entw. e. Darstellung d. zw. d. Embryonalzustande d. hohern
Tliiere n. d. perman. d. niedern stattfindenden Parallele." Beitr. z. vergleicli.
Anat., II., Leipzig, 1811, pp. 1-148. Meckel speaks of his publications as only
preparatory to more extended researches.

t I desire to express my indebtedness to Miss M. F. Slack, Librarian of Mus.
Comp. Zo()logy, for her assistance in finding this important reference.

212 PROCEEDINGS OF THE AMERICAN ACADEMY.

erst spater iu die Reibe der Wirbelthiere, wenn gleich Spuren der
Wirbelsiiule schoa in den friihesten Perioden seinen Anspruch auf diese
Stelle in der Reilie der Tliiere beglaubigten."

It is very obvious, from these statements of Meckel, that the correla-
tion of embryology and the epembryonic stages of the individual vpith
the permanent modifications of animals of simpler construction was under-
stood, as far as was possible with existing knowledge, from the time of
Aristotle, and that it was to a greater or less extent a working hypothesis
at that time, and, as declared by him, had been helpful in giving a clearer
understanding of the development of the individual and of the relations of
the individual to the whole animal kingdom.

The next step was taken by Von Baer, in dividing the animal kingdom
into four types and in limiting this general statement to animals occurring
within each of these types. He also considered it highly probable (not
barely possible, as it is quoted by some writers) that the earliest stages
of the embryo resemble in aspect the adult stages of the lowest grade of
forms in the animal kingdom. He had in mind in this statement the
modern view of the affinities of the earliest stages of the embryo or its
repetitions of the characteristics of Protozoa,* so far as the knowledge
of his time permitted.

Von Baer endeavored to prove that each of the four types had similar
embryos, and that the type characters were determinable at early stages
in the onto'^eny. Both Von Baer and Louis Agassiz were pupils of
Ignatius Bollinger, an embryologist who published nothing. Both of
these eminent men have recognized him as their master in embryology,
but have given no definite statement of what they were taught by him.
Louis Agassiz accepted Von Baer's opinions, and subsequently enlarged
them, when he published his work on fossil fishes, by the introduction
of the element of succession in time, and thus laid the basis for all
more recent investigations.

Agassiz gave the fullest exoression of his views in " Twelve Lectures
on Comparative Embryology," Lowell Institute, Boston, 1848-49, subse-
quently published in pamphlet form. One wonders, as he reads, how any
man holdins such views could have held his mind closed to the conclusion
that animals were evolved from simpler or more primitive forms. The
effect of theoretical preconceptions in closing the mind to the reception
of new ideas never had a stronger illustration. Louis Agassiz, in 1849,
had all the facts essential for building up an hypothesis of evolution that

* Entwickelung. d. Thiere, Scholion V. pp. 119, 120, etc.

HYATT. — ORGANIC CYCLES. 213

would have placed him iu the history of science on the same line with
Lamarck and Darwin.

He states in fourth lecture, p. 26, as follows : " The results thus far
obtained in the lectures which I have delivered can be expressed as
follows. There is a gradation of type in the class of Echiuoderms, and
indeed in every class of the animal kingdom, which, in its general out-
lines, can be satisfactorily ascertained by anatomical investigation ; but it
is possible to arrive at a more precise illustration of this gradation by
embryological data. The gradation of structure in the animal kingdom
does not only agi-ee with the general outlines of the embryonic changes.
The most special comparison of these metamorphoses with full-grown
animals of the same type leads to the fullest agreement between both,
and hence to the establishment of a more definite progressive series than
can be obtained by the investigation of the internal structure. These
phases of the individual development are the new foundations upon
which I intend to rebuild the system of zoology. These metamorphoses
correspond, indeed, in a double sense, to the natural series established in
the animal kingdom : first, by the correspondence of the external forms,
and secondly, by the successive changes of structure, so that we are here
guided by the double evidence upon which the progress in zoology has,
up to this time, generally been based.

"Their natural series again correspond with the order of succession of
animals in former geological ages, so that it is equally as true to say that
the oldest animal^ of any class correspond to their lower types in the
present day as to institute a comparison with the embryonic changes, and
to say that the most ancient animals correspond with the earlier stages of
growth of the types which live in the present period. In whatever point
of view we consider the animal kingdom, we find its natural series agree
with each other ; its embryonic phases of growth correspond to its order
of succession in time, and its structural gradation, both to the embryonic
development and the geological succession, corresponds to its structure ;
and if the investigations had been sufficiently matured upon this point, I
might add that all these series agree also in a general way with the geo-
graphical distribution of animals upon the surface of our globe, but this is
a point upon which I am not yet prepared to give full and satisfactory
evidence. So much for the views referring to embryology in its bearing
upon zoological classification."

And again on page 27 : —

" However, another step had to be made to show a real agreement
between the earlier types of animals and the gradual development of the

214 PROCEEDINGS OP THE AMERICAN ACADEMY.

animal kingdom, which has been the last progress in our science of fossils,
namely, to show that these earlier types are embryonic in their character ;
that is to say, that they are not only lower in their structure when
compared with the animals now living upon the siirface of our globe, but
that they actually correspond to the changes which embryos of the same
classes undergo during their growth. This was first discovered among
fishes, which I have shown to present, in their earlier types, characters
which agree in many respects with the changes which young fishes under-
go within the egg. Without entering into all the details of these re-
searches, I will conclude by saying it can now be generally maintained
that earlier animals correspond not only to lower types of their respect-
ive classes, but that their chief peculiarities have reference to the modifi-
cations which are successively introduced during the embryonic life of
their corresponding representatives in the present creation. To carry
out these results in detail must now be, for years to come, the task of
paleontological investigations." *

* Prof. George Baur of Chicago University has, since tlie above was written,
called my attention to Carl Vogt's " Embryologie des Salmones," publisiied un-
der the general title of " Poissons d' Eau Douce," par L. Agassiz, Neuchatel, 1842.
I take the following quotations from his pages.

Vogt says, on the second page of the Preface, after stating that Louis Agassiz
had handed this part of the work over to him: "En me confiant une tache aussi
honorable, mon ce'lebre ami n'est cependant point reste ctranger a mes recherches.
Nous avons discute' ensemble les faits cajiitaux, a mesure que I'observation me
les revclait; souvent meme nous les avons examines de nouveau en commun, et
lorsque j'eus re'dige mon travail, c'est encore lui qui a bien voulu le revoir."

Again, on pages 256, 257 : —

" Depuis longtemps on a discute duns des sens tres-divers la question de I'analogie
entre les phases du deveJoppement des animaux vivants maintenant et les changements qui
sent su7-i-enus dans Vordre de succession des especes fossiles ; mais faute de renseigne-
mens precis sur Ton ou I'autre des cotes de la question, ces generalisations sont
reste'es dans un vague tres-facheux pour les vrais progres de la science. Sans en-
trer ici dans des considerations hasardees, sans aborder le domaine encore trop peu
cultive de la plus grande analogic qu'ofFrent entre elles les diffcrentes parties du
corps des poissons fossiles les plus anciens et que Ton pourrait paralleliser avec
riiomogeneite des tissus primitifs de I'embryon, je me bornerai a faire ressortir
quelques points qui ne sauraient plus etre conteste's et qui, je I'espere, feront
faire de nouvelles recherches sur I'ensemble de la question."

Vogt concludes, on page 260, as follows : —

"On pourra done dire a I'avenir, en restant rigoureusement dans les limites de
I'observation, qu'Ii certains c'gards, les especes fossiles d'une classe parcourent dans
leur succession historique des metamorphoses semblables a celle que subissent les
embryons en se developpant ; ou vice-versa, que les embryons des animaux de
notre c'poque passent, dans les differentes epoques de leur dcveloppemeut par des

HYATT. — ORGANIC CYCLES. 215

Perliai^s in consequence of pressure of other work or of his theoretical
views, Louis Agassiz seemed to have lost sight of the great importance
of continuing his researches upon the meaning and correlations of the
epembryonic stages. These were referred to in his publications, but
were not made as prominent as they deserved after the lectures at the
Lowell Institute in 1849, and in his personal talks with his students or

etats analogues a ceux que pre'sentent les especes fossiles dans leur succession ; ou
en d'autres teriiies enfin, que le de'veloppement d' une classe dans 1' liistoire de la
terre offre, a divers o'gards la plus grande analogie avec le de'veloppement d'un
individu aux differentes e'poques de sa vie. La demonstration de cette ve'rite' est
un dcs plus beaux rc'sultats de la pak'ontologie raoderne."

These quotations antedate Louis Agassiz's direct statements since in liis " Dis-
cours sur la Succession et le De'veloppement des Etres organises a la Surface du
Globe Terrestre," etc., Neuchatel, 1841, he does not give any embryological laws.
It is of course quite possible that he did not consider it necessary to mention these
in his " Discours,'' which, notwithstanding its title, was merely a general statement
of the paleontological evidences of a plan in nature, and ended with an argument
for tlie existence of an intelligent creator. Whether this was an omission, or that
he did not at that time recognize the importance of the law of palingenesis, does not
appear from the text of his publications. The first definite statement under Louis
Agassiz's own hand appears to be in his " Recherches sur les Poissons Fossiles,"
Neuchatel, 1833-43, Vol. I. Chap. IV. p. 93 : —

" J'aurai par conse'quent souvent recours a I'embryologie d'une part pour recher-
cher les analogies entre le squelette des embryons et celui des poissons cartilagineux
par rapport aux poissons osseux ; d'autre part pour cclaircir les rapports qui ex-
istent entre les formes primitives que nous trouvons chez les poissons des anciennes
couches de la terre, et celles qui se voient dans les premiers temps de la formation
de I'embryon."

Tins was certainly written subsequently to the "Poissons d'Eau Douce," since
Agassiz quotes Vogt upon tlie following pages 94, 95. He also speaks of the same
on page 169, referring both to zoological rank and occurrence in time.

These are extracts from the Introduction, which was probably written and pub-
lished in 1843, judging by tlie immediately following Preface to the first volume,
which is dated 1843.

Vogt's candid admission in his Preface of constant consultations with his master,
and his failure to state this law as his own discovery or to claim the authorship
subsequently, lead one to think that it might not have originated with him.

Vogt fully understood the meaning and application of this law, and restates it
clearly in his "Zoologische Briefe," Frankfurt am Main, 1851, p. 16, and also in his
"Lehrbuch der Geologic und Petrefactenkunde," 1854, Vol. I. p. 471, and tries to
apply it to the classification of different groups of the animal kingdom and to tlie
history of their fossil remains in " Paleontologische Entwickelung," pp. 423-541.
He seems to have been tlie first to have seen its true meaning as a law of evolution,
since he completely rejects the doctrine of special creations. It may be that the
only way out of this difficulty is to call this Agassiz and Vogt's law, just as it is
now not infrequent to speak of Darwin's and Wallace's laws of natural selection.

216 PROCEEDINGS OF THE AMERICAN ACADEMY.

in his lectures I cannot remember that they were ever treated directly
by anything more than incidental references, although embryology was
very often the principal theme.

Nevertheless, I must have got directly from him, subsequently to 1858,
the principles of this branch of research, and through this and the abun-
dant materials furnished by the collections he had purchased and placed
so freely at my disposal, I soon began to find that the correlations of the
epembryonic stages and their use in studying the natural affinities of
animals were practically an infinite field for work and discovery.

Although within a year after the beginning of my life as a student
under Louis Agassiz I liad become an evolutionist, this theoretical change
of position altered in no essential way the conceptions I had at first
received from him, nor the use we both made of them in classifying and
arranging forms. This experience demonstrated to my mind the absurd-
ity of disputing the claims of any author to the discovery of a series of
facts and their correlations because of his misinterpretation of their more
remote relations or general meaning. It is of some importance to notice
this, because it is the rule now to attribute Von Baer's and his predecessors'
and Louis Agassiz's discoveries in this line to Haeckek This eminent
author has, indeed, given one of the most modern definitions of this law,
and has named it the ' law of biogenesis.' Ilaeckel's discoveries in em-
bryology are sufficiently great without swelling the list with false entries,
but it will probably be a long time before naturalists realize and acknowl-
edge this error. Some of the most eminent embryologists in this country
have ado[)ted the Haeckelian nomenclature without sufficient critical
examination of the term under discussion. The so called Haeckelian
'law of biogenesis' is really Agassiz's law of embryological recapitula-
tion restated in the terms of evolution.

It has surprised me that serious objections to the use of the word
' biogenesis ' in this connection have not been made. This word has
been long employed in another sense, as antithetical to ' abiogenesis.'
The latter has been for many years applied to the theory of the genera-
tion of living from inorganic matter, and tlie former to the theory assert-
ing that living matter can originate only from living matter ; the use of
the phrase ' the law of biogenesis ' is consequently inappropriate, since
neither did Agassiz's nor Haeckel's discoveries cover so much ground.
The former gave us a law for the correlations of the earlier stages of
ontogeny with phylogeny. Tliis cannot be called ' the law of biogene-
sis,' since that has been long ago stated as the law of the origin and
continuity of organism, or, in other words, the genesis and continuity of

HYATT. — ORGANIC CYCLES, 217

life from and through living matter only. There are two different
manifestations of Awassiz's law, which Haeckel defined and named
'palingenesis' and ' oaMiogenesis,' the former referring to the ordinary
or regular mode in which the characteristics of ancestors are repeated
in the development of the individual, and the other to what is frequently
called the abbreviated mode, etc.

These two modes are by no means all, but at present only the first
or simplest manifestations of the phenomena need be treated of. This,
or what Haeckel very appropriately calls ' palingenesis,' was what Louis
Agassiz had studied, and, so far as all the essential facts were concerned,
thoroughly understood, and it was this that he taught his students, so that
it became, at any rate in my own case, the foundation of all my subse-
quent work in determining the mutual relations of forms. If then, as I
have proposed in former publications, the term ' law of palingenesis ' be
adopted, this expressly states just what Louis Agassiz discovered.

Observations upon this ground made especially upon Cephalopoda have
led to the discovery of correlations between the latter or epembryonic
stages and the adult stages of extinct ancestors, which have greatly en-
larged the field of application of Agassiz's law of palingenesis, and given
it an exactitude that has made it of surpassing importance in the study
of evolution. Beecher has been able to point out the single species of
Brachiopod from which the whole of the vast number of distinct forms
of this great group have originated. He has established this fact not
only by showing that the young of the existing and fossil forms all repeat
more or less at one stage the form of the adult of the initial species, but
has also found a very near affine of this single ancestral species as a
fossil appearing in one of the earliest of fossil-bearing formations.

Dr. R. T. Jackson has done the same work for the Aviculoids among
the Pelecypoda, tracing all to one genus, JV^iicula, and has treated the
Echinoidea in the same way, tracing them by the use of Agassiz's law to
the senus JBothriocidaris.

Although the evidence is perhaps less conclusive with reference to the
ancestor of Cephalopoda as a whole, this class has furnished the means
of showing the action of this law in smaller groups with great accuracy.
It has been possible to trace the origin of a number of smaller groups to
single ancestors within the class by carefully studying the correlations of '
the epembryonic stages with the adults of the same group that have pre-
ceded them in time, and this study has also led to further discoveries. It
has been found that the new characters were first introduced in the later
stages of ontogeny, usually in the full-grown stage ; then, as old age

218 PROCEEDINGS OF THE AMERICAN ACADEMY.

approached, certain losses of the characters of the adult took place,
or, if additional growths were acquired, these were of a peculiar kind.
These senile stages had been noticed by D"Orbigny and Quenstedt,
but these authors did not attempt to show that any correlations existed
between any stages of the ontogeny and the gradations occurring in the
full-grown forms during their evolution iu time, or what is called phylo-
geny. The oldest stage of the shell in Cephalopoda, Brachiopoda, and
Pelecypoda is commonly marked by a series of retrogressive changes,
which have been fully described elsewhere. These changes have a
similar nature to those found in the old age of man, but they are more
noticeable because they are recorded in the permanent characters of
the hardened shell. The old man returns to second childhood in mind
and body, and the shell of the cepbalopod has in old age, however dis-
tinct and hin'hly ornamented the adult, very close resemblance to its own
young. This resemblance is a matter of form and aspect only, since
there can be no close comparison in minute structure, nor functions
between organs and parts, at these two different ends of life. Such
analogies, however, have their own meaning, and are o-f great importance
when properly translated.

In the first place they show that the cycle of life as manifested in man
is found also in the ontogeny of other animals and more perfectly in pro-
portion to the perfection of the record. They are consequently among
shell-bearing animals, especially those that carry their embryonic shells
and all their subsequent stages of development throughout their lives,
more perfect, more decisive, as well as more obvious, than in animals, like
the Vertebrata, which carry no such burden of hard dead parts, upon and
in which their stages of development are recorded. The cycle of the
ontogeny is, therefore, not only physiological, but it is also a definite
series of structural changes and is often accompanied by transformations
of remarkable and sometimes startling character.

These retrogressive transformations in old age of the shells of Cepha-
lopoda, Brachiopoda, and Pelecypoda have been found to have decided
correlations with the adult characters of species that appear simultane-
ously or later in time. If one traces any group through its evolution
in time, it has, as stated by many authors, a period of rise called the
epacme, a second period of greatest expansion in numbers of forms and
species called the acme, and then usually a movement towards contraction
called the paracme. All three of these terms were first proposed by
riaeckel, who used them largely in a physiological or dynamical sense.
The epacme of any group, large or small, is usually a process of evolu-

HYATT. — ORGANIC CYCLES. 219

tioii by addition of new structures or characteristics based on older struc-
tures and thus leading to greater and greater complication of the primitive
organization. The acme represents the time of greatest complication in
structure and greatest expansion in numbers of forms for any group,
large or small. The paracme is the decline, and this takes place througii
the reduction and actual loss of structures and characteristics that have
been built up by evolution during the epacme. This is no ideal picture,
but a simple statement of the experiences of those paleontologists who
have patiently traced the history of groups through geologic time.
Agassiz's law enables one to follow the epacme of the evolution of a
species, or genus, or order, or larger group, but further correlations be-
tween the cycle of individual life and those in the evolution of its own
genetic group must be sought in the correlations existing between the
older retrogressive stages of the ontogeny and the paracme of each
group.

The importance and peculiar nature of these correlations led me, in
one of my papers, to introduce, for this branch of research, the term
* Bioplastology,' which will be found convenient by those interested
in this class of work.

The following table of terms is useful here to explain the relations of
the cycle of development in the individual to that of the group to which
it belongs.

Terms of Bioplastology explaining the Correlations between Stages
OF the Ontogeny and those of Phylogeny.

Ontogeny or Development. Phijlogenij or Evolution of the Phylum.

Structural «!fn„oo Structural „, t^ • ,

Conditions. ^*^Ses. Conditions. S*^Ses. Dynamical.

I Embryonic Embryo or Foetal I Phylembrvonic

Anaplasis j Nepionic Baby Phylanaplasi.s \ Phy lonepionic Epacme

( Neanic Adolescent ( Pliyloneanic

Metaplasis Ephebic Adult Phylometaplasis Phylephebic Acme

Paraplasis Gerontic Senile Phyloparaplasis Phylogerontic Taracme

The dynamical terms are quoted from Haeckel, and were used by him
to designate the phenomena of the rise and decline of types, and also the
terms anaplasis and metaplasis. He, however, used ' cataplasis ' in
place of paraplasis, which is here preferred on account of the faulty deri-
vation of cataplasis.

He realized the importance of these phenomena, and also the signifi-
cance of the structural characteristics of decline, but did not trace out
the distinct correlations which are claimed as fundamentals in bioplas-
tology.

220 PROCEEDINGS OF THE AMERICAN ACADEMY,

The terms anaplasis, etc., and their correspondents, phylanaplasis, are
the structural correlatives of dynamical terms, epacme, etc., and will be
found useful when the statical jjhenoraena or structures are mentioned or
contrasted with the dynamical phenomena, or with periods of time in
which they occur, since the terms epacme, acme, and paracme also refer
to time. Terms of the ontogeny are placed opposite to their correlatives
in the column of i)hylogenetic terms, but in reading the table it should be
clearly understood that the individual whose life history is represented
by the first three columns is supposed to have been taken from the midst
of those that lived during the acme of the phylum and belonged to a
phylephebic species. In studying the development of such an individual
it has repeatedly been observed that tlie embryo repeated the adult char-
acters of the most ancient representatives of the 2:)hylum, which are here
called, in accordance with this evidence, phylembryonic.

It has also been ascertained that there are full-grown types in the
epacme and acme of groups which correspond to the transient uepionic or
baby stage of those that occur later in time ; these are the phylone-
pionic ; others have similar correspondences with the neanic stages, and
are properly designated as phyloneanic types or forms. The structures
of the ephebic (adult) stage are essentially the ditferentials of the time
and fauna in which they occur, and necessarily have no correlations with
the past. Their relations are obviously and wholly with the present,
except in so far as they represent the consummations of evolution in
structures. The structural changes in the gerontic stage of the individual
are repeated with sufficient accuracy in the adult, and often even in the
neanic stages of types that occur in the paracme of the evolution of a
phylum, so that one is forced seriously to consider whether they may not
have been inherited from types that occur at the acme of the same group.
The fact that these changes occur first in the ontogeny during the geron-
tic stage does not necessarily imply that they first make their appearance
after the reproductive period. No gerontic limit is known to the repro-
ductive time in the lower animals, arid it may well be that the continual
recurrence of gerontic stages in individuals during the epacme of groups
may lead to their finally becoming fixed tendencies of the stock or heredi-
tary in the phylum, and thus established as one of the factors that
occasion the retrogression or paracme of groups. The paracme may also
be considered as occasioned by changes in the surroundings from favor-
able, as they must have been up to acmatic time, to unfavorable during
the succeeding paracmatic period in evolution. Still a third supposition is
also possible, namely, that the type, like the individual, has only a limited

HYATT. — ORGANIC CYCLES. 221

Store of vitality, and both must progress and retrogress, complete a cycle
and finally die out, in obedience to the same law.

All of these views can be well supported, but, whatever may be the
true explanation, it is obvious that there are plenty of paracmatic types,
which, in their full-grown and even in their neanic stages, correlate in
characters and structures with the characters and structures that one
first finds in the transient gerontic stages of acmatic forms of the same
type. These can, therefore, be truthfully and accurately described as
phylogerontic in the phylum.

In other words, one is able to apply gerontic changes in the ontogeny
to the deciphering of the true relations, the arrangement and classification
of forms occurring in the paracme, just as Agassiz's law of palingenesis
can be used to explain the relations of the links in the chain of being
forming the epacme of groups.

The cycle of the ontogeny is, therefore, the individual expression and
abbreviated recapitulation of the cycle that occurs in the phylogeny of
the same stock; and, while the embryonic, nepionic, and neanic stages-
give us, in abbreviated shape, the record of the epacme, the gerontic
stages give, in a similar manner, the history of the paracme.

The difference between the nature of the two records is, however, neces-
sarily as great as between the beginnings and the endings of existence.
The successive stages of the individual are derived from the past, and
simply point backwards along the track traversed by the phylum ; the
changes of the gerontic stage, on the other hand, point to the future,
and are prophetic of what is to come in the decline of the type. The
retrogressive decline of the individual and that of its type are along
parallel lines, and the two are in direct correlation, so that the former
becomes an abbreviated index of the latter.

One of the most useful results of these studies has been the method
of work developed, the mode of study by series. To follow it out suc-
cessfully, one must trace the terms of series from the first or most
primitive grade to the last, through perhaps long periods of time, and, if
upon the same level, through many gradations of structure.

The histologist or embryologist picks out a convenient form here and
there for thoi'ough investigation, but does not seem as yet to see the
importance of the point of view here insisted upon, viz. that the only
method of getting at the correlations of ontogeny and phylogeny is by
following out the history of representative series of genetically connected
embryos, and the same is true of the experimentalist. "While, conse-
quently, their results have been in the highest degree instructive and

222 PROCEKDINGS OF THE AMERICAN ACADEMY.

progressive along other lines of research, they throw no very strong
light on the laws of evolution, and the best modern works on embryology,
zoology, and experimentation neglect the only 2:)roper and efficient mode
of studying one very important side of their subject.

One of the results of this mode of study has been the discovery of the
law of acceleration in the inheritance of characters, or tachyo'enesis.
Thus it has been found that characteristics are inherited in successive
species or forms in a given stock at earlier and earlier stages in the onto-
geny of each member of the series. These characteristics, as a rule,
disappear from the ontogeny altogether in the terminal, or last occurring,
members of a series, and terminal forms thus become very distinct in
their development. This law I habitually illustrate as the crawling,
walking, hopping, skipping, and jumping law.

Auotlier result of this mode of study is the discovery that, in most
genetic series, primitive forms exhibit much greater iiulifference to
geologic changes, persist with comparatively unchanged structures
through longer periods of time than those that occur at the acme of
groups, and paracmatic forms, if widely distributed, are apt to be par-
ticularly short lived, and are very often narrowly localized in origin and
duration. Primitive forms are also less changeable in their ontogeny ;
the adult differs less from either the young or the old than in acmatic
forms. The same is true of phylogerontic forms ; their old age and
youth are less distinct from each other as stages than in acmatic forms.
Primitive forms are less affected by gerontic changes in their ontogeny,
that is, they have shorter old-age stages, than acmatic forms. Parac-
matic forms have much longer old-age or gerontic stages than acmatic
forms.

Lastly, it has been found that at the beginning of the evolution of any
stock the progress was not only very rapid, but the departures in struc-
tures much more marked between the diverging lines of different species,
genera, or families, and so on, than those that subsequently occurred in
any one of these. This rapidity of expansion is also marvellously sudden
in every series near its point of origin, and it is equally so in the whole
animal kingdom, which appears with the larger proportion of all its
principal divisions in the earliest known fossil-bearing rocks. Each series
or type appears to have had a more or less free field, and its first steps
in evolution were obviously not affected by natural selection. Subse-
quently, in the acme of the same series or type, the departures became
less marked, and the divergences took place in less important structures;
in other words, as stated above, the evolution is slower.

HYATT. — ORGANIC CYCLES. 223

Ou the other liuiul, after tlic acme is passed and the paracme sets in,
there is a sensible quickening of evolution during decline.

Phylogerontic forms become more and more numerous, and there are
wider departures in the structures from the acmatic forms than any of the
divergences that occur within the acmatic forms themselves.

The hopping, skipping, and at last the jumping, begins in the extremes
of the series, so that it becomes difficult, as has been shown by the author
in a number of series, and by Cope when giving illustrations of the action
of the law of tachygenesis, to connect one of these extreme forms with
its nearest conireners.

The characters of the cycle in the ontogeny are here again similar to
those of the phylogeny ; thus the final substages of the gerontic stage are
wider departures from the ephebic substages than these are among them-
selves and when compared with each other. The analogy of the old with
the young shows this most conclusively, and the similarity of phylogeron-
tic and pliylonepionic forms in the same stock occurs in the phylogeny.

In fact, there is no end to the homolocjical and analosfical similarities
and parallelisms of ontogeny and phylogeny wherever both are found
complete.

There are types in which the ontogeny is incomplete, as among insects
and other jjurely seasonal animals, and in these it becomes difficult, if not
impracticable, to study the gerontic stages, and thus translate the phylo-
gerontic types if they occur. These same types, and others also, present
difficulties in their larval stages, owing to their indirect modes of develop-
ment, which have been discussed by the author in -'lusecta" and other
publications, and need only be referred to here.

One of the bearings of these researches is of interest on account of the
discussions between biologists, geologists, and mathematicians with regard
to the length of time that life has existed on this planet, and the bearing
of this upon calculations with regard to the age of the earth. It cannot
be assumed that the time ratio was the same during the Eozoic or Pre-
Paleozoic as during the Paleozoic or the Mesozoic, so far as the evolution
of forms is concerned.

The author in other publications has claimed that this must have been
the law, and explained the phenomena as parallel with that which takes
place at the beginning of every series arising in the Paleozoic and
Mesozoic, and also according to Minot's law of growth and other phe-
nomena of the earlier stages in the ontogeny of every animal.

The evidence is very strong that great structural differences were
evolved much more quickly in these early times, and the probabilities

224 PROCEEDINGS ,0P THE AMERICAN ACADEMY.

are that the progressive steps of the evolution of the primitive types of
organisms took place with a rapidity unexampled in later ages. If the
laws of bioplastology are ti-ue, the evolution of these forms must have
occurred more quickly than those of their descendants, except some iso-
lated phylogerontic types and phylopathic forms.* Man being the most
remarkable of these phylogerontic types, we can at once realize what
this statement means. If his remains in all their vastness were a part of
geological history, and could be contemplated separated from the artificial
halo of idealism, it would be seen that they were the direct results of the
law of tachygenesis acting upon the basis of a simian organization and
exceptional, as compared with other phylogeronts only because of this ex-
ceptional basis.

All inferences with reference to the length of time that life has existed
upon the earth are consequently defective, since, as far as known to the
author, the}^ do not take into consideration this law of variable pro-
portions of time in evolution of organisms at different stages in their
history.

* The plirase ' evolution by saltation ' has been used for the sucWen appearance
of divergent types by several authors, first by Dr. W. H. Dall; but this seems to
me to be simply a mode of expressing a general fact, or series of facts that occur
everywliere, and in all series more or less through the action of the law of tachy-
genesis.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. No. 11. — Apuil, 1897.

CONTRIBUTIONS FROM THE PHYSICAL LABORATORY

OF THE MASSACHUSETTS INSTITUTE OF

TECHNOLOGY.

XLVIIL — 7W^ VISCOSITY OF MERCURY VAPOR.
By A, A. NoYEs and II. M. Goodwin.

Intestigations on Light and Heat, made and published wholly or in part with Approprlation

from the rumford fond.

VOL. XXXII. — V)

THE VISCOSITY OF MERCURY VAPOR.
By a. a. Noyes and H. M. Goodwin.

Presented October 14, 1896, by Charles R. Cross.

The uncertainty which attaches to the specific heat ratio of gases as a
means of distinguishing between monatomic and polyatomic molecules
has been recently made evident by the extended discussions of the sig-
nificance of that property in connection with the atomic weights of argon
and helium. It is therefore of great interest to investigate other proper-
ties which may be expected to be related to the atomicity of the molecule.
Of such properties those dependent on the volume or cross-section of the
molecules seem most promising. We have therefore undertaken the
investigation of one of these, the viscosity or internal friction, in order
to determine whether a marked difference in its value exists in the case
of gases with monatomic and those with polyatomic molecules. To this
end we have made comparative measurements of the viscosity of hydro-
gen, carbon dioxide, and mercury vapor at the boiling temperature of the
last named substance.

According to the Kinetic Theory of Gases the viscosity coefficient has
the theoretical significance expressed by the following equation,*

77 = - NmLc,

in which N is the number of molecules in the unit of volume, m the
mass of a single molecule, L the mean free path, and c the mean ve-
locity. Moreover, the free path L is dependent solely on the number of
molecules A^ and the mean cross-section 5' of a single molecule, or its
sphere of action. f

* O. E. Meyer, Kinetische Theorie der Gase, 1st ed., pp. 130, 139.

t Ibid., pp. 206, 218. The symbol Q used by the author represents the total
cross-section of all molecules in the unit of volume, and is therefore evidently
equal to Nq.

228 PROCEEDINGS OF THE AMERICAN ACADEMY.

whence it follows that

1 mc

'^ ~ 4 77^2 y

or for any two different gases,

miC, mnC

But, since for any two gases at constant temperature,

the above proportion may be simplified to the following equation :

Hl = Vi .M. (1)

from which it is evident that the relative mean cross-sections of the
molecules of the two gases are readily calculated from their molecular
weights and viscosity coefficients. It was thought by us that monatomic
molecules might prove to be much smaller than polyatomic ones, since it
seems a 'priori not improbable that the spaces between the atoms of the
latter are large in comparison with the dimensions of the atoms them-
selves. The experiments to be here described show, however, that no
marked distinction exists between monatomic and polyatomic gases in
this respect.

Experiments on the viscosity of mercury vapor, and especially on the
effect of temperature upon it, have been already made by S. Koch, * who
calculated that at 300° the volume of mercury molecule is 4.4 as great
as that of the hydrogen molecule. As this calculation was not based on
direct comparative experiments made by passing the two gases through
the same capillary, but was an indirect one involving the measurements
of different experimenters, and the dimensions of the capillaries used by
them, it seemed desirable to subject the matter to further investigation in
the direct manner indicated. Moreover, the author does not discuss the
significance of his result in its bearing on the relative magnitude of atoms
and molecules.

The method used by us in determining the relative viscosity consisted

* Wied. Ann. Pliys. Clicm., XIX. 857, 1883.

NOYES AND GOODWIN. — VISCOSITY OP MBUCURY VAPOR. 229

in measuring the quantities of the different gases which under a constant
difference of pressure passed in a given time through the same capillary
kept at a definite constant temperature. O. E. Meyer* has derived the
following formula for calculating the viscosity coefficient of a gas from
its rate of flow through a capillary tube :

"^ 16 A. joiKi

where A is the length and R the radius of the tube, t the time, ;3i the
pressure at which the gas enters and p2 at which it leaves the tube, and
Vi the volume of the transpired gas measured at the pressure p^. In
case of comparative experiments made with the same capillary on two
different gases, the following proportion holds true :

Vi '■ V-2

(pr - p^'h h , {po^- pz-y.h^ (2)

Wl »2

in which 7ii, n^, represent respectively the number of gram molecular
weights of the two gases transpired, — since n is proportional to the pro-
duct p V.

The apparatus and experimental method that we employed were
necessarily quite different from the usual ones, and they will therefore
be briefly described. The capillary used in the most complete series
of experiments consisted of a glass tube about 74 cm. in length and
0.34 mm. in internal diameter (determined by measuring the volume of a
known length by means of mercury). A smaller capillary about 49 cm.
in length and 0.22 mm. in diameter was used in a preliminary series.
The capillary was bent in the manner shown in Figure 1, except that, as
actually constructed, it was made much more compact. To its ends were
fused pieces of ordinary glass tube, as shown in the figure : one of these
was provided at the point A with a ground glass joint. The capillary
was placed in a vertical position in a heavy steel cylinder (see A, Fig. 2)
30 cm. high, 2.8 cm. internal diameter, having a small orifice at the side,
through which the ground joint protruded for a distance of about one
centimeter. The capillary was held in position in the orifice by packing
with loose asbestos. Although the capillary was vertical, the influence
of gravity was eliminated by reason of the fact that the ascending and
descending parts were made equal in length. The top of the cylinder
was closed by an iron plate screwed down with a nut, N; the nut and

* Fogg. Ann., CXXVII. 269.

230 PROCEEDINGS OP THE AMERICAN ACADEMY.

the plate were perforated in the centre, and into the latter was welded
an open iron tube B, projecting 25 cm. in length and 1-i cm. in diameter.
The cylinder was covered, except on the bottom, with a jacket of asbes-
tos about 5 cm. thick, and the projecting tube B, which was to serve as a
condenser, was wound with spirals of copper wire to increase the cooling
surface. Pure mercury was placed in the cylinder and boiled vigorously
by means of a number of lamps beneath. The capillary was thus kept
at the boiling temperature of mercury under atmospheric pressure. No
regard was paid to the variations of temperature arising from changes in
barometric pressure, as their effect would evidently be entirely negligible.

Any desired difference of pressure at the two ends of the capillary
was attained by inserting a tube in the ground joint and connecting it
with a large air reservoir, R, which was itself connected through the
cock ^S" with a suction pump and with an open mercury manometer, M.
The gas or vapor entered at the other end of the capillary always under
atmospheric pressure. The whole apparatus in the form used for
measuring the rate of transpiration of the mercury vapor is shown in
Figure 2.

In making the experiments, the rate of flow of the mercury vapor was
first determined in the following manner. While the cylinder was being
heated, carbon dioxide was forced through the capillary to prevent the
condensation in it of liquid mercury and the formation of its oxide.
After the mercury was boiling actively, and its vapor entirely enveloped
the capillary, as shown by a mercurial thermometer inserted into the tube
B (Fig. 2), it was connected with the suction pump, and mercury vapor
drawn through for half an hour. The carefully ground end of a
weighed bulb, W, was then inserted in the ground joint, its other end
being connected by means of a clamped rubber tube, C, with the air reser-
voir, in which the desired reduction of pressure had been produced. At
a definite moment the clamp G was opened and the time noted. As the
volume of the condensing bulb W was very small compared with the
volume of the air reservoir, no sensible change in the pressure was thus
produced. The mercury vapor was found to be completely condensed in
W about two to three centimeters from the ground joint. It was found
that a slight and unavoidable leakage * through the ground joint occurred,
and it was therefore necessary to readjust the pressure occasionally. It

* In tlie case of the mercury experiments no error could arise from this source,
as the leakage was inward. In tlie case of those of carhon dioxide and hydrogen,
it was proved by bLank experiments that the amounts of carbon dioxide and water
which leaked in were less tlian one per cent of the total weiglit.

NOYES AND GOODWIN.

VISCOSITY OF MliRt'UUY VAPOU. 231

Fig. 1.

Fig. 2.

232 PROCEEDINGS OF THE AMERICAN ACADEMY.

could easily be maintained constant to 0.2 or 0.3 mm., or even closer.
After a sufficient time, usually sixty minutes, the clamp was closed, and
at a noted instant the bulb was removed and subsequently weighed.
Check experiments were made in this way at each of two or three dif-
ferent pressures.

The capillary was now removed from the cylinder and the opening
^(F'ig- 1) carefully closed by fusion. A glass tube long enough to project
beyond the upper end of B (Fig. 2) was fused into the end G, and the
capillary was then ready for the experiments with carbon dioxide and
hydrogen. It was replaced in the cylinder as before, and the glass tube
projecting through B connected through suitable wash-bottles with the
gas generator. The carbon dioxide was made in a Kipp generator by
the action of dilute sulphuric acid on lumps of pure fused sodium car-
bonate, and was dried by passing through two Allihn gas wash-bottles
containing strong sulphuric acid. The hydrogen was prepared from pure
Bertha zinc and dilute sulphuric acid, was washed with caustic soda solu-
tion and dried by sulphuric acid, as in the case of the carbon dioxide.
In order to maintain the gas entering the capillary at atmospheric pres-
sure, a T-tube was inserted between the wash-bottles and the capillary,
and its perpendicular arm was turned downwards and caused to dip into
sulphuric acid barely below its surface. The cock of the generator was
opened sufficiently to cause the gas to bubble out steadily through the
sulphuric acid.

The transpiration measurements were made as in the case of mercury.
The carbon dioxide flowing through in a definite time was determined by
absorption in weighed tubes filled with lumps of soda lime. The hydro-
gen was burnt by passing it over hot copper oxide contained in hard glass
tubes from which the air was previously displaced by carbon dioxide, and
the water collected in weighed calcium chloride tubes.

The results are presented in the following table. In the first column
is given the symbol of the substance ; in the second, the atmospheric
pressure pi ; in the third, the diflference in pressure (jOi — jOo) ; in the
fourth, the time t expressed in hours ; in the fifth, the weight w in grams
of the substance weighed ; in the sixth, the mean weight transpired in
one hour as computed from the separate check experiments ; and in the
last, the quotient obtained by dividing this weight by the molecular
weight m of the substance, the time, and the pressure function (pi^-po^).*

* In the calculation of tliis quantity tlie snme mean value of ;i, wafs used in all
the experiments of each series, namely, 760 for those with the smaller capillary,
765 for those with the larger.

NOYES AND GOODWIN. — VISCOSITY OP MERCURY VAPOR. 233

(See expression (2), page 229.) As the headiugs indicate, the first of
these two series of experiments was made with the smaller, and the sec-
ond with the larger capillary. It will be seen that in the second series
two sets of determinations are given for mercury vapor : the first of these
was made before and the second one after the experiments with carbon
dioxide and hydrogen, and their agreement shows that the capillary had
undergone no change by stoppage or otherwise during the course of the
experiments with it.

SERIES I.
Small Capillary.

jh

Pi - Pz

t

IV

to : t

w X lO")

Hg

753
753
756

200

U

1 1

3

0.738
0.740
0.745

0.494

93.6

755
755

300

0.685
0.686

0.686

93.6

754
754

400

0.834
0.831

0.833

93.9

cOo

752
752
752

200

|1

0.237
0.239
0.357

0.238

205

759

759
76G

300

0.329
0.327
0.324

0.327

203

766

766
759

400

0.400
0.396
0.396

0.397

202

Attention may be first called to the agreement of the values of the
]a^<t column in the case of the ti'anspiration of the same substance under
different ditFerences of pressure, thus proving that the effect of pressure
is in close accordance with that required by the formula, and conse-
quently that the capillaries are of sufficient length and small enough
bore to give the true values of the viscosity coefficients.

Of the two series of experiments the first one made with the smaller
capillary is to be regarded as the less reliable by reason of the fact that,
owing to an accident to the capillary, check experiments with mercury

234

PROCEEDINGS OF THE AMERICAN ACADEMY.

SERIES II.
Large Capillary.

Pi

P1—P2

t

IV

w : t

w X low

<TO(Pi2-p22

Hg

769

769

150

1.548
1.548

1.548

374

769
769
769

300

2.763
2.764
2.739

2.755

873

CO.,

765
765

150

0.704
0.706

0.705

774

765
765

300

1.267
1.264

1.265

779

H2

766
706

150

0.557
0.557

0.557

1517

766
766

300

1.009
1.006

1.008

1496

Hg

766

766
764

300

1

2.728
2.746
2.743

2.739

371

vapor after completion of those with carbon dioxide could not be made
as in the former case, in consequence of which it is not certain that a
stoppage did not occur in the course of the series. While there was
nothing to indicate that such was the case, yet on account of its ex-
tremely small bore it was much more liable to stoppage than the larger
capillary. We consider the experiments with the latter to be certainly
accurate within two per cent.

Nevertheless the agreement between the results of the first and second
series is perhaps as close as could be expected with capillaries so different
from each other in character. The relative viscosity coefficients of the
different gases were calculated from the values of the last column in the
table by means of formula (2). The results are as follows.

First Series.

V^ = 2.17.

'/CO,,

NOYES AND GOODWIN. — VISCOSITY OF MERCURY VAPOR. 235

Second Series.
^ = 0.08. '!h^ = 4.04. V^ = 1.94.

The relative values for mercury and carbon dioxide agree within about
four per cent. That of the second series is, however, more reliable, for
the reason above stated, and will be used in the subsequent calculations.
It may be stated that the value of the ratio T/cog ■ Vnz ^* ordinary temper-
ature is 1.72,* somewhat smaller than that found by us at 357°, a result
which is in accordance with the greater temperature coefficient of carbon
dioxide established by several investigators.

The corresponding values of the relative mean cross-sections as cal-
culated by formula (1) are

?li = 1.02; ^-2?- = 2.48.

That is to say, the average cross-section of the mercury molecule or
atom is very nearly the same as that of the carbon dioxide molecule, and
is about 2^ times as large as that of the hydrogen molecule. This last
result does not differ very greatly from that (2.68) corresponding to the
relative molecular volumes of mercury and hydrogen as calculated for
300° by Koch.

These results indicate that atoms and molecules are of the same order
of magnitude, and that the spaces between the atoms within the mole-
cule, if any exist, are not large ip comparison with those occupied by the
atoms themselves, and consequently the viscosity of gases, or any other
property which like it is dependent only on the size or form of the
molecules, is not adapted for distinguishing between monatomic and
polyatomic molecules.

In considering the significance of the above values of the cross-section,
the different masses of the various molecules, to be sure, ought not to be
entirely disregarded. The mercury and carbon dioxide molecules have,
as we have seen, the same cross-section, and therefore, assuming both to
be of the same general form, they occupy the same volume. The mass
of the former is, however, 4.55 times as great as that of the latter. The
density of the mercury molecule is consequently greater in this same
proportion. But this difference is not marked enough to make it neces-

* O. E. Meyer, Kinetische Theorie der Gase, p. 142.

236 PROCEEDINGS OP THE AMERICAN ACADEMY,

sary to attribute it to free spaces within the carbou dioxide molecule.
For it is not improbable that the inherent density of massive atoms like
those of mercury may be considerably greater than that of light atoms,
such as those of carbon and oxygen.

In closing, we desire to point out that the principle here established,
that atoms and molecules are of the same order of magnitude, and that
no considerable free interatomic spaces exist within the molecule is in
accordance with the remarkable fact that the molecular cross-section of
most comparatively simple molecules is approximately an additive prop-
erty calculable from certain constant values of the atomic cross-section.*
This fact would be unintelligible, were the principle not correct ; for if
considerable space existed between the atoms, it is not to be supposed
that those spaces would be the same in entirely dissimilar molecules, —
that, for example, the space between the hydrogen and chlorine atoms
in hydrochloric acid would have any relation to the space between the
atoms in the elementary gases hydrogen and chlorine.

* See 0. E. Meyer, Kinetische Tbeorie der Gase, p. 209.

Rogers Laboratory of Physics,
September, 189G.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. No. 12. — April, 1897.-

CONTRIBUTIONS FROM THE CHEMICAL LABORATORY
OF HARVARD COLLEGE.

ON THE ACTION OF AMMONIA UPON CUPRI-
AMMONIUM ACETOBROMWE.

By Theodore William Richards and Robert Jay Forsythe.

ON THE ACTION OF AMMONIA UPON CUPRI-
AMMONIUM ACETOBROMIDE.

By Theodore William Ricuards and Robert Jay Forsytiie.

Presented May 13, 1895.

It is well known that there is a large class of cupriammouium com-
pounds which contain two molecules of ammonia for each molecule of the
cupric salt. The compounds of this class are -by far the most stable of
the cupriammonium compounds ; they are usually permanent in the air,
and many of them may be heated above 100° without decomposition.
They have, as a general rule, the power of absoi'bing more ammonia if
exposed to an atmosphere of this gas ; but the additional ammonia is ex-
pelled with the greatest ease. For this reason the compounds containing
little ammonia have been assumed in the series of papers of which this is
one to be the normal cupriammonium compounds, and any extra ammonia
in any other class of compounds has always been designated in the name
of this new class. For instance, " tetrammon-cupriammonium bromide "
indicates Cu(NH3),Br . 4 NH. or CuBiv . 6 NHg.

To the class of stable normal cupriammonium compounds belongs the
salt Cu(NH3)2BrC2H302, which was recently discovered by one of us.*
The possible absorption of more ammonia by this double salt promised to
be a question of much interest ; and it is with this question that the
present paper is concerned.

A weighed amount of finely powdered pure cupriammonium aceto-
bromide was introduced into a weighed glass tube, so drawn out and bent
that the part containing the salt could be immersed in a freezing mixture.
Through this tube was passed dry ammonia until the substance, which
absorbed a large quantity of gas, remained constant in weight. In two
experiments the weights were as follows : —

(1) (2)

(2) Weight of substance taken =0.9904 1.4801

Weight of substance found = 1.2157 1.8090

Gain in weight 0.2253 0.3289

* Berichte der deutsch. ch. G,, XXV. 1492.

240 PROCEEDINGS OF THE AMERICAN ACADEMY.

The weights gained were equal to 22.7 per cent and 22.2 per cent
of the original weight, while a gain of 3 NHg would have corresponded
to 21.51 per cent. The error of this result is not greater than might
have been expected, considering the fact that the deep blue powder was
extremely hygroscopic, and that it was necessary to weigh the tube while
this was still ice-cold, in order to avoid loss of ammonia. A single
analysis, although not exact, served to confirm sutRciently the formula
of the resultinor substance.

(3) 1.809 grams of the substance required 29.6 cubic centimeters of
normal acid to neutralize the ammonia contained in it. This corresponds
to 28.2 of ammonia instead of the theoretical 30.3. The loss occurred
during the transferrence from the tube to the retort, for the substance is
excessively unstable.

The new substance, empirically CuBrC2H30o5NH3, dissolved easily in
water with the formation of a fine deep blue solution, which lost ammonia
upon exposure to the air, and soon deposited a basic salt of copper.
Exposed in a dry state to dry air, the substance lost ammonia with great
rapidity, and turned distinctly green in color. The only green compound
which could be formed under these conditions was Cu(NH3)2Br.2, and
hence the color change was proof that the ammonia had decomposed the
original cupriammonium acetobromide according to this reaction : —

2 Cu(NH3)2BrCoH302 + 6 NH3 = Cu(Br)26 NH3 + Cu(CoH30.).,.4NIl3.

Upon exposure to the air the compound CuBroGNHg, which has re-
cently been discovered in this laboratory,* is known to lose ammonia
with conversion into the likewise newly discovered CuBr2 . 2 NH3.

The other compound represented above, Cu(C2H30o)4 NH3, has not
heretofore been made. Foerster, who tried to prepare it, was unable to
make cupric acetate absorb anything like its full complement of ammonia,
because he did not cool the mixture. f The only step necessary for the
final proof of the reaction given above was then the preparation of
diammon cupriammonium acetate, Cu(NH3)2C2H30.2 . 2 NHg, by the action
of ammonia gas upon cupriammonium acetate at very low temperatures.

Under conditions precisely like those of Experiments 1 and 2, the
two following syntheses were made : —

(4) (5)

Weight of Cu(NH3)2(C2H302)o taken = 0.4380 1.0324

Weight of substance formed = 0.5160 1.2082

Gain in weight 0.0780 0.1758

* Berichte der deutsch. cli. G., XXIII. 3790. t Ibid., XXV. 3416.

RICHARDS AND FORSYTHE. — ACTION OF AMMONIA. 241

The gains were then 17.8 per cent and 17.0 per cent of the weight of
substance taken, while the formula Cu(C2H302)o4 NH^, requires a gain
of 15.8 per cent. Here again the error is due to the great hygroscopic
power of the resulting compound, — which absorbed all of the traces of
water in the ammonia, — as well as to the fact that the tube had to be
weighed while very cold. Since the errors in an analysis of the final
compound must necessarily, as before, be in the opposite direction to the
errors just cited, the following analysis was made: (6) 1.208 grams of
substance required for neutralization 18.75 cubic centimeters of normal
acid, thus having contained 26.5 per cent of ammonia. The theoretical
amount corresponding to Cu(NH3)2 (C2H302)o2NHs is 28.0 per cent.
Greater exactness than this would have demanded great elaboration of
experimental detail, on account of the excessive instability of the com-
pound.

These analyses and syntheses afford conclusive evidence that the com-
pound diammou cupriammonium acetate exists, and that it, together with
the most highly aramoniated cupriammonium bromide, is formed by the
action of ammonia upon cupriammonium acetobromide.

VOL. XXXII. — 16

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. No. 13. — May, 1897.

ON OBTAINING METEOROLOGICAL RECORDS IN THE
UPPER AIR BT MEANS OF KITES AND BALLOONS.

By a. Lawrence Rotch.

ON OBTAINING METEOROLOGICAL RECORDS IN THE

UPPER AIR BY MEANS OF KITES AND

BALLOONS.

By a. Lawrence Rotch.

Presented January 13, 1897.

A KNOWLEDGE of the pliysical conditions which prevail up to the
highest cIo«d levels, live to nine miles above the earth, is of great im-
portance to meteorologists, who until recently have been studying prin-
cipally the conditions existing near the floor of the aerial ocean, and
from that standpoint have endeavored to formulate the laws which
control the i^ressure, temperature, humidity, and currents in the great
volume of air above them. Continued and systematic observations on
mountains in different parts of the world latterly have contributed
much to our knowledge of the approximate conditions of the atmosphere,
under various circumstances, up to a height of more than three miles
above sea level, but the mass and surface of the mountain, even when
this is an isolated peak, influence very considerably the surrounding
air. Recognizing, then, the value of the determination of the true
conditions of the free air, let us consider what methods are available
for this investigation, which must necessarily be sporadic and of shorter
duration than if conducted on mountains. In the writer's opinion, free
balloons with aeronauts cannot be recommended on account of the large
cost in money, and sometimes the loss of life, which attend their frequent
use, while without artificial aids to respiration the aeronaut cannot rise
much above five miles. Captive balloons, with observers, have been
used in England, and more recently, with self-recording instruments,
in Germany, but their height is limited to about two thousand feet by
the weight of the lifted cable, and a wind which is sufficient to overcome
tlieir buoyancy drives them down and occasions violent shocks to the sus-
pended instruments. A kite-balloon on trial in the German army is
intended to combine the advantages of a kite and a balloon, but the cost
and the moderate height attainable render it inferior to the simple kite

216 PROCEEDINGS OF THE AMERICAN ACADEMY.

for meteorological researches, except during calms which sometimes occur
at the earth's surface, but rarely exteud aloft.

There remain kites aud unmanned balloons, both recording graphically
and continuously the chief meteorological conditions, and these it is my
intention to describe in this paper. The recent development of the kite
for meteorological purposes has taken place in the United States, while
the use of the automatic balloon for obtaining data at very great altitudes
has hitherto been confined to Europe.

Kites appear to have been first applied in meteorology by Alexander
Wilson, of Ghisgow, who in 1749 raised thermometers attached to the
kites into the clouds.* Three years later, Franklin performed in Phila-
delphia his celebrated experiment of collecting the electricity of the thun-
der-cloud by means of a kite.f Although kites haVe served a variety of
purposes, their first systematic use in meteorology was probably in Eng-
land, between 1883 and 1885, when E. D. Archibald mad« differential
measurements of wind velocity by anemometers raised by kites fifteen
hundred leet.t In 1885, A. McAdie repeated Franklin's kite experiment
on Blue Hill, using an electrometer,§ and in 1891 and 1892 made simul-
taneous measurements of electrical potential at the base of Blue Hill,
on the hill, and several hundred feet above it with kites as collectors. 1|
The invention of light-weight self-recording instruments made it possible
to obtain graphic records in the air by means of kites, and after W. A.
Eddy iiad introduced tailless kites into this country, and had attached a
minimum thermometer in 1891,^ a thermograph reconstructed of lighter
materials by S. P. Fergusson, of the Blue Hill Observatory, was raised
on August 4, 1894, 1430 feet above the hill.** It was no doubt the first
instrument, recording continuously and graphically, to be lifted by kites,
and it permitted simultaneous observations to be obtained in the free air
and near the ground. This method of studying the meteorological con-
ditions of the free air has ever since been in regular use at the Blue Hill
Observatory ; but notwithstanding the general interest which has recently
been aroused in kites, it is not known by the writer that meteorographs
have elsewhere been raised by them.

* Trans. Roy. Soc. of Edinburgh, Vol. X. Part II., pp. 284-286.
t Sparks's Works of Benjamin Franklin, Vol. V. p. 295.
t Nature, Vol. XXXI.

§ These Proceedings, Vol. XXI. pp. 129-134.

II Annals Astr. Obs. Harv. Col., Vol. XL. Parts Land II., Appendices A and C
H Am. Met. Journal, Vol. VIII. pp. 122-125.
** Il.id.,Vol. XI. pp. 297-303.

ROTCU. — EXPLORATION OF UPPER AIR. 247

The details of the work, as now carried on at Blue Hill, are as fol-
lows. The kites, which have no tails, are of Eddy's Malay, or of liar-
grave's cellular ty[)cs, the former presenting a convex surface to the wind,
and the latter two {)air of superposed planes, each pair being connected
by side planes. In addition to the two leading kites, others are attached by
independent cords to various points of the line, which is a steel music wire,
0.033 inch in diameter, having a tensile strength of three hundred pounds,
and weighing fifteen pounds per mile. The meteorogniphs are composed
mostly of aluminium and weigh less than three pounds each, the one con-
structed by J, Richard, of Paris, recording on a single clock cylinder atmos-
pheric pressure, air temperature, and relative humidity, while that made
by Mr. Fergusson similarly records wind velocity and air temperature.
One of these instruments is hung to the wire between two kites, in order to
insure its safety in case of breakage of the wire or of one kite, or the fail-
ure of the wind to support the latter. The wire is coiled upon the drum
of a windlass, which may be turned by two men, and a measuring device
registers the amount of wire uncoiled, while the angular elevation of the
meteorograph, when not hidden by clouds, is observed from time to time
with a surveyor's transit at the windlass or at the ends of a base line.
From these data, or from the barometric record, the altitude of the
meteorograph is calculated. Kites may be flown in all kinds of weather,
whenever the wind's velocity is between fourteen and thirty-five miles an
hour ; and since on Blue Hill the average velocity is more than eighteen
miles an hour, days are frequent when flights are jDOSsible.

Probably the greatest elevation yet attained by kites, and certainly the
highest level to which kites have lifted a meteorograph, is 8,740 feet above
Blue Hill. This was accomplished, October 8, 1896, by the aid of nine
kites, having a total area of 170 square feet, which gave a maximum pull
at the ground of about 100 pounds.* The meteorograph remained during
several hours higher than a mile, and good records of the indications of
the barometer, thermometer, and hygrometer were brought down. More
than one hundred records of atmospheric pressure, temperature, and rela-
tive humidity of the air, or wind velocity, at intermediate heights up to
the extreme altitude just stated, have been obtained, and they are being
discussed for publication with the Blue Hill observations for 1896, in the
" Annals of the Astronomical Observatory of Harvard College." A few
general conclusions may be mentioned. At the height of about a mile the
diurnal changes of temperature in the free air nearly disappear, although

* Science, Nov. 13, 1896, p. 718.

248 PROCEEDINGS OF THE AMERICAN ACADEMY.

in fair weather the days are damper than the nights. " Cold and warm
waves " commence in the upper air, as is proved by the temperature de-
creasing faster than normal, or even increasing abruptly, with altitude
before the fall or rise of temperature commences at the earth's surface.
Several ascents through clouds have shown the air above them to be
usually warmer and drier than the air below. Kites furnish a ready
and accurate method of measuring the heights of certain low and
uniform clouds, which could not easily be measured otherwise in the
daytime. It is interesting to note that this method was used by Espy,
about 1840, to verify his calculations of the height at which conden-
sation begins.* Changes of wind direction in the different air strata are
determined from the azimuths of the kites, and this change sometimes
amounts to 180°. The wind velocity usually increases with altitude,
and vertical currents commonly prevail near cumulus clouds. During
high flights the wire is s*^^rongly charged with electricity, but no measure-
ments of its kind or potential have lately been attempted.

The writer is glad to acknowledge his indebtedness to his assistants,
Messrs. Clayton and Fergusson, who have devised and constructed im-
proved kites and apparatus, and during his absence have taken entire
charge of the work. To them and to another assistant, Mr. Sweetland,
is largely due the success which has been attained in this novel branch of
research. For still higher ascents there will be required a steam engine
to operate the windlass, and a meteorograph with a lower pressure scale.
With these appliances, for whose purchase a grant has been asked from
the Hodgkins Fund of the Smithsonian Institution, it is probable that
records can be obtained three miles above Blue Hill, and possibly
higher.

To reach much higher altitudes, unmanned free balloons, or " ballons
sondes " as they are called, have been considerably used both in France
and Germany. These balloons, which carry self-recording apparatus, rise
until equilibrium is attained in the rarefied air. When they lose their
buoyancy and fall to the earth, most of them have been recovered, with
the instruments and records uninjured, by the senders, who have been
notified by the finder of the place of descent, which is often at a great
distance from the starting point. The altitudes are calculated from the
barometric pressure, according to Laplace's formula, but the impossi-
bility of knowing the mean temperature of the whole mass of air makes
the determination inexact. Theoretically, in order to ascend ten miles

* Philosophy of Storms, 1841, p. 75.

ROTCH. — EXPLORATION OF UPPER AIR. 249

above the earth, where the pressure is about one ninth that at the earth,
the balloon must lift itself from the ground when one ninth filled with
gas. Therefore a relatively large balloon is required, and its initial
velocity of ascent is great, because it is found advantageous to fill the
gas-bag completely. The greatest difficulty has been to protect the ther-
mometers from insolation, and to insure records being made, notwith-
standing the great cold to which the instruments are exposed.

The first systematic experiments of the kind were made in Paris, in
1893, by G. Hermite, who was later associated with G. Besan^on.
There have been six high ascents from Paris of the three balloons
called V Acrophile. The second one of the name had an envelope of
gold-beaters' skin, with a capacity of 6,3G0 cubic feet, which when nearly
filled with coal-gas gave an initial lifting power of 235 pounds, in ex-
cess of its own weight of 49 pounds, and the instruments and screens,
which weighed 12 pounds. With this balloon, in October, 1895, at an
approximate height of 46,000 feet, a temperature of — 94° Fahrenheit
was recorded, which is the lowest noted in a balloon, and probably the
lowest natural temperature observed on the earth. The average decrease
of temperature was 1° Fahrenheit for 320 feet of height. The instru-
ments used are of the well known Richard type, and have been tested in
a chamber whose pressure and temperature are lowered to the limits
which it is expected may be reached by the balloon. They are placed
below the balloon in a wicker tube six feet high, lined with silvered paper
to ward ofi the sun's rays. It is believed by Hermite, that during the
rapid ascent of the balloon the draught of air through the tube is suffi-
cient to neutralize the heating of the enclosed air by the sun. It is
admitted that when equilibrium is nearly reached this may not be true,
and that the temperature recorded near the highest point may be too
high. To avoid freezing of the ink the registration is now made on
smoked paper, and to protect the instrument from shocks it is hung by
springs in a closed basket, which is itself suspended in the tube already
mentioned. An apparatus for obtaining samples of air at high altitudes
has been carried by the balloon, but as yet without success, owing to
difficulties in hermetically closing the receiver after the air has entered,
since mechanically closing the inlet tube and sealing it by heat generated
chemically have each proved ineffectual at great heights.

By means of a grant from the German Emperor to the Deutsche
Verein zur Forderung der Luftschiffahrt, R. Assmann, A. Berson, and
others in Berlin, have been able to carry on an extensive series of mete-
orological investigations with manned balloons, and also with a captive

250 PROCEEDINGS OF THE AMERICAN ACADEMY.

and a free balloon, both equipped with self-recording instruments. The
latter, called the Cirrus, of 8,830 cubic feet capacity, when inflated with
coal-gas had a lifting force of about 290 pounds, besides its envelope
weighing 93 pounds, and the meteorological apparatus weighing nearly
6 pounds. This is more complicated than the French instruments, since
the registration is photographic, and a continuous ventilation of the alco-
hol thermometer in Assmann's aspiration aj^paratus is effected by allow-
ing a weight to drive the aspirator. Even with these precautions, the
temperatures are probably too high, and the registration is often defective.
There have been seven fli";Iits of the Cirrus, one of the hisrhest occurring
in September, 1894, when the uuprecedentedly low barometric pressure
of about two inches of mercury was recorded, giving a computed height
of 60,500 feet. The lowest tempei-ature, which was registered at a
somewhat less altitude, was not below — 88° Fahrenheit, giving rise to
the supposition that the thermometer was heated by insolation. Hence
the average decrease of temperature appears to have been but 1° in
409 feet. This balloon rose from Berlin with the great velocity of about
30 feet [ter second, and travelled 5G0 miles in an east-northeast direction
at a velocity of 83 miles per hour.

For some time past negotiations have been in progress between the
French and the Germans for simultaneous ascents of unmanned balloons
at night, using identical instruments, whereby the errors due to insolation,
and the discrepancies which might be attributed to different instruments,
would be avoided. By this co-operation the simultaneous conditions of
the upper air over a wide extent of country can be ascertained, just as
these conditions near the earth's surface are daily obtained at the meteoro-
logical stations in the different countries. The desired result was brought
about by the International Meteorological Conference which met last
September in Paris. Resolutions were adopted favoring scientific ascents
with manned balloons, as well as simultaneous flights of unmanned regis-
tration balloons in the different countries. The successful use of kites at
Blue Hill to lift self-recording instruments over a mile into the air, led to
expressed desire that similar experiments should be tried elsewhere. An
international committee was appointed to carry out these resolutions,
consisting of Messrs. de Fonvielle and Hermite for France; Assmann,
Erk, and Hergesell for Germany ; Pomortzeff for Russia ; and the writer
for the United States. In accordance with the first named resolutions, a
flight of four manned and four registration balloons occurred in France,
Germany, and Russia on the night of November 13-14 last. Owing to
hurried preparations, only the registration balloon liberated from Paris

ROTCH. — EXPLORATION OF UPPER AIR. 251

reached a great height ; but in presenting a summary of the results to
the French Academy * E. Mascart, the director of the French Meteoro-
logical Ollice, remarks that there is reason to hope that this international
co-operation will contribute valuable data to our knowledge of the varia-
tions of temperature and wind in the upper atmosphere.

As the American representative of the International Aeronautical
Committee, the writer hopes that in this country a similar exploration
of the high atmosphere with registration balloons will be attempted,
and he is now preparing an estimate of the cost to submit to the trustee
of the Hodgkins Fund. Since it should supplement his own researches
with kites which are described first in this paper, he has taken the occa-
sion to bring the subject of free registration balloons to the attention of
the Academy.

* Comptes Rendus, Vol. CXXIU. No. 22, pp. 918, 961.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. No. 11. — April, 1897.

THE ENERGY CONDITIONS NECESSARY TO PRODUCE

THE RONTGEN RAYS.

By John Trowbridge.

Investigations on Light and Heat, made and published wholly or in part with Appropriation

FROM THE KuilFORD FUND.

THE ENERGY CONDITIONS NECESSARY TO PRODUCE

THE RONTGEN RAYS.

By John Tkowbridge.

Presented March 10, 1897.

This paper is a preliminary study of the conditions which exist in
hif^hly rarefied media under discharges of electricity, conducted by means
of the large storage battery of the Jefferson Physical Laboratory.

The value of a large storage battery for the study of the discharge of
electricity through gases has long been recognized, and such batteries of
1,800 to 2,000 cells have been constructed by Zehnder, by Quincke,
and others. Quincke devotes a large portion of a recent article* to
a description of the details of construction of such a battery of 1,200
cells. The battery of the Jefferson Physical Laboratory consists of
10,000 cells, and it is of such practical construction that I believe a
detailed account of it will be of advantage to those contemplating the
installation of a similar one. When I was considering the cost of such
a battery, Professor B. O. Peirce, my colleague, expressed the opinion
that dry wood would serve perfectly well for insulating material ; and
the mechanician of the Laboratory, Mr. G. "W. Thompson, coinciding in
this opinion and deprecating the use of vulcanite or any of the forms of
insulators in the market, on account of the loss of insulation due to sur-
face action, I decided to adopt wood for the supports of the cells of the
battery.

There are 10,000 cells, each one consisting of a test tube 5 J inches
long and 25 mm. diameter, shown in elevation and plan in Figure 1
and Figure 2. The plates of the cell are strips of lead 1 mm. thick,
12 mm. broad, which have been run through a peculiar mill to give
them a corrugated surface. The strips do not reach to the bottom of
the tubes, in order to avoid short circuiting due to a possible falling
off of the peroxide of lead. They are separated by rubber bands, B, and

* Annalen der Physik und Chemie, No. II. p. 417, 1896.

256 PROCEEDINGS OF THE AMERICAN ACADEMY. .

the terminals of the cells are made of thick lead wire, which are led
through wooden supports far from any possibility of corrosion where
they connect to the copper wires of the main circuits. The unit of the
battery consists of three test tubes, mounted in holes bored in blocks
of wood 5 X 5 X 15 cm. {U, Figs. 1 and 2). These blocks of wood
are boiled in parafline, and the tubes are held upright by means of
paraffine, which is poured into the holes in which the tubes are inserted,
these holes being made slightly larger than the tubes. On solidification
the tubes are held upright. There are twenty of such blocks on each
shelf of one upright case, thus making sixty cells to a shelf ; and there are
seven shelves to a case, making four hundred and twenty cells to a case.
On the back of such cases are arranged knife-edge switches, which enable
me to arrange the cells in multiple or in series, — to employ one shelf or
the entire number. At the extremity of each row of six cases is a switch
board with similar switches, which enable me to use one case or any com-
bination of the entire number of six cases. These switch boards consist
of dry whitewood, the insulation of which has been found to answer per-
fectly. Since the practical success of such a large battery depends upon
the ease with which the cells can be charged in multiple and discharged
in series, I have represented the scheme of connections of two shelves,
sixty cells on each case. The remaining cells on successive shelves of
each case are connected in a similar manner. A and A' (Fig. 3) repre-
sent the terminals of each line of twenty cells. Their lead terminals are
led to the back of the case, where there are switches which have pivots
at the points E and D. These switches, revolving about the pivots E
and D, connect the terminals of the cells to a wire running through the
pivots E, and through the points L ; in which case the cells are in mul-
tiple. If, however, the points N' and N", are thrown out by revolving
the switches about E, and the points L of the switches are connected to
N' and N" , the cells are thrown into series. This operation can be
quickly accomplished, the points N' and the points N" moving to-
gether, and also the points L moving together about the pivots D. In
arrangement for series, therefore, the current goes from D to N', then
to A, then through the row of twenty cells to A', thence to D, to N',
and to A, again to A', and then to the next row of switches correspond-
ing to the next shelf, and so on.

A similar plan of switch boards has been erected for each half of the
entire battery, consisting of five thousand cells, in order that one half of
the battery may be used, or the entire number of cells. At the ex-
perimental table to which the terminals of the battery are led there are

TROWBRIDGE. — RoNTGEN RAYS.

257

M Qm

F., I

^ — v^ , ^

A , A-

aT^ ^^ 1 A-

A A-

u u

u

F.o2

c

JoL

m

K^

M

VOL. XXXII. — 17

258 PEOCEEDINGS OF THE AMERICAN ACADEMY.

two distilled water resistances which enable one to control the strength
of current of the battery. The cells are charged in multiple with only
twenty in each branch of the divided circuit; and they are always left
on the multiple circuit when not in use. The resistance of each cell is
about one fourth of an ohm, and the electromotive force about 2.1 volts.
A study of the voltage of each case by means of a voltmeter showed that
the voltage could be closely estimated by knowing the number of cells.

The Plante rheostatic machine, by means of which one can rise from
the voltage of the battery (20,000) to 500,000 volts,* is a practical
modification of that described by its author. Instead of mica plates, the
mechanician of the Laboratory, Mr. Thompson, selected glass plates 8 X 10
inches, one tenth of an inch thick : these were coated with tinfoil to
within one inch of the edges of the plate. At first it seemed doubt-
ful whether glass of this thickness would stand the high voltages to which
it would be subjected. Preliminary experiments, however, showed that
this thickness of glass would stand a steady stress of twenty thousand volts,
— although it would break down under much less voltage arising from
alternating stresses. I have employed thirty of such glass condensers or
Franklin plates, which are charged in multiple by the battery of from five
thousand to ten thousand cells ; that is, all the coatings of one side of
the plates are connected with the — pole of the battery, and the coat-
ings of the other side with the -f pole of the battery. The change
from multiple to series is accomplished by a series of brushes which are
arranged on two vulcanite rods. These brushes are connected with the
coatings of the glass plates. A revolving drum provided with pins and
connecting wires is driven by an electric motor, and serves rapidly to
change the connection of the condensers from multiple to series.

Starting from the point reached by De La Rue and Muller, about
15,000 volts, the Plante rheostatic machine shows that the length of
spark is closely proportional to the electromotive force. On plotting
lengths of sparks as ordiuates, and JVV as abscissas, H being number
of Franklin Plates, a straight line is obtained.

The method I have employed for studying the energy conditions in
Crookes tubes, and in the production of electrical discharges in general,
may be termed the damping of an additional spark method, or the com-
parison of resistances by the estimation of the damping of electrical oscil-
lations-t The electrical circuit is provided with two spark gaps. One

* Comptes Rendus, Tom. LXXXV. p. 794, Oct. 29, 1877.

t Damping of Electric Oscillations, These Proceedings, Vol. XXVI.

TROWBRIDGE. — RONTGEN RAYS. 259

of these is placed in a gas or under the conditious which are to be ex-
amined, while the other is protographed according to Feddersen's method
by a revolving mirror. A previous study of the behavior of various
electrodes had led me to select cadmium for the terminals of the spark
gap which are examined by the revolving mirror. The light from cad-
mium is very actinic ; and with carefully pointed terminals shielded from
the currents of air produced by the revolving mirror, very sharp photo-
graphs of electrical oscillations can be obtained. The resistance of a
spark in air or in rarefied media can be estimated by this method to
one half an ohm.

The revolving mirror which I employed was the one I have used in
previous researches.* It is a glass concave mirror of ten feet focal
length, silvered on its concave surface, and corrected for the use to
which it was put. The improvements in electric motors and storage
batteries enables one to obtain great steadiness of rotation. The accora-
panying figure represents the revolving miri-or, together with the electric
motor and the chronograph, and the diagram explains the parts. Fig-
ure 4 is the apparatus in elevation, and Figure 5 in plan. M represents
the mirror, J5J the motor, (7 the chronograph, S the stylus, ^the cutting
tool for obtaining electrical contact at the instant the mirror reaches a
definite angle in its revolution. This cutting tool passes through a strip
of type metal, T. This strip is adjusted up and down by the screw 0,
and is adjusted in a direction at right angles by the spring bolt P. The
catch R releases P, at any desired moment, by means of a tension on a
string connected to it. The stylus S is drawn along guides by a string
which is connected through clockwork to a weight, W. A fan, F, serves
to control the movement of S. The stylus is also released by a catch
which can be tripped at any moment by the operator.

The camera consisted of a box ten feet long, made of a trussed frame
covered with black cloth. This camera is shown in plan and eleva-
tion in Figure G. 3f is the revolving mirror, G the camera, S the
spark gap, P a partition which extends one quarter way through the
interior of the camera to shield the photographic plate at iV from the
direct light from the spark gap S. The photographic room in which this
camera is placed was about twenty-five feet square, and was provided
with shutters of orange fabric. In one corner of the room is the de-
veloping closet. The same room contains a mercury pump for exhaus-

* Phil Mag., 1894 ; Am. Journal of Science, 1894 ; Velocity of Electric Waves,
Am. Journal of Science, 1895.

260 PROCEEDINGS OF THE AMERICAN ACADEMY.

tion, and the wires from the storage battery are led to a switch board
in this room. A connecting room contains a Rowland grating.

With the facilities at ray command, I set to work to investigate the
conditions which seem to be necessary to produce the Rontgen rays.
I have said that my previous experience in studying oscillatory dis-
charges had led me to select cadmium for the material of the terminals
of a spark gap. The light is liighly actinic and with not too strong
discharges, and vrith sharply pointed terminals well defined images of
even minute evidences of oscillations can be obtained. I have also ex-
perimented with various developers, and finally adopted Rodinol. This
developer works with great intensity, and quickly brings out the latent
images. The study of rapid electrical oscillations enables one to esti-
mate with considerable accuracy the merits of different developers for
instantaneous photography. With a definite electrical circuit consisting
of a known capacity (Leyden jar) and a known self-induction, together
with constant rate of speed of the revolving mirror, one can obtain
the time of exposure ; and the number of oscillations brought out by the
various developers is a guide to their relative values.

Having at my command a battery giving a voltage of twenty thousand,
with an internal resistance of only one quarter of an ohm per cell,
and therefore capable of giving a very powerful current, I first studied
the behavior of Crookes tubes, which were directly connected to the
terminals of this battery. I speedily discovered that no Rontgen raj-s
could be obtained with a voltage of twenty thousand. On strongly heat-
ing the Crookes tubes they were filled with a pale white light, which
showed very faint bands in the green when examined by a spectroscope.
After a short interval, the entire strength of the battery appeared to be
manifested in the tubes, the electrodes became red-hot, the medium
apparently broke down, and offered little resistance to the battery cur-
rent. This white discharge showed, even at its culminating point, no
Rontgen rays, and appeared to be of the nature of a voltaic arc discharge.
I then employed the Plante rheostatic machine. I found tiiat at least
one hundred thousand volts were necessary to produce the Rontgen
rays, and that they were produced more intensely as I increased the
voltage, — certainly to the point of five hundred thousand volts. In order
to ascertain how great a loss occurred in this machine on discharging in
series, I investigated the length of spark obtained by varying the number
of Leyden jars in this machine. Experiments to be described later in
this paper had shown me that increase of electromotive force diminished
the resistance offered to a spark in air. The higher the electromotive

TROWBRIDGE. — RONTGEN RAYS. 2G1

force, therefore, employed to charge the jars, the less resistance would
occur at the brushes where the change from multiple to series occurred,
and therefore the less the loss of energy due to Joule's heat. I there-
fore investigated the length of spark obtained from the Plante machine
by continuing the curves plotted by De La Rue and Miiller,* and found
that, starting with 10,500 volts, and increasing the Ley den jars pro-
gressively, the length of sparks plotted as ordinates and the rise in
voltage (iV^O gave a straight line which was an extension of those
obtained by them. Furthermore, as will be shown later, differences
between points and planes, and points and paraboliform surfaces dis-
appeared with high voltages. In order to ascertain if the discharges
through Crookes tubes when the Rontgen rays were apparently pro-
duced most strongly were oscillatory, I first placed a Geissler tube in
tlie circuit with the Crookes tube, and carefully observed the appearances
at the two electrodes of the Geissler tube. The electrodes were quite
alike in appearance, and indicated an oscillatory discharge. I then re-
placed the Geissler tube by a small spark gap, and photographed it in
the rapidly revolving mirror.

The photograph showed ten clearly defined oscillations, with a period of
about one ten-millionth of a second, with the Crookes tube and the cir-
cuit 1 employed. Furthermore, applying the method of estimating re-
sistances by the method of damping, I found that the resistance of
the rarefied medium was less than five ohms. The energy therefore at
the moment of the emission of the Rontgen rays was not far from
3,000,000 horse power acting for one millionth of a second. The
Crookes tube which I employed was of the focus tube pattern (King's
College, London). I also employed a Crookes tube with an aluminium
mirror of about two centimeters' focus. The resistance of this tube to
the discharge was approximately the same as that in which the mirror
had a focal length of about five centimeters. There seemed to be no
advantage in shortening the distance between the anode and the kathode
in a Crookes tube. Struck by the fact that the distance between the
electrodes did not appear to make any appreciable difference in the re-
sistance offered by the Crookes tube to oscillating currents, I replaced
the tube by a spark gap in air of six inches in length, and photographed
the spark in another gap iu air in the same circuit. This latter gap was
6 mm. in length. The photographs showed on the average the same
number of oscillations, both when the additional spark gap was six inches

* Proc. Roy. Soc, Vol. XXXVI. p. 151, 1883-84.

262 PROCEEDINGS OF THE AMERICAN ACADEMY.

in length and when it was one inch in length. I found, moreover, on
increasing the electromotive force, that the resistance of the sparks in air
decreased. By quickly drawing apart the terminals of the large storage
battery of ten thousand cells, a flaming arc discharge can be produced
in air of nearly three feet in length. Rhigi has shown also that sparks
from an electrical machine or Leyden jars can be greatly increased in
air by quickly drawing apart the spark terminals. We thus see that
very little resistance is encountered by more than quadru2:)ling the length
of discharge.

I next placed the additional spark gap in a receiver connected with an
ordinary air pump, and studied the resistance offered by rarefied air at
the point when long ribbon-like white disruptive discharges can be ob-
tained. This point is about 100 cm. pressure in the receiver. The
resistance of such discharges of about six inches in length in a receiver
containing air at this pressure is two or three ohms more than sparks
one quarter of an inch in air. The latter offer a resistance of from
two to three ohms. On measuring by the damping method the resistance
of sparks of different lengths in the receiver at this pressure, no difference
in resistance could be perceived between a spark of six inches in length
and one of three inches in length. The method would have detected a
difference of half an ohm.

The additional spark gap was next placed in a chamber of air which
was compressed to four atmospheres. This amount of compression
made no difference in the resistance to the disruptive discharges. It
would be interesting to push this research to the amount of compression
reached by Professor Dewar in the case of liquid oxygen. He has ob-
tained a dielectric constant for liquid oxygen of 1.45. When this dielec-
tric, however, is broken down by a disruptive spark, I am inclined to
believe that it would show little more resistance than air under the same
circumstances.

The additional spark gap was next placed in hydrogen gas generated
at atmospheric pressure by electrolysis. No appreciable difference at
this pressure was noticed between the resistance offered by this gas and
air at the same pressure. The length of spai-k which could be obtained
with a given voltage was somewhat more in hydrogen than in air. It has
been shown by Professor T. W. Richards and myself that the resistance
of gases at low pressures diminishes with the increase of electromotive
force.* I was interested to test this question by the employment of the

* Am. Journal of Science, April, 1897.

TROWBRIDGE. — RONTGEN RAYS. 263

high voltages which can be obtained by the use of the Plantc machine.
A Geissler tube containing hydrogen at 1 mm. pressure was placed be-
tween the terminals of the Plante machine together with a spark gap.
This tube would give only four half oscillations with ten thousand volts
and gave twenty half oscillations with approximately 300,000 volts,
the electrical circuits beinij the same in the two cases.

I next placed the additional spark gap in the flame of a Bunsen
burner. It is well known that the spark length can be greatly increased
in an atmosphere of heated air. On photographing a spark produced in
the same circuit, the resistance appeared to be slightly increased by the
heated air; doubling the length of the spark, however, made no change
in the resistance that was encountered in the heated medium. The phe-
nomenon was very analogous to that observed in the receiver exhausted
to 100 cm. It is well known that lightning follows currents of heated
air, striking into barn doors from which arise hot air currents from hay,
and passing into chimneys from which issue heated air. The striking
fact is presented that the medium breaks down more easily when it is
heated ; but it offers during the oscillations of the liehtninor somewhat
more resistance than cold air.

I was interested to discover whether heating the air in which the spark
in the primary of the Thomson Tesla transformer is produced would
have any marked effect on the high tension spark of the secondary of
such a transformer. It was immediately evident that such heating of the
air was detrimental. The high tension sparks immediately ceased to
jump at the extreme sparking distance of the terminals. Following this
train of thought, I next placed a spark gap of the primary of the above
mentioned transformer between the poles of a powerful magnet, giving a
field of certainly ten thousand lines to the centimeter. It is well known
that when such a field is excited the primary spark appears to be blown
out with a loud report, and a great increase of length of spark is obtained
in the secondary of the transformer. Applying the method of estimation
of resistance by damping to an additional spark gap which was placed in
this magnetic field, I found no difference in resistance offered to the
spark, whether the magnetic field was excited or not, or whether the
spark jumped across the direction of the magnetic lines or in the same
direction. Is it possible that, the ether being already under a mag-
netic stress, the addition of a powerful electrostatic stress serves suddenly
to break down the ether? It is well known that a blast of air imi-
tates the action of a magnetic field, and produces also a great increase of
spark in the secondary circuit of a Thomson Tesla transformer. It prob-

264 PROCEEDINGS OF THE AMERICAN ACADEMY.

ably does so by blowing out the voltaic arc which tends to form. It may
be that the electrodynamic repulsion compels the oscillations of the spark
not to follow, so to speak, the voltaic arc and its current of heated air.
It seems as if the oscillations of the spark were true voltaic arcs, and that
the electrodynamic repulsion blows these out. There are, however, just
as many oscillations in the magnetic field as outside of it. The field
exerts no influence on the number of oscillations, or on their appai'ent
duration. The loud report which is produced when a spark is formed in
a magnetic field, notably when the primary circuit of an ordinary Ruhm-
korf coil is broken in a strong magnetic field, may indicate a sudden
stress in the medium ; in the case of the Crookes tube, the highly rare-
fied medium within it would effectually jirevent our hearing a similar
report.

In order to see if the radiations from a Crookes tube emitting Ront-
gen rays could produce any effect upon the primary spark of the Thomson
Tesler transformer, I produced it near a Crookes tube, and examined it
by the method of damping. No change in resistance could be perceived,
and no eflFect was observed upon the length of the spark produced by the
secondary of such a transformer. In the next place, I resolved to deter-
mine whether differences in the materials of the spark gap made any
appreciable difference in the resistances observed in disruptive discharges.
I accordingly employed terminals of platinum, iron, aluminium, brass,
cadmium, zinc, and carbon. No difference arising from difference of
metals could be noticed. These experiments confirm the results obtained
by Rhigi * and by De La Rue and Hugo MuUer.f

Moreover, no difference of resistance between spheres, between pointed
terminals, or between a point and a plaue^ could be perceived. With
powerful discharges such differences disappear. The employment of a
powerful storage battery together with a Plante rheostatic machine shows
conclusively that the dischai'ge in a Crookes tube, when on the point of
emitting the Rontgen rays most intensely, is an oscillatory one, and
that such discharge encounters a resistance less than five ohms. An
estimate of the great amount of energy thus developed in an exceedingly
small interval of time can be obtained if we suppose that Ohm's law holds
for individual oscillations. This reservation is an important one, for the
investigations I have described in this paper show that a discharge of
six inches in length encounters no more resistance during its oscillations

* Nuovo Cimento, [2.], Vol. XVI. p. 97, 1876.
t Phil. Trans., Vol. CLXIX. Pt. 1, p. 93, 1878.

TROWHRIIXIE. — RONTGEN RAYS. 265

than one of two inches in length. In popular language, it can be main-
tained that a discharge of lightning a mile long encounters no more
resistance than one of a foot in length. Ohm's law does not hold for
electrical discharges in air and rarefied gases. It is well known that a
voltaic arc can be started in a vacuum. My experiments lead me to
believe that in every case the arc is started by a spark which breaks down
the medium, and the arc follows. I am led to believe that electrical
oscillations are of the nature of voltaic arcs, and that the discharges in
Crookes tubes are voltaic arcs. I am thus forced to the conclusion that
under high electrical stress the ether breaks down and becomes a good
conductor.

Jeffeuso.v Physical LAnoRAxoRT,
Harvard University.

Proceeding's of the American Academy of Arts and Sciences.
Vol. XXXII. No. 15. — May, 1897.

CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF
THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD
COLLEGE, E. L. MARK, DIRECTOR, No. LXXIX.

A MEASURE OF VARIABILITY, AND THE RELATION

OF INDIVIDUAL VARIATIONS TO SPECIFIC

DIFFERENCES.

By Edwin Tenney Brewster.

CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE

MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD

COLLEGE, E. L. MARK, DIRECTOR, NO. LXXIX.

A MEASURE OF VARIABILITY, AND THE RELATION

OF INDIVIDUAL VARIATIONS TO SPECIFIC

DIFFERENCES.

By Edwin Tenney Brewster.

Presented by E. L. Mark, April 14, 1897.

Problem.

This paper, prepared under the supervision of Dr. C. B. Davenport,
deals with an inquiry into the relation between those small variations
which distinguish individuals of the same group, and those larger differ-
ences which separate species and genera.

For the prosecution of such an inquiry, it is, first of all, necessary to
devise a method of measuring variability. Quetelet ('46), Stieda ('82),
Galton ('91), and Weldon ('93) have shown that variations in organisms
follow the well known laws of the distribution of error. Thus the ordi-
nary methods of treating problems in error of observation may be made
to furnish a measure of variability ; it is upon these methods that the
methods of this paper are based.

Method.

Any measurable quality of an object has a value, which is expressible
by a number. A series of such numbers, expressing the varyino; value
of a quality throughout a group of similar objects, will have a mean,
about which the quantities are arranged in accordance with the law of
distribution of error. Such a series is, for the present purpose, essen-
tially like a series of slightly erroneous observations of a single quantity
distributed about the true value. What, therefore, would be the proba-
ble error in the latter case, is a measure of variability in the former.

Suppose such a series of numbers obtained by measuring some single
quality in each individual of a natural group of organisms. Let there be

270 PROCEEDINGS OF THE AMERICAN ACADEMY.

n individuals, and let di, (L ... d,^ represent the difference between each
number of the series and the mean of all.
Then

0.8453 (c?i + «?2 ••• d„)

's/n {ii — 1)

which is the common working formula for probable error,* is a measure
of the variability of the given quality in this particular group.

Or, in symbols,

T. 0.8453 2 d

Foe

^Jn (n — 1)

where V stands for variability, 2 c? = 6?i + c?2 . . . (f^,, and oc may be read

" is measured by." This is, appproximately, Galton's Q.
Obviously,

Vcc ^^. (Formula 1.)

^/n {n — 1) ^ ^

In Formula 1,

Vn {71 - 1) = V»' - n = yjn" (\-]-\ = nJ I --.

n

V n

Consider the expression, — .

n

Here 2 c? is the sum of all differences between single numbers of the

series and the mean of the series. Since these are n in number, — is

n

the average deviation of single values from the mean value.
Let ad he the symbol for this average deviation. Then,

V (X. ad — . (Formula 2.)

V n

Suppose, however, that the number of cases measured is large ; that
is to say, that n is made indefinitely large. As n approaches infinity,

- approaches 0, and consequently — ^^=^ approaches 1.
n

0^

V n

* See any tex^book treating of such subjects; for example, Merriman ('84), p. 93.

BREWSTER, — RANGE OF VARIABILITY. 271

If, then, n be taken sullicieutly large,

Fa ad. (Formula 3.)

Even if n is not large, Formula 3 may usually be employed in practice.
For in any single investigation n is likely to be constant, — when, as
will often be the case, different qualities of the same individuals are to be
compared it is necessarily so, — but the introduction of a constant factor
will not affect the correctness of the formula.

Practically, a d may be found accurately enough without the hibor of
subtracting each quantity from the mean.

Let

a + -»---- a ■\- d, a -\- c, a -\- b, a — b, a — c, a — d • • • • a — -

be a series of quantities, n in number, distributed according to the law of
probability. The mean is evidently a, and the differences between the
several quantities and the mean are

-, d, c,b,b, c, d, .

"^ +d+c + b + b + c + d.... + l
.*. ad^ —

n
2b+ 2c + 2d .... + n

n

(Formula 4.)

Or, again,

« + y a -\- d, a -\- c, a -\- b. n — b, a — c, a — d, « — k*

being the series of n measurements as before, the sum of all terms greater
than the mean is

(« + 0 • • • • + (« + rf) + (a + c) + (a + 5),

and of these the mean is

n
2

272 PROCEEDINGS OF THE AMERICAN ACADEMY.

In like manner, the sum of all terms less than the mean is

. — d — c — b

na n

n
2

One half the difference between these is

25 + 2c + 2d....n

n

(Formula 5.)

But Formula 5 is identical with Formula 4, and is therefore the formula
for the average deviation.

Hence the average deviation may be found by separating the numbers
into two groups, — one of which shall contain all quantities greater than
the mean, the other all those less than the mean, — and taking half the
difference between the means of each group.

In practice there will usually be several observed values equal to the
mean ; these may be distributed to make the two groups of equal size.
If the number of observed values is odd, one of these mean values may
be neglected.

In order to compare the average deviation of one group of numbers
with that of another which has a diiferent mean, it is necessary to reduce
the two to a common measure. This is most simply done by dividing
each average deviation by the corresponding mean.* It is proposed to
call this ratio the coefficient of variability and to designate it by the
symbol 0. V.

* Tlie justification for this procedure is found in the following considerations:
The relative size of the average deviation of two organs depends very largely upon
the relative size of these organs. Where the mean dimension is large, we expect
a greater average deviation than where it is small. Thus the average deviation
of the stature of adult British males from the mean is about 2 inches. An aver-
age deviation of 2 inches in the length of the nose, in any race, would clearly
indicate a much greater variability in the nose length than in stature. In compar-
ing the variability of two such diverse measures as stature and nose length, it is
better to compare the ratios of tlie average deviations to the mean dimension.
Thus, since the mean stature of adult British males may be taken at 67 inches,
variability in stature may be expressed by the ratio ^^ = .02985. This number
indicates that the average deviation from the mean stature is about three one-
hundredths of the mean stature; which is clearlj' more important than to say that
it is 2 inches. Moreover, this method of expression has the advantage that it is
independent of the unit in which the dimension is measured, whether feet, milli-
meters, grams, degrees, or ergs. — C. B. Davenport.

BREWSTER. — RANGE OF VARIABILITY. 273

In practice C. V. may be found most readily by separating the given
numbers into two groups, as discussed above, and finding the mean of
each. The difference between these two means, divided by their sum, is
the C. V. of the structure under consideration.

Application of the Method of Pleasuring Variability to the Problem of the
Relation between Individual Variations and Specific Differences.

Thesis.
While it is generally a<rreed that specific characters are more subject to
strikinsr variations in individuals than are the characters common to allied
species, it is not clear how far this relation extends, I wish to show that
what is true of obvious variations and sports is also true of those minute
differences between individuals which only careful measurements can
detect ; or, in other words, that any measurable quality is, in general,
variable in individuals in proportion as it is a distinguishing character
of the group to which the individuals belong.

Evidence.

Table A is based on the extensive tables of body measurement of
twenty races of men, which are given by Weisbach ('78) in the Appendix
to his *' Koroerraessunsen," etc. Each of the numbers of the first eight
columns is the CV. of a single dimension of a single race, and the ninth
column gives the average for eight races of the C. F,'s of each dimension.
The values of these C. F.'s only approximate to the true values, for the
reason that the number of individuals measured is not sufficiently large
to eliminate all accidents of age and sex.

An examination of Table A shows that these eighteen dimensions are
of nearly the same relative variability in each of the eight races. This
fact is well brought out by Table B, in which the largest number of each
column of Table A is replaced by 1, the next smaller by 2, and so on up
to 18, From Table B it appears that certain dimensions — as, for ex-
ample, the height of the forehead — are always decidedly variable ; and,
on the other hand, it appears that other dimensions — such as the length
and breadth of the head — are more constant.

The last column of Table A, " Mean of 20 Races," gives the coef-
ficient of variability, not of any individuals, but of the means of each of
the twenty races of "Weisbach's tables. That is, the mean value of a
dimension in each race is treated as are the dimensions of individuals in
other columns. This column, therefore, shows the distribution of racial
differences in the same way in which the remainder of the table shows

VOL. XXXII. — 18

274

PROCEEDINGS OF THE AMERICAN ACADEMY.

saoirjj

on JO i|o«ajo

snoiBUrfuiip auau

•rr t — p o !M -^ cc --^ (M c; o 00 Ci cr. c c c. CO
(^1 -M c; 1- '-■0 c -r >-t! cc -r id ■:■: .~ i^ r^ o ui o
c; c. q = I— . ^. w o o o c c o o o c: c o
o o o o o o c o o o o o o o' o o o o

'aSGJ3Ay

o

CI

•suEiaBumoji gj;

^ CO CO X) O lO -M ~ i~ Tti ^ lO -f CO <— ' — X O
(.0 CO O -r O ~. CT —r CO CC CO CO CO O-J CO CC 0^) CO

O O p O C^_ O O O C O O C: O C O O O O

ooooooooooocoo'oo'oo

o

o

C3

CO

o

OD

is

•bjbXSbi^ 03

CC(M ':DOCr:i.OO'MCCO'Mr^-^t:OOt^t—
c: CM -^ oi — CO o — . o-j — ^ en. CO — 1.0 " r;

0-1 CO t~- >ro " X -r- o CO o CO -r oi c : -r c; -oj co

p O t— C; p p C: p O C' O O C C: C O' O p

o o o o o o o o o o o o c c; o o o o

•SMaf 61

o

X

o

Bi

o
o

d

CI.

u

'a

CO

e3

OCO>-Ot-i.O — OC:.— I — oOC'—ii-'O XiO •—!'*<
C w CO -1- 1- P CO 1— ' rO I— (0) -* C ~ .— l^ uO ^^

C'O CO -f -r i-o o co o CO o -I" CO »o cc co co co -f
p p p O' p p p O p p p p o c c c c o
ooooooooo'ooooooc'oo

s

CO

•S'B3[Bn'B3 g

_ _ ^ ■* "M 00 CI O CO t^ CO CO CO

O"— i-H:r)C;t^oc^coi~cooo-)c;-*ooo-io

CO O) -r ~. -f X i-O CO -f cc -ti O'l O -.1' CO O -* r~

a — ~, c:^ s^ c:_ iC^^ C:' ^ c^ c:> c: CD cz C^ a c: a>
o o o o o o o c; o o c o o c o o o o

c

•asaurqo qz

•asauBdBf j;i

CO t^ iC-^c:tOO'— ooo

00 t^ CO -^ CO t^ -JO X O OD f- C' Ol C: C ~ CO -f
tOl X X CO O X O ^r CO cc O uOi -^ O CO iC C'J lO
OO^ipo— 'OCOOOCOOCOOOO

oooooo'odooooooooo'o

X

CO

o X o ^- t^ o oi '-0 1— ' t^ cc X c; C". c; 'O -J •-H
o~i o) 'O lo X ~ -f X o -f -f 0-; cc oi co -f co -fi

COCCOO'-'COOCaCOCCOCrOCOO

o o o o o o o o o o o o o o o o o d

to
o

•esanrcig g

CO ici>-o osco-^"— >ooo>i

1— I ■»tl >— < lO) CO X 05 O: CO OO i-O i-O 1 ~ -1- OC ^ 0-1 CO
(M ri lO C5 "-" I— Ol X 'O CO ^ 0-1 X CO cc 0-1 "O CO

p p p p — . p p p p O C p O C C: O O O

d d d d d d d d d c d d d d d d d d

X

cc

•BABIS fS

-t >0 (M CO CO ^- 05 05 t— X O I- t- C -t< 05 — • C5

^- -f CO -r X 01 i-O ^- "O X I-- f CO "-O OJ ^- *+ 05

CO CO '-O -f CO i^ — r CO -t" o-i -+I -r uo -t> CO -r -f -1'
cooooosoooococoocoo
dddddddddddddddddd

•2 "5; - i i !'■'- J

fct S£

si ?

4^ w^

■3 5

CU 0)

C3

— S

: — 0

^ t^

t^

O O

M W

0) "o

^ ti

01 CU

0. >

0/

O) CJL^ ■ — ^

— =r~ U tl. tjr rt

^ CC — 0/ o ;r- 0)
c > — — ' p- 1-,

Qi

ri C C ^

o o

^=3:2;?'.

;^ fe 5 S 5*1^ t^f^ ^X^t^'^^

O

c:

*->

c

0

e

0

OJ

0

ai

fci

be

<

1— t

CS

BREWSTER. — RANGE OF VARIABILITY.

275

the distribution of individual variations. To sliow the connection between
individual variations and racial ditfei'ences, I take all possible pairs of
coefficients in the first nine columns of Table A and compare them with
the corresponding pairs in the last colunui. When, in one of the tirj;t
nine columns, the member of a pair which is larger is larger in the tenth
column also, it counts one in the line at the bottom of Table A marked
" Cases of Agreement." When, however, the larger member of a pair
in one column is the smaller in the other, it counts one in the line marked
'• Cases of Disagreement." For example, in the column of Slavs, the
G.V. of the head length is 0.0314, that of the head breadth is 0.0345,
the head breadth being the more variable. In the column of racial dif-
ferences also, the C. V. of the head breadth is greater than that of the
head length; therefore this gives in column of Slavs one "Case of
Agreement." On the other hand, for the Slavs, the heights of the fore-
head and of the nose do not agree in relative variability with the means
of the twenty races given in the last column ; consequently this compari-
son gives a " Case of Disagreement."

TABLE B.
Showing Order of Magnitude of C. V. of Table A.*

6

CO

i

o5

1

>

1

1

3

s

w

DQ

?
s

3
O

so
ci

00

CO

-f

c-i

o

Ci

o

CO

>

00

<^l

o

CI

si

18

18

Head length

17

17

18

17

16

18

17

16

Head breadth . .

15

16

16

16

18

13

15

10

17

17

Nose lentfth . . .

5

8

6

3

18

8

1

2

3

3

Nose breadth . .

12

2

5

4

1

7

5

5

4

5

Nose Iieight . . .

2

1

2

1

9

3

3

3

2

1

Forehead lieight .

1

5

1

2

2

1

2

1

1

2

Under jaw length

8

14

10

7

7

5

7

11

9

14

Moutli breadtli

3

3

3

12

14

2

4

4

5

10

Upper face breadth

9

7

4

15

11

14

12

6

11

16

Lower face breadth

18

12

7

14

4

4

6

8.5

7

13

Upper arm length

7

9

9

9

10

9

11

13.5

10

6

Forearm length

11

15

17

11

17

16

9

7

10

15

Middle finger length

4

4

13

13

6

6

16

8.5

6

8

Hand width . . .

10

11

15

10

8

11

14

18

14

4

Upper leg length .

16

10

12

5

15

17

8

13.5

13

12

Lower leg length .

14

18

8

6

5

12

13

12

12

11

Foot length . . .

i;^

6

14

18

12

15

18

17

15

9

Foot breadth . . .

6

13

11

8

3

10

10

15

8

7

* I. e. in column "20 Races," head length has C. V. smallest; head breadth, next
larger; C. V. of nose height is largest.

276

PROCEEDINGS OF THE AMERICAN ACADEMY.

It results, then, from the two lines of numbers at the bottom of Table A
that, in spite of the somewhat inaccurate values of this coefficient, in from
two tiiirds to four fifths of the comparisons, that dimension in any pair
■which is the more variable among individuals of the same race is likewise
the dimension which is the more diverse in the various races. Inciden-
tally it may be noted, too, that the measurements of the face have, in
general, the largest coefficients, and thus the high value for personal
identification which has long been accorded to photographs of the face
alone is justified.

TABLE C.

C. V. OF Various Dimensions of Skulls of a Genus (Lepus) of Rodents.
(Based on Cones and Allen, 77, pp. 222-226.)

Total length

Greatest width

Distance between orbits

Nasal bones, length

Nasal bones, width behind

Nasal bones, width before

Upper incisors to molars

Upper incisors to hinder margin of palate

Upper incisors, height

Upper incisors, width

Length of upper molars

Width between upper molars

Lower jaw length

Lower jaw height

q; O

Cases of Agreement

Cases of Disagreement

0.121

O.O'.iSu

0.0825

0.122

0.150

0.178

0.125

0.0799

0.146

0 146

0.0958

0.14

0.124

0.122

■«
p.

3
P.

1^

in

0.0438

0 0272

0.0691

0.0732

0.0084

0.0700

0.0527

0.0477

0.0364

0.0502

0.0459

0.0348

0.0491

0.0329

55

30

0.0178
0.0239
0.0475
0.0348
0.0327
0.0405
0.0280
0.0238
0.0568
0.0434
0.0393
0.0452
0.0240
0.0174

Q.
CO

s

59

0.0308
0.0255
0.0583
0.0.540
0.0505
0.0582
0.0403
0.0357
0.0466
0.0408
0.0426
0.0400
0.0365
0.0251

60

31

Table C shows for a genus of rodents what Table A shows for races of
mankind. The first column is computed from the means of twelve species
and subspecies of the genus Lepus. This gives, therefore, the relative
value of these dimensions as specific and subspecific differences. The

BREWSTER.

RANGE OF VARIABILITY.

277

remaining columns give the relative variability of the several dimensions
in individuals of the two species. The numbers at the bottom of the table
— " Cases of Agreement " and " Cases of Disagreement " — are found as
in Table A. Table C is much less satisfactory than Table A, because
the differences in variability between different dimensions are here small,
and therefore the relative sizes of the coefficients are more at the mercy
of accident. Even here, however, there is a decided preponderance of
cases in which individual variation is directly correlated with specific
difference.

TABLE D.

C. V. OP Four Measurements of Two Genera of Rodents.
(Based on Miller, '93 and '93*.)

Tail.

Hind Foot.

Ear.

Tail and
Body.

21 Zapus insignia

34 Zapus hudsonius

0.033
0.054

0.027
0.034

0.028
0.070

0.025
0.013

Mean

0.043

0.030

0.049

0.034

Tail.

Hind Foot.

Ear.

Body.

105 Sitomys americanus

90 Sitomys americanus canadensis

0.067
0.070

0.028
0.031

0.050
0.052

0.056
0.049

Mean

0.068

0.029

0.051

0.053

Table D is computed from the measurements given with the descrip-
tion of two new s{)ecies of rodents.* Zapus insignis and Z. hudsonius
are reported to resemble each other closely, but to differ in the length of
the ear. Sitomys americanus and S. americanus canadensis are described
as nearly alike except in the length of the tail. Table D shows that, of
the four dimensions which are recorded for these species, the mean C. V.
of the ear is greatest in Zapus, and of the tail greatest in Sitoinys. The
0. V. in each single species, too, is greatest for the characteristic dimen-
sion, except in the case of Z. insignis. Possibly Z. insignis is the con-

* Miller ('93), pp. 1-8 and 55-70.

278

PROCEEDINGS OP THE AMERICAN ACADEMY.

stant parent species from which Z. hudsonius, with its remarkably high
coefficient for the ear, has separated. At any rate, in each genus, the
characters which mark the species have the highest mean coefficient.

TABLE E.

C. V. OF Measurements of Three Carnivora.
(Measured by the Author.)

Length of nasal bones

Length of frontals

Length of snout, incisors to margin of palate .

Total basal length

Length from posterior nares to occipital foramen
Length of parietal and occipital bones . . .

Length of tooth series

Width between canines

Width between maxillary bones

Width of zygoma

a

0.050
0.038
0.039
0.028
0.024
0.030
0.044
0.041
0.045
0.049

o

■a

0.054
0.056
0.062
0.054
0.058
0.046
0.050
0.055
0.040
0.054

St

0.039
0.031
0.033
0.029
0.026
0.027
0.024
0.045
0.028
0.044

to .

0.052
0.047
0.050
0.041
0.041
0.038
0.047
0.048
0.042
0.051

O to

c S

0.048
0.041
0.045
0.037
0.036
0.034
0.039
0.047
0.038
0.049

Table E gives the G. F.'s of ten dimensions of the skulls of the cat,
fox, and lynx. A comparison of the skulls of these three forms shows
that the representatives of the cat and dog families are very similar, ex-
cept for the length of the muzzle and the shape of the zygoma. A short
muzzle and a wide zygoma are characteristic of all the cats,* and nowhere
is the difference between the skulls more marked than in the length of
the nasal bones. Examination of Table E shows that it is in just these
dimensions — namely, those of the front part of the skull, the zygoma,
and especially the nasal bones — that individuals are most variable ; while,
on the other hand, the " total basal length " and the dimensions of the
back part of the skull, which are alike in the two families, give the
smallest coefficients of variability.

* Flower (70), p. 142.

BREWSTER. — RANGE OP VARIABILITY. 279

Conclusion.

These four cases are the only oues to which I have applied this method
of measuring variability. They are, however, taken entirely at random,
and are in no wise selected cases. The conclusion to which they all
point is that which, on general grounds, seems most likely to be true.
This conclusion is, that there is so intimate a causal connection between
the characters of individuals and those of the allied groups into which
they are combined, that, in proportion as any character is variable in the
individuals of one group, it is different in the allied groups.

Finally, I have to express my great indebtedness for assistance, criti-
cism, and suggestion to Dr, Charles B. Davenport and Mr. Frederick H.
Satford, both "of Harvard University.

WOLFBOROCGH, N. H.,

March 5, 1896.

280 PROCEEDINGS OP THE AMERICAN ACADEMr.

LITERATURE CITED.

Coues, E., and J. A. Allen.

'77. Monographs of North American Rodentia. Rept. U. S. Geol. Surv.
Territories, Vol. XL 1091 pp.
Flower, W. H.

'70. An Introduction to the Osteology of the Mammalia. 344 pp.
Loudon.
Galton, F.

'89. Natural Inheritance. 259 pp. London.
Merriman, M.

'84. A Text-book on the Method of Least Squares. New York.
Miller, G. S., Jr.

'93. A Jumping Mouse (Zapus insignis Miller) new to the United

States. Proc. Biol. Soc. Washington, Vol. VIII. pp. 1-8. April 22.

'93^ Description of a new White-footed Mouse from the Eastern United

States. Proc. Biol. Soc. Washington, Vol. VIII. pp. 55-70. June 20.

Quetelet, A.

'46. Lettres sur la The'orie des Probabilites, appliqude aux sciences
morales et politiques. 450 pp. Bruxelles.
Stieda, L.

'82. Ueber die Anwendung der Wahrscheinlichkeitsrechnung in der
anthropologischen Statistik. Arch. f. Anthrop., Band XIV. pp.
167-182.

Weisbach, A.

'78. Korpermessungen verschiedener Menschenrassen. Berlin.
Weldon. W. F. R.

'92. Certain Correlated Variations in Crangon vulgaris. Proc. Roy.
Soc. London, Vol. LI. pp. 2-21.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. No. 16. — June, 1897.

CONTRIBUTIONS FROM THE GRAY HERBARIUM OF

HARVARD UNIVERSITY.

New Series. — No. XL
By J. M. Greenman.

I. Revision of the Mexican and Central American Species of
Houstonia.

II. Key to the Mexican Species of Liabum.

III. Descriptions of New or little known Plants from Mexico.

^

CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD
UNIVERSITY, NEW SERIES, No. XI.

By J. M. Greenman.

Presented by B. L. Robiason, 14 April, 1897.

I. — REVISION OF THE MEXICAN AND CENTRAL
AMERICAN SPECIES OF HOUSTONIA.

HOUSTONIA, Gronov. (Named in honor of Dr. William Hous-
toun, born in Scotland, 1695, died in Jamaica, 1733.) — Calyx-tube
usually a little compressed ; lobes 4, often with one or more minute
teeth in each sinus, erect or spreadic^, --»ersistent, later becoming widely
separated. Corolla funnelform or salverfbrm, 4-lobed, glabrous or pilose
in the throat; lobes valvate. Stamens 4, inserted in the throat of the
corolla ; filaments short or elongated ; anthers dorsally affixed, oblong or
linear. Disk inconspicuous. Ovary 2-celled ; style terete, slender,
short or elongated ; stigmas 2, linear or linear-oblong ; ovules in each
cell few or many, centrally attached to the median placenta. Capsule
^ to f inferior, more or less subglobose, obovate, obovate-oblong, or
turbinate ; seeds crateriform, acetabuliform, or oblong, peltate and
slightly concave on the hilar surface, usually scrobiculate ; albumen
horny ; embryo clavate. — Gronov. in L. Syst. Nat. ed. 1 ; L. Gen.
no. 70; Benth. & Hook. f. Gen. ii. 60; Gray, Syn. Fl. N. A. i. pt. 2,
24 ; Schumann in Engl. & Prantl, Nat. Pflanzenf. iv. Ab. 4, 27. Poire-
tia, Gmel. Syst. 263. Panetos, Raf. Ann. Gen. Sci. Phys. v. 227, vi. 81.
Chamisma, Raf. in Steud. Nom. ed. 2, i. 776, in syn. — Low slender
annuals, herbaceous perennials, or suflfruticose plants with reciprocally
dimorphous flowers.

§ 1. EuHOusTONiE^. Low herbaccous plants often much branched
from a woody perennial base : leaves ovate, oblong, linear or subterete,
not rigid-setaceous nor acerose-linear : capsule globose, obovate, or obloug-
obovate, more or less compressed and usually emarginate : seeds cra-
teriform, acetabuliform, or oblong, peltate and slightly concave on the
hilar surface, scrobiculate. — Gray, Proc. Am. Acad. iv. 313.

284 PROCEEDINGS OP THE AMERICAN ACADEMY.

* Low slender creeping or erect annuals or perennials : flowers usually small on
axillary or terminal pedicels or disposed in terminal cymes : capsule obcordate-
depressed or oblong, emarginate, one half to three fourths inferior ; seeds deep-
crateriform, open-crateriform, or oblong, peltate and slightly concave on the
hilar surface, usually scrobiculate.

•<- Perennial by filiform rootstocks or creeping stems : flowers axillary or on

slender terminal pedicels.

H. c^ RULE A, L,, Sp. PI. i. 105, accredited to Mexico by Hemsley
(Biol. Centr.-Am. Bot. ii. 30), is essentially a more northern plant, and
probably does not extend into Mexico.

H. serpyllacea, C. L. Smith, in herb. A low prostrate or creeping
perennial : stems quadrangular, suffrutescent : branches ascending, 3 to
5 cm. high, glabrous : leaves elliptic-ovate, short-jDetiolate, acute, gla-
brous on either surface, margin revolute, hispidulous ; stipules 1-2-long-
setiferous with several shorter glandular setse : flowers on short axillary
pedicels : calyx-divisions lanceolate, acute, somewhat foliaceous, 2 to 4 mm.
long ; corolla fuunelform, about 1 cm. long ; lobes above and the throat
pubescent : capsule oblong-obovate, 4 mm. long, one half as broad, about
three fourths inferior ; seeds (6 or more) oblong-rotund or a little oblique,
peltate, slightly concavo-convex, prominently scrobiculate. — Hedyotis
serpyllacea, Schlecht. Linnaea, ix. 599. Mallastoma Shannoni^ Donnell
Smith, Bot. Gaz. xviii. 203. — Jalapa, Schiede ; C. L. Smith at Cruz
Verde near Jalapa, altitude 2,150 m., no. 1486; Chiapas, Ghiesbreght,
no. 814; Guatemala at San Miguel Uspantan, Depart. Quiche, altitude
2,150 m., Heyde and Lux (no. 3176 of Donnell Smith's sets).

•*- -^ Delicate annuals with minute white flowers and erect or spreading corolla-
lobes ; seeds open-crateriform, oblong-rotund with a short hilar ridge.

H. croftise, Britton & Rusby. A depressed annual : stems 1 to 2
cm. long, hirtellous with a few scattered hairs, especially near the nodes:
leaves oblanceolate, 5 to 15 mm. long, glabrous or with a few scattered
hirsutish hairs on the midrib above, margins revolute, ciliate : stipules
scarious, laciniate-dentate : flowers minute on short axillary pedicels :
calyx including the ovary 2 mm. long, hirsute-pubescent ; lobes acute :
corolla white, 3 mm. long ; lobes above short-pubescent : capsule de-
pressed-globose, subdidymous, about one fourth inferior, the free portion
short-hirsute-pubescent : seeds open-crateriform, oblong-rotund with a
short hilar ridge. — Trans. N. Y. Acad. Sci. vii. 10. — San Diego, Duval
Co., Texas, Miss Mary B. Croft, no. 85.

H. parviflora, Holzinger, in litt. A low annual, branching from a
single slender root : branches ascending, quadrangular, minutely hispidu-

GREENMAN. — GENUS HOUSTONIA. 285

lous upon the angles : lower leaves oblanceolate ; upper gradually nar-
rowed to linear, 7 to 17 mm. long: stipules broadly ovate, scarious,
minutely denticulate : flowers minute in the axils near the ends of the
branches : pedicels 1 to 6 mm. long : calyx glabrous ; lobes linear-oblong,
acutish, spreading or recurved, 1-^ to 2 mm. long : corolla white, equalling
or slightly exceeding the lobes of the calyx ; lobes erect, glabrous, about
equalling the tube: capsule three fourths inferior; seeds open-crateriform,
oblong-rotund, with a median hilar ridge. — Collected by J. E. Bodin,
Bound Rock, Texas, 10 March, 1890, no. 24.

•»--»-+- Erect or low-spreading herbaceous perennials, 1 to 4 dm. high : leaves
linear, linear-oblong, or ovate : inliorescence in terminal cymes : capsule
obcordate-compressed, or oblong and a little nerved below, one half to three
fourths inferior.

++ Erect herbaceous perennials : divisions of the calyx acute.

H. PURPUREA, L., Sp. PI. i. 105, attributed to Mexico by Hemsley
(Biol. Centr.-Am. Bot. ii. 30), apparently does not extend so far south.
Gregg's no. 66, cited by Hemsley under the above species, is Houstonia
angustifolia, Michx., var. rigidiuscida, Gray.

H. angustifolia, Michx. Erect, usually much branched from a
woody perennial base : stems strict, quadrangular, glabrous : leaves linear
to oblanceolate-linear, 1 to 5 cm. long, 1 to 5 mm. broad, often hispidu-
lous upon the recurved margins and upon the midrib above : inflorescence
cymose, corymbosely or paniculately branched ; pedicels slender, erect,
2 to 8 mm. long: flowers dimorphous : calyx-lobes 1 to 1|- mm. long,
acute : corolla somewhat funnelform, about 6 mm. long ; tube pubescent
within ; lobes pubescent above : capsule oblong-obovate, about three
fourths inferior, the free portion usually a little exceeded by the persist-
ent calyx-lobes ; seeds slightly concavo-convex, distinctly scrobiculate. —
Fl. i. 85 ; Gray, Syn. Fl. N. A. i. pt. 2, 26. Hedyotis stenophylla^ Torr.
& Gray, Fl. ii. 41. Oldenlandia angustifolia, Gray, PI. Wright, ii. 68.
— Northern Mexico, Texas, northward to Iowa and eastward to Florida.
Mexico, Pringle, no. 2260; Texas, Lindheimer, no. 620; Berlaudier,
nos. 432, 681, 1062, 2413, 2560, etc.; Wright, nos. 239, 240, 1390;
E. Hall, no. 285; Palmer, no. 396; Heller, no. 1G61. Thurber's
specimen, without number, collected at Indianola, seems to be a depau-
perate form of this species.

Var. filifolia, Gray. Stems profusely branched ; leaves linear-
filiform : capsule somewhat smaller. — Syn. Fl. N. A. i. pt. 2, 27. Ol-
denlandia angustifolia, Michx., yrt. flifolia, Chapm. Fl. 181. — Texas,
between Fredericksburg and San Sabra, Thurber, no. 67 ; Florida, at

286 PROCEEDINGS OP THE AMERICAN ACADExMY.

Key "West, Blodgett ; also Curtiss, no. 1137. Drummond's no. 116, and
Havard's nos. 231, 232, from Texas, are somewhat intermediate between
the species and variety.

Var. scabra, Wats. Flowers short-pedicellate : calyx somewhat
liirsute-scabrous : corolla-lobes near the apex externally covered with
hirsute hairs (especially in bud) : capsule rather larger than in the species.

— Proc. Am. Acad, xviii. 97. — Caracol Mountains, Coahuila, Palmer,
no. 410; Monterey, Nueva Leon, Palmer, no. 2116.

Var. rigidiuscula, Gray. From 1 to 3 cm. high: leaves linear
to linear-oblong, 1 to 2 cm. long, 1 to 5 mm. broad, rather rigid: pedi-
cels somewhat shorter than in the species proper : calyx-divisions usually
exceeding the free portion of the capsule. — Syn. Fl. N. A. i. pt. 2, 27,
in part. — Cameras Pass, Coahuila, Pringle, no. 2855; Gregg, no. 66;
Valley of San Luis Potosi, Schaffner, no. 614; Texas, Guadalupe Moun-
tains, Havard, no. 233 ; Rio Brasses, Drummond, no. 73 (Florida, Rugel,
no. 322, from Dr. Gray).

H. mpicola. Low, profusely branching from a woody perennial
base : stems erect, 8 to 14 cm. high, 4:-angled, glabrous : leaves linear,
5 to 15 mm. long, acute or submucronate, rather rigid : stipules triangular,
acute or short acuminate, usually unequally setiferous-denticulate : flowers
dimorphous, sessile or short-pedicelled, disposed in small cymules terminat-
ing the stems and branches : calyx-divisions small, acute, scarcely 1 mm.
long : corolla narrowly funnelform, about 5 mm. long ; lobes pubescent
above : capsule globose, H mm. or less in length, about one half inferior ;
seeds oblong, peltate, slightly concave on the hilar surface, scrobiculate.

— H. fasciculata, Gray, Syn. Fl. N. A. i. pt. 2, 27, as to plant Wright.
Hedyotis (Amphiotis) stenophylla, \ a.r. jmi-vijlora, Gray, PI. Wright, i. 81.

— Crevices of rocks on the San Pedro River, Arizona, Wright, no. 238.
•*-v -w Low, branching from the base : stems spreading or prostrate : divisions of

tbe calyx obtuse.

H. salina, Heller. Low, spreading or prostrate: stems 8 to 10 cm.
long, branching from a woody perennial base, glabrous : leaves oblong-
linear, 5 to 15 mm. long, 2 to 5 mm. broad, essentially glabrous, mar-
gin volute : inflorescence in close subcapitate cymules terminating the
stems and branches : flowers dimorphous : calyx-lobes short, scarcely
1 mm. long, obtusish, glabrous or hirtellous : corolla funnelform, 5 mm.
long, pubescent in the throat ; lobes pubescent above : capsule short-
oblong, a little narrowed below, about three fourths inferior, the free
portion about equalling the calyx-lobes ; seeds as in the preceding spe-
cies. — Contrib. Herb. Frank. & Marsh. ColL no. 1, 96. Houstonia

GREENMAN. — GENUS HOUSTONIA. 287

angnstifolia, Michx., var. Hgidinscula, Gray, Syn, Fl. N. A. i. pt. 2,
27, iu part. — Corpus Christi, Texas, Palmer, without number, collected
in 1879; Heller, no. 1812.

* * Low-tufted leafy species, 2 to 12 cm. high : pedicels reflexed in fruit : cap-
sule one fourth inferior ; seeds open-crateriform, oblong-rotund, with a median
longitudinal ridge on the hilar surface.

•t- Distinctly perennial, multicipital from a deep ligneous root.
H. "Wrightii, Gray. Stems quadrangular, pruinose-puberulent :
leaves linear or the lower narrowly spatulate, thickish, glabrous ; stipules
rather pronounced, broadly deltoid, entire or minutely denticulate : flow-
ers dimorphous, 5 to 7 mm. long, in terminal leafy cymes : calyx-divisions
lance-linear, acute, 1^ to 2 mm. long, glabrous: corolla salver-shaped;
tube 2^^ to 4 mm. long (depending upon the long- or short-styled form
respectively), slightly enlarged above : seeds scrobiculate. — Proc. Am.
Acad. xvii. 202, & Syn. Fl. N. A. i. pt. 2, 26. H. humifusa^ Hemsl.
Biol. Centr.-Am. Bot. ii. 30, not Gray. Hedyotis {Houstonia) humi-
fusa, Gray, PI. Wright, i. 82. Oldenlandia (^Houstonid) humifusa^ Gray,
PI. Wright, ii. 68, excl. syn. — State of Mexico, in meadows on the
Sierra de las Cruces, elevation 3,000 m., Priugle, no. 5744; on the
Serrania de Ajusco, altitude 3,000 m., Pringle, no. 6468; San Luis
Potosi, elevation 1,800 to 2,460 m.. Parry &, Palmer, no. 301 ; in the
vicinity of the city of Durango, Palmer, no. 312 ; Chihuahua, on the
summits of Sierra Madre, Pringle, no. 2284 ; and on dry mountain sides,
Chuchuichupa, in the same State by C. V. Hartraan, no. 705 ; New
Mexico, Wright, no. 241, Thurber, nos. 717, 1121 ; Santa Magdalena
Mountains, Vasey ; near Santa Rita del Cobre, E. L. Greene ; Arizona,
at Fort Whipple, Coues & Palmer, no. 75 ; in the San Francisco Moun-
tains near Clifton, E. L. Greene ; Prescott, Dr. Smart, no. 460 ; Lem-
mon, nos. 2726, 512, 512|^; in the Rincon and Santa Rita Mountains,
Pringle ; at Willow Spring, Palmer, no. 523.

H. rubra, Cav. Much depressed, rather dense, 2 to 10 cm. high:
stems and leaves hispidulous or nearly glabrous : leaves linear, or the
lower oblanceolate-linear, 5 to 25 mm. long; stiixiles short with one or
more setge : flowers \\ to 2|^ cm. long, axillary: calyx-divisions linear,
acute, in anthesis 2 mm. long, in fruit somewhat elongated : corolla red
or purple varying to white ; tube slender : capsule depressed, subro-
tund ; seeds scrobiculate. — Ic. v. 48, t. 474, f. 1 ; Benth. PI. Hartw. 15 ;
Hemsl. Biol. Centr.-Am. Bot. ii. 30; Gray, Syn. Fl. N. A. i. pt. 2, 25.
Oldenlandia {Hoiistonid) ruhra^ Gray, PI. Wright, ii. 68. — Aguas
Calientes, Hartweg, no. 93 ; Valley of San Luis Potosi, Schaflfner,

288 PROCEEDINGS OF THE AMERICAN ACADEMY.

no. 26 ; region of San Luis Potosi, altitude 1,800 to 2,460 m., Parry &
Palmer, no. 300 ; hills and plains near Chihuahua, Pringle, no, 274 ;
Sonora, south of San Luis, Parry, no. 6; Thurber, nos. 319, 707 ; with-
out locality, Coulter, no. 196 ; New Mexico, Fendler, no. 291, Wright,
no. 1119, Palmer, no. 93; Arizona, Wolf & Rothrock, no. 865; in the
Huachuca Mountains, Lemmou, no. 2725, at Bisbee, F. E. Lloyd.

-t- -I- Distinctly annual : stems dichotomously branched : corolla funnelform.

H. hnmifusa. Gray. A low annual : stems dichotomously branched,
hispidulous : leaves linear or the lower oblanceolate, 1 to 1|^ cm. long, 1
to 5 mm. broad, mucronate, hispidulous above, margins revolute : stip-
ules broad, scarious, and setiferous : flowers axillary, rather numerous
above : calyx-divisions setaceous-subulate, 3 to 4 mm. long : corolla
open-funnelform, 6 to 7 mm. long ; lobes pubescent above : pedicels
strongly recurved in fruit : capsule depressed-globose, subdidymous. —
Proc. Am. Acad. iv. 314. Hedyotis {Houstonia) humifasa, Gray, PI.
Lindh. ii. 216. — New Mexico, Thurber, no. 292; Texas, Lindheimer,
nos. 621, 622, 377, 378 ; E. Hall, no. 286 ; on sandy prairies near
Austin, Wright ; Belknap. Sutton Hayes, no. 333 ; Colorado City, Rever-
chon, no. 60 ; on sandy hills and plains near Sweetwater, Curtiss, no.
1140; also specimen cultivated in the Cambridge Botanic Garden in 1849.

» * * A low slender leafy stemmed annual, with small white flowers on spread-
ing axillary pedicels : capsule three fourths inferior.

H. snbviscosa, Gray. A delicate herb, 4 to 15 cm. high : stems
branched from the base : branches spreading, covered with a subviscid
hirtellous pubescence: leaves linear-filiform, 5 to 15 mm. long, glabrous
or hirtellous : stipules short, scarious, minutely setulose-ciliate : pedicels
filiform, 3 to 8 mm. long, horizontally spreading but scarcely reflexed in
fruit : calyx hirtellous-pubesceut ; divisions acute, about 1 mm. long :
corolla salverform, about 3 mm. long : capsule subdidymous, broader
than long ; seeds open-crateriform, minutely scrobiculate. — Proc. Am.
Acad. iv. 314. — Oldenlandla subviscosa, Wright in Gray, PI. Wright.
ii. 68. — Southern Texas, Berlandier, nos. 991, 2421, and Wright, with-
out number.

* * * * Slender diffuse herbs, or perennials, 5 to 25 cm. high : stems more or less

dichotomously branched, often becoming slightly woody below : leaves linear,

subterete or filiform, shorter than the internodes : pedicels erect or spreading

in fruit.

-t- Low delicate annuals : calyx adnate nearly to the top of the ovary : capsule

clavate-turbinate or subglobose.

H. Brandegeana, Rose. Erect or ascending, 5 to 10 cm. high,
freely branching from the base : stems filiform, hispidulous-puberulent :

GREENMAN. — GENUS HOUSTONIA. 2S9

basal leaves narrowly spatulate, 5 to 8 mm. long; the upper gradu-
ally narrowed, hispidulous-puberuleut above, glabrous beneath : pedicels
elongated, filiform: flowers small, 4 to 6 mm. long: calyx-divisions
short, scarcely 1 mm. in length, obtuse : corolla funnelform : capsule
subglobose, Ih mm. long. — Coutrib. U. S. Nat. Herb. i. 70. — Lower
California, La Paz, Palmer, no. 31.

H. aspernloid.es. Gray. Much branched from the base, 5 to 18
cm. high : branches divaricately spreading and ascending, inconspicu-
ously hispidulous-puberulent or quite glabrous : leaves linear or a little
dilated above the middle, 5 to 24 mm. long, glabrous : flowers on slender
filiform pedicels: calyx-lobes lance-linear, li to 2 mm. long, acute,
slightly unequal : corolla funnelform, about 8 mm. long ; tube below pu-
berulent on either surface: capsule clavate-turbinate, tapering below;
seeds oblong, peltate, somewhat concavo-convex, minutely scrobiculate.
— Proc. Am. Acad. v. 158; Rose, Contrib. U. S. Nat. Herb. i. 70.
Hedyotis (Ericotis) asperuloides, Benth. Bot. Sulph. 19, t. 13. — Lower
California, Cape St. Lucas, Xantus, no. 43 ; at La Paz, Palmer, nos. 24,
31"; and by Brandegee, 5 February, 1890, the latter distributed as
Houstonia Brandegeana, Rose.

H. arenaria, Rose. Erect or ascending, glabrous throughout : stems
much branched : upper leaves linear or linear-lanceolate, 1^ to 2 cm.
long : stipules short, scarious, laciniate : inflorescence in loose terminal
dichotomous cymes with the terminal flower sessile : calyx-lobes about
1 mm. long, obtusish, inconspicuously serrulate on the margins : corolla
about 4 mm. long, funnelform ; lobes spreading, broadly ovate : capsule
obcordate-compressed, 1^ mm. long ; seeds oblong, peltate, slightly con-
cavo-convex, scrobiculate. — Contrib. U. S. Nat. Herb. i. 70. — Lower
California, La Paz, Palmer, no. 28 ; San Jose del Cabo, Brandegee,
no. 261.

M- -i- Low (lichotomously branched perennials : stems becoming sh'ghtly woody
below : ovary one half to three fourths inferior : capsule short-oblong or sub-
globose.

++ Stems rather leafy : capsule about 3 mm. long.

H. Palmeri, Gray. Erect, profusely branched : branches hirtellous-
puberulent or glabrous: leaves linear-filiform, glabrous, 5 to 18 mm.
long ; stipules small, subtruncate, minutely dentate, or triangular-acute :
pedicels slender, erect or ascending : flowers about 1 cm. long, dimor-
phous : calyx-divisions lance-linear, 2 mm. long, ciliate : corolla salver-
shaped or slightly funnelform, upper portion of tube pubescent within :
free portion of capsule often hirtellous ; seeds open-crateriform, oblong-

VOL. XXXII. — 19

290 PROCEEDINGS OF THE AMERICAN ACADEMY.

rotund with a longitudinal hilar ridge on the ventral surface, scrobicu-
late, — Proc. Am. Acad. xvii. 202. — Northern Mexico, Coahuila, in
the region of Saltillo, Palmer, nos. 397, 398, and in the same State on
limestone hills, Carneros Pass, Pringle, no. 2381.

++ i-f Leaves much reduced : capsule about 2 mm. long.
H. longipes, Watson. Erect, much branched, 1 to 2^ dm. hi^h :
stems and branches subterete, slender, glabrous or below granular-hirtel-
lous: leaves linear-filiform, 4 to 10 mm. long, glabrous ; stiimles minute:
flowers small, " whitish and yellowish," 8 mm. long, on slender filiform
pedicels : calyx-divisions linear, about 1 mm. long, obtusish and submu-
cronate : corolla f unnelform, upper portion of tube pubescent within ;
lobes pilose-pubescent above : capsule small, subrotund ; seeds saucer-like
or subcrateriforra with a longitudinal ridge on the hilar surface, scrobic-
ulate. — Proc. Am. Acad, xviii. 97. — Near Mouclova, Coahuila, Palmer,
no. 394, and by the same collector at Monterey, Nueva Leon, no, 395 ;
also by Dr. Gregg at Cerralbo, collected 29 May, 1847.

H. brevipes, Rose. A low much branched distinctly suffrutescent
plant, 2 to 3 dm. high, glabrous throughout, somewhat glaucous : branches
slender, terete: leaves linear-filiform, 10 to 15 mm. long; stipules short,
1-4-setiform : flowers sessile or on slender pedicels : calyx deeply 4-
lobed, with a stipitate gland in each sinus : divisions in anthesis about
1 mm. long: corolla salverform, 10 to 15 mm. long ; tube slender, slightly
enlarged above ; lobes glabrous above : capsule subglobose ; seeds oblong,
peltate, cancavo-convex, minutely scrobiculate. — Contrib. U. S. Nat.
Herb. i. 83, 132. — Lower California, Santa Rosalia, Palmer, no. 202 ;
Carmen Island, Palmer, no. 836 ; also specimen in Gray Herbarium from
Lower California, apparently from the herbarium of Dewey, without
further data.

§ 2. Ereicotis. Fruticose or fruticulose plants with linear subterete
rigid-setaceous or acerose-linear leaves : capsule short-oblong, obtuse or
retuse ; seeds open-crateriform or oblong, peltate, minutely scrobiculate. —
Gray, Proc. Am. Acad. xvii. 203.

* Low, 2 to 18 cm. high, much branched from a woody perennial base : leaves

rigid-setaceous or acerose-linear.

H. acerosa, Gray. A low diminutive cespitose shrub : branches as-
cending or erect, strict, striate, hispidulous: leaves opposite or in threes
or fours, 5 to 12 mm. long, connected at the base by a stipular setiferous
cup, more or less hispid and ciliate : flowers sessile at the ends of thi
branches and branchlets : calyx deeply 4-parted ; divisions setaceous,
4 mm. long, hispidulous : corolla-tube slender, about 1 cm. long, dilated

GREEiVMAN. — GENUS HOUSTONIA. 291

at the throat, almost fuunelform : capsule about two thirds inferior,
globose or short-oblong, obtuse or slightly retuse, pubescent, much ex-
ceeded by the erect persistent calyx-divisions ; seeds more or less crateri-
form, somewhat roughened on the back. — Proc. Am. Acad. xvii. 203,
& Syn. Fl. N. A. i. pt. 2, 27 ; llemsl. BioL Centr.-Am. Bot. iv. 47.
Hedyotis {Ereicotis) acerosa, Gray, PI. Wright, i. 81. Oldenlandia
{Ereicotis) acerosa, Gray, PI. Wiight. ii. 67, Mallostoma acerosa,
Hemsl. Biol. Centr.-Am. Bot. ii. 31. — Northern Mexico, Parry, nos. 7,
302, Gregg, no. 72, Palmer, nos. 400, 401, 402, 403 ; adjacent Texas,
Wright, nos. 237 (type), 1118, Thurber, no. 91, and Bigelow.

H. polypremoides, Gray. Low, cespitose, 6 to 15 cm. high, much
branched from a woody perennial base: stems strict, ascending or erect,
usually branched above, glabrous or densely hirtellous-puberulent : leaves
less fascicled than in the preceding species, acerose-linear, 5 to 10 mm.
long, glabrous or hirtellous, margin revolute, often ciliate, midrib thick-
ened : flowers dimorphous, sessile or distinctly pedicellate, terminating
the stems and branches ; pedicels elongated in fruit, reaching a length of
13 mm. : calyx-divisions awl-shaped, 3 to 4 mm. long : corolla salver-
form or slightly funnelform, " white changing to purple," 8 to 10 mm.
long : capsule globose, about two thirds inferior ; seeds open-crateriform,
obloncr-rotuud with a short hilar ridge. — Proc. Am. Acad. xxi. 379. — ■
Chihuahua, Santa Eulalia Mountains, Pringle, nos. 16,356; on route
from Leavenworth to El Paso, F. R. DifFenderfFer ; at junction of Dela-
ware Creek and Pecos River, Texas, Pope.

Var. Bigelovii. A form with the calyx-lobes H to 2 mm. long,
otherwise as in the type. — Florence Mountains, Bigelow, in the Mexican
Boundary Survey, June, 1852, no. 437.

* * Distinctly shrubby, 2J dm. or more high, much branched : leaves thickish,
more or less fascicled : inflorescence in cymes terminating the stems and
branches.

■^ Leaves 3 to 10 mm. in length, much fascicled : corolla 3 to 4 mm. long.
H. fasciCTllata, Gray. A low bushy shrub, much branched: stem
and branches covered with a grayish bark : branchlets tetragonal, hirtel-
lous-puberulent : leaves linear, subterete, thick, rigid, nearly equalling the
internodes, decidedly fascicled in the axils, glabrous or hirtellous, margin
revolute ; stipules short, scarious, 1 to 3 setiferous-dentate : cymes few-
flowered : flowers small, " white, drying purple," dimorphous : calyx-
divisions short, about 1 mm. long, obtuse, usually puberulent : corolla
deeply 5-lobed ; lobes recurved ; tube long-pilose-pubescent on the inner
surface : capsule oblong ; seeds oblong, peltate, concavo-convex, minutely

292 PROCEEDINGS OP THE AMERICAN ACADEMY.

scrobiculate. — Proc. Am. Acacl. xvii. 203. — In the mountains, 24
miles N. E. of Monclova, Coahuila, Palmer, nos. 404, 406 ; Santa
Eulalia Mountains, Chihuahua, Priugle, nos. 151, 354; adjacent Texas,
,it Presidio, Bigelow, in Mexican Boundary Survey ; foothills of Chisos
Mountains, Havard, No. 31 ; and in the Organ Mountains, New Mexico,
Vasey, 1881.

+- -I- Leaves 3 to 18 mm. in length : corolla 10 to 14 mm. long.

H. fruticosa, Rose. Low much branched shrub, glabrous through-
out : stem and branches covered with a rough yellowish brown or grayish
bark : leaves linear or somewhat thickened and subterete, submucronate,
more or less fascicled in the axils ; stipules short, glandular-setiferous :
inflorescence in terminal rather few-flowered dichotomous cymes ; flowers
subsessile or short pedicellate : calyx deeply 5-lobed with a single stipi-
tate gland in each sinus; lobes linear-lanceolate, about 3 mm. long, folia-
ceous : corolla about 3 times longer than the calyx ; tube narrowly
funnel-shaped, glabrous : capsule short, oblong, subtruncate or slightly
emarginate ; seeds oblong, peltate, concavo-convex, minutely scrobiculate.
— Contrib. U. S. Nat. Herb. i. 132, 239. Hedyotis {Ericotis) mucronata,
Benth. Bot. Sulph. 19. — Magdalena Island, Lower California, Brandegee,
17 January, 1889; Carmen Island, Palmer, no. 885.

§ 3. Macrohoustonia. Erect shrubby plants with ovate leaves and
rather large flowers. — Gray, Proc. Am. Acad. iv. 314. A section
scarcely to be retained.

H. bouvardioides, Benth. & Hook. f. Shrub 3 to 4 feet high :
stems terete, granular- jDuberulent : leaves opposite or ternate, short-jDetio-
late, ovate or elliptic-lanceolate, acute or short-acuminate, entire, finely
puberulent upon either surface, especially upon the midrib and veins :
inflorescence in terminal compound minutely pubescent cymes : calyx
deeply 4-parted ; divisions linear, 3 mm. long : corolla funnelform, about
12 mm. long, glabrous, rather deeply 4-lobed ; tube with a few scattered
hairs on the inside toward the base ; lobes oblong, obtuse : anthers long-
exserted : style included (probably dimorphous) ; ovules rather numer-
ous : seeds not seen. — Gen. ii. 60 ; Hemsl. Biol. Centr.-Am. Bot. ii. 30.
Hedyotis (§ Anotis) bouvardioides, Seem. Bot. Herald, 296, t. 64. — N. W.
Mexico, Seemann. A plant having much more in common with Bouvardia
than with Hoiutonia, but without fruit or mature seeds it seems unwise to
make the transfer. More complete specimens, however, may justify its
removal to the former genus.

H. (Macrohoustonia) longiflora, Gray, Proc. Am. Acad. iv. 314,
Hemsl. Biol. Centr.-Am. Bot. ii. 30, is Bouvardia longijlora, HBK.,

GREENMAN, — GENUS LIABUM. 293

Nov. Gen. & Spec. iii. 38G ; Bot. Mag. 4223. u^ginetia longijiora^ Cav.
Ic. vi. 51, t. 257, f. 1.

H. (Macrououstonia) triflora. Gray, 1. c, Hemsl. 1. c, is a good
Bouvardia.

From the technical characters of complete specimens of the last two
species, now at hand, there seems no reason for retaining them in the
genus Houstonia.

11. KEY TO THE MEXICAN SPECIES OF LIABUM.

* Leaves ovate or ovate-deltoid : heads radiate.

t- Leaves ovate-deltoid, more or less deeply sinuate-dentate and mucronate-denticu-

late : heads few, large (3 to 5 cm. in diameter), terminating the stem.

L. Andrieuxii, Benth. & Hook. f. Gen. ii. 437. Vernonia Andrieuxii,
DC. Prodr. V. 16.

•»- -t- Leaves ovate, entire or mucronate-denticulate : heads disposed in terminal

panicles.
++ Pappus of slender unequal setae, not biseriate.
= Involucral scales lance-attenuate : petioles winged and often dilated about the
stem : ligules narrow : achenes pubescent.

L. asclepiadeum, Schz. Bip. Linnaea, xx. 521.

= = Involucral scales oblong, obtuse : ligules ovate : achenes glabrous.
L. polyanthum, Ivlatt, Bull. Soc. Bot. Belg. xxxi. (1892), 209.
++ *+ Pappus distinctly biseriate, outer series short, paleaceous.
= Achenes pubescent.
L. andromachioides, Benth. & Hook, f. Gen. ii. 43G. Vernonia
andromachioides^ Less. Linntea, vi. 397, 645.

L. Deppeanura, Hemsl. Biol. Centr.-Am. Bot. ii. 232. Andro-
machia Deppeana, Less. Linncea, vi. 401.

= = Achenes glabrous.
L. platylepis, Schz. Bip. Flora, 1856, 160; Klatt, Leopoklina,
xxiii. (reprint, p. 7).

L. discolor, Benth. & Hook. f. Gen. ii. 436. Sinclairia discolor,
Hook. & Arn. Bot. Beech. 433 ; Hook. Ic. PL v. t. 451.

* * Leaves ovate or ovate-lanceolate : heads discoid.

1- Inflorescence in loose terminal pyramidal panicles : involucre 3-5-seriate :

pappus biseriate.

t-f Immature achenes stipitate-glandular : involucral scales obtuse.
Li. glabrum, Hemsl. 1. c.

294 PROCEEDINGS OF THE AMERICAN ACADEMY.

"Var. hypolencum. Leaves ovate-oblong, acuminate, glabrous
above, white-tomentose beneath, more or less glabrescent. — Collected
by C. G. Pringle, in canons near Guadalajara, 8 December, 1888, no.
2169, and on rocky banks of river below the Falls of Juanacatlan near
Guadalajara, 25 October, 1889, no. 2372, the latter distributed as Liabum
discolor^ Benth. & Hook. f.

The companion number cited by Hemsley, Bourgeau, no. 1401, as
represented in the Gray Herbarium, shows an arachnoid pubescence still
adhering to the under surface of the leaves. Pringle's no. 2169 shows
the tomentum on the lower surface of the leaves to be evanescent ; this
number, however, certainly represents the same species as Pringle's
no. 2372, which has a dense tomentum on the lower surface of the
leaves. It seems best therefore to regard these two specimens collected
by Mr. Pringle as a tomentose variety of Liabum glabrujn, Hemsl. Mr.
Pringle's no. 6182, collected in the type locality, is probably more typical
of the species proper.

++ ++ Involucral bracts acute : immature achenes setulose at the apex.

L. sericolepis, Hemsl. 1. c.

•i-H- Inflorescence in thyrsoid panicles, terminating the stems and branches : invo-
lucral scales lance-attenuate : pappus of uiicqual setas, not biseriate.

Li. Klattii, Rob. & Greenm. Am. Jour. Sci. ser. 3, L. 156.

.1- ^- •(- Heads few, large, 2 cm. or more in diameter : involucre multiseriate ;
scales lanceolate, acute : petioles short, about 4 mm. in length.

L. Pringlei, Rob. & Greenm. Proc. Am. Acad, xxxii. 49,

* * * Leaves sessile, linear or palmately lobed : heads discoid, few, on long pedun-
cles : involucral scales linear-attenuate : achenes sericeous.

I- Leaves amplexicaul, palmately lobed.

L. cervinum, Rob. Proc. Am. Acad. xxix. 317.
L. Palmeri, Gray, Proc. Am. Acad. xxii. 432.

■*- H- Leaves linear, subverticillate.

L. angustissinmm, Gray, 1. c.

* * * * Species of doubtful aiBnity.
L. Liebmannii, Klatt, Leopoldina, xxiii. (reprint, p. 7). From the
very meagre characterization of this species it is impossible to give its
exact affinity without seeing Liebmann's specimen.

GREENMAN. — MEXICAN PLANTS. 295

III. — DESCRIPTIONS OF NEW AND LITTLE KNOWN

PLANTS FROM MEXICO.

Tradescantia macropoda. Roots tuberous, fascicled : stems erect
or suberect, sparingly branched, bilineate-pubescent, otherwise glabrous
or puberulent : leaves oblong-ovate, acute, 5 to 8 cm. long, 1|^ to 2^ cm.
broad, abruptly narrowed below to an unequal base, glabrous or some-
times inconspicuously puberulent on either surface ; margin, except at the
ciliated base, hispidulous : sheaths 5 to 8 mm. long, puberulent, espe-
cially along the line of pubescence on the stem, strongly ciliate : inflores-
cence umbellate on rather long slender peduncles (2 to Ak cm. long) :
bracts orbiculai'-ovate, obtuse, glabrous, o-7-nerved: flowers several,
fugitive in the bracts ; calyx-divisions oblong, navicular, thin or scarious,
1-uerved, 3 to 4 mm. long, one half as broad ; the outer sepal often
sparingly long hirsute-pubescent on the keel : corolla light purple in the
dried state ; petals about 6 mm. long, two thirds as broad : stamens 6,
equal; filaments somewhat exceeding the corolla, rather scantily covered
toward the base with multicellular hairs ; anther-cells globose, equal,
widely separated by a long narrow connective : ovary glabrous or with a
few scattered glandular hairs at the apex : seeds light brown, transversely
rugose. — Collected by C. G. Pringle, on moist banks of mountains
above Cuernavaca, altitude 2,000 m., 3 August, 1896, no. 6402. A
species nearly related to T. comyneluioides, Roem. & Schult., but readily
separated by the nearly glabrous character throughout, and by the long
slender glabrous peduncles.

Schcenocanlon Pringlei. Bulbs ovoid, 1 to 2 cm. in diameter:
caudex erect, cylindrical, 3 to 10 cm. long, surrounded by a covering of
dark brown fibres, the remnants of the bulb-scales and outer leaves : foliar
leaves linear-attenuate, 1 to 5 dm. long, 2 to 6 mm. broad, 9-lO-nerved,
smooth on either surface, margin slightly roughened : naked scape S^ to
6 dm. high, terete or slightly flattened, smooth : inflorescence dense, 3 to
8 cm. long, 1 cm. broad : bracts broadly ovate, acute, scarious, 2| mm.
long : divisions of the perianth oblong, about 4 mm. long, 3-nerved, ob-
tuse or rounded at the slightly thickened apex, margins scarious, irregu-
larly subdenticulate, or entire : stamens in anthesis about equal or slightly
exceeding the divisions of the perianth (even in the later stages exceeding
the perianth divisions by about 1 mm. only) : immature capsules about
12 mm. long, erect, glabrous, rather dense, and somewhat appressed to the
rhachis. — Collected by C. G. Pringle, on lava beds, Serrania de Ajusco,
altitude 2,400 m.. 23 August, 1896, no. 6415.

296 PROCEEDINGS OF THE AMERICAN ACADEMY.

Agave (Euagave) coUina. Acaulescent: leaves 30 to 40 in a
rosette, linear-attenuate, G to 8 dm. long, 2 cm. broad just above the base,
2|- cm. broad at the middle, gradually narrowed to the apex and termi-
nated by a reddish brown spine (2 cm. or more long), margin narrowly
cartilaginous bearing reddish brown straight or curved teeth (3 to 5 mm.
long) at intervals of 8 to 18 mm., upper surface concave, below convex,
bluish green: peduncles 3 to 4 m. long: panicle 6 to 8 dm. long, the
lower branches about 3 dm. long : flowers greenish yellow, includiug the
stamens, 9 cm. long : tube of the perianth 1^ cm. long, funnel-shaped ;
segments linear-oblong, obtuse, thickened at the apex, about 2 cm. long :
stamens more than twice as long as the perianth-segments ; filaments
flattened, glabrous, about 5 cm. long from the point of insertion ; anthers
2^ to 3 cm. in length : style overtopped by the anthers : fruit a loculici-
dal capsule, 4^ to 5 cm. long, 2| to 3 cm. in diameter ; seeds obliquely
triangular, about 1 cm. long. — Collected by C. G. Pringle, on hillsides
near Cuernavaca, altitude 1,500 m., 1896, no. 6349. This may prove
eventually to be identical with the imperfectly known species A. serra-
tula, Karw., but as the flowers and fruit of the latter are unknown, and
our plant does not agree in all details with the fragmentary description
of Karwinski, it seems best to regard Mr. Priugle's plant as a distinct
species.

Nemastylis caerulescens. Bulb ovoid, 1| to 2 cm. in diameter,
covered with dark brown scales : stem or scape 14 to 20 cm. high,
branched above, bearing 1 to 3 pedunculate spathes : the outer radical
leaves much reduced, the inner a single linear-attenuate leaf, 10 to
15 cm, long, usually exceeding the scape ; stem leaves reduced to lance-
attenuate bracts, 3 to 4^ cm. long, subtending the inflorescence or pe-
dunculate spathes : peduncles 1 to 2^ cm. in length : bracts of the spathe
linear-oblong, acuminate, the outer 2 to 3 cm. in length, the inner a little
longer : pedicels slender, 2^ to 3^ cm. long : flowers white, sometimes
with a faint shade of blue, 3|- to 4|- cm. in diameter, becoming reflexed ;
divisions of the perianth flabellate-nerved, hirtellous-puberulent above
near the base, the outer oblong-elliptic, about 6 mm. broad, the inner
somewhat smaller and a little narrowed toward the base : filaments united
at the base, free portion flattened, about 2 mm. long ; anthers 1 cm. in
length, curling with age: capsule smooth, obovoid, subtruncate above,
7 to 10 mm. long, 5 to 7 mm. in diameter: seeds irregularly angled,
yellowish, minutely pitted, 2 mm. long. — Collected by C. G. Pringle,
on dry gravelly soil near Cuernavaca, altitude 1,600 m., June, 1896, no.
6324. A species most nearly related to M. acuta, Herb. (N. geminijiora,
Nutt.), but a much smaller plant and with different flowers.

GREENMAN. — MEXICAN PLANTS. 297

Eletia macristlimochila. Bulbs terrestrial, subglobose or some-
what gibbous : scape 6 to 6^ dm. high, surrounded at the base by 4 to 6
purplish striated acute sheaths, naked portion smooth, 1-2-vaginate, ter-
miuating above iu a few-flowered raceme : leaves plicate, broadly lance-
acuminate, smootli, about 7-nerved, 12 to 30 cm. long, 2 to 4 cm, wide :
bracts triangular-acuminate, 10 to 14 mm. long: flowers 5 to 8 cm. in
diameter, rose-purple in the dried state : sepals oblong, short-acuminate,
a little narrowed at the base, 3 to 4 cm. long, 10 to 12 mm. broad : petals
oblong, obtuse or rounded at the apex, rather abruptly narrowed below,
considerably broader than the sepals ; labellum deeply 3-lobed, lateral
lobes broad and obtusely triangular, strongly flabellate-nerved, the upper
nearly horizontal margins subcreuate, lateral margins entire, median lobe
deeply obcordate, apiculate in the terminal sinus, the two spreading sub-
creuulate or erose lobes narrowed below, and connecting with the lateral
lobes by a long narrow neck or isthmus ; lamella3 5, strongly undulate,
continuous from near the apex almost to the base of the labellum : gynoe-
cium narrowly winged, 2 to 2^ cm. long. — Collected by Dr. E. Palmer,
on a barranca, near Guadalajara, June, 1886, no. 127, also by C. G.
Pringle, in the same locality, 2 July, 1889, no. 2875. This plant has
been confused with B. campanulata, Llav. & Lex., to which species it
was referred by Dr. Sereno Watson. It may however be distinguished
from B. campanulata, as well as from B. fulgens, Reichb., a large flowered
species growing in similar localities, by the long narrow neck or isthmus
of the labellum.

MicuosTTLis STREPTOPETALA, Rob. &, Greenm. Proc. Am. Acad,
xxxii. 36. A large flowered form of this species was collected by
C. G. Pringle, on lava beds, Serrania de Ajusco, at an elevation of
2,400 m., August, 1896, no. 6410. Except in the size of the flowers, which
are about one third larger than in the type, the specimens agree in every
detail with the above species.

Aristolochia longecaudata, Watson, var. virescens. Perennial
from a ligneous base : stems decumbent or prostrate, sulcate, some-
what retrorsely hirsute-pubescent : leaves ovate to ovate-oblong, often
acuminate, pubescent on either surface, 3-5-nerved, 3 to 9 cm. long, 1|-
to 5^ cm. broad : petioles 1 to 3^ cm. long. — Collected by C. G. Pringle,
about Cuernavaca, altitude 1,500 m., 23 July, 1896, no. 6383. A form
differing from the type of the species by its larger thinner and less pu-
bescent leaves, but having the same technical characters in flower and
fruit.

Euphorbia ramosa, Seaton, var. villosior. Stems decumbent,

298 PROCEEDINGS OP THE AMERICAN ACADEMY.

much branched from a suffrutescent base, rather densely covered with a
spreading villous pubescence : involucre purple, glabrous : capsule pu-
bescent. — Collected by C. G. Pringle, Pedrigal (lava beds), Valley of
Mexico, altitude 2,300 m., 25 August, 1896, no. 6436.

Mentzelia Conzattii. Stems whitish, covered with a thin scarious
bark, below glabrous, above pubescent : lower leaves subopposite, upper
alternate, simple, including the petioles 5 to 14 cm. long, 1^ to 3^ cm.
broad, oblong-lanceolate, acuminate, acute, finely dentate, gradually nar-
rowed at the entire base into a short petiole, scabrous above, tomentose
beneath, midrib somewhat depressed above, prominent beneath ; petioles

4 to 12 mm. long : inflorescence cymose-paniculate : flowers large, about

5 cm. in diameter, showy : calyx deeply 5-lobed ; tube turbinate, barbel-
late-pubescent ; lobes lance-ovate, acuminate, 12 mm. long, 4 mm. broad
at the base, pubescent : corolla pentapetalous ; petals ovate-oblong, short-
acuminate, a little narrowed at the base, 2^ cm. long, 1 to 1^ cm. broad,
yellow in the dried state : stamens disposed in phalanges opposite the
petals, the filaments of the 3 outer stamens of each phalanx somewhat
flattened, the others filiform : style single, filiform ; stigma terminal :
mature fruit not seen. — Collected by C. Conzatti, Oaxaca, altitude 1,500
m., 5 January, 1896, no. 62.

Streptotrachelus, nov. gen. oi Apocynacece {EuecMtidea;) . Calyx
5-parted, naked at the base within. Corolla salverform, coronate, esqua-
mate, tube cylindrical not dilated at the throat, soon becoming strongly
contorted at the middle; lobes 5, dextrorsely convolute in the bud, and
sinistrorsely twisted. Stamens affixed to the upper half of the corolla-
tube, included, free portion of the filament short ; anthers sagittate,
acuminate, connivent about the stigma, the cells projecting below into
two rigid appendages. Disk of 5 oblong fleshy persistent scales about
equalling the ovary. Ovary of two distinct carjiels with a common fili-
form style ; stigma fleshy, oblong-cylindrical, short-acuminate, somewhat
2-cleft at the apex, dilated at the base, adherent above and below to the
anthers ; ovules in each carpel numerous. Follicles subterete. Seeds
oblong-linear, furrowed, not contracted into a beak at the coma-bearing
apex ; albumen scanty ; cotyledons plane. Woody stemmed twining
plants with opposite leaves and cymose inflorescence. The generic
name is taken from o-Tpe7rT09, twisted, and TpdxrjXo's, throat, in reference
to the twisting of the corolla-tube.

Streptotrachelus Pringlei. Stems woody, more or less covered
with lenticels, twining and clambering over shrubs and small trees to 4 or

6 meters ; brauchlets reddish brown, puberulent : leaves opposite, petio-

GREENMAN. — MEXICAN PLANTS. 299

late, oblong-ovate to broadly ovate, 4 to 8 cm. long, 2 to 5 cm. broad,
short-acumiuate, entire, usually cordate at the base, finely pubescent upon
either surface, especially upon the prominent midrib and the strongly
reticulate veins beneath, above dark green, paler beneath : petioles canalic-
ulate, 1 to 3 cm. long, puberulent or finely pubescent, bearing at, their
bases 2 or more small subulate stipule-like structures : inflorescence in ax-
illary pedunculate minutely pubescent cymes ; jDcduncle 2 cm. or more
long ; pedicels 1 to 1^ cm. long ; calyx deeply 5-parted ; divisions lance-
linear, acute, about 4 mm. in length : corolla salverform, greenish yellow or
sometimes purple ; tube 22 mm. long, strongly contorted at the middle,
provided at the orifice with an aduate subcuneate crown and bearing on
the inside just above the attachment of the stamens a tuft of long villous
hairs ; lobes broadly obovate, 8 to 9 mm. long : anthers pubescent on the
outer surface: carpels villous-pubescent : fruit reaching a length of 3 dm.;
seeds about 2 cm. long. — Collected by C. G. Priugle, on lava beds near
Cuernavaca, altitude 1,600 m., 23 September, 1896, no. 6554.

Astephanus pubescens. Stems slender, twining, finely striate,
pubescent : leaves short-petiolate, lanceolate to ovate-lanceolate, l-J- to
6 cm. long, 4 to 20 mm. broad, acute at the apex or sometimes submu-
cronate, obtuse at the base, entire, sparingly pubescent on either surface :
petioles 2 to 4 mm. long, pubescent : inflorescence in subumbellate short-
pedunculate several flowered axillary clusters ; peduncles 1 to 2 mm.
long : flowers on slender pubescent pedicels, the latter about 3 mm. in
length : calyx minute, 5-parted, pubescent on the outer surface, provided
on the inside with 5 small oblong glands alternating with the lobes : divis-
ions ovate, acutish : corolla subcampanulate, about 3 mm. in diameter,
deeply 5-lobed, white or tinged with brownish purple : lobes dextrorsely
convolute in the bud, oblong, emarginate at the apex, spreading, later
becoming somewhat reflexed, the upper inner surface, as well as the
corolla-tube inside, granular or almost lepidote : fruit not seen. — Col-
lected by C. G. Pringle, on a wet barranca above Cuernavaca, altitude
2,000 m., 21 September, 1896, no. 6507, and in the mountains near
Cuernavaca, 3 August, 1896, no. 7203. A plant with the habit of
Metastehna, or of Vincetoxicum, but by the entire absence of a crown, or
at most by a very rudimentary one, its affinity is rather with the genus
Astephanus, notwithstanding the presence of the minute glands at the
sinuses on the inside of the calyx.

Gonolobus chrysanthus. Stems twining, greenish purple, granu-
lose, interspersed with spreading or subreflexed hirsute pubescence : leaves
ovate-oblong, acuminate, obtusish at the apex, a deep narrow open sinus

300 PROCEEDINGS OP THE AMERICAN ACADEMY.

at the base with rounded basal lobes on either side, margin entire, finely
ciliate, hirsute or hirsutish pubescent above, hirsute on the midrib and
veins beneath, 5 to 9 cm. long, 2 to 4 cm. broad; petioles hirsute, 1 to
5 cm. long: inflorescence in axillary pedunculate subumbellate racemes,
granulose or glandular, and interspersed with spreading hirsute hairs :
peduncles 1 to 2 cm. long : bracts linear, subulate, hirsute : pedicels
slender, 10 to 17 mm. long : buds oblong, oval, obtuse, pubescent: calyx-
divisions oblong-linear, acutish, ciliate, hirsute on the outer surface,
glabrous within, 3 to 4 mm. long, about 1 mm. broad: corolla 1|- to
2^ cm. in diameter, orange yellow, or yellow streaked with parallel
greenish veins, outer surface pubescent, inner surface glabrous ; lobes
oblong-linear, obtuse or rounded, usually oblique and often with a slight
notch at the apex, 6 to 10 mm. long, about 3 mm. broad ; crown 5-
lobed, lobes denticulate. — Collected by C. G. Pringle, on hills near
Patzcuaro, State of Michoacan, 3 August, 1892, no. 5277 ; on a wet
wooded barranca above Cuernavaca, altitude 2,000 m., 4 August, 1896,
no. 6373; Pedigral (lava beds), Valley of Mexico, altitude 2,500 m.,
25 August, 1896, no. 6437.

Lithospermnm (Batschia) oblongifolium. Herbaceous from
a woody perennial base : stems erect, sparingly branched, 3^ to 5 dm.
high, covered with a spreading hirsute pubescence : leaves sessile, ob-
long or oblong-ovate, acute, entire, tuberculate-hispid above, with a
scattered substrigose pubescence beneath (somewhat more dense upon the
veins), 4 to 8^ cm. long, 1^ to 2|- cm. broad : flowers axillary, disposed in
leafy one-sided racemes or scorpoid cymes, in fruit about 15 cm. in length :
pedicels 4 to 8 mm. long: calyx deeply 5-parted, persistent; divisions
linear, 1-nerved, hirsute-pubescent, 10 to 14 mm. long: corolla tubular,
5-lobed, gibbous at the throat, about 3 cm. long ; tube pubescent on the
outer surface, inner surface glabrous except along five lines in continuance
with the short filaments ; lobes erect, subreniform, 3 mm. long, 5 mm.
broad : style filiform, glabrous, more or less persistent ; stigma capitate-
2-lobed : nutlets ovate, acute, smooth, white, or white below and brown-
ish above. — Collected by C. G. Pringle, on the Serrania de Ajusco,
altitude 3,000 m., 18 August, 1896, no. 6451. A species with the habit
of L. Palmeri^ Wats., but readily distinguished by the pubescence of the
stem and by the characters of the corolla and stigma.

Citharexylnm glabrum. A tree 6 to 8 m. in height, glabrous
throughout : branches and branchlets subterete, striate, covered with a
brownish bark: leaves oblong to oblontj-lanceolate, 4 to 10 cm. Ions, 1 to
3 cm. broad, rounded, obtuse or acute at the apex, narrowed at the base

GREENMAN. — MEXICAN PLANTS. 301

into a petiole (10 to 16 mm. long), punctate especially on the under sur-
face, dark green above, paler beneath, glandular below on either side of
the midrib at the base of the blade with 1 to 2 or 3 oblong-elliptic glands :
inflorescence in terminal and axillary nodding spike-like racemes, 4^1- to
1 1 cm. long ; bracts minute, equal to or exceeding the pedicels, the latter
about 1 mm. long, and jointed just below the flowers : calyx 2 mm. long,
subtruncate, 5-nerved, somewhat 5-angled and ciliate : corolla about
5 mm. in length, tubular, 5-lobed, white or pinkish ; tube broad, glabrous
on the outer surface below, puberulent above, pubescent in the throat ;
lobes broad, oblong or rounded, pubescent on either surface : style
glabrous. — Collected by C. G. Pringle, in the mountains near Lake
Chapala, Jalisco, 16 December, 1889, no. 24-42. Flowers fragrant. Dis-
tributed as Gonzalea glabra^ Wats., n. sp., Proc. Am. Acad. xxv. 152.

Citharesyl-am. ovatifolium. A soft woody shrub, 2 to 3 m. in
height : branches tetragonal, minutely striate, pubescent : leaves mem-
branaceous, ovate-acuminate, 8 to 1 2 cm. long, 3^ to 6|- cm. broad, ob-
tuse, submucrouate, abruptly narrowed at the base into a more or less
winged pubescent petiole, entire or irregularly crenate-dentate, ciliate,
pubescent upon either surface, dark green above, slightly paler beneath,
usually provided with one or more large glands on the lower surface at
the base of the blade on either side of the midrib : inflorescence in pubes-
cent spike-like nodding or flexuous racemes, 5 to 12 cm. long, terminating
the branches and branchlets : bracts about 1 mm. long, minute, subulate,
equalling or slightly exceeding the short pubescent pedicels : calyx 3 mm.
long, subtruncate or minutely and irregularly 5-dentate, ciliate, 5-uerved,
and somewhat 5-angled, pubescent on the outer surface : corolla about
7 mm. long, tubular, o-lobed, white ; tube broad, about 4 mm. long,
glabrous on the outer surface, densely pilose-pubescent in the throat ;
lobes subrotund, glabrous on either surface, strongly ciliate : style
glabrous. — Collected by C. G. Pringle, in a wet wooded barranca
above Cuernavaca, altitude 2,000 ra., August-September, 1896, no. 6540.
This species has much in common with the imperfectly known C. Sessei,
D. Don, but as the latter species is so very obscure it seems best to regard
Mr. Pringle's plant as distinct, especially as it does not accord in all details
with the meagre description of Don's species.

SolanTim (Morella) deflesum. An annual erect or ascending
herb becoming slightly woody at the base : roots fibrous : stems 1 to
4|- dm. high, simple or much branched, terete, covered especially above
with a spreading hirsute pubescence: leaves simple, solitary or in pairs
of subequal size, 2 to 6 cm. long, one half as broad, ov^ate, acute or

302 PROCEEDINGS OF THE AMERICAN ACADEMY.

obtusish, entire or subrepand, ciliate, strigose-pubescent upon either sur-
face, and hirsute upon the midrib and veins beneath, abruptly contracted
below into a narrowly winged petiole ; petioles 5 to 22 mm. long, hirsute-
pubescent : inflorescence extra-axillary of 1 to 4 slender reflexed pubescent
pedicels, becoming in fruit about 2 cm. long : calyx deeply 5-parted,
densely covered with a long spreading hirsute pubescence ; divisions
linear-oblong, acute, 3 to 7 mm. long, more or less persistent : corolla
about 1 cm. in diameter ; lobes broadly ovate, short-pubescent at the
acute apex, externally covered with a few long scattered jointed hairs :
stamens 5, equal; anthers oblong: ovary glabrous, 2(-4)-celled : fruit
smooth, about 8 mm. in diameter ; seeds oblique-ovate, corrugated, 3 mm.
long. — Collected by Lucius C. Smith, at Cuicatlan, 15 July, 1895,
no. 403 ; E. W. Nelson, between Topana, Oaxaca, and Tonala, Chiapas,
altitude 61 to 150 m., 1-3 August, 1895, no. 2876^; and by C. G. Prin-
gle, on shaded hillsides near Cuernavaca, altitude 1,500 m., 26 July, 1896,
no. 6400. A species apparently well characterized by the infloresce ce,
calyx, and pubescence. Most nearly related to the Solanum nigrum
group.

Solanum Mitlense, Dun. Specimens agreeing in all detailed char-
acters with the original description of the above species were collected
in the State of Oaxaca by Lucius C. Smith in the Calderon, San Juan
del Estado, 1 June, 1894, no. 37; by E. W. Nelson, in the Valley of
Oaxaca, altitude 1,540 to 1,600 m., 8 and 24 September, 1894, no. 1234;
by C. G. Pringle, in ravines of hills near Oaxaca, altitude 1,750 m.,
15 September, 1894, no. 4907; and by C. Conzatti and V. Gonzalez,
about the city of Oaxaca, altitude 1,550 m., 2 April, 1896, no 98. From
Mr. Prino-le's full and careful notes, the following additional characters
may be given. Entire plant 3 to 4^ m. in height, tree-like in habit ;
trunk 2 to 2^ m. high, 1 to 2 dm. in diameter, trichotomously branched
above, covered with a grayish brown bark, stout spines, and a scattered
stellate pubescence : mature fruit green, globose, about 2 cm. in diameter :
seeds 3 mm. long, oblique, compressed and scrobiculate.

Dicliptera Pringlei. Annual, herbaceous, 5 to 8 dm. high : roots
fibrous : stems hexangular, somewhat farrowed, lineolate, puberulent
upon the angles, swollen or constricted just above the nodes : leaves
ovate to ovate-lanceolate, acuminate, acute, narrowed below into a slen-
der petiole, entire, ciliate, lineolate with a few scattered hairs on either
surface ; petioles pubescent, 2\ to 'dh cm. long, becoming gradually
shorter above : inflorescence much branched, open: heads on long slen-
der pedicels, 1 (-2)-flowered : outer bracts equal, elliptical, muticous,

GREENMAN. — MEXICAN PLANTS. 303

glabrous, resupinate, 5 to 6 mm. long, two thirds as broad ; inner bracts
liuear-linceolate, nearly 5 mm. long, 1 mm. broad, margins scarious ;
bractlets linear-lauceolate, diaphanous, about 4 mm. long with a rudi-
mentary flower in the axil : calyx deeply 5-parted ; divisions lance-atten-
uate, ciliate: corolla bilabiate, 1|^ to 2 cm. long, yellowish white striped
with purple in the dried state, externally pubescent: stamens with fila-
ments and anthers purple : capsule smooth, about 6 mm. long ; seeds
orbicular, flattened, muricate. — Collected by C. G. Pringle, on lava
beds near Cuernavaca, altitude 1,500 m., 3 November, 1S9G, no. G602 ;
and by E. W. Nelson, near Tlalixtaquilla, Guerrero, 10 December, 1894,
no. 2256.

Bnceragenia, nov. gen. of Acanthacece {EnjusticieoR). Calyx 5-
parted, segments linear, subequal. Corolla tubular; tube cylindrical,
erect, not ampliated above ; limb short, 2-labiate, posterior lip interior,
incurved, deeply emarginate or shortly 2-Iobed, anterior erect or scarcely
spreading, 3-lobed, external. Stamens 2, anterior, inserted at the mid-
dle of the corolla-tube, included ; anthers 1-celled, oblong, medio-dorsally
afiBxed, obtuse. Staminodea 2. Disk annular, inconspicuous. Style
slightly thickened above, stigma minutely and unequally 2-lobed ; ovules
2 in each cell. Capsule oblong, narrowed below into a solid stipe ; seeds
4, suborbicular, flattened, strongly muricate or roughened ; retinacula slen-
der ; embryo normal. — Herbaceous perennials having their affinity with
the genus Habracanthus, but differing by the shallowly bilabiate corolla,
and the presence of staminodea. The generic name is taken from ySov?,
Kepa?, and jLyvofxai, in reference to the Spanish name Cuernavaca, from
which place it comes.

Buceragenia minntiflora. Stems erect, about 1 m. high, simple,
nearly naked below, branched and leafy above, glabrous ; branches slen-
der, internodes 5 to 12 cm. in length: leaves ovate, usually acuminate
with an obtusish apex, rather abruptly narrowed at the base and attenu-
ated into a long narrowly winged petiole, entire or subrepand, ciliate,
minutely liueolate on the upper surface with a few scattered hirsutish
hairs, pubescent on the midrib and prominent veins beneath, dark green
above, paler beneath : inflorescence in slender interrupted spikes or
spikoid racemes terminating the stem and branches, puberulent and stipi-
tate-glandular ; spikes (4 to 13 cm. long) on slender peduncles ; bracts
ovate, acute, about 1 mm. long ; bractlets minute : calyx 2 to 5 mm. long,
5-parted, persistent ; divisions subequal, linear, acute, puberulent and
stipitate-glandular : corolla minute, 2 to 3 mm. long, tubular, 5-nerved,
slightly enlarged below ; lobes oblong, rounded at the apex, covered with

304 PROCEEDINGS OF THE AMERICAN ACADEMY.

several erect or spreading hairs on the outer surface near the tip : capsule
13 to 15 mm. long, glabrous. — Collected by C. G. Priugle, in a wet
barranca above Cuernavaca, altitude 2,000 m., 21 September, 1896, no.
6506.

Justicia Clinopodiuni, Gray, in herb. Stems ascending or erect,
usually branching, terete, 2 to 5 dm. high, lineolate pubescent, also
covered with long spreading villous pubescence, often interspersed with
glandular hairs : leaves short-petiolate, obtuse, entire, subcordate at the
base, with a scattered villous pubescence on either surface, especially
upon the veins beneath, 2 to 5 cm. long, 1 to 2^ cm. broad : petioles
short, 1 to 5 mm. in length ; inflorescence in terminal pedunculate
sjjikes becoming 8 cm. long ; bracts and bractlets lanceolate, acute, 10
to 15 mm. long, covered on the outer surface and marii;iu with longf
villous hairs intermixed with glandular hairs : calyx deejDly 4-parted,
the equal or subequal divisions linear-acute, about equalling the bracts,
externally villous-pubescent : corolla glabrous, about 2 cm. long, jiurple,
the upper lip shortly bifid, the lower deejily 3-lobed, segments subequal,
oblong, rounded at the tip : stamens of the genus, but with the adnate
portion pubescent : style smooth : capsule glabrous, about 13 mm. long ;
seeds compressed, more or less densely covered with shaggy armed
trichomes. — Collected by Bourgeau, in the region of Orizaba, 25 August,
1866, no. 2901 ; also by Botteri in the same locality without number ;
and by Ghiesbreght, in Chiapas, 1864-70, nos. 80, 684.

BouvARDiA OBOVATA, HBK. A stout glabrous herb nearly a me-
ter in height : stems simple, unbranched, angular, striate-suleate ; iuter-
nodes 5 to 13 cm. in length : leaves verticillate, 5 to 8 in each whorl,
narrowly obovate or oblanceolate, 4 to 12 cm. long, 1 to 3 cm. broad,
usually obtusish and submucronate, sometimes acute, entire, revolute,
subrepand and often hispidulous on the margins toward the apex, gradu-
ally narrowed below the middle to a subpetiolate base, rather dark green
to olive green in color (in the dried state), slightly paler beneath ; mid-
rib and veins rather prominent, especially beneath ; interpetiolar stipules
subdeltoid, acute or short-acuminate, unequally serrulate or sublaciuiate :
inflorescence in a terminal close subtrichotomous cyme : peduncles 9 to
12 mm. long; the ultimate divisions pruinose-puberulent : calyx deeply
4(-5)-parted ; divisions linear, about 3 mm. long, acute, often deeply
colored, ciliate or subhispidulous on the margins : corolla tubular, 4(-5)-
parted, 2^- to 3 cm. long, deep red or crimson, minutely pruinose or puberu-
lent on the outer surface, glabrous within ; lobes ovate, acute, about 4 mm.
long : anthers subsessile in the upper portion of the corolla-tube : style

GREENMAN. — MEXICAN PLANTS. 305

slender, filiform, exceeding the corolla ; stigma 2-2:)arted : capsule subglo-
bose, somewhat compressed laterally, 8 to 9 mm, long, equally broad, and
G to 7 mm, thick, — Collected by C. G, Pringle, on a mountain side near
Cuernavaca, State of Morelos, 18 November, 1895, no. 70G2, and in the
same locality in 1896, no, 6330, The above plant is in all probability
the one described in HBK. Nov. Gen. & Spec, iii, 385, notwithstand-
ing the little discrepancy as to the number of leaves in the whorl, and
the quadrangular stem ; the former character is a variable one, as shown
by many species of the genus, and the angulation of the stem depends to
a great extent upon the number of leaves in the whorl. It seems best,
therefore, in view of the very meagre description of the above species,
which was doubtless drawn from a very incomplete specimen, to regard
the above cited numbers of Mr. Pringle as representing the Bouvardia
obovata, HBK., and to amplify the characters of the same.

Hamelia nodosa, Mart. & Gal. Specimens agreeing in all essen-
tial characters with "W^alper's description of the above species (Walp.
Eep. vi. 51) were collected by Rev, Lucius C. Smith at Ojitlan, Oaxaca,
altitude 198 m,, 21 August, 1895, no, 605. The plant differs, however,
in having the fruit elliptic-oblong (6 to 9 mm. long, one half as broad)
instead of globose.

Crnsea, calcicola. An herbaceous annual : stems erect, branchinor
from the base, tetragonal, glabrous or hispidulous on the angles : leaves
lanceolate or lance-oblong, 2 to 4|- cm. long, 6 to 13 cm. broad, acute,
somewhat narrowed at the base, margin revolute, scabrous, midrib and
veins prominent beneath, glabrous on either surface, slightly paler be-
neath ; stipules puberulent, 3 to 4 mm, long bearing 3 to 7 setas on either
side : inflorescence in axillary or terminal pedunculate involucrate heads :
peduncles above puberulent and often hispidulous on the angles : floral
leaves (subtending the heads) 1 to 2 cm. long, ciliated at the broadened
base : bracts strongly dilated at the base, short-acuminate, shorter than
the floral leaves : flowers short-pedicellate, intermixed with numerous
laciniate pales : calyx small, about 2 mm. long, 4-lobed ; lobes slightly
exceeding the tube, narrow, acute, long-ciliate : corojla tubular, narrowed
below, somewhat ampliated above, 4-lobed, about 4 mm. long, white ;
lobes oblong, obtuse, usually bearing on the outer surface near the apex
a few spreading hairs : stamens exserted : stigma overtopping the anthers.
— Collected by C. G. Pringle, on dry calcareous hills. Las Sedas, Oaxaca,
at an elevation of 1,800 m., 8 September, 1894, no. 4869. Distributed
as C. cruciata, "Wats,, but differing from the latter species in the char-
acter of the stem, the stipules, and the bracts,
VOL. xxxii. — 20

806 PROCEEDINGS OF THE AMERICAN ACADEMY.

Crusea coccinea, DC, var. pubescens. Leaves ovate, acute,
often short-acumiuate, covered above with a scattered hirsute pubescence :
corolla-tube, especially the lower slender portion, pubescent on the outer
surface. — Collected by E. W. Nelson, on the west slope of Mt. Zempo-
altepec, altitude 2,100 to 2,400 m., 5-13 July, 1894, no, 581 ; also on
the northwestern slope of the same mountain, altitude 2,400 to 3,000 m,,
' 10 July, 1894, no. 699. Differs fi'om the species proper by the smaller
leaves, the pubescence of the same, and by the pubesceut corolla.

Crusea cruciata, Watson, var. villosior. Stem f to 1 m. in
height: leaves lanceolate, acuminate, 9 to 12 cm. long, 1 to 2 cm. broad;
floral leaves villous-pubescent on the lower surface near the base : flowers-
white to purple. — Collected by Dr. E. Palmer, at Jalisco, 1886, no.
901, and by C. G. Pringle, near Cuernavaca, altitude 1,600 m., 17 Sep-
tember, 1896, nos. 6508, 7225. With habit, laciniate stipules, and
essential characters of the type, but somewhat more robust, and with a
rather striking villous pubescence on the broadened basilar portion of the
floral leaves.

Crusea villosa, Watson, founded upon Pringle's no. 2448 from
Jalisco, collected in 1890, also Pringle's no. 3257 from the same locality,
collected in 1890 and distributed under the above name, may be referred
to Crusea Palmeri, Gray. Notwithstanding the somewhat remote locali-
ties, both of Mr. Pringle's numbers are identical with Dr. Gray's type,
collected by Dr. Edward Palmer in Southwestern Chihuahua.

Galium prsetermissum. Perennial : the slender woody roots
containing a red coloring matter : stems 3 to 4| dm. long, slender, weak,
decumbent, purplish at the base, green above, glabrous or hispidulous on
the angles : leaves in fours, sessile, linear-lanceolate, 8 to 14 mm. long,
1 to 3 mm. broad, acute, margin revolute (hispidulous on the older leaves),
upper surface usually hispidulous, shining, glabrous beneath : flowers in
subtrichotomous cymes : pedicels 3 to 10 mm. long, glabrous or minutely
hispidulous : corolla rotate, yellow, 4(-3)-lobed ; lobes ovate-oblong, ob-
tuse or short-acuminate and obtusish: fruit long-hirsute with uncinate
hairs. — Collected by C. G. Pringle, in pine woods, base of Sierra de
Ajusco, altitude 2,460 m., 19 September, 1896, no. 6596. A species with
the habit of G. Mexicnnum, HBK., but with leaves in fours, not pun-
gent-acuminate, and with much longer uncinate hairs on the fruit.

Viburnum elatum, Benth. Excellent flowering and fruiting speci-
mens, apparently belonging to this species, have been secured by Mr.
Pringle, and the following additional characters may be noted. Drupe
ovoid, 12 to 14 mm. long, black, covered with a blue bloom ; stone flat,

GREENMAN — MEXICAN PLANTS. 307

broadly ovate or ovate-oblong, 9 to 11 mm. long, nearly as broad, convex
ou one surface, flat on the other with a short ridge near the base. — Col-
lected by streams. Valley of Mexico, 10 October, 1896, no. 6226.

Stevia clinopodioides. An herbaceous perennial : roots fibrous,
clustered at the base of the stem: stems simple or nearly so, about 30 cm.
high, giving off at the base several slender horizontal subterranean
root-stocks, closely pubescent above and purple : leaves opposite below,
alternate above, spatulate, 24- to 4 cm. long, 6 to 10 mm. broad, obtuse
at the apex, gradually narrowed below to the subpetiolate base, callously
serrate along the upper half, entire below, appressed-puberulent on the
veins and on the upper surface near the margin, glandular-punctate : in-
florescence a compact dense umbel terminating the stems : heads about
1 cm. long, o-flowered ; involucral scales linear-acuminate, often slightly
unequal, appressed-puberulent on the outer surface, greenish purple, 6 to
7 mm. long : flowers considerably exceeding the involucre : corolla with
a slender greenish tube below, ampliated above, the upper ampliated
portion and the limb purple ; pappus paleaceous, short, exaristate :
achenes glabrous, 3 to 4 mm. long. — Collected by C. G. Pringle, on the
Serrania de Ajusco, altitude 3,000 m., 22 October, 1896, no. 6594.

Stevia diffusa. Annual : roots fibrous : stems herbaceous, erect, te-
rete, greenish purple, hirsutish pubescent with jointed hairs, interspersed
above with a stipitate glandular pubescence : leaves (including the petiole)
3 to 7 cm. long, 1 to 4 cm. broad, opposite below, becoming alternate above,
membranous, broadly ovate to ovate-lanceolate (the uppermost often
acuminate), acute at the apex, contracted below into a narrowly winged
and ciliated petiole, crenate-serrate, or the more reduced leaves with a
subentire margin, essentially glabrous upon either surface, except on
the veins, where they are sparingly hirtellous : heads widely separated
on long slender glandular pedicels, the latter from 1 to 1^ cm. long;
scales of the involucre 5, linear, acuminate, 6 mm. long, rather strongly
2-nerved, the outer surface appressed-puberulent, very rarely glandular :
flowers one third longer than the involucre: corolla discolorous, tube
purplish, puberulent, limb white : achenes puberulent, 3 mm. long: pap-
pus paleaceous and 3-aristate. — Collected by C. G. Pringle, on lava beds
near Cuernavaca, altitude 2,100 m., 3 November, 1896, no. 6608. Per-
haps related to S. micrantha. Lag., but with longer corollas, shorter
achenes, and a much more diffuse habit.

Stevia trachelioides, DC. It is interesting to note that Mr.
Pringle has rediscovered this apparently rare species on the Serrania
de Ajusco, altitude 3,000 m., 22 October, 1896, no. 6593, where specimens

308 PROCEEDINGS OF THE AMERICAN ACADEMY.

were secured which agree in every detail with the original plant collected
by Berlandier in the Valley of Toluca, no. 1164. Our plant as well as the
type shows the inflorescence and the scales of the involucre to be glandular-
pubescent, a character not mentioned in De Candolle's description, Prodr.
V. 115.

Eupatorium oreithales. Perennial from a horizontal root-stock :
roots fibrous, rather stout : stems herbaceous, erect, purple, pubescent,
5 to 6 dm. high, leafy below, nearly naked above, frequently branching
at the base : leaves opposite, broadly ovate, obtuse or rounded at the
apex, subtruucate or short-cuneate at the base, crenate, hirtellous on the
upper surface, puberulent on the veins beneath, becoming essentially
glabrous, 2 to 5 cm. long, nearly or quite as broad : petioles pubescent,
purple, 1 to 3 cm. long : heads few, disposed in loose terminal corymbs,
about 10 mm. high, 50-60-flowered : involucral scales about 2-seriate,
nearly equal, greenish purple, pubescent, linear-oblong, acute, ciliate,
about 6 mm. long : corollas white, 4 to 5 mm. long, tubes slender below,
about equalling the ampliated upper portion, the lobes pubescent on their
outer surfaces : pappus about as long as the corolla-tube : achenes 3 mm.
long, puberulent. — Collected by C. G. Pringle, on the Serrania de
Ajusco, altitude 2,400 ra., 28 September, 1896, no. 6563. Nearly related to
E. Saltivarii, Schz. Bip., but readily distinguished by its broader more
rotund and crenate leaves, and also by the nearly naked stem and longer
petioles.

Eupatorium euontjiifolium, Greene, Pittonia, iii. 31 = ^. Lem-
moni, Rob. Proc. Am. Acad, xxvii. 171.

EuPAToniUM KCELLT^FOLiuM, Greene, 1. c. = E. hyssopinum, Gray,
Proc. Am. Acad. xv. 28.

Gnaphalinm linearifolium. Annual, becoming somewhat ligneous
below, herbaceous above : stems terete, white-lanate, 2 to 4 mm. in diam-
eter, simple or branched, 7 to 10 dm. high : leaves sessile, linear-attenuate,
1 (-3)-nerved, margin revolute, arachnoid-pubescent above, white-tomen-
tose beneath, later becoming reflexed, 6 to 11 cm. long, about 2 mm. wide :
inflorescence corymbose : heads sessile or nearly so, 6 mm. high ; invo-
lucre about 4-seriate ; scales stramineous, the outer ovate-acute, inner
oblong-obtusish : flowers numerous, perfect flowers about ten : achenes
terete, glabrous. — Collected by C. G. Pringle, on rocky hills near
Guadalajara, 7 October, 1889, no. 2342 (distributed as G. leptophyl-
lum, DC.) ; on the Sierra de San Felipe, at an altitude of 2,150 m., 17
November, 1894, no. 5685 ; and also in the latter locality by C. L. Smith,
no. 592.

GREENMAN. — MEXICAN PLANTS. 309

Sclerocarpus Schiedeanus, Benth. & Hook., van elongatus.
A slender much branched herb, 1 to LV m. high: leaves lanceolate, o to
12 cm. long, 5 to 12 mm. broad : ray-llowers 5 to 8. — Collected by L. C.
Ervendberg, at Wartenberg near Tantoyuca, Prov. Huasteca, 1858,
nos. 98, 99, and by C. G. Pringle, in fields around Cuernavaca, altitude
1,600 m., 31 October, 1896, no. 6606. A variety readily recognized by
its slender habit and long narrow leaves.

Verbesixa oncopiiora, Rob. & Seaton? Specimens collected by
C. G. Priugle, under bluffs of barranca above Cuernavaca, altitude 2,000 m.,
1 November, 1896, no. 6600, agree in inflorescence and floral characters
with this species, but differ conspicuously by having opposite or suboppo-
site and more remotely denticulate leaves, also by the absence of any
excrescences on the stems at the base of the leaves. Further material
may show this plant to be worthy of specific rank, but for the present at
least the writer prefers to regard it as a form of the above named species.

"Verbesina stenophylla. Stems herbaceous, winged, erect from a
ligneous base, about 1 m. in height, essentially glabrous below, hirtellous-
puberulent above : leaves alternate, narrowly spatulate, 7 to 18 cm. long,
6 to 10 mm. broad, obtuse or acute at the apex, gradually narrowed below
to a sessile decurreut base, entire or inconspicuously denticulate, some-
what roughened on the margins, glabrous above, below glabrous or in the
earlier stages hirtellous, midrib prominent beneath, veins semitransparent,
rather strongly reticulated ; the upper leaves gradually reduced to linear-
attenuate bi'acts : inflorescence in a single terminal head or disposed in a
single cyme: heads few, about 12 mm. in diameter (including the rays
3 to 3^ cm. in diameter) ; involucral scales 2-3-seriate, linear, hirsute-
pubescent ; ligules yellowish white, about 10 in number, 10 to 12 mm.
long ; chaff oblong-linear, attenuate above to an acute apex, margin some-
what erose near the tip : achenes 6 mm. long, 4 mm. broad, strongly
winged with subciliated margins, and scattered hirsute pubescence on
either surface. — Collected by C. G. Pringle, on moist slopes above
Cuernavaca, altitude 2,000 m., 18 September, 1896, no. 6503.

Ballia Pringlei. Perennial by slender rhizomes : stems branching
from the base : branches ascending, 25 cm. or more high, hoary pubes-
cent : leaves opposite or alternate, triternately compound, segments lin-
ear, obtuse, white hirsute-pubescent: heads 12 mm. high, nearly 2 cm. in
diameter, usually on long slender peduncles : involucre about 3-seriate,
hirsute ; scales several, oblong-linear to subobovate, the inner ones with
scarious margins : rays yellow, 6 to 7 ram. long, 2 mm. broad : disk-
flowers numerous : corolla 5 mm. or more long, externally glandular-

310 PROCEEDINGS OP THE AMERICAN ACADEMY,

pubescent : pappus of about 8 broadly obovate scarious scales, 2 mm.
long, nearly or quite as broad : achenes puberulent, 4 to 5 mm. long. —
Collected by C. G. Pringle, on calcareous bluffs, near Tula, altitude
2,100 m., 6 August, 1896, no. 6407. Allied to B. Schaffneri, AVats., from
which it is distinguished by its perennial base, large heads and flowers.

Tagetes triradiata. An annual erect branching glabrous herb, 4 to
5 dm. high : stems sulcate, striate, greenish purple : lower branches oppo-
site, upper alternate, ascending : leaves imparipinnately divided to the
narrowly winged midrib ; divisions subopposite in 7 to 10 pairs, narrowly
lanceolate, dentate with acute and often setigerous teeth, 10 to 16 mm. long,
3 to 6 mm. broad, much reduced near the base of the leaf: inflorescence pan-
iculate, formed by the branches terminating in slender pedunculate heads:
peduncles somewhat thickened above, 2 to 3 cm. long : heads cylindrical or
somewhat spindle-shaped, 1 to 1|- cm. long, about 12-flowered ; involucre
greenish purple, obtusely 3-4-dentate, often split down the side to one
third its entire length ; rays three, bright yellow, including the achenes

14 mm. long, tube somewhat angular, puberulent on the angles, the ex-
panded portion unequally 3-lobed, about twice as broad as long ; disk-
flowers polygonal, puberulent : pappus variable, usually of 6 unequal
scales, 2, 3, 4, or even 5 of the scales hispidulous, long acuminate-aris-
tate, the others short linear-oblong or spatulate, obtuse ; achenes appressed-
puberulent, about 7 mm. long. — Collected by C. G. Pringle, Pedigral
(lava beds), Valley of Mexico, altitude 2,300 m., 12 October, 1896,
DO. 6592. In general habit closely resembling T. foetidissima, DC,
differing, however, in the number of the ray-flowers, the character of the
same, and also in the character of the pappus.

Cacalia suffulta. Stems 2 to 3 m. high, terete, sulcate-striate,
purple, glabrous, leafy throughout : lower leaves orbicular-ovate, cordate,

15 to 18 cm. long, 18 cm. or more broad, palmately nerved, 9-11-lobed,
mucronate-denticulate, above glabrous, below hirsutish on the strongly
reticulated veins; petioles 10 to 12 cm. long; upper cauline leaves con-
siderably smaller, 5 to 8 cm. long, equally broad, puberulent on the veins
above (becoming glabrous), hirsutish on the veins and nerves beneath,
becoming gradually reduced above to the sessile foliar bracts of the sub-
corymbose inflorescence : outer clusters of the corymb on long naked
striated purple peduncles: heads large, 2 to 2^ cm. high, about 2 cm. in
diameter, subtended and rather closely enveloped by ovate-oblong obtuse
sessile bracts : involucre about 2-seriate ; scales oblong, obtuse, 10 to
14 mm. long, 3 to 4 mm. broad, greenish purple, glabrous, somewhat
keeled on the back, much thickened at the base, apex often ciliated :

GREENMAN. — MEXICAN PLANTS. 311

flowers numerous (about 60), iucluding the achenes 2 cm. or more in
length : achenes short, columnar, striate, puberulent, 3 mm. long in tlie
flowering stage : pappus silvery white. — Collected by C. G. Pringle, on
a wet barranca above Cuernavaca, altitude 2,000 m., 1 November, 1896,
no. 6626. A very distinct species of Cacalia, the younger leaves much
resembling the foliage of Senecio reticulatus, DC, and with the inflores-
cence of C platylepis, Rob. & Seaton.

Ctnaua Carduxcellus, L. This very showy plant, native of the
Old World, appears not to have been hitherto recorded from Mexico. It
was collected by AYislizenus, at Saltillo, Coahuila, no. 310, and by E. W.
Nelson, at Celaya, Guanajuato, no. 3866.

Proceedings of the American Academy of Arts and Sciences.
Vol. XXXII. No. 17. — July, 1807.

PROCEEDINGS OF THE ACADEMY, 1896-1807.

REPORT OF THE COUNCIL: BIOGRAPHICAL NOTICES.
Francis James Child. By C. E Norton.
Thomas Tracy Bouv^. By W. O. Crosby.
Francis Amasa Walker. By C. F. Dunbar.
Benjamin Apthorp Gould. By Seth C. Chandler.

LIST OF THE FELLOWS AND FOREIGN HONORARY
MEMBERS.

STATUTES AND STANDING VOTES.

INDEX.

(Title Page and Table of Contents.)

PROCEEDINGS.

Eight hundred and eighty-third Meeting.

May 13, 1896. — Annual Meeting.

Vice-President Geoi-Q-e L. Gooclale in the chair.

The chair aunounced the death of Atticus Green Haygood,
Associate Fellow.

The Corresponding Secretary read the following letters : from
Charles Lane Poor and T. R. Lounsbury, acknowledging their
election as Associate Fellows ; from the Librarian of the Boston
Public Librar}', announcing the preparation of a list of serial pub-
lications now taken in the principal local libraries, and inviting
correspondence from persons owning periodicals not to be found
in public collections and willing to offer them for occasional
consultation ; from the College of New Jersey, inviting the
Academy to send a delegate to the sesquicentennial celebration
of the founding of the College ; and from the Principal of the
University of Glasgow and the Lord Provost of Glasgow, in-
viting the Academy to appoint a representative to attend the
jubilee of Lord Kelvin on the 15th and 16th of June next.
The chair appointed William Everett delegate to the celebra-
tion at Princeton.*

The Corresponding Secretary read the Report of the Council.

The Treasurer presented his annual report, of which the fol-
lowing is an abstract : —

* Mr. Everett being unable at tlie last moment to attend, his place was filled by
the appointment of Horace E. Scudder, as substitute.

316

PROCEEDINGS OF THE AMERICAN ACADEMY

General Fund.

Receipts.

Balance, May 1st, 1S95 S2,675.24

Sale of bonds 15,233.75

17,908.99

Assessments $1,050.00

Sale of publications 9.10 §1,059.10

Income from investments 4,677.00

Retm-n of bank tax 51.51

Gift of E. C. Clarke 100.00

Gift of F. H. Storer 5.00 5,892.G1

§23,801.60

Expenditures.

General expenses $2,039.00

Publishing expenses 1,939.14

Library expenses 1,157.78 §5,135.92

Investments 16,550.00

Balance 2,115.68

§23,801.00
RuMFOKD Fund.

§5,863.70
3,550.00

§11,457.4-

Receipts.

Balance May 1st, 1895

Sale of bonds

Income §1,952.00

Retm-n of bank tax 91.72 2,043.72

Expenditures.

Books and binding §140.17

Kent 10.00

Investigations 1,000.00

Publishing expenses 71.42

Riimford ]Medal 327.50 1,549.09

Investments 8,000.00

Balance 1,908.33

§11,457.42

OP ARTS AND SCIENCES. 317

Warren Fund.
Receipts.

Balance, May 1st, 1S95 $374.89

lucome 840.00

$1,214.89

Expenditures.

Investigations $600.00

Balance 614.89

$1,214.89
Building Fund.

Receipts.

Balance, May 1st, 1895 $843.08

Income 445.00

$1,288.68

Expenditures.

Investment $1,000.80

Balance 288.68

$1,288.68

The Librarian presented his annual report, which showed that
3,096 books and pamphlets had been added to the Library during
the year, 2,106 of which were obtained by gift and exchange,
748 by purchase with the appropriation from the General
Fund, and 242 with the appropriation from the Rumford Fund.
345 volumes were bound at an expense of $883.55, of which
822.35 were charged to the Rumford Fund. The total expendi-
ture for books, periodicals, and binding amounted to 81,297.95,
of which 8140.17 were charged to the Rumford Fund. 196
books were borrowed by thirty-five persons, of whom twenty-
three were Fellows of the Academy.

The following reports were also presented : —

Report of the Rumford Cojimittee.

The Rumford Committee has considered since the last annual meeting
of the Academy all possible candidates for the Rumford Medal, but has
been unable to reach a conclusion, and therefore does not recommend any

818 PROCEEDINGS OF THE AMERICAN ACADEMY

one for the medal at this time. The Rumford Committee has been pre-
sented by Professor Marcou with copies of forty-eight letters written by
Count Rumford to Professor Pictet of Geneva. The committee voted to
recommend to the Academy that it authorize the printing at the expense
of the Rumford Fund of the following four papers, and that a sum of
$25 be appropriated for illustration of these papers.
The titles of these papers are as follows : —

1. Thermo-electric Interpolation Formula;, by S. W. Ilolman.

2. The Melting Points of Aluminum, Silver, Gold, Copper, and Plati-
num, by S. W. Holman with R. R. Lawrence and L. Barr.

3. Pyrometry : Calibration of the Le Chatelier Thermoelectric Pyro-
meter, by S. W. Holman.

4. Calorimetry : Methods of Cooling Correction, by S. W. Holman.
The committee voted to authorize the publication in the American

Journal of Science of a preliminary account of an investigation on the
thermal conductivities of certain kinds of stone, by Prof. B. O. Peirce
and Dr. R. W. Willson. Professor E. H. Hall's paper on the Thermal
Conductivity of Mild Steel was presented to the Academy, January 8,
1896, and has been published by the aid of the Rumford Fund.

It is the desire of the Committee that the vote of the Academy at its
last annual meeting to appropriate from the Rumford Fund the sum of
one thousand dollars, to be expended at the discretion of the committee
in aid of investigations in light and heat, — payments from the sum to
be made on the order of the Chairman of the Committee, — should be
reaffirmed at the present meeting.

John Trowbridge,

Chairman Rumford Committee.

Report of the C. M. Warren Committee.

I have to report that the recipients of the money granted by the
Academy from the Cyrus M. Warren Fund have been doing good
work.

This statement is borne out very emphatically by the publications of
Professor Mabery, which are of a high order of excellence.

At the last annual meeting it was voted by the Academy " that the
C. M. Warren Committee be requested to consider the propriety of np
propriating $50 from the income of the C. M. Warren Fund for the
purchase of books relating to chemical research."

OF ARTS AND SCIENCES. 319

This question has been carefully considered by the committee. After
full debate, it was voted that '' lu the opinion of the Committee no gen-
eral approj^riatiou for the purchase of books can properly be made from
the C. M. Warren Fund."

F. H. Stoker, Chairman.

On the recommendation of the Committee of Finance it
■was

Voted., To make the following appropriations from the income
of the General Fund for the ensuing year : —

For general expenses $1,950.00

For the Library . 1,350.00

For publications 2,300.00

Voted, That the assessment for the ensuing year be five
dollars (85.00).

In accordance with the recommendation of the Rumford
Committee it was

Voted., That one thousand dollars (-^1,000) from the income
of the Rumford Fund be placed at the disposal of the Rumford
Committee, to be expended in aid of investigations in light and
heat, payments to be made on the order of the Chairman of the
Committee.

In accordance with the recommendation of the C. M. Warren
Committee, it was

Voted, To appropriate six hundred dollars (8600) from the
income of the C. M. Warren Fund to C. F. Mabery, of
Cleveland, Ohio, in aid of his researches on the chemistry of
petroleum.

Upon motion it was

Voted, That the thanks of the Academy be tendered to
Charles Loring Jackson for his faithful and efficient services
as Corresponding Secretary.

The annual election resulted in the choice of the following
officers and committees : —

320 PROCEEDINGS OF THE AMERICAN ACADEMY

Alexander Agassiz, President.
Benjamin A. Gould, Vice-President for Class I.
George L. Goodale, Vice-President for Class II.
Augustus Lowell, Vice-President for Class HZ
Samuel H. Scudder, Correspoyiding Secretary.
William Watson, Recording Secretary.
Eliot C. Clarke, Treasurer.
Henry W. Haynes, Librarian.

Councillors.

Henry Mitchell,

Leonard P. Kinnicutt, \ of Class L

Edwin H. Hall,

Henry P. Bowditch,

William M. Davis, \ of Class II.

Roland Thaxter,

Andrew M. Davis,

Horace E. Scudder, \ of Class HL

John E. Hudson,

3Iemher of Committee of Finance.
Augustus Lowell.

Run ford Committee.

John Trowbridge, Charles R. Cross,

Erasmus D. Leavitt, Amos E. Dolbear,

Edward C. Pickering, Benjamin A. Gould,

Arthur G. Webster.

C. 31. Warren Committee.

Francis H. Storer, Henry B. Hill,

Charles L. Jackson, Leonard P. Kinnicutt,
Samuel Cabot, Arthur M. Comey,

Robert H. Richards.

OP ARTS AND SCIENCES. 321

The chair appointed the following Standing Committees : —

Committee of Publication.

Samuel H. Scudder, Seth C. Chandler,

Crawford H. Toy.

Committee on the Library/.

Amos E. Dolbear, G. Stanley Hall,

SxVMUEL Henshaw.

Auditing Committee.
Henry G. Denny, John C. Ropes.

In accordance with a vote adopted at the meeting of the 8th
of April, the Rumford Premium, awarded at the last annual
meeting to Thomas Alva Edison, was presented, John Trow-
bridge acting as Mr. Edison's proxy.

Vice-President Goodale, in presenting the medals, made the
folio winsf remarks : —

'&

It would be highly presumptuous for one whose knowledge of physics
is of the most elementary character to occupy the time of the Academy
by any statement of his own in conveying these medals. Happily such
a course is unnecessary. The Chairman of the Rumford Committee has
placed at our command a brief statement which makes clear the ground
of the award.

"The Rumford Committee voted, June 22, 1893, that it is desirable to
award the Rumford Medal to Thomas Alva Edison, in recognition of his
investigation in the field of electric lighting, and they confirmed this vote
on October 9, 1893, in the following words : ' Voted for the second time
to recommend to the Academy that the Rumford Medal be awarded to
Thomas Alva Edison for his investigations in electric lighting.'

" The Committee reached the conclusion expressed by these votes after
lonf deliberation and after careful sifting of all the evidence which was
at their disposal in regard to Mr. Edison's claim for priority in the con-
struction of the incandescent lamp, the conception of the central lighting
station, too'ether with the multitude of devices, such as the three-wire
circuit, the disposition of the electric current feeders, and the necessary
methods for maintaining the electric potential constant.

VOL. XXXII. — 21

322 PROCEEDINGS OF THE AMERICAN ACADEMY

" The Committee felt that they could not decide upon Mr. Edison's
claims for priority in any particular invention in this new industry. In-
deed, courts of law after prolonged litigation have found it difficult to de-
cide how far Mr. Edison was in advance of contemporary workers. The
task given to the Rumford Committee to decide who is the most worthy
of the Rumford Medal, especially in the field of the application of elec-
tricity for the production of light and heat, is not an easy one. The
number of investigators is now so large that it is no longer possible in
general for one man to claim to be the first to apply electricity to a new
field. The successful application is the result of many minds working
on the same problem. Although the Committee did not feel justified in
expressing the opinion that Mr. Edison invented the incandescent carbon
filament lamp, or that he was the first to arrange such lamp in multiple
on the circuit, thus producing what is popularly termed a subdivision of
the electric light, or that the Edison dynamo had greater merits than the
machine of Gramme and Siemens and others ; still they are convinced
that Mr. Edison gave a great impulse to the new industry, and that he was
the first to successfully install a central electric lighting plant with the
multitude of practical devices which are necessary. They believe that this
impulse was due to his indefatigable application, to his remarkable in-
stinct in whatever relates to the practical application of electric circuits,
and to his inventive genius. They therefore have unanimously recom-
mended to the Academy to bestow the Rumford Medals upon him, feel-
ing that the work of Mr. Edison would especially appeal to the great
founder of the medals, Count Rumford, if he were living."

The Academy has accepted the report of the Rumford Committee,
and has voted to confer the gold and the silver medal upon Mr. Edison.
The recipient finds it impossible to be present at this meeting of the
Academy, and has requested Professor Trowbridge to act as his proxy,
and to receive the medals for him.

In the name of the Academy I beg you. Professor Trowbridge, to
accept the charge of conveying these medals to Mr. Edison's hands.
It would be most ungracious for us who are assembled in this room,
which is flooded by this steady and brilliant electric light, to withhold
our personal thanks for what Mr. Edison's investigations and practical
activities have done for us all. And, hence, I may venture to say that
our thanks and all good wishes are to be conveyed with the Rumford
Medals.

Professor Trowbridge replied as follows : —

OF ARTS AND SCIENCES. 823

Mr. President and Gentlemen of the Academy, — I accept the medals
for Mr. Edison, and at his request I wisli to express his deep sense of
the great honor the Academy has conferred upon him. His work in the
tield of electric lighting has been the subject of prolonged litigation, and
at times he has had doubts in reading the opinions of learned experts
whether this work has been original or whether he has really contributed
anytliiiig to the world's progress. The recognition of his labors by the
American Academy of Arts and Sciences, regarded by Count Rumford
in his gifts as the coequal of the Royal Society of London, is therefore
especially grateful to him. Acting as his proxy, I thank the members of
the Academy for the distinction which they have by their votes conferred
upon him.

The following papers were presented by title : —

A Revision of the Atomic Weight of Magnesium. By Theo-
dore William Richards and H. G. Parker.

A Revision of the Atomic Weight of Strontium. Second
Paper : The Analysis of Strontic Chloride. By T. W. Pdchards
and H. G. Parker.

Contributions from the Gray Herbarium of Harvard Uni-
versity. New Series, No. 10 : — 1. Revision of the Genus Tridax.
2. Synopsis of the Mexican and Central American Species of the
Genus Jlilcania. 3. Revision of the Genus Zinnia. 4. Revision
of the Mexican and Central American Species of the Genus
Calea. 5. A Provisional Key to the Species of Porophyllum
ranging north of the Isthmus of Panama. 6. Descriptions of
new and little known Phanerogams, chiefly from Oaxaca. By
B. L. Robinson and J. M. Greenman.

On the group of real Linear Transformations whose Invariant
is a real Quadratic Form. By Henry Taber.

David Wo Cheever read an obituary notice of Richard M.
Hods:es.

o

Eight hnndred and eighty-fourth Meeting.

October 14, 1896. — Stated Meeting.

The President in the chair.

The following deaths were announced : Thomas Tracy Bouvd
and Francis James Child, Resident Fellows ; George Brown

324 PROCEEDINGS OF THE AMERICAN ACADEMY

Goode and Hubert Anson Newton, Associate Fellows ; and
Heinrich Ernst Beyrich, Ernst Curtius, Friedrich August Ke-
kule, Baron Ferdinand von Mueller, and Jules Simon, Foreign
Honorary Members.

In the absence of the Corresponding Secretary, Charles Lor-
ing Jackson was requested to act in his place, and read the
following letters : from Karl Weinhold, Alexandre Kovalevsky,
M. Foster, and B. L. Gildersleeve, acknowledging their election
into the Academy ; from a committee of the Society of Colonial
Wars in the District of Columbia, proposing a joint memorial
meeting in memory of George Brown Goode ; from the Inter-
national Geological Congress, enclosing the programme for its
seventh session ; from the Northwest Mining Association, sug-
gesting the creation of a Department of Mines in the United
States Government as a cabinet office ; from the widow of
Heinrich Ernst Beyrich, announcing the death of her husband ;
from the Academy " Gioenia " of Natural Sciences at Catania,
announcing the death of its President, Giuseppe Zurria ; from
General-Major Rykatchew, reporting his appointment as Direc-
tor of the Central Physical Observatory at St. Petersburg in
place of Heinrich Wild, resigned ; and from the committee of
the British Association for the Advancement of Science on
Zoological Bibliography and Publication, asking for the co-
operation of learned societies in the method of publishing sci-
entific papers.

James Coolidge Carter, of New York, was elected an Asso-
ciate Fellow in Class III., Section 1, in place of the late Atticus
Green Haygood.

Alpheus Hyatt read a paper entitled, Cycle in the Life of the
Individual (Ontogeny) and in the Evolution of its own Group
(Phylogeny).

The following papers were presented by title : —

The Constituents of Pennsylvania, Ohio, and Canadian
Petroleum between 150° and 220°. By Cliarles F. Mabery.

Refractive Power of the Hydrocarbons and Clilorine Deriva-
tives described in the Preceding Paper. By Charles F. Mabery
and Edward J. Hudson.

OP ARTS AND SCIENCES. 325

On the Butanes and Octanes in American Petroleum. By
Charles F. Mabery and Edward J. Hudson,

On the Sulphur Compounds and Unsaturated Hydrocarbons
in Canadian Petroleum. By Charles F. Mabery and William
O. Quayle.

On the Composition of a South American Petroleum. By
Charles F. Mabery and Arthur S. Kittelberger.

On the Specific Refractive Power of Petroleum Hydrocar-
bons, and some of their Derivatives. By Charles F. Mabery
and Edward J. Hudson.

The Viscosity of Mercury Vapor. By A. A. Noyes and
H. M. Goodwin.

Contributions from the Zoological Laboratory of the Museum
of Comparative Zoology at Harvard College. E. L. Mark, Di-
rector. No. 73. — Studies in Morphogenesis. 6. A Contribution
to the Quantitative Study of Correlated Variation and the
Comparative Variability of the Sexes. By C. B. Davenport
and C. Bullard.

Voted, To meet on adjournment on the second Wednesday
in November.

Eight hundred and eighty-fifth Meeting.

November 11, 1896. — Adjourned Stated Meeting.

Vice-President George L. Goodale in the chair.
As a quorum for business was not present it was
Voted, To meet on adjournment on the second Wednesday
in December.

Eight hundred and eighty-sixth Meeting.

December 9, 1896. — Adjourned Stated Meeting.

The Academy met at the house of the President in Cambridge.
The President in the chair.

326 PROCEEDINGS OP THE AMERICAN ACADEMY.

The chair announced the death of Benjamin Apthorp Gould,
Vice-President of the Academy for Class I., and of Hugo Gyl-
d^n, Foreign Honorary Member.

The Corresponding Secretary read a letter from James C.
Carter, acknowledging his election as Associate Fellow.

On the motion of Charles L. Jackson it was

Voted, That the money received from the sales of publica-
tions in any year be appropriated for the publications of that
year.

Voted, That the sum of four hundred and sixty-one dollars
be added to the appropriation for publications this year, as this
is the amount of the subscription raised to be used for publica-
tions, but not yet appropriated for this purpose.

On the motion of S. C. Chandler it was

Voted, To grant the use of the Hall of the Academy to the
Colonial Society of Massachusetts on the third Wednesday of
each of the five months December to April, in the same way as
during the past year.

In view of the expected absence of the Recording Secretarj^
the President appointed William R. Livermore Recording
Secretary pro tempore.

John E. Wolff exhibited a crystal of tourmaline from Paris,
Maine, which had just been given to the Harvard Mineralogical
Museum and is remarkable for its size and color. It is seven
inches long, with a diameter varying from two to three inches,
and weighs twelve hundred and twelve grammes, or two pounds
eleven ounces avoirdupois. The color is a clear, transparent
green, except at one end, where the crystal is tipped with red
(rubellite) for half an inch. This is said to be the largest crystal
of gem quality ever taken out from that famous locality.

William R. Livermore read a paper entitled, The Growth of
the Roman Dominions in Europe from 650 b. c. to 200 A. D.,
illustrated by one hundred Sketches for an Atlas of Historical
Geography.

The President described his recent explorations of the Florida
elevated reefs, in connection with those of L. S. Griswold of
Southern Florida.

OP ARTS AND SCIENCES. 327

Eight hundred and elRhty-seventh Meeting.

January 13, 1897. — Stated Meeting.

The President in the chair.

The chair aunounced the death of Francis Amasa Walker,
Resident Fellow ; Henry Newell Martin, Associate Fellow ; and
Emil Heinrich Du-Bois Reymond, Foreign Honorary Member.

The Corresponding Secretary read letters from the Secretary
of the Pasteur Committee in Washington, announcing the open-
ing of subscriptions in America, and inviting the Academy to
participate, and from the Observatory of San Salvador announ-
cing the death of its director, Don Alberto Sanchez.

The vacancies occasioned by the death of Benjamin A. Gould
were filled by the election of

John Tno\y bridge. Vice President for Class I.

Theodore W. Richards, Member of the Rumford Committee.

On the recommendation of the C. M. Warren Committee it
was

Voted, That the sum of two hundred dollars (8200) from the
income of the C. M. Warren Fund be granted to H. O. Hofman,
of Boston, for the prosecution of a research on the formation
temperatures of slags in the smelting of ores of lead and copper
or other metals.

S. H. Scudder read by title a paper on The Genera of North
American Melanopli.

A. Lawrence Rotch read a paper on Obtaining IVIeteorological
Records in the Upper Air by Means of Kites and Balloons.

Eight hundred and eighty eighth Meeting.

February 10, 1897.

The Academy met at the house of A. Lawrence Rotch.
The President in the chair.
The following papers were read : —

The Hebrew and Muhammedan Calendars. By William E.
Story.

Forts Henry and Donelson. By John C. Ropes.

328 PROCEEDINGS OF THE AMERICAN ACADEMT

Eight hundred and eighty-ninth Meeting.

March 10, 1897. — Stated Meeting.

Vice-President John Trowbridge in the chair.

The Corresponding Secretary read a letter from the Royal
Academy of Sciences of Turin, announcing the death of Luigi
Schiaparelli ; also, one giving notice of the terms for competi-
tion for the eleventh Bressa prize.

The chair appointed the following committee to nominate
officers for the ensuing year : —

Leonard P. Kinnicutt, of Class I., Henry P. Bowditch, of
Class II., and Andrew M. Davis, of Class III.

The following gentlemen were elected Members of the
Academy : —

Charles Henry Fernald, of Amherst, to be a Resident Fellow
in Class II., Section 3 (Zoology and Physiology).

Ludwig Boltzmann, of Vienna, to be a Foreign Honorary
Member in Class I., Section 2 (Physics), in place of the late
Friedrich August Kekule.

Wilhelm Pfeffer, of Leipsic, to be a Foreign Honorary Mem-
ber in Class II., Section 2 (Botany), in place of the late Baron
Ferdinand von Mueller.

Wilhelm Dorpfeld, of Athens, to be a Foreign Honorary
Member in Class HI., Section 2 (Philology and Archaeology),
in place of the late Ernst Curtius.

The following papers were read : —

International Bimetallism. By George S. Boutwell.

On the Minimal Nutrients of Bacteria in Water. By W.
T. Sedgfwick and D. D. Jackson.

An Investigation of some of the Bacteriological Aspects of
the Art of Tanning. By S. C. Prescott. (By invitation.)

The Energy Conditions necessary to produce the Rcintgen
Rays. By John Trowbridge. (By title.)

OF ARTS AND SCIENCES. 329

Eight hundred and ninetieth Meeting.

April 14, 1807.

The Academy met at the house of John C. Ropes.

Vice-President George L. Goodale in the chair.

The chair announced the death of James Joseph Sylvester
and Karl Weierstrass, Foreign Honorary Members in Class I,
Section 1.

The following papers were read : —

The study of English at Harvard College. By Barrett
Wendell.

Electricity and Ether. By Arthur G. Webster.

The following papers were presented by title : —

A Measure of Variability and the Relation of individual
Variations to specific Differences. By Edwin Tenney Brew-
ster. Presented by E. L. Mark.

Contributions from the Gray Herbarium of Harvard Univer-
sity. New Series, No. XI. — 1. Revision of the Mexican and
Central American Species of Houstonia. 2. Key to the Mexi-
can Species of Liahum. 3. Descriptions of new and little
known Plants from Mexico. By J. M. Greenman. Presented
by B. L. Robinson.

AMERICAN ACADEMY OF ARTS AND SCIENCES.

Eeport of the Council. — Presented May 12, 1897.

BIOGRAPHICAL XOTICES.

Francis James Child By C. E. Xortox.

Thomas Tracy Bouv]^ W. O. Crosby.

Francis Amasa Walker Charles F. Dunbar.

Benjamin Apthokp Gould Seth C. Chandler.

Proceedings of the American Academy, Vol. 32.

ERRATUM.

On page .333, line five, of the Memoir of Francis James Child, for
"The hoy was the youngest," etc., read: "The boy was the third in
a family of eight brothers and sisters."

EEPORT OF THE COUNCIL.

Since the Annual Meeting of "Slay 13, 1896, the Academy
has lost bv death seventeen members : — four Resident Fellows,
Thomas Tracy Bouv^, Francis James Child, Benjamin Apthorp
Gould, and Francis Amasa Walker ; five Associate Fellows,
George Brown Goode, Atticus Green Ha3'good, Matthew Carey
Lea, Henry Newell Martin, and Hubert Anson Newton ; and
nine Foreign Honorary Members, Heinrich Ernst Beyrich,
Ernst Curtius, Eniil Heinrich Du-Bois Reymond, Hugo Glyden,
Friedrich August Kekule, Baron Ferdinand von Mueller, Jules
Simon, James Joseph Sylvester, and Karl Weierstrass.

FRANCIS JAMES CHILD.

Francis James Child was born in Boston, on the 1st of February,
1825. His fatlier was a sailmaker, one of that class of intelligent and
independent mechanics wliich has had a large share in determining the
character of our democratic community, as of old the same class had in
Athens and in Florence. The boy was the youngest of five brothers and
sisters. He was sent to the public schools. His unusual capacities Avere
early displayed. He stood first in his classes, and was a favorite with his
schoolfellows. At the English High School he won all the prizes, and
having by chance attracted the attention of our venerable fellow citizen,
Mr. Epes S. Dixwell, then the Master of the Latin School, his father was
induced, at Mr. Dixwell's suggestion, to allow him to proceed to the Latin
School, that he might continue his studies and be prepared for entrance
to college. He speedily caught up with the boys who had already made
progress in the study of Greek and Latin, and soon took the first place
here, as he had done in the schools which he had previously attended.
The sweetness of his disposition, the pleasant mingling in his nature of
gav spirits and serious purpose, his high principles, his unaffected modesty
won the affection of his teachers and of his comrades. His superiority in

334 FRANCIS JAMES CHILD.

his classes was so unmingled with pretension or conceit, that it was ad-
mitted without question or envy. Mr. Dixwell became strongly attached
to him, and, in view of the great promise of his talents and his character,
secured the means for his support in college, which he entered in the
autumn of 1842,* Harvard was then still a comparatively small institu-
tion, with no claims to the title of University ; but she had her traditions
of good learning as an inspiration for the studious youth, and still better
she had teachers who were examples of devotion to intellectual pursuits,
and who cared for those ends the attainment of which makes life worth
living. Josiah Quiucy was approaching the close of his term of service
as President of the College, and stood before the eyes of the students as
the type of a great public servant, embodying the spirit of patriotism, of
integrity, and of fidelity in the discharge of whatever duty he might be
called to perform. Among the Professors were "Walker, Felton, Peirce,
Channing, Beck, and Longfellow, men of utmost variety of temperament,
but each an instructor who secured the respect no less than the gratitude
of his pupils.

The Class to which Child belonged numbered hardly over sixty. The
prescribed course of study which was then the rule brought all the mem-
bers of the Class together in recitations and lectures, and every man soon
knew the relative standing of each of his fellows. Child at once took
the lead and kept it. His excellence was not confined to any one special
branch of study, he was equally superior in all. He was the best in the
classics, he was Peirce's favorite in mathematics, he wrote better Eng-
lish than any of his classmates. His intellectual interests were wider
than theirs, he was a great reader and his tastes in reading were mature.
He read for amusement as well as for learning, but he did not waste his
time or dissipate his mental energies over worthless or pernicious books.
He made good use of the social no less than of the intellectual oppor-
tunities which college life affords, and became as great a favorite with his
classmates as he had been with his schoolfellows.

The close of his college course was marked by the exceptional distinc-
tion of his being chosen by his classmates as their Orator, and by his
having the first part at Commencement as the highest scholar in the
Class. His Class Oration was remarkable for its maturity of thought
and of style. Its manliness of spirit, its simple directness of presentation

* The pecuniary debt thus incurred was afterwards paid with interest. But
thoutjh only thus could Mr. Child's spirit of independence he satisfied, he clierished
through life the most grateful affection for the friend who had thus served him.

FRANCIS JAMES CHILD. 835

of the true objects of life, and of the motives by which the educated man,
whatever might be his chosen career, should be inspired, together with
the serious and eloquent earnestness with which it was delivered, gave
to his discourse peculiar impressiveuess and effect.*

Immediately upon his graduation he was appointed Tutor in the Col-
lege, with duties of instruction in English. To the study of the English
language and literature he was led by taste, and his knowledge was
already considerable in this wide field, to the cultivation of which the
remainder of his life was to be in great part devoted, and in which he was
to become an acknowledged master. In 1848 he published his first work,
an edition in one volume of " Four Old Plays," all of the sixteenth
century, and of interest to the student of the development of the English
drama as exhibiting its conditions immediately before its splendid mani-
festation in the works of the Elizabethan playwrights. Nothing of the
kind had l)een done previously in America. The volume appealed to
but a small class of readers, but, with those who were competent to
judge ofit, it established the reputation of its editor as a scholar of more
than usual competence of learning and sobriety of judgment.

In 1851, on the resignation of Professor Channing as Boylston Pro-
fessor of Rhetoric, Child was appointed his successor with leave of ab-
sence for study in Europe, before assuming the duties of the position.
The opportunities which Europe then afforded to the young American
scholar were diligently made use of. He obtained the degree of Doctor
of Philosophy at Gottingen in 1854, and in the autumn of the same year
he returned to his work at Harvard. A great part of his time was em-
ployed in the teaching of English composition, and the drudgery of cor-
recting students' exercises, but he had an indefatigable industry and a
steady ardor of learning, and he found time to carry on his own special
studies. He undertook the general superintendence of a series of the
works of the chief British Poets, and himself prepared for it the edition
of Spencer (1855) in five volumes, which for the use of the general
reader still remains the best. For the same series he compiled a Col-
lection of Ballads in eight volumes, published in 1857-58, which in
extent of range, in judgment in selection, and in thoroughness of liter-
ary and historical illustration, was far superior to any preceding similar

* An eminent living graduate of Harvard, who was present on the occasion,
having come to Cambridge to take his entrance examination, has said that he re-
ceived from that oration his first vivid sense of the dignity of intellectual pursuits,
and his first strong impulse to devote himself to them.

«

336 FRANCIS JAMES CHILD.

work. But a more important piece of work, one of original investiga
tiou and displaying " wonderful industry, acuteness, and accuracy,"
was the treatise issued in the Memoirs of our Academy in 1862 under
the modest title of " Observations on the Language of Chaucer," f
which was followed in 1868 by a Supplement, entitled "Observations
on the Language of Gower's Confessio Amantis." t " It is dilficult
at the present day," says Professor Kittredge, " to imagine the state of
Chaucer philology at the moment when this paper appeared. Scarcely
anything, we may say, was known of Chaucer's grammar and metre
in a sure and scientific way. Indeed, the difficulties to be solved had
not even been clearly formulated. . . . Mr. Child not only defined the
problems, but provided for most of them a solution which the researches
of younger scliolars have only served to substantiate. He also gave a
perfect model of the method proper to such investigations, — a method
simple, laborious, and exact." §

For many years after this Mr. Child published little, but with steady
purjaose devoted such leisure as his incessant professional task allowed to
the extension of the vast stores of his learning, and to the accumulation
of the material for the main work of his life, a complete critical edition
of " The English and Scottish Ballads." At length in 1882 appeared
the first part of his work. The character of the undertaking was set
forth in a prospectus. The popular Ballads existing in the English lan-
guage had never been collected into one body ; a large portion of the
remains of the ballads was unprinted ; the text of much that was in print
was vitiated by editorial changes ; it was now proposed to publish all in
their entirety and their purity ; to include every independent version of
every ballad, and to record all important variations of different copies,
both printed and manuscript ; each ballad was to have a proper Preface,
and in the case of those ballads which the English have in common with
other nations an account was to be given of related traditions. The
work was to be completed by a general introduction, a glossary, and in-
dexes. The vast scale of this matured design became obvious on the
publication of the first part. The large range of the themes of the bal-
lads, the immense variety of local, histoi'ical, and romantic tradition

* These are the terms used by Mr. A. J. Ellis, the learned author of the History
of English Pronunciation.

t Memoirs of the American Academy, New Series, Vol. VIII. pp. 445-502.

t Ihid., Vol. IX. pp. 2G5-315.

§ From the admirable appreciation of Professor Child's character and works in
the Atlantic Monthly for December, 1896.

FRANCIS JAMES CHILD. 337

exhibited by them, the wide diffusion among the nations of Europe of
the legends which many of them embodied, opened a field of investiga-
tion of enormous extent, requiring acquaintance alike with many lan-
guages and many literatures. The task was one with which only a
scholar possessed of exceptional acquisitions could hope to accomplish
satisfactorily, and from which even the most industrious might have
shrunk. It had hardly a parallel in the variety of learning which it
exacted for its due performance, but this was not all ; it demanded in no
less measure fine critical acumen and poetic appreciation, — the gifts of
taste and culture as well as of scholarship. Mr. Child possessed them all.

" It was my wish," he said, in the Advertisement prefixed to the first
part, " not to begin to print The English and Scottish Popular Ballads,
until this unrestricted title should be justified by my having at command
every valuable copy of every known ballad. . . . What is still lacking is
believed to bear no great proportion to what is in band. . . . Meanwhile
the uncertainties of the world forbid a longer delay to publish so much as
has been got together."

From year to year the parts followed in rapid succession, — rapid in
view of the character of the work, — and in the Advertisement to
Part IX., which appeared in the spring of 1894, Mr. Child had the
satisfaction of saying that, to the extent of his knowledge of sources
the collection was complete, with the exception of a single ballad, " which
is probably a variety of one or another here given in several forms."
Such had been the extraordinary success of his research, in which he had
been aided by many English and foreign scholars glad to assist in the
perfecting of a work which was of interest to them in itself, and which
roused their admiration by the manner in which it was executed.

The body of the work was complete ; but one more part was needed,
to contain a general Introduction, a Glossary, and Indexes. In spite of
failing health Mr. Child set himself resolutely to the drudgery involved
in this task. To the last month of his life he labored steadily, but with
a sense of weariness to which he had been unused. With the exception
of the Introduction the task was mainly accomplished, and the work was
left in such condition that it could be taken up and carried through by
the most competent hands next to Mr. Child's own, those of his discijile,
assistant, and friend, Professor Kittredge.

The year 1895-96 completed the fiftieth year of Mr, Child's service in

the University. It was a matter of satisfaction to him that during its

course he had been able, in spite of physical infirmity, to meet his classes

without the omission of a single lecture. The College year ended in

VOL. XXXII. — 22

338 FRANCIS JAMES CHILD.

June. He took no vacation, but busied himself in his study, and found
pleasure in his garden of roses. In August he became seriously ill, and
on the 11th of September he died.

Mr. Child's fame as a scholar is secure. His work is so done that it
can never be superseded. But to those who had the happiness of in-
timacy with him, his learning and all that he accomplished seem but as
secondary and accessory to the essential qualities of his character and his
manner of life. The sedentary nature of his occupations, and their nar-
row material confines within the limits of a University helped to preserve
the strongly marked and altogether delightful originality of his nature
from the pressure and attrition of the world, which speedily wear down
the marks of distinctive individuality and shape the mass of men into a
general dull uniformity. He had a most sympathetic and tender heart,
so easily touched that its impulses might sometimes have overcome the
restraint of good judgment, had they not been encountered by his keen,
kindly, and lively humor, which, while it generally saved him from senti-
mental extravagances, yet became often the inciter and ally of his liberal
sympathies. 'His charity might be abused, but his pity included even
the most open of impostors, and, taking a humorous enjoyment in the suc-
cess of deceits practiced upon himself, he chose rather to aid the unde-
serving than to let a single deserving needy man go by unhelped.

The same liberality of disposition was manifest in his relations to all
whom he could assist in literary or scholarly work. He made a friend
of every young scholar who sought from him advice or direction, and
gave his time willingly to serve interests not his own. He could be
merciless with pretenders, but he was marvellously patient with unpre-
tending and innocent incompetence.

With the highest sense of the duties and the privileges of his calling,
he did not regard them as exempting him from the discharge of the com-
mon duties of a citizen. He did not bury himself in his books, and he
had nothing of the indifference of a recluse to the affairs of the commu-
nitv in which he lived. His feedincrs were stronjj and his iudiiment was
sound in regard to the matters affecting public interests. His opinions
carried weight, for they were based on principles and maintained with
clear intelligence and ready wit. If roused by argument, no one was his
match in the flash of wit and the j^lay of humor. He took the part of a
good citizen in local politics ; he was for many years an active member
and officer in local charities, and he served his term as a member of the
School Committee. At the time of the civil war he threw himself with
ardor into the service of the cause to which so many of the youth of Har-

FRANCIS JAMES CHILD. 839

vard, dear to him, were devoting themselves. He cherished with pecu-
liar teuderness the memory of those who fell in the war. He was the
main promoter of the two precious volumes of Harvard Memorial Biogra-
phies ; and on the walls of his study, always before his eyes, close by the
portraits of his old masters in learning, the brothers Jacob and Wilhelm
Grimm, were those of his young heroic friends, the brothers Charles and
James Lowell.

His fidelity in the discharge of the exacting duties of his professorship
was complete. For far too many years far too much of his time was
occupied in the correction of students' themes. He never shirked this
wearisome drudgery. No teacher was ever more exacting of himself iu
the discharge of his regular duties. In the later years of his life, when
he suffered much from gout and rheumatism, he did not allow pain or de-
pression of spirits to interfere with the regular discharge of his task as
instructor.

Even the dullest and most careless vmdergraduate could hardly fail to
be quickened and improved by such teaching as Mr. Child's. Here was
a master of most accurate and extensive learning, a scholar of unwearied
diligence and exact method, with the faculties and sympathies which
enabled him to impart his learning to his pupils, and to inspire in the
more capable among them something of his own enthusiasm for the best
in literature and life.

It is impossible not to regret that Mr. Child should not have done
more independent literary work. The several introductions to the Bal-
lads in his great collection, excellent as they are in their kind, very
seldom afford him free space for the display of his own genius ; but they
abound in touches which light up the page with gleams of fancy or of
humor, and more rarely with a flash of poetic imagination that reveals
the restraint which the editor had imposed upon himself. His style when
at freedom was of tiie best, — for it was the simple expression of the
man himself.

Original, quaint, humorous, sweet, sympathetic, tender-hearted, faith-
ful, — these are the terms which first come to mind in describing him ;
the traits that these terms imply included all his intelligence, gave char-
acter to his work, and made his learning the least part of him.

Those who knew him best think of him mainly as one who had the
gift of love. He was a lover of nature, of poetry, of roses, of all that
was fair and sweet and good ; above all he was a lover of his fellow men.
When he died the world lost much more than one of its great scholars.

C. E, Norton.

340 THOMAS TRACY BOUVE.

THOMAS TKACY BOUVE.

Thomas Tracy Bouve, Resident Fellow of the Academy, was born
on Prince Street, in Boston, January 14, 1815, and died in the city of
his birth, June 3, 1896. He was of Huguenot descent. His parents
were Ephraim Osborn and Lydia (Tracy) Bouve, and his paternal
grandfather was Jonathan Bouve, a soldier of the Revolution. In
1839 he married Emily G. Lincoln ; and his wife and five children sur-
vive him.

Very early in life he attended a private school ; at the age of seven
years he entered the Eliot public school, received the Franklin medal
when twelve years of age, and then entered the English High School.
It being necessary for him to do something for self-support, he left the
school and entered a dry goods store. At fifteen he entered the service
of an iron manufacturing company, where he became bookkeeper and chief
accountant ; later, he entered the office of the Hamilton, Appleton, and
Lowell Manufacturing Companies to take charge of the books of these
respective companies, and he also kept the books of the Great Falls
Manufacturing Company. When about twenty-six years of age, Mr.
Bouve became a partner in a commission iron house, remaining in the
business thirty years ; and in 1870 he became treasurer of the Glendon
Iron Company, retaining this office until his death.

Although his life was thus burdened to the end with the cares and
responsibilities of a successful business career, they could not fetter his
mind or stifle his interest in more distinctly intellectual pursuits. After
leaving school he devoted much time to the study of chemistry, French,
and Latin, and received the degree of A. M. from Harvard University in
1850. But the great central fact of Mr. Bouve's life was his deep and
abiding love of nature. This was manifested at a very early age, and
throughout his long and busy life he was an ardent and steadfast student
of natural history. His lifelong friend and associate. Dr. James C AVhite,
justly describes him as a fine example of the amateur naturalist, and
what this class has done for the advancement of science in this country.

Mr. Bouve was only fifteen years old when the Boston Society of
Natural History was founded. Four years later he became a member
of the Society ; and his labors in its behalf extended over more than
threescore years, the period of his membership being the longest in its
history. In fact, it was in connection with the Boston Society of Natural
History, and especially in the building up of its collections, that his
interest in natural history was chiefly manifested.

THOMAS TRACY BOUV^. 341

After serving the Society efficiently in almost every subordinate ca-
pacity, he was, in 1870, elevated to the presidency. The success of his
administration is attested by the spontaneous and earnest expressions of
regret when, in 1876, he desired to resign his high office. At the meet-
ing prior to the annual election, to use his own words, "AVhat was read
as a valedictory was listened to with great attention, after which a call
to proceed to the business of the meeting was made. Instead of respoud-
inof to this call, one after another of those whom the writer most
respected addressed him in such terms of affectionate remonstrance
against his resignation, as to induce him not only to withdraw it, but to
feel that henceforth what had been regarded as a burden would be a joy,
that the performance of the duties of his office would be sweetened as
never before by the recognition that the respect and regard which he felt
towards all the members were fully reciprocated by them." In 1880. on
the completion of the fiftieth year of the society, and his tenth year as
president, he declined re-election. But he performed no greater service
to the Society or science than is represented by the admirable history of
the Society's first half-century contributed in this year to its memorial
volume. We search in vain, however, through this painstaking record,
for any adequate account of his own labors. Nevertheless, although
belonging to a later generation, I have had abundant opportunity to
learn with what rare zeal and fidelity he served the Society ; and I feel
justified in stating that its extensive and efficient collections of minerals,
rocks, and fossils were in a large measure created by him. It was
throughout evidently a labor of love on his part ; and that fact alone
can explain how he was able to find in these busiest years of his life the
great amount of time which this work must have required. Nor did his
devotion to these departments cease with his accession to the presidency.
It was my privilege, during the greater part of the decade when he was
at the head of the Society's affiiirs, to work with him or under his
direction in the care of these collections ; and I can, therefore, speak
from intimate personal knowledge of his methods. What most strongly
impressed me were his affectionate regard for the specimens, and his
absolute conscientiousness in every detail of the work. I often recall
with much profit his painstaking investigations of individual specimens.
He spared no pains to remove the last modicum of doubt as to the
authenticity of species, variety, locality, composition, or formation, nor hes-
itated to use the query mark when the desired result could not be attained.
Working in this methodical manner, progress could not be rapid when
measured by hours. But what was done was well done ; and he came

342 • THOMAS TRACY BOUV]^.

again the next day and the next, for half an hour or an hour in the morn-
ing and as long as daylight lasted in the afternoon, making this work,
which he loved so well and called his recreation, a part of his daily
routine. Thus month by month, without haste or faltering, important
and enduring results were accomplished.

Although specially interested, as his published papers and the records
show, in several branches of natural history, Mr. Bouve's first love
among the natural sciences was mineralogy. In the mineralogical annals
of the Society three names are especially prominent, — Francis Alger,
Charles T. Jackson, and Thomas T. Bouve. These three men were
contemporaries and lifelong friends ; and each in his own way con-
tributed in an imjDortant degree to the development of this department
of the Museum. Mr. Bouve was the youngest of the trio, and the period
of his activity extended well into the modern era in the history of educa-
tional methods. He was thus led to a deeper appreciation of the value
of minerals as a factor in elementary education. It is therefore prob-
ably well within bounds to say, that for this reason, and because of his
greater length of service, Mr. Bouve has done more than any other one
man to make this collection what it is to-day, an important adjunct of
the educational system of the community.

Durino- the years when we worked together upon the minerals, I had
the pleasure of hearing from his lips the history, both in general and in
detail, of a large part of the collection ; and whether the specimens had
been collected by himself or a fellow member, or obtained by purchase
or exchange, his evident love for them made me feel that I was being in-
troduced to his dear friends, whose care he was reluctant to relinquish
to another, 4

His appreciation of the beautiful in minerals culminated in his well
known fondness for gems ; and these he did not value commercially, but
only in proportion to their intrinsic beauty and scientific interest. This
study forced upon his attention the unsatisfactory nature of the criteria,
such as color and lustre, commonly relied upon in the identification of
cut stones; and led him to test more thoroughly than had been done
before the relatively fundamental property of specific gravity. In his
valuable paper on this subject he demonstrated to the satisfaction of
Professor Dana, and other high authorities, that while each of the mineral
species to which the gems belong varies notably in specific gravity, so
that the several species overlap and are indistinguishable by this char-
acter, the gems, being in every instance the purest and most ideal forms
of their respective species, are essentially constant, and only rarely so

THOMAS TRACY BOUVl 343

nearly of the same density as not to be readily and certainly distinguished
by careful weigliing.

Mr. Bouve's later contributions to scientific literature relate chiefly to
the geology of the Boston Basin, and especially of the South Shore.
These papers include: (1.) "The Genesis of the Boston Basin and its
Rock Formations," in which for the first time the importance of the ante-
cedent chemical decay of the older rocks in explaining the origin of the
conglomerates and slates is adequately recognized. (2.) " The Indian
Potholes or Giant's Kettles of Foreign Writers," a carefully illustrated
and in every way admirable account of the most interesting group of
glacial potholes yet discovered in this region. (3.) The chapter on
the geology of Hingham, prepared for the town history. This was by
far his most important study, and it was my good fortune to be intimately
associated with him in this work, which extended over a period of half
a dozen years.

I have never found field-work more enjoyable or profitable ; and, as
in the Museum, I was deeply impressed by Mr. Bouve's unselfish single-
ness of purpose, his high appreciation of the educational value of the
local geology, and his painstaking thoroughness. He labored, not for
fame, but for the advancement of the community in which he lived.
lie never counted time or strength while there seemed to be any possi-
bility of verifying a fact or testing a conclusion ; and in this contribution
to the town history he realized in gratifying measure his ideal, which
was not merely to set forth the geologic structure of the town, but to
show his fellow townspeople that in its geological history Hingham is in
a large degree an epitome of the world, and that within the narrow
limits of the town are presented problems as stupendous and full of in-
terest as any tnat have ever engaged the attention of geologists.

The part of this work relating to the surface geology or drift phe-
nomena was subsequently amplified and published as a separate paper in
in the Proceedings of the Natural History Society. Not content with
giving to Hingham and the world the intellectual results of his labors,
Mr. Bouve placed in the town library, as an additional stimulus to inter-
est in the local s:eolo(iy, a well selected and carefully labelled series of
the Hingham rocks. To these he subsequently added the main part of
his extensive collection of minerals. The arrangement and labelling of
this generous gift to the town, almost his last public work, was frequently
interrupted by his failing health ; and he experienced no little satisfaction
in not beinjr obliged to leave it unfinished.

Nothing connected with his work on the geology of the South Shore

844 FRANCIS AMASA WALKER.

has afforded the writer more pleasure than the privilege, in association

with Mr. A. W. Grabau, of giving the name of our friend to the glacial

lake which once adorned that region, and was a controlling factor in the

development of the surface geology. Mr. Bouve lived long enough to

note the general and cordial acceptance of this name ; and I venture to

hope that it may long endure as a fitting memorial of this earliest and

most devoted student of the geology of the district which embraces Lake

Bouve.

W. 0. Crosby.

FRANCIS AMASA WALKER.

Francis Amasa Walker, late President of the Massachusetts Insti-
tute of Technology, and a Fellow of this Academy from October, 1882,
was born in Boston, July 2, 1840, and died of apoplexy in that city,
January 5, 1897.

His father, the late Araasa Walker of North Brookfield, was a well
known figure in the political life of Massachusetts for many years. He
was a leader in the Free Soil movement of 1848, and in the subsequently
combiued opposition to the Whig party. He served in each branch of the
Legislature, was for two years Secretary of the Commonwealth, was a Pres-
idential Elector in 1860, and a member of the lower House of Congress
for the session of 1862-63. From 1842 to 1848 he lectured on pohti-
cal economy in Oberlin College, and was afterwards a frequent wi-iter for
periodicals, especially upon topics connected with finance and banking, in
which he also showed special interest when in Congress. From 1859 to
1869 he was Lecturer upon Political Economy in Amherst College, pub-
lishing during that time his well known book, the " Science of Wealth,"
and died in 1875.

Francis Amasa Walker, the son, thus grew up with an inherited predilec-
tion and aptitude for economic study, strengthened by the associations of
boyhood and youth. When he graduated from Amherst College in 1860,
however, his first step was to enter as a student of law the office of
Charles Devens and George F. Hoar of Worcester, — both gentlemen des-
tined, like himself, soon to attain national reputation. On the breaking
out of the Civil War in 1861, Mr. Devens at first took the field as
an officer of militia, and, when later he raised the Fifteenth Regiment
of Massachusetts Infantry in Worcester County, young Walker en-
listed and was mustered into the service as Sergeant Major, August 1,
1861. Ten days later, he was commissioned and assigned to the staff of

FRANCIS AMASA WALKER. 3-45

General Couch. From that time he was upon duty with the Army of the
Potomac, serving with advancing rank upon the staff of Generals Warren
and Hancock through some of the severest campaigns of the war. He
resigned his commission in January, 18G5, from illness contracted while
a prisoner within the Confederate lines, received the brevet rank of
Brigadier General " for distinguished service and good conduct," and re-
turned to civil life bearing the honorable scars of the brave. It after-
wards fell to his lot, in his " History of the Second Army Corps" (188G),
and his ''Life of General Hancock" (1894), to write the narrative of
events no small part of which had passed before his eyes. Little of
his own history is to be found in those glowing pages, but every line bears
witness to the intense enthusiasm with which he never failed to kindle
when he recalled his army life, and to his devotion to the great captains
under whom he served.

Like many other young men, who, as soldiers in the "War for the Union,
drank the wine of life early. General Walker came home with his charac-
ter matured, his capacities developed, his intellectual forces aroused and
trained, — a man older than his years. The career in which he was to
win new distinction did not open for him at once upon the sudden return
of peace. For three years he was a teacher of the classics in "Williston
Seminary, and in 1868, being compelled by an attack of quinsy to seek a
change of occupation, he became an assistant of Mr. Samuel Bowles,
editor of the Springfield Republican. From this place he was drawn
into the public service at "Washington, by the agency of Mr. David A.
Wells, who was then Special Commissioner of the Revenue, and in search
of a new Chief for the Bureau of Statistics. The work of the Bureau had
fallen into some discredit, and was far in arrears, and the inability of the
former Chief of the Bureau to command the confidence of Congress seri-
ously embarrassed the continuance of an important work. By Mr. Wells's
advice General Walker was made Deputy Special Commissioner and
placed in charge of the Bureau, and a new career was at once opened
before him, for which he was fitted in a peculiar manner both by his
intellectual interests and his administrative capacity. The Bureau was
reorganized and its reputation was regained. The monthly publications
were resumed, and soon showed that progressive improvement which has
made them one of the most valuable repositories in existence for the
study of the commercial and financial activity of a great country.

From his appointment to the charge of the Bureau of Statistics
the steps in General Walker's new career followed in rapid succession.
In 1870 he was appointed Superintendent of the Ninth Census of the

346 FRANCIS AMASA WALKER.

United States ; in 1871 he was appointed Commissioner of Indian Affairs ;
in 1872 he was made Professor of Political Economy and History in the
Sheffield Scientific School of Yale College; in 1876 he was Chief of the
Bureau of Awards for the Centennial Exposition in Philadelphia ; in 1878
he was sent as a Commissioner for the United States to the International
Monetary Conference at Paris ; in 1879 he was appointed Superintendent
of the Tenth Census of the United States; in 1881 he was made President
of the Massachusetts Institute of Technology; in 1882 he was elected
President of the American Statistical Association; in 1885 he was elected
first President of the American Economic Association; in 1891 he was
elected Vice-President of the National Academy of Sciences ; in 1893
he was President-adjunct of the International Statistical Institute, at its
session in Chicago.

General Walker's successive apjwintments as Superintendent of the
Census of 1870 and of that of 1880 were the direct result of the enerory
and skill with which, during the mouths of his service in the Bureau of
Statistics, he had effected the reorganization of that office and its work.
The opportunities given to him as a statistician, by having charge of these
two censuses, were of a remarkable kind. The census of 1870, being the
first taken after the Civil "War, was for that reason by far the most inter-
esting and important since 1790. It was to show the social and economic
changes wrought by four years of prodigal expenditure both of life and of
resources, and by the unparalleled revolution in the industrial organiza-
tion of the former slave States. It was also to ascertain and record the
conditions under which the nation entered upon anew and wonderful stage
of its material growth. The census of 1880 was the unique occasion for
what General Walker designed as a " grand monumental exhibit of the
resources, the industries, and the social state of the American people,'
made approximately at the close of a century of national independence.

The Census of 1870, to the great regret of all who had any scientific
interest in the sul)ject, was left by Congress to be taken under the provis-
ions of the Census Act of 1850, by persons neither selected nor con-
trolled by the Census Office. In the still disturbed condition of some
of the Southern States, the work was thus thrown into the hands of men
notoriously unfit for such employment, and the returns, especially of the
black population, were vitiated at their source. In his Report of 1872,
and in his Introduction to the " Compendium of the Census of 1880," Gen-
eral "Walker described in strong lansuaije the difficulties which thus beset
the work in 1870 ; and again in the Publications of the American Statis-
tical Association for December, 1890, writing upon the " Statistics of the

FRANCIS AMASA WALKER. 347

Colored Race in the United States," he used his freedom from official
relations in exposing the mischief done by legislative failure to provide
intelligently for an important public service. As a whole, however, the
Census of 1870 was the best and the most varied in its scope that had
yet been obtained for the United States. It was, after all, a signal proof
of what can be done by a competent head, even with imperfect legislation,
and established the reputation of the Superintendent as an administrative
officer, at the same time that his fresh and vigorous discussion of results
secured him high rank among statistical writers. Great interest was
excited, moreover, by the remarkable use made of the graphic method
in presenting the leading results of this census, in his " Statistical Atlas
of the United States" (1874).

The Act providing for the Census of 1880 was greatly modified, by
General Walker's advice, and the working force was for the first time
organized upon an intelligent system, by the employment of specially
selected enumerators in place of the subordinates of the United States
marshals, to whom the law had previously intrusted the collection of
returns. Highly qualified experts were also employed for the historical
and descriptive treatment of diiferent industries and interests, as de-
manded by the monumental character of the centennial census. Various
causes delayed the completion of this gigantic undertaking. Those to
whom a census is merely a compendious statement of passing facts became
impatient at the slow issue of the twenty-two stately quartos, and com-
plained that the work was on such a scale as to be obsolescent before its
appearance. General Walker, in an article in the Quarterly Journal of
Economics for April, 1888, explained some of the special causes of the
delay in publication and took upon himself perhaps an undue share of re-
sponsibility for the difficulties caused by an original underestimate of the
total cost of the census. But notwithstanding: its misfortunes, the Census
of 1880 is a great work of enduring value and not excessive cost, —
great in its breadth of design, worthy of the nation and of the epoch,
and a lasting monument of the power of its Superintendent to conceive
and to execute. Following the Census of 1870, it won for him univer-
sal recognition as one of the leading statisticians of his time.

In the article to which reference has just been made, General Walker,
in his discussion of future arrangements for the national census, offered
as the fruits of his own experience some valuable suggestions, which
deserve more attention than they have yet received. It is hardly neces-
sary, however, to enter upon them here, except to recall the fact that he
advised the organization of the Census Office as a permanent establish-

348 FRANCIS AMASA WALKER.

ment, in order to secure the improved service and economy of a trained
force of moderate size, constantly employed. Upon an office thus organ-
ized could be laid, at the regular intervals, the duty of collecting and
preparing the returns of population and of agriculture for the decennial
census required by the Constitution, and perhaps for an intermediate
fifth year enumeration, and also in the intervals the systematic prose-
cution of other statistical investigations, to be charged upon the office
from time to time as occasion might require.

General Walker's appointment as Professor in the Sheffield Scientific
School, in 1872, carried him beyond the boundary of statistics into the
general field of political economy. His training for this extended range
of work, although obtained by a less systematic process than is now
usual, had begun early, and as opportunity offered was carried on effect-
ively. In one of his prefaces, he remarks that he began writing for the
press upon money in 1858, probably having in mind a series of letters to
the National Era of "Washington, beginning soon after the crisis of 1857,
and continued for some months, noticeable for sharply defined views on the
subjects of banking and currency, and also as to the merits of Mr. Henry
C. Carey as an economist. In 1865, before going to Williston Seminary,
he lectured upon political economy for a short time at Amherst in his
father's absence, and in 1866 his father recognized with pride his important
assistance in finishing the " Science of Wealth." From the close of the
war, he is otherwise known to have been a keen student of economics, al-
though a student under such limitations and so hampered by pressing occu-
pations as to make it difficult for him to do equal justice to all parts of his
outfit. It was perhaps from this cause, in part, that his earliest important
publications as an economist were two treatises on widely separated topics,
'• The Wages Question " and " Money."

The earlier of these two books, "The Wages Question" (1876), in-
stantly attracted the attention both of economists and of the general pub-
lic by its lively and strong discussion of the central topic of the day, then
more commonly treated either as a matter of dry theor^^, or as a problem to
be settled by sentiment. Following Longe and Thornton, the author made
an unsparing attack upon the wages fund theory, and, arguing that wages
are paid from the product of labor and not from accumulated capital,
he set forth with great vigor the influences which affect the competition
between laborer and employer in the division of this product. Gen-
eral Walker's earliest public statement of his now familiar opinions touch-
ing the wages fund, and the payment of wages from the product, was
made, it is believed, in an address delivered before the literary societies

FRANCIS AMASA WALKER. 0-19

of Amherst College, July 8, 1874, and he further developed the subject
in an article contributed to the North American Review for January,
1875. Few books in political economy have taken a place in the fore-
ground of scientific discussion more (juickly than " The Wages Question."
Many economists followed the author's lead with little delay, and those
who were slower to admit that the object of his attack was in fact the
wages fund of the older scliool, recognized his assault as by far the most
serious yet made. Unquestionably it compelled an immediate review
of a large body of thought by the great mass of economic students in the
English speaking countries.

In "The Wages Question," General Walker drew the line clearly be-
tween the function of the capitalist and that of the employer, or entre-
preneur, and between interest, which is the return made to the former, and
profits, which are the reward of the latter. It was however in his
" Political Economy" (1883), that he worked out his theory of the source
of business profits and of the law governing the returns secured by the
employing class. This enabled him to lay down a general theory of distri-
bution, to be substituted for that associated with the wages fund theory,
which he regarded as completely exploded, and indeed "exanimate."
Of the four parties to the distribution of the product of industry, three,
the owner of land, the capitalist, and the employer, in his view, receive
shares which are determined, respectively, by the law of Ricardo, by the
prevailing rate of interest, and by a law of business profits analogous to
the law of rent. These shares being settled, each by a limiting prin-
ciple of its own, labor becomes the " residual claimant," be the residue
more or less, and any increase of product resulting from the energy, econ-
omy, or care of the laborors " goes to them by purely natural laws, pro-
vided only competition be full and free." So too the gains from invention
enure to their benefit, except so far as the law may interfere by creating
a monopoly. This striking solution of the chief problem of economics
attracted wide attention, and was further expounded and defended by its
author in the discussions which it provoked, as may be seen by reference
to the earlier volumes of the Quarterly Journal of Economics. Indeed,
in his last published work, " International Bimetallism " (page 283), he
prefaces a statement of his theory by saying, " I have given no small
part of my strength during the past twenty years to the advocacy of
that economic view which makes the laborer the residual claimant
upon the product of industry."

General Walker published his treatise, " Money " (1878), at a mo-
ment sino-ularly opportune for the usefulness of the book and the ad-

350 FEANCIS AMASA WALKER.

vancing reputation of its author. Public opinion in the United States
was iu extreme confusion on the questions involved in the return to
specie payment ; there was a formidable agitation for the repeal of the
Eesumption Act, and Congress was entering upon its long series of
efforts to rehabilitate silver as a money metal. At this juncture, when
every part of the theory of money was the subject of warm discussion,
scientific and popular alike, General Walker, using the substance of a
course of lectures delivered by him in the Johns Hopkins University
in 1877, laid before the public an elaborate and broad-minded survey
of the whole field, claiming little originality for his work, but giving
material help in concentrating upon scientific lines a discussion which
was wandering in endless vagaries. On the general subject his views
had no doubt been formed early, under the influence of his father, to
whom, in more than one passage of this book, he makes touching
allusion, and later in life he found in them little to change, although
the long regime of paper money and its consequences suggested many
things to be added. In 1879 he published, under the title of "Money
in its Relations to Trade and Industry," what was in some sense an
abridgment of the larger work, made for use in a course of lectures in
the Lowell Institute ; and in his " Political Economy " he again con-
densed his arguments and conclusions as to money, as part of his dis-
cussion of the grand division, Exchange.

When the International Monetary Conference met in 1878, by invita-
tion of the United States, General Walker went to Paris as one of the
commissioners for this country. His discussion of bimetallism had not
been carried in " Money " much beyond a careful statement of the ques-
tion and of the arguments on each side, but it was carried far enough to
show that international bimetallism, and not the simple remonetization of
silver by the United States, was, in his view, the proper method of secur-
ing what he deemed an adequate supply of money for this country and
for the commercial world. Great emphasis was laid, in " Money, Trade,
and Industry," upon the necessity for "concerted action by the civilized
states," and this ground was consistently held by him until his share in
the discussion ended with the publication of " International Bimetal-
lism" (1896), a few months before his death. In this book, which was
the outcome of a course of lectures delivered in Harvard University,
after reviewing the controversy over silver, which had more and more
engaged his attention as time went on, he declared more vigorously than
ever his opinion of the futility of the policy of solitary action, adopted
by the United States in the Act of 1878. " International Bimetallism"

FRANCIS AMASA WALKER. 351

appeared iu the midst of a heated Presidential canvass, in which the
issues had taken such form that some, who like himself were supporters
of " sound money," found a jarring note ia what they regarded as need-
less concessions to " free silver," and iu the sharp phrase in which his
ardor and deep conviction sometimes found expression. But the book
was not written for effect upon an election ; it was the last stroke of a
soldier, in a world-wide battle, — soon to lay aside his arms.

It was General Walker's good fortune to enter the Held as an econo-
mist when the study of economics was gaining new strength in the
United States from the powerful stimulus of the Civil War, and of the
period of rapid material development and change which followed. The
revision of all accepted theories which set in did not displease him, and
he took his share in the ensuing controversies, whether raised by himself
or others, with equal zest. His own tendency, however, was towards a
rational conservatism, and his modes of thought never ceased to show
the influence of writers, French and English, of whom he appeared to the
superticial observer to be the severe critic. " A Ricardian of the Ri-
cardians "■ he styled himself in his Harvard lectures on laud, published
under the title of "Land and its Rent" (1883). His theory of distri-
bution, if enunciated by one of narrower sympathies than himself, might
have been thought to be designed as a justification of the existing order
of things. In his monetary discussions he contended for a return to
what he deemed the safe ways of the past. As for his view of the
future, in a public address in 1890, after a remarkable passage describ-
ing the sea of agitation and debate which had submerged the entire do-
main of economics, and threatened to sweep away every landmark of
accepted belief, he said, " I have little doubt that in due time, when these
angrv floods subside, the green land will emerge, fairer and richer for
the inundation, but not greatly altered in aspect or in shape."

The election of General Walker as the first President of the Ameri-
can Economic Association, in recognition of his acknowledged eminence,
deserves a passing notice at this point. The Association was organized
at Saratoga in 1885, under circumstances which threatened to make it
the representative of a school of economists rather than of the great
body of economic students in America, and with a dangerous approach
to somethino- like a scientific creed. General Walker cannot be said to
have represented any particular school. He was both theorist and
observer, the framer of a theory of distribution, and also an industrious
student of past and current history. By a happy choice the new Asso-
ciation strengthened its claim upon public attention by electing him its

352 FRANCIS AMASA WALKER.

President, iu his absence ; and be wisely took bis place at its Lead, with
the conviction that its purposes were better than the statement made of
them, and that the membership of the new organization gave promise of
good results for economic science. Under his administration, which
lasted until 1892, the basis of the Association was broadened, all appear-
ance of any test of scientific faith disappeared, and American economists
found themselves associated in catholic brotherhood. In part this change
was no doubt due to the marked subsidence of the debate as to the de-
ductive and the historical methods, but in part also it was due to the
good judgment, personal influence, and perhaps in some instances the
persuasive efforts of the President, who thus rendered no small service
to economic science.

Which of General Walker's contributions to economic theory are likely
to have lasting value, is a question not yet ready for decision. The
subjects to which he specially devoted his efforts are still under dis-
cussion. His theory of distribution is not yet established as the true
solution of the great problem ; the wages fund has not yet ceased to
be controversial matter ; it is not yet settled whether the advocates or
the opponents of bimetallism are to triumph in the great debate of this
generation. But whether as a theoretical writer he is to hold his pres-
ent place or to lose it, there can be no question as to the importance of
his work, in imparting stimulus and the feeling of reality to all economic
discussions in which he had a part. His varied experience and wide
acquaintance with men had made him in a large sense a man of affairs.
He watched the great movements of the world, not only in their broad
relations, but as they concern individuals. He was apt to treat economic
tendencies, therefore, not only in their abstract form, but also as facts
making for the happiness or the injury of living men. Economic law was
reasoned upon by him in much the same way as by others, but he never
lost his vivid perception of the realities among which the law must work
out its consequences. In his pages, therefore, theory seemed to many to
be a more practical matter and nearer to actual life than it is made to
appear by most economists. His words seemed to carry more authority,
his illustrations to give more light, the whole science to become a lively
exposition of the trend and the side movements of a world of passion
and effort. A great English economist has said that Walker's explana-
tion of the services rendered by the entrepreneur remind one of passages
of Adam Smith. A great service has been rendered to the community
by the writer who, in our day, has been able thus to command attention
to political economy as a discussion belonging to the actual world.

FRANCIS AMASA WALKER. 353

General "Walker's election to the Presidency of the Massachusc^tts
Institute of Technology, in 1881, placetl him at the head of an institution
badly in need of a vigorous, confident, and many-sided administrator, for
the development of its great possibilities. The plan on which it should
work had been prepared and its foundations laid broad and deep by
President Rogers, but the work itself was still languishing, endowment
and equipment were scanty, and the luunber of students declining.
General Walker's administration was signalized by a sudden revival of
the school. Funds were secured, new buildings were built, the confi-
dence of the public won, and at General Walker's death the school of
barely two hundred students, still maintaining the severe standard of
work set by its founder, had upon its register nearly twelve hundred
students and maintained a staff of one hundred and thirty professors
and instructors of different grades. Of the qualities as an educator and
administrator of a great technical school displayed by General Walker
in this brilliant part of his career, a striking description, made from
close observation, has been given by Professor H. W. Tyler of the
Faculty of the Institute, in the Educational Review for June, 1897.

There was doubtless much in the circumstances attending the founda-
tion of the Institute of Technology which any disinterested friend of
scientific education must now regret. But time has healed wounds and
removed jealousies which divided a former generation, and none can
now be found to question either the practical or the scientific value of the
great institution conceived by Rogers, and brought to its present de-
served eminence under the successor of whose day he lived to see little
more than the dawn.

At no period of General Walker's life did he fail to take an active
interest in the work of the community in which he lived. That he was
already charged with great responsibilities was a reason, both with his
fellow citizens and with himself, for increasing the load. An early in-
stance of this was his service as Commissioner of Indian Affairs for one
year while still in charge of the census of 1870, — a service marked by
an annual report remarkable for its thorough review of the whole subject,
and by the appearance of his book, " The Indian Question" (1874). At
different times, in New Haven and in Boston, he was a member of the
local School Board and of the State Board of Education. He was a
Trustee of the Boston Public Library and of the Museum of Fine
Arts, one of the Boston Park Commissioners, and an almost prescrip.
tive member of any more temporary board or committee. In some of
these capacities his labors have left their traces in his written works,
VOL. XXXII. — 23

354 FRANCIS AMASA WALKER.

in others his name gave weight to organizations in which he was not
called upon for active effort. The number and variety of the appoint-
ments thus showered upon him marked not only the unbounded range of
his own interests, but the confidence of others that every appeal to
public spirit would stir his heart.

The bibliography of his written work, prepared at the Institute of
Technology and revised with great care since his death, will be found in
the Publications of the American Statistical Society for June, 1897.
It is a remarkable record of intellectual activity, maintained for nearly
forty years, and resulting in a series of important contributions to the
thouiiht of his time, — a manifold claim to eminence in the world of

o

science and letters.

A complete list of the honorary degrees and other marks of distinc-
tion conferred upon General Walker by public bodies, at home and
abroad, cannot be undertaken here. It is enough to say that he was
made Doctor of Laws by Amherst, Columbia, Dublin, Edinburgh, Har-
vard, St. Andrews, and Yale, and Doctor of Philosophy by Amherst
and Halle ; that he was a member, regular or honorary, of the National
Academy of Sciences, the Philosophical Society of Washington, the
Massachusetts Historical Society and this Academy, of the Royal Sta-
tistical Society of London, the Royal Statistical Society of Belgium, the
Statistical Society of Paris, the French Institute, and the International
Statistical Institute ; and that be was an officer of the French Legion of
Honor.

General Walker was endowed by nature with peculiar gifts for a
career of distinction. In any company of men he instantly drew atten-
tion by his solid erect form and dignified presence, by his deep and
glowing eye, and by his dark features, cheerful, often mirthful, always
alive. His instant command of his intellectual resources gave him the
confidence needed for a leading place, and his friendly bearing, strong
judgment, and easy optimism made others welcome his leadership. His
convictions were deep, and his opinions, once formed, were shaken with
difficulty, for in discussion he had the soldier's quality of not know-
ing when he is beaten. His ambition was strong, and he liked to feel
the current of sympathy and approval bearing him on, but he did not
shrink from his course if others refused to follow. From Qrst to last,
he grappled with large undertakings and large sul)jects, conscious of
powers which promised him the mastery. Sucli as his contemporaries
saw him he will live for the future reader in many a sentence and page, —

cheerful, courageous, hopeful.

Charles F. Dunbae.

BENJAMIN APTHORP GOULD. 355

BENJAMIN APTHORP GOULD.

Benjamin Aptiiokp Gould was born on September 27, 1824, at
No. 5 Wiuthrop Place, Boslou. The years of bis preparatory education
were passed at the Chauncy Hall and Boston Latin Schools. He grad-
uated from Harvard College in lS4i. It is interesting to note, as evi-
dence of his wonderful versatility of miud, and in view of the subsequent
choice of a career which has made his name so illustrious in the annals
of science, that, although he attained in college high distinction in mathe-
matical and physical branches, the earliest studies of his predilection
were classical rather than scientific. After leaving collese he took
charge of the Boston Latin School. At the end of a year, however, his
intellectual energies took their final direction towards science. What
were the springs of the influence which turned him to astronomy, and
bow long the impulse had been gathering force, cannot now be told, but
the passionate devotion with which he entered upon this career, and the
tenacity with which he held to it, — at times under circumstances of ex-
treme discouragement which never caused him to falter or swerve from
his purpose, — are characteristic of the man. This initial decision of a
young man of twenty-one, as we look back upon it from a point where
we can measure the results that have flowed from it, seems not so much
an episode in an individual career as an epoch in our national scientific
history. For the date, Jul}', 1845, when Gould placed his foot upon
the steamer from Boston, with the avowed and definite purpose to devote
himself to a life purely of scientific research, marks quite distinctly a
most important phase in American astronomy. Up to that time the in-
stance of a man doing this as his only earthly aim, while unassured of a
professor's chair or other similar appointment, and not as a means of
livelihood, was in this country absolutely unknown. With a steadiness of
purpose singular in so young a man, he pursued diligently the oppor-
tunity he had himself created, for three years at the fountain-heads of
astronomical learning. Beginning his life of preparation, study, and
work with Airy at Greenwich, he then went to Paris. But it was after-
wards, in Germany, that he secured his true education in the regenerated
methods of modern astronomical research ; at Berlin, where he spent a
year with Eucke ; at Altona, Pulkowa, and Gotha, under Peters, Struve,
and Hansen ; and at Gottingen, where for a year he was a pupil of
Gauss, and where he took his Doctorate of Philosophy at the University
in 1848. He then returned home, full of early honors and flushed
with lofty hopes and honorable ambitions.

356 BENJAMIN APTHORP GOULD.

From this poiut Dr. Gould's life became one of incessant activity, im-
pressing its mark in many ways upon the intellectual life of the com-
munity ; but the line of intensest force naturally took the direction of
his own beloved science, to which he communicated an impulse not meas-
ured merely by what he accomplished for it by his direct investigations,
great as that is, but also by the force which always emanates from so
earnest a nature. He inspired a new breath into American astronomy.
The new atmosphere which he brought with him from Germany, where
he had caught the spirit of the great masters under whom he studied,
became gradually transfused upon this side the sea. His enthusiasm for
the introduction of better means and methods of research was caught
by his compatriots, their courage to regenerate our science was sustained,
and transmitted through various channels to the next and to the present
generation. Thus we may say without fear of being controverted that
American astronomy to-day is a different thing from what it would have
been without Gould's predominant influence, deep and quiet but strong,
to upbuild it and to free it from the clumsiness and imperfections which
still impede it, even in some of the otherwise most enlightened nations of
the world. It is under his leadership that American astronomy has
climbed to where it looks with steady and level eye uiDon that of Ger-
many, which occupies perhaps a larger, but not a loftier plane.

Let us now glance at Dr. Gould's more prominent labors, passing by
his earlier important investigations in applied theoretical and in practical
astronomy, as well as his numerous and valuable coutributions to the
literature of science, education, and other departments of thought, which
we find scattered through the long range of his career. In 1852 he was
appointed to take charge of the longitude determinations of the Coast
Survey. He organized, developed, and extended this service, retiring in
1867. Meanwhile, in 1855, he became director of the Dudley Observa-
tory in Albany, equipped and organized the institution, and carried it on
without remuneration and at his private expense. He left it in 1859,
after a severe struggle to preserve the institution for purposes of scien-
tific investigation.

In 1859 he published his discussion of the places and proper motions
of circumpolar stars, for use as standards in the Coast Survey. These, as
revised by him in 1861, together with his similar list of clock-stars, were
adopted as the standards for the American Ephemeris, and, as to the
circumpolars, remain in such use to this day. In 1866 he published his
reduction of D'Agelet's observations. About the same time he per-
formed a similar service for the greater part of the observations made at

BENJAMIN APTIIORP GOULD. 357

the United States Naval Observatory siuce its establishment, as he had
done also, several years previously, tor the expedition to Chili to deter-
mine the solar parallax. In 18G6 he planned and executed the work of
establishing, by the Atlantic cable, the relation in longitude between
European and American stations, involving, as a part, interesting re-
searches on the velocity of the galvanic current in submarine cables,
similar to those he had already made on land-lines. As actuary of the
United States Sanitary Commission, he conducted, and published in a
large volume, extensive and important researches upon Military and
Anthropological Statistics and the Distribution of Population. About
the same time he undertook the reduction of Rutherfurd's photographs
of the Pleiades. The results, partially published in 1866, were sub-
mitted completely, in an elaborate memoir, to the National Academy in
1870, together with a second memoir on the Prassepe. He was, indeed,
a pioneer in the utilization of photography for exact astronomical meas-
urement. About 1864 he built an observatory in Cambridge, equipped
with an eight-foot transit instrument, and, until 1867, carried on a de-
termination of the right ascensions of all the stars to the tenth magnitude
within one degree of the pole. The work was completely reduced,
but the discussion and publication were postponed by his removal to
Cordoba.

In 1865 he became intensely impressed with a desire to explore the
southern celestial hemisphere. The opportunity to do so soon came.
This project assumed at first the form of a private astronomical expedi-
tion, for which his friends in Boston had promised the pecuniary means;
but, under the enthusiastic support of Mr. Sarmiento, at first as Argen-
tine Minister to this country, and later as President of that republic, it
rapidly broadened, and finally led to the establishment by Dr. Gould of
a permanent National Observatory at Cordoba. This marks an epoch in
modern astronomy, the equalization of our knowledge of the two celestial
hemispheres. The institution and its work form an impressive monu-
ment to his memory.

It is impossible, in brief space, to describe or characterize the marvel-
lous work here undertaken and so faultlessly pushed to completion by
Dr. Gould, during the fifteen years of self-imposed exile from his native
land, with unfaltering devotion and energy, in the face of difficulty and
domestic bereavement. The work on the uranography of the southern
heavens was finished in 1874, and was published under the title of the
"Uranometria Argentina," which will remain a classic for all time. The
zone observations of the stars between 23° and 80° south declination,

358 BENJAMIN APTHORP GOULD.

which were the original and always the dominant object of the enter-
prise, were begun in 1872, substantially completed in 1877, and revised
in 1882-83. This work was embodied in the Zone Cataloo-ues, con-
taining 73,160 stars, which appeared in 1884. Parallel with this, and
almost overshadowing it in importance, was carried on the independent
series of meridian-circle observations for the General Cataloirue of
32,448 stars, completed in 1885. Dr. Gould, besides, left the manu-
script of the remainder of his series of fifteen volumes, not then publislied,
— containing the observations and the annual catalogues, incorporated in
the General Catalogue, — complete to the minutest detail, ready for the
printer. These have since appeared, from time to time, — the last vol-
ume, rounding out his work, reaching Cambridge but a few hours before
his death.

Another part of the work for the Cordoba Observatory, planned by
Dr. Gould as a fitting extension of it. was a " Durchmusterung " of the
southern sky. For this, indeed, he had provided the instrumental means
and trained the assistants, it being his purpose to be ready to begin it at
any time in case of any unforeseen delay or accident to the other work.
On leaving Cordoba he confided it to the care of Dr. Thome, by whom,
with Mr. Tucker, it has been so worthily conducted.

Dr. Gould also established, under the auspices of the Argentine gov-
ernment, a meteorological service, second in extent, it is believed, only
to that of the United States. Upon leaving South America he intrusted
this charge to the hands of his worthy successor, "Walter G. Davis.

The earliest to recognize and demonstrate the capabilities of photog-
raphy to render service to the astronomy of precision, Dr. Gould, by his
experience with the Rutherfurd plates of the Pleiades and the Prassepe,
was incited to arrange to carry forward at Cordoba, on an extensive
scale, a similar work upon the southern stellar clusters. His other
labors there were so onerous that he confined his attention to securing
plates suitable for precise measurement. Of these he accumulated about
1400, and brought them home with him for measurement and reduc-
tion. Without permitting himself a well earned retirement, he turned
at once, tirelessly, to this labor, wliich has been the principal occupation
of the last ten years of his life. Tliis is substantially complete, and will
be given to the world as it came from his hand.

Dr. Gould had an enthusiasm for the advancement of his beloved sci-
ence for wider than the limits of what he could by personal investigation
accorapli;<h. Early in his career he keenly realized that astronomy had
reached a stage of development in America which entitled it to a higher

BENJAMIN APTHORP GOULD. 359

claim than had yet been accorded to it ; and that a journal worthily sup-
porting the dignity of a pure science would have very great influence
upon its future progress. Accordingly, without ostentation, be estab-
lished the " Astronomical Journal " in November, 1849, offering it to the
use of astronomers, for the publication exclusively of original investiga-
tions. He edited and supported it until, at the end of the sixth volume,
in 18G1, its issue was suspended, first by the war for the preservation of
the Union, afterward by his absence in Cordoba, A long nurtured hope
was realized when he was enabled, in 1885, to resume its publication, and
to continue it, at the rate of nearly one volume annually, to the present
time. Of all the great enterprises of his life, this is the one which he has
most cherished. "With careful forethought, he has made due provision
for its continuance.

Dr. Gould took a deep interest in the work of the International Commit-
tee of AVeights and Measures, of which he was a most influential member.
He represented both the United States and the Argentine Republic on
that Committee, and was a constant attendant upon its annual conferences
in Paris. He was President of the American Metrological Society and
of the Colonial Society of Massachusetts from their inception. He was
one of the original charter members of the National Academy of Sciences,
and has been a member of the American Academy of Arts and Sciences
since 1847, and one of its Vice-Presidents since 1895. He received the
degree of Ph. D. from Gottingen in 1848, and that of LL. D. from
Harvard in 1885, and from Columbia in 1887. During his illustrious
career he was the recipient of the highest honors that Europe has to
bestow, to an extent scarcely vouchsafed to any other American. A
few only will be named here. Mem. Roy. Soc. (London) : For.
Assoc. Roy. Astr. Soc. (London) : Cor. Mem. Acad. Sci. (Institut de
France) : Acad. Imp. Sei. (St. Petersburg) : Kon. Akad. Wiss. (Berlin) :
Kon. Ges. Wiss. (Gottingen): Kais. Akad. Wiss. (Vienna): Bur. d.
Long. (Paris). He was also knighted, of the Order Pour le Merite, by
Prussia.

Dr. Gould came, on both paternal and maternal sides, of old Colonial
stock. The interest which he took throughout life in genealogy and the
history of the early founders of New England was the outcome, not so
much, perhaps, of pride of ancestry, as it was of his intense patriotism,
which was one of his most marked characteristics. Next to the sacred
domestic ties and the innumerable personal friendships which he so ten-
derly cherished, even before his devotion to his beloved science, came his
affection for and pride in his country. No more vivid exemplification of

860 BENJAMIN APTHORP GOULD.

this can be given than his remarkable utterance, prefacing the sixth vol-
ume of the " Astronomical Journal," on the occasion of its temporary
suspension by the war in 18G1 : " There is but one mode of laboring
for Cisatlantic science to-day, namely, by struggling for the maintenance
of civilization against barbarism in the Western Hemisphere. And this
leaves little opportunity for astronomical research. Those who may not
fight against armed treason may at least assume the burdens of those
who do."

Dr. Gould married, in 1861, Mary Apthorp Quiucy, daughter of the
Hon. Josiah Quincy. She died in 1883. Never was there a nobler
example of a wife's self-sacrifice than hers, in the endurance of privation
and self-imposed exile, the severance of family and social ties, for the
purpose of making possible the accomplishment of a husband's life-
work.

Dr. Gould died on Thanksgiving day, November 26, 1896. His
memory and achievements are a precious heritage to his country and his
science.

Seth C. Chandler.

The Academy has received an accession of one Resident
Fellow, one Associate Fellow, and three Foreign Honorary
Members.

The roll of tlie Academy, corrected to date, includes the
names of one hundred and ninety-three Resident Fellows,
ninety-three Associate Fellows, and sixty-six Foreign Hon-
orary Members.

May 12, 1S97.

LIST

OF THE

FELLOWS AND FOREIGN HONORARY MEMBERS.

(Corrected to June 1, 1897.)

RESIDENT FELLOWS. — 195.

(Number limited to two hundred.)

Class I. — Mathematical and Physical Sciences. — 75.

Section T. — 17.
Mathematics and Astronomy.
Solon I. Bailey, Arequipa, Peru.
Seth C. Chandler, Cambridge.
Alvan G. Clark, Cambridgeport.
J. Rayner Edmands, Cambridge.
Francis M. Green, Boston.
Gustavus Hay, Boston.

Henry Mitchell, Boston.

Edward C. Pickering, Cambridge.
John Ritchie, Jr., Boston.
John D. Runkle, Brookline.

T. H. Safford, Williamstown.

Edwin F. Sawyer,

Brighton.

Arthur Searle,

Cambridge.

William E. Story,

AVorcester.

Henry Taber,

Worcester.

0. C. Wendell,

Cambridge.

P. S. Yendell,

Dorchester.

Section H.

— 23.

Physics

.

A. Graham Bell,

"Washington

Clarence J. Blake,

Boston,

Francis Blake,

Weston.

John H. Blake,

Boston.

Charles R. Cross,

Boston.

Amos E. Dolbear,
H. M. Goodwin,
Edwin H. Hall,
Hammond V. Hayes,
Silas W. Holman,
William L. Hooper,
William W. Jacques,
Alonzo S. Kimball,
Frank A. Laws,
T. C. Mendenhall,
Benjamin O. Peirce,
A. Lawrence Rotch,
Wallace C. Sabine,
John S. Stone,
Elihu Thomson,
John Trowbridge,
A. G. Webster,
Robert W. Willson,

Somerville.

Boston.

Cambridge.

Cambridge.

Boston.

Somerville.

Newton.

Worcester.

Boston.

Worcester.

Cambridge.

Boston.

Cambridge.

Boston.

Lynn.

Cambridge.

Worcester.

Cambridge.

Section IIL — 22.
Chemistry.
Samuel Cabot, Boston.

Arthur M. Comey, Cambridge.
Thos.M. Drown, So. Bethlehem, Pa.
Charles W. Eliot, Cambridge.
Thomas GafBeld, Boston.

Henry B. Hill, Cambridge.

362

RESIDENT FELLOWS.

Henry M. Howe, Boston.
Charles L. Jackson, Cambridge.
Walter L. Jennings, Worcester.
Leonard P. Kinnicutt, Worcester.
Charles F. Mabery, Cleveland, O.
Arthur Michael,
George D. Moore,
Charles E. Muuroe
John U. Nef,
Robert H. Richards, Boston.
Theodore W. Richards, Cambridge.
Charles R. Sanger, St. Louis.
Stephen P. Sharpies, Cambridge.
Francis H. Storer, Boston.
Charles H. Wing, Ledger, N. C.
Edward S. Wood, Boston.

Boston.
Worcester.
Washington.
Chicago.

Section IV. — 13.

Technology and Engineering.

EHot C. Clarke, Boston.

Gaetano Lanza, Boston.

E. D. Leavitt, Cambridgeport.

William R. Livermore, Boston.
Hiram F. IMills, Lowell.

Cecil H. Peabody, Boston.
Alfred P. Rockwell, Manchester.
Andrew H. Russell, Rock Island, 111.
Peter Schwamb, Arlington.

Charles S. Storrow, Boston.
George F. Swain, Boston.

William Watson, Boston.

Morrill Wymau, Cambridge.

Class IL — Natural and Physiological Sciences.

63.

Section I. — 13.

Geology, Mineralogy, and Physics of
the Globe.

II. H. Clayton,

Milton.

Algernon Coolidge,

Boston.

William 0. Crosby,

Boston.

William M. Davis,

Cambridge.

Benj. K. Emerson,

Amherst.

0. W. Huntington,

Cambridge.

Robert T. Jackson,

Boston.

Jules Marcou,

Cambridge.

William H. Niles,

Cambridge.

John E. Pillsbury,

Boston.

Nathaniel S. Shaler,

Cambridge.

Warren Upham,

St, Paul, Minn.

John E. Wolff,

Cambridge.

Section H. — 9.
Botany.
William G. Farlow, Cambridge.
Charles E. Faxon, Boston.
George L. Goodale, Cambridge.
H. H. Ilunnewell, Wellesley.
B. L. Robinson, Cambridge.

Charles S. Sargent, Brookline.
Arthur B. Seymour, Cambridge.
Charles J. Sprague, Boston.
Roland Thaxter, Cambridge.

Section HI. — 26.

Zoology and Physiology.
Alexander Agassiz,
Robert Amory,
James M. Barnard,
Henry P. Bowditch,
William Brewster,
Louis Cabot,
Sanmel F. Clarke,
W. T. Councilman,
Charles B. Davenport, Cambridge.
Harold C. Ernst, Boston.

Charles H. Feruald, Amherst.
-T. Walter Fewkes, Boston.
Edward G. Gardiner, Boston.

Cambridge.
Boston.
Milton.
Boston.
Cambridge.
Brookline.
Williamstown.
Boston.

Cambridge.
Cambridge.
Somerville.
Brookline.

Samuel Ilenshaw,

Alpheus Hyatt,

John S. Kingsley,

Theodore Lyman,

Edward L. Mark,

Charles S. IMinot,

Edward S. Morse,

George H. Parker,

James J. Putnam,

Samuel II. Scudder,

William T. Sedgwick, Boston.

James C. White, Boston.

William M. Woodworth, Cambridge.

Cambridge.

Boston.

Salem.

Cambridge.

Boston.

Cambridge.

Section IV. — 15.
Medicine and Surgery.

Samuel L. Abbot, Boston.
Edward H. Bradford, Boston.
Arthur T. Cabot, Bostou.

David W. Cheever, Boston.
Frank W. Draper, Boston.
Thomas Dwight, Boston.

FELLOWS.

363

Reginald II. Fitz,

Boston.

Charles F. Folsom,

Boston.

Frederick I. Knight,

Boston.

Francis ^Ihiot,

Hyde Park.

Samuel J. Mixter,

Boston.

W. L. Richardson,

Boston.

Theobald Smith,

Boston.

Henry P. Walcott,

Cambridge.

John C. Warren,

Boston.

Class III. — 3Ioral and Political Sciences. — 57.

Section I. — 9.
Philosophy and Juris^prudence.

James B. Ames,
Charles C. Everett,
Horace Gray,
John C. Gray,
G. Stanley Hall,
Nathaniel Holmes,
John E. Hudson,
Josiah Royce,
James B. Thayer,

Cambridge.

Cambridge.

Boston.

Boston.

Worcester.

Cambridge.

Boston.

Cambridge.

Cambridge.

Section H. — 21.

Philology and Archceology.

AVilliam S. Appleton, Boston.

Charles P. Bowditch, Boston.

Lucien Carr,

Franklin Carter,

Joseph T. Clarke,

Henry G. Denny,

Epes S. Dixwell,

William Everett,

William W. Goodwin, Cambridge.

Henry W. Haynes, Boston.

David G. Lyon,

Bennett H. Xash,

Cambridge.
Williamstown.
Boston.
Boston.
Cambridge.
Quincy.

Cambridge.
Boston.

Frederick W. Putnam,
Edward Robinson,
F. B. Stephenson,
Joseph H. Thayer,
Crawford H. Toy,
John W. White,
Justin Winsor,
John H. Wright,
Edward J. Young,

Cambridge.

Boston.

Boston.

Cambridge.

Cambridge.

Cambridge.

Cambridge.

Cambridge.

Waltham.

Section III. — 16.
Political Economy and History.

Charles F. Adams,
Edward Atkinson,
Mellen Chamberlain,
John Cummings,
Andrew M. Davis,
Charles F. Dunbar,
Samuel Eliot,
John Fiske,
A. C. Goodell, Jr.,
Henry C. Lodge,
A. Lawrence Lowell,
Augustus Lowell,
John C. Ropes,
Denman W. Ross,
Charles C. Smith,
F. W. Taussig,

Lincoln.

Boston.

Chelsea.

Woburn.

Cambridge.

Cambridge.

Boston.

Cambridge.

Salem.

Nahant.

Boston.

Boston.

Boston.

Cambridge.

Boston.

Cambridge.

364

RESIDENT FELLOWS.

. Section IV. — 11.
Literature and the Fine Arts.
Francis Bartlett Boston.

John Bartlett,
George S. Boutwell,
J. Elliot Cabot,

Cambridge.

Groton.

Brookline.

T. W. Higginson, Cambridge.

S. R. Koehler, Boston.

Charles G. Loring, Boston.

Percival Lowell, Brookline.

Charles Eliot Norton, Cambridge.

Horace E. Scudder, Cambridge.

Barrett Wendell, Boston.

ASSOCIATE FELLOWS.

365

ASSOCIATE FELLOWS. — 96.

(Number limited to one hundred. Elected as vacancies occur.)

Class L — Mathematical and Physical Sciences. — 36.

San Francisco.
Baltimore.
Wasliington.
Washington.

Section I. — 15.

Mathematics and Astronomij.

Edward E. Barnard, Chicago.
S. W. Burnhani, Chicago.
George Davidson,
Fabian Franklin,
Asaph Hall,
George W. Hill,
E. S. Holden, Mt. Hamilton, Cal.
James E. Keeler, Allegheny, Pa.
Emory McClintock, New York.
Simon Xewcomb, Washington.
Charles L. Poor, Baltimore.
William A. Rogers, AVaterville, Me.
George M. Searle, Washington.
J. N. Stockwell, Cleveland, O.
Chas. A. Young, Princeton, N.J.

Section II. — 7.
Physics.

A. M. Mayer,
A. A. Michelson,
Ogden N. Rood,
H. A. Rowland,

Hoboken, N. J.
Chicago.
New York.
Baltimore.

Section HI. — 7.
Chemistry.

Wolcott Gibbs, Newport, R.L
Frank A. Gooch, New Haven.
S.W.Johnson, New Haven.
J. W. Mallet,
E. W. Morley,
J. M. Ordway,
Ira Remsen,

Charlottesville, Va.
Cleveland, O.
New Orleans.
Baltimore.

Carl Barus,

J. Willard Gibbs,

S. P. Langley,

Providence, R.I.
New Haven.
Washington.

Section IV. — 7.

Technology and Engineering.

Henry L. Abbot, New York.
Cyrus B. Comstock, Washington.
W. P. Craighill, Washington.
F. R. Hutton, New York.

George S. ]\Iorison, Chicago.
William Sellers, Philadelphia.
Robt. S. Woodward, New York.

Class II. — Natural and Physiological Sciences. — 31.

' Section I. — 15.

Geology, Mineralogy, and Physics of
the Globe.

Cleveland Abbe,
George J. Brush,
Edward S. Dana,
Walter G. Davis,
Sir J. W. Dawson,
G. K. Gilbert,

Washington.
New Haven.
New Haven.
Cordova, Arg.
Montreal.
Washington.

James Hall,
Clarence King,
Joseph LeConte,
J. Peter Lesley,
S. L. Penfield,
J. W. Powell,
R. Pumpelly,
A. R. C. Selwyn,
G. C. Swallow,

Albany, N.Y.
New York.
Berkeley, Cal.
Philadelphia.
New Haven.
Washington.
Newport, R.I.
Ottawa.
Columbia, Mo.

]QQ

ASSOCIATE FELLOWS.

Section II. — 3.

Botany.

A. W. Chapman, Apalachicola, Fla.
W. Trelease, St. Louis.

John D. Smith, Baltimore.

Section III. — 6.

Zoology and Physiology.

Joel A. Allen,
W. K. Brooks,
O. C. Mar.sh,

New York.
Baltimore.
New Haven.

S. Weir Mitchell,
A. S. Packard,
A. E. Verrill,

Philadelj^hia.
Providence.
New Haven.

Section IV. — 7.

Medicine and Surgery.

John S. Billings, New York.
Jacob M. Da Costa, Philadelphia.

W. A. Hammond, New York.

William Osier, Baltimore.

Alfred Stille, Philadelphia.

Wm. H. Welch, Baltimore.

li. C. Wood, Philadelphia.

Class III. — Moral and Political Sciences. — 29.

Section I. — 5.

Philosophy and Jurisprudence.

T. M. Cooley, Ann Arbor,Mich.
D. R. Goodwin, Philadelphia.
Charles S. Peirce, New York.
T. R. Pynchon, Hartford, Conn.
Jeremiah Smith, Cambridge.

Section II. — 7.

Philology and Archceology.

A. N. Arnold,
Timothy Dwight,

B. L. Gildcrsleeve,

D. C. Oilman,

T. R. Lounsbury,

E. E. Sahsbury,
A. D. White,

Pawtuxet, R.I.
New Haven.
Baltimore.
Baltimore.
New Haven.
New Haven.
Ithaca, N.Y.

Section III. — 9.
Political Economy and History.
Henry Adams, Washington.

G. P. Fisher, New Haven.

M. F. Force, Cincinnati.

H. E. von Hoist, Chicago.
Henry C. Lea, Philadeljihia.

Edward J. Phelps, Burlington, Vt.
W. G. Sumner, New Haven.
J. H. Trumbull, Hartford, Conn.
David A. Wells, Norwich, Conn.

Section IV. — 8.

Literature and the Fine Arts.
James B. Angell, Ann Arbor, Mich.
L. P. di Cesnola, New York.
F. E. Church, New York.

H. H. Furness, Philadelphia.

R. S. Greenough, Florence.
Augustus St. Gaudens, New York.
E. C. Stedman, Bronxville, N.Y.
W. R. Ware, New York.

FOREIGN HONORARY MEMBERS.

367

FOREIGN HONORARY ME MB E RS. — 65.

(Number limited to seventy-five. Elected as vacancies occur.)

Class I. — Mathematical and Physical Sciences. — 22.

Section I. — 7.

Mathematics and Astronomy.

Arthur Auwers, Berlin.

Francesco Brioschi, Milan.

H. A. E. A. Faye, Paris.

Charles Hermite, Paris.

William Huggius, London.

Otto Struve, Karlsruhe.

H. C. Vogel, Potsdam.

Section II. — 4.
Physics.

Boltzmann, Vienna.

A. Cornu, Paris.

Lord Rayleigh, Witham.

Sir G. G. Stokes, Bart., Cambridge.

Section III. — 8.
Chemistry.
Adolf Baeyer, Munich.

Marcellin Berthelot,
Robert Bunsen,
J. H. van't Hoff,
Mendeleeff,
Victor Meyer,
Sir H. E. Roscoe,
Julius Thomseu,

Paris.

Heidelberg.
Amsterdam.
St. Petersburg.
Heidelberg.
London.
Copenhagen.

Section IV. — 3.
Technology and Engineering.
Sir Henry Bessemer, London.

Lord Kelvin,
Maurice Levy,

Glasgow-
Paris.

Class II. — Natural and Physiological Sciences. — 23.

Section I. — 5.

Geology, Mineralogy, and Physics of
the Globe.

Alfred Des Cloizeaux, Paris.
A. E. Nordenskiold, Stockholm.
C. F. Rammelsberg, Berlin.
Henry C. Sorby, Sheffield.

Heinrich Wild, Zurich.

Section H. — 6.

Botany.

J. G. Agardh, Lund.

E. Bornet, Paris.

Sir .Joseph D. Hooker, Sunningdale.

W. Pfeffer, Leipsic.

Solms-Laubach, Strasburg.

Eduard Strasburger, Bonn.

368

FOREIGN HONORARY MEMBERS.

Section UI. — 8.
Zoology and Physiology.

Michael Foster,
Carl Gegenbauer,
Ludimar Hermann,
Albrecht Kolliker,
A. Kovalevskij,
Lacaze-Duthiers,

Cambridge.
Heidelberg.
Konigsberg.
WUrzburg.
St. Petersburg.
Paris.

Rudolph Leuckart,
J. J. S. Steenstrup,

Leipsic.
Copenhagen.

Section IV — 4,

Medicine and Surgery,

W. Kiihne, Heidelberg.

Lord Lister, London.

Sir James Paget, Bart., London.
Rudolph Virchow, Berlin,

Class III. — Moral and Political Sciences. — 20.

Section I. — 3.
Philosophy and Jurisprudence.

James Martineau, London.

Sir Frederick Pollock, Oxford.
Henry Sidgwick, Cambridge.

Section H. — 7.

Philology and Archceology.

Ingram Bywater, Oxford.
Sir John Evans, Hemel Hempstead.

Pascual de Gayangos, Madrid.

J. W. A. Kirchhoff, Berlin.

G. C. C. Maspero, Paris.

Max IMiiller, Oxford.

Karl Weinhold, Berlin.

Section III. — 7.

Political Economy i

2nd History

Due de Broglie,

Paris.

James Bryce,

Oxford.

W. Dbrpfeld,

Athens.

AV. E. Gladstone,

Hawarden

Hermann Grimm,

Berlin.

Tlieodor Mommsen,

Berlin.

William Stubbs,

Oxford.

Section IV

. — 3.

Literature and the

Fitie Arts.

Jean Leon Gerome, Paris.
John Ruskiu, Coniston.

Leslie Stephen, London.

STATUTES AND STANDING VOTES.

STATUTES.

Adopted May 30, 1854 : amended September 8, 1857, November 12, 1862,
May 24, 1864, November 9, 1870, May 27, 1873, January 26, 1876,
June 16, 1886, October 8, 1890, January 11 and May 10, 1893, April
11, iia?/ 9, and October 10, 1894, and March 13, ^/)ri7 10, a?i(Z May
8, 1895.

CHAPTER I.

Of Fellows and Foreign Honorary Members.

1. The Academy consists of Felloivs and Foreign Honorary Mem-
bers. They are arranged in three Classes, according to the Arts
and Sciences in which they are severally proficient, viz. : Class I.
The Mathematical and Physical Sciences ; — Class II. The Nat-
ural and Physiological Sciences ; — Class III. The Moral and
Political Sciences. Each Class is divided into four Sections, viz. :
Class I., Section 1. Mathematics and Astronomy ; — Section 2.
Physics ; — Section 3. Chemistry ; — Section 4. Technology and
Engineering. Class IL, Section 1. Geology, Mineralogy, and
Physics of the Globe ; — Section 2. Botany ; — Section 3. Zoology
and Physiology ; — Section 4. Medicine and Surgery. Class III.,
Section 1. Philosophy and Jurisprudence ; — Section 2. Philol-
ogy and Archseology ; — Section 3. Political Economy and
Historv ; — Section 4. Literature and the Fine Arts.

2. Fellows, resident in the State of Massachusetts, only, may
vote at the meetings of the Academy.* Each Eesident Fellow
shall pay an admission fee of ten dollars and such annual assess-
ment, not exceeding ten dollars, as shall be voted by the Academy
at each Annual jNIeeting.

* The number of Resident Fellows is limited by the Charter to 200.

VOL. XXXII. — 2^.

3T0 STATUTES OF THE AMERICAN ACADEMY

3, Fellows residiug out of the State of Massachusetts shall be
known and distinguished as Associate Fellows. They shall not
be liable to the payment of any fees or annual dues, but on remov-
ing within the State shall be admitted to the privileges,* and be
subject to the obligations, of Resident Fellows. The number of
Associate Fellows shall not exceed one hundred, of whom there
shall not be more than forty in either of the three Classes of the
Academy.

4. The number of Foreign Honorary Members shall not
exceed seventy-five ; and they shall be chosen from among per-
sons most eminent in foreign countries for their discoveries and
attainments in either of the three departments of knowledge
above enumerated. And there shall not be more than tliirty
Foreign Members in either of these departments.

CHAPTEE II.

Of Officers.

1. There shall be a President, three Vice-Presidents, one for
each Class, a Corresponding Secretary, a Recording Secretary, a
Treasurer, and a Librarian, which officers shall be annually elected,
by ballot, at the Annual Meeting, on the second Wednesday in
May.

2. At the same time, and in the same manner, nine Councillors
shall be elected, three from each Class of the Academy, but the
same Fellows shall not be eligible on more than three successive
years. These nine Councillors, with the President, the three
Vice-Presidents, the two Secretaries, the Treasurer, and the
Librarian, shall constitute the Council. It shall be the duty of
this Council to exercise a discreet supervision over all nomina-
tions and elections. "With the consent of the Fellow interested,
they shall have power to make transfers between the several
Sections of the same Class, reporting their action to the
Academy.

3. If any office shall become vacant during the year, the
vacancy shall be filled by a new election, and at the next stated
meeting, or at a meeting called for this purpose.

* Associate Fellows may attend, but cannot vote, at meetings of the Academy.
See Chapter I. 2.

OF ARTS AND SCIENCES. 371

CHAPTER III.

Of Nominations of Officers.

1. At the stated meeting in March, the President shall appoint
from the next retiring Councillors a Nominating Committee of
three Pellows, one for each class.

2. It shall be the duty of this Nominating Committee to
prepare a list of candidates for the offices of President,
Vice-Presidents, Corresponding Secretary, Recording Secretary,
Treasurer, Librarian, Councillors, and the Standing Committees
■which are chosen by ballot ; and to cause this list to be sent by
mail to all the Resident Fellows of the Academy not later than
four weeks before the Annual Meeting.

3. Independent nominations for any office, signed by at least
five Resident Fellows and received by the Recording Secretary
not less than ten days before the Annual Meeting, shall be in-
serted in the call for the Annual INIeeting, which shall then be
issued not later than one week before that meeting.

4. The Recording Secretary shall prepare for use, in voting
at the Annual Meeting, a ballot containing the names of all
persons nominated for office under the conditions given above.

5. When an office is to be filled at any other time than at the
Annual Meeting, the President shall appoint a Nominating Com-
mittee, in accordance with the provisions of Section 1, which shall
announce its nomination in the manner prescribed in Section 2
at least two weeks before the time of election. Independent
nominations, signed by at least five Resident Fellows and received
by the Recording Secretary not later than one week before the
meeting for election, shall be inserted in the call for that
meeting.

CHAPTER IV.

Of the President.

1. It shall be the duty of the President, and, in his absence,
of the senior Vice-President present, or next officer in order as
above enumerated, to preside at the meetings of the Academy ;
to summon extraordinary meetings, upon any urgent occasion ;
and to execute or see to the execution of the Statutes of the

372 STATUTES OF THE AMERICAN ACADEMY

Academy. Length of continuous membership in the Academy
shall determine the seniority of the Vice-Presidents.

2. The President, or, in his absence, the next officer as above
enumerated, is empowered to draw upon the Treasurer for such
sums of money as the Academy shall direct. Bills presented on
account of the Library, or the Publications of the Academy, must
be previously approved by the respective committees on these
departments.

3. The President, or, in his absence, the next officer as above
enumerated, shall nominate members to serve on the different
committees of the Academy which are not chosen by ballot.

4. Any deed or writing to which the common seal is to be
affixed shall be signed and sealed by the President, when thereto
authorized by the Academy.

CHAPTER V.

Of Standing Committees.

1. At the Annual Meeting there shall be chosen the following
Standing Committees, to serve for the year ensuing, viz. : —

2. The Committee of Finance, to consist of the President,
Treasurer, and one Fellow chosen by ballot, who shall have
charge of the investment and management of the funds and trusts
of the Academy. The general appropriations for the expendi-
tures of the Academy shall be moved by this Committee at the
Annual Meeting, and all special appropriations from the general
and publication funds shall be referred to or proposed by this
Committee.

3. The Rumford Committee, of seven Fellows, to be chosen
by ballot, who shall consider and report on all applications and
claims for the Rumford Premium, also on all appropriations from
the income of the Rumford Fund, and generally see to the due
and proper execution of this trust.

4. The C. M. "Warren Committee, of seven Fellows, to be
chosen by ballot, who shall consider and report on all applica-
tions for appropriations from the income of the C M. Warren
Fund, and generally see to the due and proper execution of this
trust.

5. The Committee of Publication, of three Fellows, to whom
all memoirs submitted to the Academy shall be referred, and to

OP ARTS AND SCIENCES. 373

whom tlie printing of memoirs accepted for publication shall be
intrusted.

6. The Committee on the Library, of three Fellows, who
shall examine the Library, and make an anniial report on its
condition and management.

7. An Auditing Committee, of two Fellows, for auditing the
accounts of the Treasurer.

CHAPTER VL
Of the Seceetaeies.

1. The Corresponding Secretary shall conduct the correspond-
ence of the Academy, recording or making an entry of all letters
written in its name, and preserving on file all letters which are
received ; and at each meeting he shall present the letters which
have been addressed to the Academy since the last meeting.
With the advice and consent of the President, he may effect
exchanges with other scientific associations, and also distribute
copies of the publications of the Academy among the Associate
Fellows and Foreign Honorary Members, as shall be deemed
expedient ; making a report of his proceedings at the Annual
Meeting. Under the direction of the Council for Nomination,
he shall keep a list of the Fellows, Associate Fellows, and Foreign
Honorary Members, arranged in their Classes and in Sections in
respect to the special sciences in which they are severally profi-
cient ; and he shall act as secretary to the Council.

2. The Eecording Secretary shall have charge of the Charter
and Statute-book, journals, and all literary papers belonging to
the Academy. He shall record the proceedings of the Academy
at its meetings; and after each meeting is duly opened, he shall
read the record of the preceding meeting. He shall notify the
meetings of the Academy, and apprise committees of their ap-
pointment. He shall post up in the Hall a list of the persons
nominated for election into the Academy ; and when any indi-
vidual is chosen, he shall insert in the record the names of the
Fellows by whom he was nominated.

3. The two Secretaries, with the Chairman of the Committee
of Publication, shall have authority to publish such of the pro-
ceedings of the Academy as as may seem to them calculated to
promote the interests of science.

374 STATUTES OF THE AMERICAN ACADEMY

CHAPTER VII.

Of the Teeasueer.

1. The Treasurer shall give such security for the trust reposed
in him as the Academy shall require.

2. He shall receive oflficially all moneys due or payable, and
all bequests or donations made to the Academy, and by order of
the President or presiding officer shall pay such sums as the
Academy may direct. He shall keep an account of all receipts
and expenditures; shall submit his accounts to the Auditing
Committee ; and shall report the same at the expiration of his
term of office.

3. The Treasurer shall keep a separate account of the income
and appropriation of the Kumford Fund, and report the same
annually.

4. All moneys which there shall not be present occasion to
expend shall be invested by the Treasurer, under the direction of
the Finance Committee, on such securities as the Academy shall
direct.

CHAPTEPt VIII.

Of the LiBRAElAN AND LIBRARY.

1. It shall be the duty of the Librarian to take charge of the
books, to keep a correct catalogue of same, and to provide for
the delivery of books from the Library. He shall also have the
custody of the publications of the Academy.

2. The Librarian, in conjunction with the Committee on the
Library, shall have authority to expend, as they may deem ex-
pedient, such sums as may be appropriated, either from the Rum-
ford or the General Fund of the Academy, for the purchase of
books, and for defraying other necessary expenses connected with
the Library. They shall have authority to propose rules and
regulations concerning the circulation, return, and safe-keeping of
books ; and to appoint such agents for these purposes as they
may think necessary.

3. To all books in the Library procured from the income of
the Rumford Fund, the Librarian shall cause a stamp or label to
be affixed, expressing the fact that they were so procured.

OF ARTS AND SCIENCES. 375

4. Every person who takes a book from the Library shall give
a receipt for the same to the Librarian or his assistant.

5. Every book shall be returned in good order, regard being
had to the necessary wear of the book with good usage. And if
any book shall be lost or injured, the person to whom it stands
charged shall replace it by a new volume or set, if it belongs
to a set, or pay the current price of the volume or set to the
Librarian ; and thereupon the remainder of the set, if the volume
belonged to a set, shall be delivered to the person so paying for
the same.

6. All books shall be returned to the Library for examination
at least one week before the Annual Meeting.

CHAPTEE IX.

Of Meetings.

1. There shall be annually four stated meetings of the Acad-
emy ; namely, on the second Wednesday in May (the Annual
Meeting), on the second Wednesday in October, on the second
Wednesday in January, and on the second Wednesday in March.
At these meetings only, or at meetings adjourned from these
and regularly notified, shall appropriations of money be made,
or alterations of the statutes or standing votes of the Academy
be effected.

2. Fifteen Fellows shall constitute a quorum for the transac-
tion of business at a stated meeting. Seven Fellows shall be
sufiicient to constitute a meeting for scientific communications
and discussions.

3. The Pvecording Secretary shall notify the meetings of the
Academy to each Fellow residing in Boston and the vicinity ; and
he may cause the meetings to be advertised, whenever he deems
such further notice to be needful.

376 STATDTES OF THE AMERICAN ACADEMY

CHAPTEE X.

Of THE Election of Fellows and Honorary Members.

1. Elections shall be made by ballot, and only at stated
meetings.

2. Candidates for election as Eesident Eellows must be pro-
posed by two or more Eesident Eellows, m a recommendation
signed by them, specifying the Section to which the nomination
is made, which recommendation shall be transmitted to the
Corresponding Secretary, and by him referred to the Council for
Nomination. No person recommended shall be reported by the
Council as a candidate for election, unless he shall have received
a written approval, signed at a meeting of the Council by at least
seven of its members. All nominations thus approved shall be
read to the Academy at a stated meeting, and shall then stasd on
the nomination list during the interval between two stated meet-
ings, and until the balloting. No person shall be elected a Eesident
Eellow, unless he shall have been resident in this Commonwealth
one year next preceding his election. If any person elected a
Eesident Fellow shall neglect for one year to pay his admission
fee, his election shall be void ; and if any Eesident Fellow shall
neglect to pay his annual assessments for two years, provided
that his attention shall have been called to this article, he shall be
deemed to have abandoned his Fellowship ; but it shall be in the
power of the Treasurer, with the consent of the Council, to dis-
pense (sub silentio) with the payment both of the admission fee
and of the assessments, whenever in any special instance he shall
think it advisable so to do.

3. The nomination of Associate Fellows shall take place in
the manner prescribed in reference to Eesident Fellows ; and
after such nomination shall have been publicly read at a stated
meeting previous to that when the balloting takes place, it shall be
referred to the Council for Nomination ; and a written approval,
authorized and signed at a meeting of said Council by at least
seven of its members, shall be requisite to entitle the candidate
to be balloted for. The Council may in like manner originate
nominations of Associate Fellows, which must be read at a stated
meeting previous to the election, and be exposed on the nomina-
tion list during the interval.

OF ARTS AND SCIENCES. 377

4. Foreign Honorary Members shall be chosen only after a
nomination made at a meeting of the Council, signed at the time
by at least seven of its members, and read at a stated meeting
previous to that on which the balloting takes place.

5. Three fourths of the ballots cast must be affirmative, and
the number of affirmative ballots must amount to eleven to effect
an election of Fellows or Foreign Honorary Members.

6. Each Section of the Academy is empowered to present lists
of persons deemed best qualified to fill vacancies occurring in the
number of Foreign Honorary Members or Associate Fellows
allotted to it ; and such lists, after being read at a stated meeting,
shall be referred to the Council for Nomination.

7. If, in the opinion of a majority of the entire Council, any
Fellow — Resident or Associate — shall have rendered himself
unworthy of a place in the Academy, the Council shall recommend
to the Academy the termination of his Fellowship ; and provided
that a majority of two thirds of the Fellows at a stated meeting,
consisting of not less than fifty Fellows, shall adopt this recom-
mendation, his name shall be stricken off the roll of Fellows.

CHAPTEK XI.
Of Amendments of the Statutes.

1. All proposed alterations of the Statutes, or additions to
them, shall be referred to a committee, and, on their report at a
subsequent meeting, shall require for enactment a majority of
two thirds of the members present, and at least eighteen affirma-
tive votes.

2. Standing Votes may be passed, amended, or rescinded, at
any stated meeting, by a majority of two thirds of the members
present. They may be suspended by a unanimous vote.

CHAPTER XII.

Of Litekaey Perfokmances.

1. The Academy will not express its judgment on literary or
scientific memoirs or performances submitted to it, or included in
its publications.

378 STATUTES OF THE AMERICAN ACADEMT

STANDING VOTES.

1. Communications of which notice had been given to the
Secretary shall take precedence of those not so notified.

2. Resident Fellows who have paid all fees and dues charge-
able to them are entitled to receive one copy of each volume or
article printed by the Academy, on application to the Librarian
personally or by written order, within two years from the date
of publication. And the current issues of the Proceedings shall
be supplied, when ready for publication, free of charge, to all the
Fellows and members of the Academy who desire to receive them.

3. The Committee of Publication shall fix from time to time
the price at which the publications of the Academy may be sold.
But members may be supplied at half this price with volumes
which they are not entitled to receive free, and which are needed
to complete their sets.

4. Two hundred extra copies of each paper accepted for publi-
cation in the Memoirs or Proceedings of the Academy shall be
placed at the disposal of the author, free of charge.

5. Resident Fellows may borrow and have out from the
Library six volumes at any one time, and may retain the same
for three months, and no longer.

6. Upon special application, and for adequate reasons assigned,
the Librarian may permit a larger number of volumes, not exceed-
ing twelve, to be drawn from the Library for a limited period.

7. Works published in numbers, when unbound, shall not be
taken from the Hall of the Academy, except by special leave of
the Librarian.

8. Books, publications, or apparatus shall be procured from the
income of the Rumford Fund only on the certificate of the Rum-
ford Committee that they, in their opinion, will best facilitate
and encourage the making of discoveries and improvements which
may merit the Rumford Premium.

9. The Annual Meeting and the other stated meetings shall be
holden at eight o'clock, P. M.

10. A meeting for receiving and discussing scientific commu-
nications may be held on the second Wednesday of each month
not appointed for stated meetings, excepting July, August, and
September.

OF ARTS AND SCIENCES. 879

RUMFORD PREMIUM.

In conformity with the terms of the gift of Benjamin, Count
Ruraford, granting a certain fund to the American Academy of
Arts and Sciences, and witli a decree of the Supreme Judicial
Court for carrying into effect the general charitable intent and
purpose of Count Rumford, as expressed in his letter of gift, the
Academy is empowered to make from the income of said fund, as
it now exists, at any Annual Meeting, an award of a gold and
silver medal, being together of the intrinsic value of three
hundred dollars, as a premium to the author of any important
discovery or useful improvement in light or in heat, which shall
have been made and published by printing, or in any way made
known to the public, in any part of the continent of America, or
any of the American islands ; preference being always given to
such discoveries as shall, in the opinion of the Academy, tend
most to promote the good of mankind ; and to add to such
medals, as a further premium for such discovery and improve-
ment, if the Academy see tit so to do, a sum of money not
exceeding three hundred dollars.

INDEX.

iEoloplus, 199, 205.

Agassiz, A., Recent Exploration of the

Florida Elevated lleefs, 326.
Agave collina, 296.
Aidemoua, 198, 20-1.
Ammonia, Action of, upon Cupri-

ammouium Acetobromide, 237.
Appropriations, 319, 326, 327.
Aptenopedes, 204, 206.
Aristolochia longecaudata, 297.
Asemoplus, 203, 206.
Astephanus pubescens, 299.
Atomic Weight of Magnesium, o3.

Bahia Pringlei, 309.

Balloons, Meteorological Records by

:Means of, 243.
Barytettix, 197, 20-1.
Berendtia laevigata, 39.
Bigelowia pyramidata, 43.
Biographical Notices, List of, 331.
"Thomas Tracy Bouve, 340.
Francis James Child, 333.
Benjamin Apthorp Gould, 355.
Francis Amasa Walker, 344.
Bioplastology, Terms of, 219.
Bletia macristhmochila, 297.
Boutwell, G. S., International Bi-
metallism, 328.
Bouvardia obovata, 304.
Bradvnotes, 199, 205.
Brewster, E. T., A Measure of Varia-
bility, and the Relation of in-
dividual Variations to specific
Differences, 267-230.
Buceragenia, 303.

minutiflora, 303.
Bullard, C. See Davenport, C. B.,

and Bullard C
Butanes in American Petroleum, 101.

Cacalia peltata, 49.
suffulta, 310.

Galea, Revision of the Mexican and

Central America Species, 20.
Galea, 20.

albida, 24.
axillaris, 27.
brachiata, 29.
cacosmioides, 29.
discolor, 30.
elegans, 22.
hypoleuca, 24.
integrifolia, 26.
Liebmannii, 24.
longipedicillata, 28.
manicata, 29.
megacephala, 21.
multiradiata, 22.
Nelsonii, 25.
Orizabaensis, 29.
Palmeri, 21.
peduncularis, 23-
pellucidinervia, 28.
prunifolia, 27.
sabazioides, 22.
salmesefolia. 24.
scabrifolia, 28.
sessiliflora, 30.
submembranacea, 28.
thysanolepis, 22.
tomentosa, 30.
trichotoma, 28.
■ Zacatechichi, 25.
Campylacantba, 198, 204.
Cardiospermum Galapageium, 38.
Carlowrightia glandulosa, 40.

Pringlei, 40.
Castilleia aurea, 39.
Cephalotettix, 197, 204.
Chemical Laboratory of Harvard Col-
lege, Contributions from, 53.
2.37.
Chemical Laboratory of the Case
School of Applied Sciencf^
Contributions from, 99, 119,
177, 183.

382

INDEX.

Chrysopsis Brandegei, 43.
Citharexyluni glabrum, oOO.

ovatifolium, 301.
Coualcfea, 196, 204.
Corniferous limestone petroleum, con-
stituents of, 156.
Correspondence. 315, 324, 326, 327, 328.
Council, Report of, 331.
Cranichis thysauochila, 35.
Crusea calcicola, 305.

coccinea, 306.

cruciata, 306.

villosa, 306.
Cycles, Organic, 207.
Cyclocercus, 197, 204.
.Cylindrolepis, 11.
Cynara CardunceUus, 311.

Dahlia tenuis, 48.

Davenport, C. B., and Bullard, C,
A Contribution to the quan-
titative Study of correlated
Variation and the comparative
Variability of the Sexes, 85-
97.

Dendrotettix, 200, 205.

Dicliptera Pringlei, 302.

Discorea coraposita, 34.

Dysodia serratifolia, 49.

Edison. T. A., Award of Rumford

Medals to, 321.
Eotettix, 198, 204.
Ereicotis, 290.
Erythrsea retusa, 38.
Eucalea, 22.
Euhoustonia, 283.
Eupatoiium eriocarpum, 42.

euonymifolium, 308.

koelligefolium, 308.

oreithales, 308.

rupicola, 42.
Euphorbia Luciismithii, 36.

Oaxacana, 37.

ramosa, 297.

subcserulea, 37.
Eupodisma, 200, 205.
Eutridax, 4.

Fellows, Associate, deceased, —
George Brown Goode, 324.
Atticus Green Ilaygood, 315.
Matthew Carey Lea, 333.
Henry NeweilMartin, 327.
Hubert Anson Newton, 324.

Fellows, Associate, elected, —

James Coolidge Carter, 324.
Fellows, Associate, List of, 365.
Fellows, Resident, deceased, —

Thomas Tracy Bouvd, 323.

Francis James Child, 323.

Benjamin Apthorj) Gould, 326.

Francis Amasa Walker, 327.
Fellows, Resident, elected, —

Charles Henry Fernald, 328.
Fellows, Resident, List of, 361.
Flaveria flaginata, 48.
Florestina pedata, 48.

platyphylla, 49.
Foreign Honorary Members de-
ceased, —

Heinrich Ernst Beyrich, 324.

Ernst Curtius, 324.

Emil Heinrich Du-Bois Reymond,

Hugo Gylddn, 326.

Friedrich August Kekule, 324.

Ferdinand von Mueller, 324.

Jules Simon, 324.

James Joseph Sylvester, 329.

Carl Weierstrass, 329.
Foreign Honorary Members elected, —

Ludwig Boltzmanu, 328.

Wilheim Ddrpfeld, 328.

Wilhelm PfefEer, 328.
Foreign Honorary Members, List of,

367.
Forsythe, R. J. See Richards, T. W.,
and Forsythe, R. J.

Galium praetermissum, 306.

Glands, Correlation in number of,
on two Sides of the Body.
93. ^

Gnaphalium linearifulium, 308.

Gochnatia Smithii, .50.

(Tonolobus chrysanthus, 299.

Goodwin, PL M. See Noyes, A. A.,
and Goodwin, H. M.

Gray Herbarium of Harvard Uni-
versity, Contributions from,
1, 281.

Greenman, J. M., Contributions from
the (Jray Herbarium of Har-
vard University, New Series,
No. 11, 281-311. See also Rob-
inson, B. L., and Greenman,
J. M.

Gymnoscirtetes, 195, 204.

INDEX.

883

Habenaria subauriculata, 34.
Hamelia nodosa, 305.
Hesperotettix, 199, 205.
Houstonia, Revision of the Mexican
and Central American Species,
283.
Houstonia acerosa, 290.

angustifolia, 285.

arenaria, 289.

asperuloides, 289.

bouvardioides, 292.

Braudegeana, 288.

brevipes, 290.

crerulea, 284.

croftife, 284.

fasciculata, 291.

fruticosa, 292.

humifusa, 288.

longifolia, 292.

longipes, 290.

Pahneri, 289.

parviflora, 284.

polypremoides, 291.

purpurea, 285.

rubra, 287.

rupicola, 286.

salina, 28G.

serpyllacea, 284.

subviscosa, 288.

triflora, 293.

Wrightii, 287.
Hudson, E. J. See Mabery, C. F. ,

and Hudson, E. J.
Hyatt, A., Cycle in the Life of the
Individual (Ontogeny) and in
the Evolution of its own Group
(Phylogenv), 207-224.
Hypochlora, 198, 204.

Imbricatse, 11.

Jackson, C. L., appointed Acting
Corresponding Secretary, 324;
vote of thanks to, 319.

Jackson, D. D. See Sedgwick, W. T.,
and Jackson, D. I).

Jacobinia candicans, 41.

Justicia Clinopodium, 304.

Kites, ^leteorological Records by

Means of, 243.
Kittelberger, A. S. See ^labery, C.

F., and Kittelberger, A. S.

Lagascea tomontosa, 43.

Leontophthalmuni, 21.

Liabum, Key to the Mexican Species,

293.
Liabum Andrieuxii, 293.

andromachioides, 293.

angustissimum, 294.

asclepiadeum, 293.

cervinum, 294.

Deppeanum, 293.

discolor, 293.

glabrum, 49, 293.

Klattii, 294.

Liebinannii, 294.

Palraeri, 294.

platylepis, 293.

polyauthum, 293.

Pringlei, 49, 294.

sericolepis, 294.
Librarian, Report of, 317.
Linear Transformations, 75.
Lithospermum oblongifolium, 300.
Livermore, W. R., appointed acting

Recording Secretary, 326.
Livermore, W. R., Growth of the
Roman Dominions in Europe
from 650 b. c. to 200 a. d., 326.

Mabery, C. F., The Constituents of
Pennsylvania, Ohio, and Cana-
dian Petroleum between 150°
and 220=, 119-176.
Livestigations in American Pe-
troleum, 99, 119, 177, 183.

Mabery, C. F., and Hudson, E. J.,
On the Butanes and Octanes in
American Petroleum, 99-118.
Refractive Power of the Hydro-
carbons and Chlorine Deriva-
tives, 177-182.
The specific Refractive Power of
Petroleum Hydrocarbons and
some of their Derivatives, 325.

Mabery, C. F., and Kittelberger, A.
S., On the Composition of a
South American Petroleum,
183-191.

Mabery, C. F., and Quayle, W. O.,
On the Sulphur Compounds
and Unsaturated Hydrocarbons
in Canadian Petroleum, 325.

Macrohoustonia, 292.

Magnesium, Atomic Weight of, .53.

Melanopli, Xorth American, 193.

Melanoplus, 202, 205.

384

INDEX.

Mentzelia Conzattii, 298.
IVIercury Vapor, Viscosity of, 225.
Meteorological records in the upper

air, 24:3.
Mexican Plants, 283, 295.
Microstylis platyglossa, 35.

streptopetala, 36, 297.
Mikania, Synopsis of the Mexican and
Central American Species, 10.
Mikania angulata, 13.

cordifolia, 13.

coriacea, 13.

denticulata, 13.

eriophora, 12.

globosa, 11.

Guaco, 12.

Hookeriana, 12.

Iloustonis, 11.

leiostachya, 12.

olivacea, 12.

pterocaula, 12.

punctata, 13.

pyraniidata, 12.

repanda, 13.

scandens, 13.

Tlalixcoyan, 13.
Montanoa niacrolepis, 44.

rosei, 45.
Morphogenesis, Studies in, 85.

Nama Pringlei, 38.
Pueblense, 39.

Nemastylis caerulescens, 296.

Netrosoma, 195, 204.

Nominating Committee, 328.

Koyes, A. A., and Goodwin, H. M.,
The Viscosity of Mercury Va-
por, 225-236.

Oaxaca, Descriptions of New and lit-
tle known Phanerogams from,
34.

Octanes in Ohio Petroleum, 111.

(Edaleonotus, 203, 206.

Officers elected, 320, 327.

Oldenlandia xestosperma, 41.

Ontogeny, 207.

Organic cycles, 207.

Oteiza, 21.

Paraidemona, 197, 204.
Paradichroplus, 196, 204.
Paratylotropidia, 201, 205.
farker, H. G. See Richards, T. W.,
and Parker, H. G.

Paroxya, 203, 205.
Perezia Cuernavacana, 50.

umbratilis, 50.
Petroleum, Canadian, 119.

Constituents of, 119.

Ohio, 111, 119.

Pennsylvania, 119.

South American, 183.
Petroleum Hydrocarbons, structure of,

173.
Phaedrotettix, 196, 204.
Phaulotettix, 197, 204.
Philocleon, 203, 206.
Phoetaliotes, 202, 205.
Phorodendron Forestierse, 36.
Phylogeny, 207.

Physical Laboratory of the Massachu-
setts Institute of Technology,
Contributions from, 225.
Podisma, 2U0, 205.
Poecilotettix, 203, 206.
Porophyllum, A Provisional Key to
the Species ranging North of
the Isthmus of Panama, 31.
Porophyllum amplexicaule, 32.

coloratura, 33.

crassifolium, 33.

decumbens, 32.

ellipticum, 31.

Ervendbergii, 31.

filifolium, 33.

gracile, 33.

Jorullense, 32.

Lin aria, 33.

Lindenii, 31.

macrocephalum, 31.

Nelsonii, 32.

nummularium, 31.

nutans, 31.

obtusifoliura, 32.

Palmeri, 32.

pausodynum, 32.

Pringlei, 32.

ruderale, 31.

scoparium, 32.

Seemannii, 32.

tagetoides, 33.

tridentatum, 33.

viridiflorum, 31.
Prescott, S. C, An Investigation of
some of the Bacteriological As-
pects of the Art of Tanning,
328.
Proceedings of Meetings, 315.
Pseudotridax, 10.

INDEX.

885

Quayle, W. O. See Mabery, C. F.,
and Quayle, W. O.

Refractive Power of Hydrocarbons,
177.

Rhabdotettix. 197, 204.

Richards, T. W., and Forsythe, R. J.,
On the Action of Ammonia
upon Cupriammonium Aceto-
bromide, '237--4:l.

Richard.s, T. W., and Parker, H. G.,
A Revision of the Atomic
Weight of Magnesium, 53-73.
A Revision of the Atomic Weight
of Strontium. Second Paper :
Tlie Analysis of Stroutic Chlo-
ride, 321.

Robinson, B. L., and Greenman, J.
]M., Contributions from the
Gray Herbarium of Harvard
University, New Series, No. 10,
1-51.

Rontgen Rays, 253.

Ropes, J. C, Forts Henry and Donel-
son, 327.

Rotch, A. L., On obtaining Meteoro-
logical Records in the Upper
Air by Means of Kites and
Balloons, 243-251.

Rumford Committee, Report of, 317.

Rumford Fund. Papers published by
Aid of, 225, 253.

Rumford Medal, Award of, 321.

Schoenocaulon Pringlei, 295.
Sclerocarpus Schiedeanus, 309.
Scudder, S. H., The Genera of North
American Melanopli, 193-206.
Sedgwick, W. T., and Jackson, D. D.,
On the ISIinimal Nutrients of
Bacteria in Water, 328.
Senecio prionopterus, 49.
Sex, Variability correlated with, 91.
Sinaloa. 197, 204.
Solanum deflexum, 301.

Mitlense, 302.
Sonchus asper, 51.
Spiranthes Oaxacana, 34.
Statutes and Standing Votes, 369.
Stevia clinopodioides, 307.

diffusa, 307.
Story, W. E., The Hebrew and Mu-

hammedan Calendars, 327.
Streptotrachelus, 298.

Pringlei, 298.

VOL. xxxii. — 25

Sulphuric Acid, Action on Petroleum
Distillates, 175.

Taber, H., On the Group of real
Linear Transformations whose
Invariant is a real Quadratic
Form, 75-83.
Tagetes triradiata, 310.
Tephrocalea, 30.
Tetrachyron, 29.
Tradescantia macropoda, 295.
Treasurer, Report of, 315.
Trenton Limestone Petroleum, Con-
stituents of, 143.
Tridax, Revision of the Genus, 3.
Tridax angustifolia, 7.

balbisioides, 5.

bicolor, 8.

brachylepis, 4.

candidissiraa, 10.

coronopifolia, 9.

dubia, 10.

Ehrenbergii, 10.

erecta, 8.

Galeottii, 8.

imbricata, 9.

lanceolata, 9.

Mandonii, 9.

obovata, 8.

Palmeri, 6.

petrophila, 5.

Pringlei, 4.

procumbens, 7.

rosea, 6.

tenuifolia, 8.

trilobata, 4.

tuberosa, 4.
Trigonospermum tomentosum, 44.
Trowbridge, J., The Energy Condi-
tions necessary to produce the
Rontgen Rays, 253-265.

Upper Air, Exploration of, 243.

Variability. Range of, 267.
Variation, Correlated, 85.
Verbesina Nelsonii, 46.

oncophora, 309.

Smithii, 46.

stenophylla, 309.

trilobata, 46.

variabilis, 47.
Viburnum elatum, 306.
Viguiera Nelsonii, 45.
Viscosity of Mercury Vapor, 225.

386

INDEX.

Warren (C. M.) Committee, Report

of, 318.
Warren (C. M.) Fund, Aid from,

!)9, 119, 177, 183.
Webster, A. G., Electricity and Ether,

Wendell, 13., The study of English at
Harvard College, 329.

Wolflf, J. E., Exhibition of a Crystal
of Tourmaline, 326.

Zinnia, Revision of the Genus, 14.
Zinnia acerosa, 14.

ambigua, 20.

angustifolia, 18.

anomala, 15.

bicolor, 16.

elegans, 18.

Zinnia grandiflora, 15.

Greggii, 16.

hybrida, 20.

juniperifolia, 15.

Liebmannii, 19.

linearis, 17.

littoralis, 16.

maritima, 17.

multiflora, 19.

Palmeri, 17.

pauciflora, 19.

pumila, 15.

Roezlii, 20.

verticillata, 20.
Zoological Laboratory of the Museum
of Comparative Zoology at
Harvard College, Contribu-
tions from, 85, 267.

MBI, WHOI I.IKKAHY

UH lAflF /

^^"11