HARVARD UNIVERSITY.

LIBRARY

MUSEUM OF COMPARATIVE ZOOLOGY.

H'^RT.

%^\^A>^^5\>u^>^^A^\0

JUL 10 1905

PROCEEDINGS

American Philosophical Society

HELD AT PHILADELPHIA

PROMOTING USEFUL KNOWLEDGE

VOLUME XLIX
JANUARY TO DECEMBER

I905

PHILADELPHIA
THE AMERICAN PHILOSOPHICAL SOCIETY

1905

Press of
The New Era printing Comp
Lancasier, Pa

PROCEEDINGS

American Philosophical Society

HELD AT PHILADELPHIA

PROMOTING USEFUL KNOWLEDGE

VOLUME XLIV
JANUARY TO DECEMBER

I905

PHILADELPHIA
THE AMERICAN PHILOSOPHICAL SOCIETY

1905

PROCEEDINGS

^A^Vt) OF THE

AMERICAN PHILOSOPHICAL SOCIETY

HELD AT PHILADELPHIA

FOR PROMOTING USEFUL KNOWLEDGE

Vol. XLIX. . January-April, 1905. No. 179.

CONTENTS.

Stated Meeting, January 6 3

Stated Meeting, January 20 4

Stated Meeting, February 3 4

Stated Meeting, February 17 4

Stated Meeting, March 3 5

Stated Meeting, March 17 5

The Filipino; His Customs and Character. By John A.

Metzger 6

Sociology of the Aborigines of Western Australia. By R. H.

Mathews, L.S 32

Stated Meeting, April y ' 35

General Meeting, April 12-14 35

A Plea for Governmental Supervision of Posts Necessitating Nor-
mal Perception of Color. By Charles A. Oliver 40

Emancipation of the Waterways. By Lewis M. Haupt 42

The Oligodynamic Action of Copper Foil on Certain Intestinal

Organisms. By Henry Kraemer 51

The Effect of Preservatives on Metabolism. By H. W. Wiley,

M.D 66

Notes on the Osteology of Sinopa ; A Primitive Member of

the Hyaenodontidae. By W. D. Matthew 69

The Marsupial Fauna of the Santa Cruz Beds. By Wm. J.

Sinclair 73

The Straight Line Concept. By P. A. Lambert .k 82

PHILADELPHIA

THE AMERICAN PHILOSOPHICAL SOCIETY

104 South Fifth Street

I905

GENERAL MEETING— 1906

The next General Meeting of the Society will he held on April 17-20.
1906; beginning on the evening of Tuesday, April 17.

Wednesday, April iS, will be devoted to the presentation and discussion
of scientific papers, and Thursday, 19 and Friday, 20, to the ceremonies
connected with the celebration of the 200th Anniversary of the Birth of
Benjamin Franklin.

Members desiring to present papers on subjects of science at the Gen-
eral Meeting are requested to communicate with the Secretaries at the
earliest possible date.

Members who have not as yet sent their photographs to the Society will
confer a favor by so doing: cabinet size preferred.

It is requested that all correspondence be addressed
To the Secretaries of the

AMERICAN PHILOSOPHICAL SOCIETY
104 South Fjktii Street

Philadelphia, 17. S A.

PROCEEDINGS

OF THE

AMERICAN PHILOSOPHICAL SOCIETY

HELD AT PHILADELPHIA
FOR PROMOTING USEFUL KNOWLEDGE

Vol. XLIV. January-April, 1905. No. 179.

Stated Meeting, January 6, 1905.

President Smith in the Chair.

The resignation of membership by Prof. Charles De Garmo,
was presented and accepted.

The decease of the following members was announced :
Prof. Benjamin W. Frazier, at Bethlehem, Pa., on January

4, 1905, aet. 64.
Edward H. Weil, at Philadelphia, on January 5, 1905,
aet. 68.
The Judges of the Annual Election of Officers and Coun-
cillors reported that an election had been held on the afternoon
of this day and that the following named members had been
elected to be the officers for the ensuing year :

President
Edgar F, Smith.

Vice-Presidents
George F. Barker, William B. Scott, Simon Newcomb.

Secretaries

I. Minis Hays, Edwin G. Conklin, Arthur W. Goodspeed,

Morris Jastrow, Jr.

3

4 MINUTES. [March 17,

Curators

Charles L. Doolittle, William P. Wilson, Albert H. Smyth.

Treasurer '

Henry La Barre Jayne.

Councillors to serve for three years

George F. Edmunds, James T. Mitchell, Joseph Wharton,

William W. Keen.

Stated Meeting, January 20, zpoj.
President Smith in the Chair.

The decease of the following members was announced :
Sir Lowthian Bell, Bart, at North Allerton, Eng., on
December 20, 1904, set. 88.
Dr. C. Juhlin Dannefeld, of Stockholm.
The following papers were read :
" Biblical Pessimism," by Prof. Paul Haupt.
"Universal Radio-Activity," by Prof. M. B. Snyder.

Stated Meeting, February 3, 1903.

President Smith in the Chair

The decease was announced of Mr. William Sellers, at
Philadelphia, on January 24, 1905, xt. 81.

Prof. Edwin G. Conklin read a paper on " Development
and Evolution."

Stated Meeting, February //, 1903.

President Smith in the Chair
The decease was announced of Prof. Alpheus Spring Pack-
ard, at Providence, on February 14, 1905, ait. 66.

'90s.] MINUTES. 5

Prof. Amos P. Brown read a paper on " The Rocky Moun-
tains."

Stated Meeting, March j, ipoj.
President Smith in the Chair.

The decease was announced of Prof. Albert Benjamin
Prescott, at Ann Arbor, Mich., on February 25, 1905, aet. 72.

Prof. Hermann V. Hilprecht read a paper on " Recent Re-
searches in the Temple Library at Nippur."

Stated Meeting, March ij, 1905.

President Smith in the Chair.

The decease of the following members was announced :
Prof. John Lyle Campbell, at Crawfordsville, Ind., on Sep-
tember 7, 1904, set. yj.
Hon. James C. Carter, at New York, on February 14,

1905, aet. 78.
Richard Somers Hayes, at New York, on March 2, 1905,

aet. 58.
James Dundas Lippincott, at Philadelphia, on March 6,
1905, set. 66.
The following papers were read :

" The Filipino, his Customs and Character," by Dr. J. A.
Metgzer. (See page 6.)

" The Sociology of the Aborigines of Western Australia,"
by R. H. Mathews. (See page 32.)

METZGER— THE FILIPINO. LMarcm7,

THE FILIPINO ; HIS CUSTOMS AND CHARACTER.

BY JOHN A. METZGER, M.D.

{Read March ij, igoj.)

The early history of the people of the Philippines can, unfortu-
nately, be none other than that of imperfect conjecture. We
do know, however, that the present-day Filipino is not the direct,
unalloyed descendent of the aboriginal inhabitants of these islands
but have lineage from some nomadic people who, through affiliation
with the aborigines, have given to the ethnologist this almost in-
comprehensible human being. The progenitorial element is un-
questionably Malayan but the source is a much debated question.
Some authorities contend that the ancestors of this great semi-
civilized people came from Chili, drifted thither by the currents
and prevailing winds, while others with as equitable reasoning
believe them to have migrated from the shores of Madagascar and
Patagonia. Neither the paleontologist nor the paleographer has thus
far been able to throw any definite light on the source or character
of the original inhabitants of th s dependency. However, the
generally accepted theory points to a mountain tribe called the
Negrito or yEtas, which is universally regarded as the surviving
remnant of this once powerful people who first populated the
archipelago.

From a paleontological and structural point of view we are wont
to believe that during the later Miocene or the very early Pliocene,
there was that progressive uplifting of the land which subsequently
became separated from Borneo and the Asiatic continent (through
Formosa) by the present China sea. The evidence which warrants
this deduction must be admitted is very fragmentary, however, the
distribution of living forms is certainly calculated to throw some
light on the more recent history of these islands and should be
made to contribute all it can, but at the same time it must not be
forgotten that the obstacles which seem geologically of small
moment may limit the extension of species. The island of Cebu
affords a striking example of this fact regarding the bird fauna and
mammalia which are regarded as the descendant forms of Borneo
and Continental Asia. This theory is corroborated by Mr. Waller,

i9°5] METZGER— THE FILIPINO. 7

who has given this considerable study, when he says: "Absence
of a large number of Malayan groups would indicate that the actual
connection with Borneo, which seems necessary for the introduction
of the Malayan types of mammalia, with the large proportion of
wide-spread continental genera of birds would seem to imply that
greater facilities had once existed for the migration from Southern
China, at which time the ancestors of that peculiar deer seen in
Samar and Cebu entered the islands." It, therefore, seems impos-
sible to understand this existing fauna unless it can be assumed that
island connection must have existed. Accepting this theory, why
then should not primitive man have made his ingress from Borneo
or Continental Asia? This question of the aboriginese is indeed a
field for research and is one for the ethnologist and not the province
of a mind inexperienced in this line of study.,

Conceding for the present the Negrito to have been the aborig-
inal inhabitants, we have as yet to discover any signs or writings
of an early day which might lead us to a solution of the origin of
this strange tribe. We have, however, characters, many of which
are hieroglyphical, of the ancient Tagalog, Visayan, Yliocano,
Pampango, Pangasinan and Tagbanua. These characters were ex-
pressed or inscribed on tubes of bamboo, with some pointed in-
strument the nature of which is as yet unknown, and like the
present-day dialects of the several tribes there seems to have been
a great preponderance of consonants and a very limited vocabulary.
A comma above a letter, should it be a consonant, gave it the sound
of having been written with an E or I, and if below as O or U.

Upon the conquest of the archipelago by the Spaniard their
alphabets were abandoned by many and the Spanish or the original
of the present mongrel dialects were adopted and after a period of
three hundred years there is scarcely a person to be found who can
either read or write in the original characters. This, however, is the
field of the paleographer but, I believe, is worthy of mention in this
connection. The adoption of the Spanish language by some of the
tribes was the first step in the domestication of these people, in
that it permitted the placing of the Doctrina in their hands with
the consequent closer affiliation. (For those wishing to further in-
vestigate these early languages of the Filipino, I would refer them
to the writings of the Agustinian father Marcilla, and especially
his " Estudio de los antigues alfabetos Filipinos.")

8

METZGER— THE FILIPINO.

[Mar

Of the fifty odd different tribes there are almost as many distinct
dialects, however, with few exceptions, there is a general similarity
which permits of mutual comprehension. There are not only many
words in common peculiar to the native tongue but Spanish words
have been adopted into most of the dialects. The Tagalog and

I. Negritos.

Yliocano are probably the most general in the northern country,
while the Visayan and the Mahratte dialect of the mother Sanskrit
predominate in the middle and south lands.

The Tagalog, Yliocano and Viscayan are gutteral languages 01
great preponderance of consonants and limited vocabularies.

The remnant of the tribe of Negritos, the supposed descendants

i9o5] METZGER— THE FILIPINO. 9

of the aboriginese, scarcely number five thousand at the present
time and are scattered widely over all the northern islands, living
in the most remote and dense parts of the hill country. They are
pygmean in stature, barely reaching four and one half feet in height
and resembling closely the Alfoor Papuan of New Guinea. Al-
though small in frame they are powerful andfleet of foot. Unlike
any of the other tribes of this archipelago they are the possessors
of a closely matted kinked head of hair. The Negrito is of very
low intellect and appreciates no conception of social order. He
is cowardly and indolent, but exhibits a marked respect for the
aged and dead such as is not seen among any of the other tribes.
Frequent attempts have been made to civilize these little people
but without success, for they will neither endure social or military
restraint but prefer to return to the mountain fastnesses and their
nomadic state. Model villages of bamboo and nepa were built in
Upper Pampanga by the Spaniards with the object of domesticat-
ing these strange people. They were supplied with food, clothing
and all the necessaries of life for a period of one year or until such
a time as they could till the soil and provide for their future but the
experiment was an utter failure and in a short time the subsidy was
discontinued. They have never been either individually or collec-
tively brought under the influence of the Church but to this day con-
tinue to worship the sun and elements as did their forefathers.
The Negrito subsists wholly upon reptiles, fish, herbs and wild moun-
tain rice. They wear no clothing except the breech-clout and
their customs and habits are those of the savage. Ablution of body
is something almost unknown to them. These little people have
no permanent abode but wander about in little bands of five to
twenty living in trees as a matter of safety. They are more or less
peacefully inclined but do occasionally make incursions into the
territory of some neighboring tribe for the purpose of carrying oft
cattle. Their means of defense is a bow made from the palma-
brava and poisoned arrows, and with these they are indeed expert
marksmen. There is no doubt but that at an early period in the
history of these islands these dwarfish-people were in great numbers
and as rulers levied heavy tribute upon the accessors of some of the
present day tribes, but as emigration increased they were gradually
forced into the background and subsequently, upon the advent of
the white-man, were forced, through terror, to take definitely to
the mountain fastness.

10 METZGER— THE FILIPINO. [March 17,

The exact number of tribes in existence on the several islands at
the present time is not definitely known, however, the following
are a few of those which we as a foster nation must deal with :
Tagalog, Viscayan, Macabebe, Yliocano, Musulman, Igorrote,
Malaneg, Pampangan, Pangasinan, Itanes, Goddan, Tinguian,
Dodayan, Idayan, Apayao, Negrito, Itugoao, Ibiloa, Zambal,
Vrigrito, Cebuano, Panayano, Munabo, Coyuro, Calamino, Agu-
tamo, and that great hybrid class, the Maestizo. Other of the
fifty-two tribes, which have thus far been determined, might be
mentioned, but I believe it to be superfluous here, as their customs
are in the main those of the aforementioned. In this ethnical
analysis I have dealt solely with generic denominations, for whilst
these tribes are subdivided, the clans show no material moral or
physical difference and the local names are apt to be confusing.
Lie,siwe, in order to avoid prejudice, it becomes necessary to
divide this great congeries of humanity into two great classes, the
domesticated Filipino and the properly termed savage. Conserva-
tive estimation elicits the fact that three hundred thousand of the
population of this archipelago are human beings in whom exotic
notions do not pertain and in whom are the instincts of the wild
animal, and of this number one fifth are to be found on the island
of Luzon, the largest and at the same time the most enlightened
from ecclesiastic and worldly standpoints.

As all uncivilized human beings have characteristics in common
and at the same time many distinctive traits characterize a people
surrounded by the same natural environments, to recount these as
they pertain to the several tribes is wholly unnecessary. It is suf-
ficient to point out a few of the characteristic features of the more
powerful of this class of untamed nomads as they pertain to the
Philippines.

Probably the most unrestrained and barbarous Filipinos are the
Gaddanes. A race occupying the extreme northwestern end of the
archipelago and entirely out of the pale of civilization. They are
the only real, war-like people of the North. They know no moral
restraint and glory in the shedding of blood. At a certain time of
the year, when the so-called fire-tree is in bloom, the young men,
as is their custom, go forth on a head hunting expedition and vie
with each other in presenting to the sachem of their tribe all the
grewsome trophies they are able to take from their enemies, as a

METZGER— THE EILIPINO.

1 I

proof of their manliness and courage. The arms used by these
people are wicked looking lances with trident tips and arrows car-
rying at the point a mesh of bats' claws which have previously been
dipped in the venom of snakes.

m

•i

teJ^^S'

i^^^^Bb

H4

h

1

N£'N

The men are magnificient specimens of physical strength, and
with a wealth of long, straight, jet-black hair reaching to the
shoulders and with a color of skin of almost equal hue, they make a
very striking picture of primitive man. Like the Negrito they
subsist on roots, game and such other food stuffs as falls in their
path. They are pagans and at no time has any attempt been made
to persuade them to embrace the western system of civilization.

\-^

METZGER— THE FILIPINO.

[March 17,

Xext to the Gaddanes in war-like propensity are the Itaves, a
tribe inhabiting the territory just to the south and adjoining that
of the head-hunters. Their manners of living and religion* are
similar to those of the Gaddanes, but they are not so fierce and are
more indolent. They are also lighter in color and wear their hair
shorter than their neighbors of the north. This tribe is one of the
few, if not the only one, which uses the war-club and executes a
war-dance preliminary to going into battle.

Fig. 3- Igorrote.

To the American, unquestionably the most interesting people of
this class of Filipinos, is the Igorrote, for the reason that they more
closely resemble the American Indian in color, configuration and
habits than any other tribe of the archipelago. These people domi-
nate the middle north country, where they live in idleness, oblivious
of time or conditions. Of all the tribes of the islands they are the
acme of physical development. Their chief ethnical characteristics
are the high cheek bones, aquiline nose, copper colored skin, long
straight black hair cut into a fringe over the forehead, naked save
for a breech-clout, and gaudily bedecked with paint, feathers and
rings. Unlike the American Indian their lips are thick and large,

i9o5] METZGER— THE FILIPINO. 13

and their gait is sluggish and toddling. They, like their brothers
of the North subsist upon that which nature is kind enough
to cast in their way, however, they do make occasional futile at-
tempts at cultivating a little sugar cane or rice. These people can-
not justly be termed a war-like people, yet revenge is one of their
strongest traits. Distrust of the white-man is a preeminent feature
of this tribe. This fixed dislike is the result of one of the many of
Spain's blunders in her sovereignty of these islands, for it was her
attempt to force western civilization on these people, who did not
wish to exchange the comforts, usages and independence of their
primitive state, for what the crown of Spain deemed a proper con-
stituent principle of good colonists. To roam at large in their
forest home, free as the wind, was to them more to be desired than
to have to wear clothes, pay taxes and incarcerate themselves in the
conventional domestic habits of the European. Foreman aptly ex-
presses it when he says "as to Christianity, it would be as hard a
task to convince them of what Roman Catholicism deems indispen-
sable for the salvation of the soul as it would be to convert all Eng-
land to the teachings of Buddha, although Buddhism is as logical a
religion as Christianity." The distrust incident to this forcible
attempt to civilize and domesticate this people will remain, I be-
lieve, one of their prime distinctive characteristics for centuries to
come.

There is a hybrid class of Igorrotes, known as the Chino-igorrote.
A people differing little from the unmixed blood except that asso-
ciated with the brutal instincts there is the cunning and astuteness
of the Mongol. This mongrel race are supposed to be the descen-
dants of an issue, the result of the affiliation of the dispersed fol-
lowers of the Corsair Li-ma-hong who attacked the city of Manila
and was routed and fled to the region of the Igorrotes. They, like
their half brothers, are confirmed infidels.

Of the uncivilized tribes of the north, there are a few who, owing
to their distinctive characteristics, are worthy of mention, the
principal of which tribes are the Tinguians, Dayapes, and a
peculiar class of tropical inhabitants known as Albinos.

The Tinguians inhabit principally the district of Al Abra and in
appearance closely resemble the Igorrote, and appear to be as
intelligent as the ordinary subdued native. They are pagans but
have no temples. Their gods are hidden in the cavities of the

14 METZGER— THE FILIPINO. [March i7l

mountain fastnesses. These idols are called Anitos and are ex-
horted when any dire calamity befalls them, and are always
appealed to when a child is to be named. In this latter ceremony
the priest to the Anito holding the new born in one hand raises a
large knife or bolo over its head and upon lowering the blade
strikes it into a nearby tree, if the tree emits sap the first name
uttered is the one the child will henceforth bear. The oozing of
the sap signifies to them the will of the deity. The Tinguians are
monogamists and generally are forced by the parent to take a mate
before the age of puberty. These people, like the Negritos, live
mostly in a baji built in trees, sometimes sixty or seventy feet from
the ground. They have a few characteristics akin to the Japanese,
principally in the manner of wearing the hair, tuft on the crown
of the head, and the custom of blackening the teeth. Their com-
mon weapon is the spear, this they use as a matter of defense as
well as a means of slaying animals for food.

The Davanese are unquestionably Hindoos and are supposed to
be the descendants of the Indian Sepoys, who deserted the British
Army when the latter occupied the city of Manila in 1763. They
are few in number and occupy principally the district lying about
the pueblo of Cainta. These people are semi-civilized, peaceful
and to an extent industrious.

There are to be seen among the natives of the north a few of the
class of people known as Albinos. These abnormalities of nature
present a marble white skin, pink white hair, and pink eyes. They
are not associated in tribes or clans but may be found scattered
about in most any of the provinces of the north.

Before taking up the analysis of the various domesticated tribes,
which go to make up two thirds of the seven million of people who
are styled Filipinos, a brief epitome of their early political history
is, I believe, essential, as it no doubt has indelibly modified and
ultimately formulated the character and customs of these people.
We are wont to believe that long before the advent of the Spaniards
in this Colony, these islands were visited by the Molaccans, for it
was from them that Hernando de Maghallanes, then a Portuguese
subject and in the service of his majesty, learned of the existence
of these supposed rich possessions in the Pacific, and had it not been
for petty jealousies and a weak and arrogant monarch, these same
Philippine Islands might have become the possession of Portugal

I905-] METZGER— THE FILIPINO. 15

and not of Spain, as they subsequently became through the public
renunciation of Maghallanes to his rights as a Portuguese citizen,
and his assumption of the fosterage of Spain, with the result of his
entering into a contract with the King of Spain to seek and discover
these islands of which he (Maghallanes) had heard. Sufficient to
say, that Maghallanes, knighted and invested with the habit of St.
James, set sail from the harbor of San Lucor de Boramida, August,
1519, in command of a fleet of five small vessels, which was to figure
in history as not only the first to formally discover the Philippine
Islands but the first to circumnavigate the globe, thus proving the
theory of Aristotle and Ptolemy.

After twenty -one months of privation, scurvy, mutiny and deser-
tion Maghallanes entered the Butan River on the Island of Min-
hanao, and effecting a landing without any opposition from the
natives, took possession in the name of King Charles of Spain,
thereby realizing his one ambition to discover those islands which
had been his constant dream for years. Thus in part he was
recompensed for the bitterness of the past, but he was not decreed
by fate to enjoy the fruits of his discovery, as he fell mortally
wounded by a poisoned arrow soon after in a conflict with the
natives on the island of Magtan. The command of this expedition
fell to Duorte de Borbosa, who also met his death soon after at the
hands of the natives of the island of Cebu. Juan Corobola, next
in command, finding his ships in a leaky condition and crews
insufficient in number abandoned all the ships except the Victoria,
and returned to Spain, first touching at Borneo and the Molaccas,
arriving in the harbor of San Lucor, September 6, 1522. Again
in 1542 a second expedition from Spain under Villalobos touched
on the island of Luzon. Here, like his predecessors, he met his
death. From 1542 to 1564 no more expeditions were sent out by
Spain. Finally, on account of the bitter jealousy existing between
Spain and Portugal over new acquisitions of territory, another
expedition was dispatched by King Philip, under Maguil Lopez de
Legaspi, in November, 1564. This expedition encountered even
more opposition from the natives than the former ones, and for a
period of five years Legaspi was busily engaged forcibly colonizing
these people. On the twenty-fourth of June, 1571, the city of
Manila was incorporated as the capital city of the archipelago, after
a treaty had been consummated with the native Rajahs, Dolumal
and Lacaubola.

16 METZGER— THE FILIPINO. [March 17.

Soon after this formal acquisition of the islands and the incorpora-
tion of its capital, Legaspi returned to Spain where he died,
destined like his predecessors to enjoy but little of the honor of
having been the first to establish real sovereignty for Spain in this
colony.

Spanish suzerainty of the Phillippines was not one long glorious
regime, neither were the islands the El Dorado they had fancied,
but instead her three hundred years of reign was but a period of
almost constant strife. Other nations strove to seize them and
rebellion followed rebellion in an effort to expel a sovereign power
whose reign was considered unjust, oppressive and tyrannical. In
truth, Spanish sovereignty was never complete except in name
only, and full domination only extended over the sea-coast towns
and a few miles into the interior. Tribal customs governed as
many, if not more, of the inhabitants as Spanish laws and Spanish
monastics.

The Spanish friar was next installed and, with the aid of the
military, set about civilizing and converting to Christianity those
tribes lying outside the Capital city.

About this time the island of Luzon was invaded by the Chinese
under the notorious pirate Li-ma-hong and the Japanese Sioco.
Early on the morning of the thirtieth of November, 1574, they
appeared in the bay of Manila and instituted a vigorous attack.
After a bloody hand-to-hand conflict the Chinese were completely
routed and, not being able to regain their fleet, fled up the coast
as far as the Province of Pangasinan, and it is through the affilia-
tion of these survivors with the natives that we accredit the mani-
fest traces of Chinese blood among some of the hill-tribes to-day.

Following the attempt of the Chinese to seize this Colony the
Emperor of Japan, learning of the European colonization, sent
one of his suite, Ferranda Kiemon, with a message to the Gover-
nor of the islands, demanding prompt surrender and threatening
invasion if refused. This, Gomez Perez Dasmarinas, the Gover-
nor, refused to do but solicited a treaty of commerce, and ex-
pressed a desire to conclude an offensive alliance for mutual pro-
tection. The Mikado consented to this proposition and thus for a
time amicable relations were assured with the Japanese.

As a result of the war with the Flanders, which terminated with
the Treaty of Antwerp in 161 9, the Dutch were obliged to seek in

i9°5.] METZGER— THE FILIPINO. 17

the far east such commodities as they were previously accustomed
to obtain on the peninsula, consequently they established trading
headquarters in the Molacca islands, and from there preyed upon
the Spanish galleons carrying provisions and silver from New Spain
to the Philippines. This state of piracy continued until 1645,
when the Dutch navy under Admiral Whitier, attacked the city of
Manila with twelve men-of-war and was defeated by General
Lorenzo Ugarte with great loss, including that of the commander
of the fleet.

The period from 1645 t0 I7I9 was one °f contention between
Church and State, as to prestige in the civil affairs of the colony.
This dissention became more marked and the bitter feeling thus
engendered finally culminated in one of the most revolting scenes
in Philippine history. Little is to be said of this most disgraceful
affair other than that a riotous mob led by the priests of the Sacred
Orders of San Fancis, San Dominic and Saint Augustine attacked
the palace, stabbed and dragged the Governor, Fernando Busta-
mente Bustillo y Riieda, through the streets of Manila, and at the
same time killed his son. The mob during their delirium, tore
down the Royal Standards and maltreated all those who in any way
offended them. A mock investigation was made in due official
form but little or no punishment was inflicted on any of the
offenders.

Early in 1561 England became involved in a war with Spain
through the so-called Family Compact — an alliance formed by the
three branches of the House of Bourbon — and this resulted in the
city of Havana and many other of the West India ports falling into
the hands of the British, and at the same time the sending of a fleet
of thirteen ships, under Admiral Carnish, to the Philippine waters.
A siege was begun on the twenty-fourth of September with heavy
cannonading from the ships and was replied to by the batteries 01
Fort Santiago and San Andres. At the same time troops, to the
number of five thousand, were landed to the south of the city and
at once engaged the Spanish allies (about five hundred native
Pangasenans) driving them back in great disorder to the fortified
city. This state of siege lasted for fifteen days, during which time
General Draper communicated freely with the Acting Governor
relative to surrender. The capitulation was finally accomplished
on the sixth day of October after great loss of life, and the British
flag soon waved over the walls of Fort Santiago.

18 METZGER— THE FILIPINO. [March i7)

By the terms of the Pacto de Paris, which reached Manila on the
twenty-seventh of August, 1763, the British evacuated the islands,
but peace and quiet did not follow. Hardly had the Spanish colors
been unfurled ere the natives of Cagayan, Ylocos and Pangasanan
provinces broke out in open rebellion under a religious fanatic
Diego de Silan, a half-caste Indian, who declaring the Spanish
sovereign a usurper, directed that no more tribute be paid to the
Spanish Treasury. This insurrection assumed considerable propor-
tions and not until many lives had been sacrificed and noteworthy
concessions made by Spain was peace established.

During this revolt in the north country, the Mussulmans under
Datto Teng-teng, attacked the Spanish garrisons on the island of
Mindanao, butchering their prisoners and destroying much of the
public property. This outbreak was, however, but one of the
many reprisals of the Mussulmans as the result of the enforcement
of a sovereignty and a religion which was to them nauseous and
antagonistic to the Mohammedan faith.

In 1872 occurred what is known as the Cavite insurrection. The
real cause of this rebellion was the native opposition to the Spanish
friars holding parochial incumbencies contrary to the decision of
the Council of Trent. However, the friars claimed to have such
authority, by virtue of papal bulls issued by Pius V, wherein they
were authorized to act as parish priests where the native clergy
were insufficient in numbers. This authority, unfortunately, was
abused, doubtless on account of the friars recognizing that full and
strict compliance meant monastic impotence politically. This
uprising of the natives was promptly suppressed and their leader,
Jose Burgos and his confederates, were duly executed, upon the
instigation of the friars, on the Luneta (Manila's famous drive) in
accordance with Spanish custom. The moral effect of these execu-
tions, however, was but temporary and only served to engender a
more bitter feeling against the friars, and at the same time, this
one act of Spain's, was the prime factor in the formation of one of
the most powerful .freemasonries in the world, the Katipunan.

This was the beginning of the end of Spanish rule in the Philip-
pine islands, for it meant the coalescence of all of the tribes, with
the common object of expelling a power (the friars) which was
not .only odious and tyrannical, but dictatorial and to which the
Spanish government of the islands was subservient. The cry of

i9°5] METZGER— THE FILIPINO. 19

the native was not against Spain as a potentate but against the
dominant power of the friars. Spain's avaricious propensity seemed
to have subverted her better judgment, and this nation, that at one
time was a power potent, was soon to experience the worst insur-
rection in the history of her Philippine dependency.

She had, by virtue of the Cortes de Cadiz, convened on the
twelfth of September, 1809, passed the first Suffrage Bill, which
permitted of the assembling of deputies from the various depen-
dencies. For twenty years the people of this colony enjoyed politi-
cal equality, but finally in 1837, their exclusion was voted as was
also the government of the islands by special laws. Spain's mis-
take was irremediable, the native had tasted of equality and suffrage
and he was apprehensive of the motive force back of this repeal and
it was this innate contempt for the timorous, so characteristic of
this people, and the hatred engendered through the treatment
accorded Jose Burgos that finally culminated in the insurrection of
1896 and '97, the result of which was the sacrifice of many lives,
especially that of Jose Rizal (a story in itself), one of Polaviejo's
most shameful acts, the imprisonment of thousands of suspects in
the dungeons of Fort Santiago, who were drowned like rats upon
the rising of the tide, the breaking of the treaty of Biac-na-bato
and finally the indelible stamp of distrust of the white-man by the
native.

With the American occupation and subsequent history, we are
all familiar and does not permit of repetition here. From this
brief summary of the political history of this colony you will have
observed the potent agencies and modifying forces the native has
been subjected to for a period of three hundred years and now we
can take up the analysis of these people who have been subjected
to this environment.

For practical purposes, we will divide the various domesticated
tribes into three great classes and endeavor to point out the char-
acteristics of the tribes which dominate the several territorial
divisions.

The Tagalog dominates the northern islands, the Visayan, the
central group and the Mussulmans, or so-called Moros, the south-
ern islands of the archipelago. There exists no mutual feeling or
harmony between these tribes, yet they may unite against a common
enemy as in the recent insurrection. The Tagalog and the Visayan

20

METZGER— THE FILIPINO.

[March 17,

listen to the teachings of the Roman Catholic church, while the
Mussulmans are the followers of Mohammed and never during the
three hundred years of Spanish sovereignty were they brought
under either her religious or political control.

Fig. 4. Tagalog Girl.

The Tagalog as a tribe, numbering about seven hundred thousand ,
are the most civilized of the three great divisions of the domesti-
cated Filipinos. This is probably due to the fact that ever since
the conquest of the islands by the Spaniards, they have been
brought in direct contact with Europeans and have felt to an
extent the influence of domesticity and social order. The Tagalog
and the Visayan differ very little in physique and configuration of

1905J METZGER— THE FILIPINO. 21

countenance, but their attitude towards strangers (Europeans) is
most distinctive. The Tagalog feigns great friendship, while the
Visayan is haughty and arrogant. From a physical point of view
they both are magnificent specimens of humanity but mentally an
anomaly which is most unfathomable. They are about five and
one half feet in height, ginger-bread in color, with high cheek
bones, flat nose and a wealth of coarse, straight, black hair pre-
senting at all times a lavish amount of cocoanut oil and surmount-
ing a placid countenance.

The innate spontaneity of moral character of these so-called civi-
lized Filipinos is that of half child and half devil. In him we see
that puerile lack of objective and simplicity, while beneath that
placid countenance and solemn gravity of feature lies deeply rooted
all the cruelty, deceit and fiendishness of a demon. He is a profli-
gate and is passionately fond of gambling. This latter foible is
gratified in the national sport of cock-fighting and the Spanish
game of monte. However, where facilities offer he is a willing
tyro to the many and varied gambling devices imported in recent
years by the Europeans. He has no sense of appreciation, neither
can he comprehend a spontaneous gift, but rather looks upon any
form of kindness as an expression of fear or weakness. Honor, in
the sense of self respect, dignity, fidelity, virtue or a just discern-
ment of right in strict conformity with duty, is to this most incom-
prehensible being virtually nil. Magnanimity and chivalry are
likewise unknown quantities in the Filipino's composition. He is
quick to borrow but slow to return, superstitious to the utmost
degree, a natural coward, a brute, and if angered does not readily
reveal it in his expression but is most unrelenting, and will await
his opportunity for revenge. Unlike the Japanese or Chinese, he
is a poor imitator and no originator. Few have any regular voca-
tion, and those few who are endowed with a spirit of self-improve-
ment are only to be found in the large cities. These, moreover,
are mostly of the hybrid class, known as msestizos, and their train-
ing is in the arts. The average full-blooded Filipino is well satis-
fied to trust to the morrow and the munificence of a bountiful nature.
He may, out of necessity, cultivate a little patch of rice or sugar-
cane, but his preference is to sit and dream in the shade of the
mango-tree.

Polity and discipline are vague institutions, and Filipino veracity,

22 METZGER— THE FILIPINO. [March ,7,

excepting the Moro, is but a myth. To lie is but the manifestation
of a second nature and to prevaricate with a nicety is an accom-
plishment with him. The native of this class is so contumacious
to all bidding and so averse to social order that, I am inclined to
believe, he understands and appreciates no law except force. Sen-
timent and honor are lost virtues, and there is nothing in which the
average male delights more than to pillage and torture. Intuitive
modesty is as foreign to the average Filipino as it is to the dumb
brutes of the jungle, while the domestic habits of many are very
little above their animal surroundings.

Early in the sixteenth century the marriage custom was estab-
lished among certain tribes through the good offices of the church,
and as a result of which nuptial vows are held very sacred, and the
husband is extremely jealous of his wife after wedlock, notwith-
standing his indifference as to any indiscretion she may have been
guilty of before entering the nuptial state. This, I believe, is but
a selfish vigilance and not a virtuous sense of chastity, for it is the
universal practice with this class of islanders, or at least a large per-
centage of them and more especially the touis, to barter their daugh-
ters. These poor creatures are virtually sold or given in exchange
for a loan to pass their youth as queridas (kept-mistresses). As
this transfer of human chattel is, in many cases, for the payment of
a gambling debt or to secure a loan for some equal moral turpitude,
the poor victim not infrequently becomes the permanent vassal of
the money-monger.

The cheapest thing in the Philippine archipelago is human life
and the dearest object to this oriental's heart is his pet game-cock.
He will risk his life many times over to save this idol of the race,
while he would tranquilly stand by and see his family in peril
rather than expose himself to possible harm in effecting a rescue.

Notwithstanding the Filipino has so many undesirable character-
istics, he is not totally devoid of good qualities. Of these I would
mention his temperance in the use of alcoholics. During my three
years of service on the islands I saw but one native inebriate, yet
these same people have liquors more powerful than the worst of
moonshine whiskey. Then again there is a certain hospitality ex-
isting among themselves which is evinced in the fact that even as
an utter stranger they are always welcomed to such food and
shelter as may be at hand and no remuneration is expected.

iyo5J METZGER— THE FILIPINO. 23

As a people they are musical, although not composers, they are,
however, in this latter respect excellent mimics. This inherent
musical talent is truly most remarkable, for not only will one find
the average native skilled in the playing of one instrument, but it
is not uncommon to see orchestral players exchange instruments
two and three times during an evening and apparently play the
various instruments with equal skill. Go where you will among
this great class of Filipinos and every community worthy the name
of town, you will find a band of musicians varying from half a
dozen to thirty pieces, and even in the isolated mountain districts,,
where conventional instruments are not obtainable, musicians are
to be found playing upon rudely constructed implements made of
bamboo of various lengths and calibre. Unlike most Oriental
music their melody is pleasing to the European ear.

Among the Tagalogs and Visayans there exists a great maestiza
genera, in consequence of which there is manifested a class of dis-
affected, arrogant and indolent people, who through appreciation
of the superiority of the Caucasian (as a race) have assumed many
of his customs, manners and dress, likewise many of his vices but
few of his virtues. This mixture of the blood has instilled an
increase of energy in some, but it has not obliterated any of the
other Malay characteristics in any.

Sunday throughout the archipelago, is the one day of the seven
in which the native throws off his state of lethargy and makes ready
to enjoy himself. True to his faith, he wends his way to the
church at the break of day, this obligation over (for it is more of
an inherent duty and superstitious fear with him than a true sense
of religious reverence), he straightway directs his steps to the
public market place to spend the day in the national sport of cock-
fighting. It is here that one gets an exemplification of a Filipino
characteristic which but goes to prove the incomprehensible anom-
alism of these people. By nature they are apprehensive of honesty,
yet according to the custom of making a stake on the combatants,
the universal practice in vogue, permits any one or any number of
persons, even though they be unknown to the keeper of the pit, to
throw their money into the arena and keep their own council as to
their choice, and should they be successful, can demand their gain
and it will be forthcoming without question. No system seems to
be practiced to prevent knavishness, and if asked as to this apparent

24 METZGER— THE FILIPINO. [March 17

laxity they simply shrug their shoulders. The Filipino fights his
cocks with monte sandwiched between pittings, until mid-day,
when he betakes himself to his house for his siesta, and when the
sun begins to dip well into the western heavens, he agains seeks
the fl/az a de gallos, where he remains reveling in this brutal sport
until the last cock has crowed over its fallen adversary. I am
wont to believe that the cock-pit is the native's club, his school
and not infrequently his only source of revenue.

Probably one of the most uninviting sights in the Colony is the
market of the so-called domesticated natives. From the amount of
filth and the myriads of flies one wonders little at the various
epidemics that so frequently scourge this archipelago. The average
Filipino market, of this class, is a combination of hasty lunch,
general merchandise and reservoir for all the bacteria known to
science. Here doubled up like a jack-knife squats the tribesman
with his wares spread out before him on the ground. The barter,
even in the city of Manila, is more an exchange of one commodity
for another than a purchase through the medium of currency.
Fabrics are exchanged for cocoa-nuts, fish for buyo, eggs for
tobacco and one of those mysterious native dulcies for personal
ornaments. The native is a true Shylock, and it is not uncommon
to see two of these tribesmen spend an hour chaffering over some
article whose value scarcely exceeds five centavos (two and one
half cents).

The buyo and betel-nut are probably the two commodities almost
indispensable to the Filipino of the lower class, as well as to many
of the elite. He can go a goodly time without food if he but has
his buyo. Properly speaking, this is the areca-nut, and which,
when cut into small pieces, dusted with the lime produced from the
oyster shell, and wrapped in the stripped leaf of the betel tree, is
marketed as an individual quid. The buyo is to this Oriental
what tobacco is to the European ; however, it is by far the more
offensive to the aesthetic, in that it stains the teeth and lips a
blood-red, exhibiting a condition most repugnant to the eye. The
effect of this, when the habit is once acquired, is most disastrous,
and in this respect closely allies itself to the results of the use of
opium.

Even though buyo plays such a prominent part in the life of
these people, everyone is a devotee to tobacco, men, women and

i9o5] METZGER— THE FILIPINO. 25

children, the high and the low, the poor and the rich, priest and
layman. The men take to the cigarette, while the women and
children prefer the cigar. It is not an uncommon sight to see a
child of some three or four years whose only adornment is a long
cigar. The cigar is to the Filipino pickaninny apparently what the
bottle is to the American youngster, a pacifier.

One may see in their marriage customs another phase of Filipino
life which characterizes this class of natives. A sort of purgatorial
preliminary exists among these people, in which the vicissitudes of
the average native swain are anything but enviable. If poor, and
this seems to be the universal state, the prospective groom must
serve the girl's parent as a catiped or house servant for a more or
less indefinite period, according to their whim, and it is not infre-
quently the case that after many months, or perchance years, of this
bondage, he is turned out and another suitor installed. Again, the
marriages are arranged by the parents without consulting the wishes
of the child, and quite frequently they are wholly obnoxious to one
or both of the contracting persons, and as a result it is not uncom-
mon for the child to force the hand of dictatorial parents by com-
pelling them to countenance his or her legitimate aspirations.
Before a marriage is consummated, a dowry is made by the girl's
parents in favor of the bride, with the understanding that it is not
transferable to the husband upon the death of the wife, but must
revert to the parents in the event of there being no offspring (which,
however, is rarely the case). In consequence of this it is not un-
common to see the children well provided while the father is a
beggar. The day of the wedding is always fixed by the ever vigi-
lant padre and the fee, which is always exhorbitant, is paid in
advance, either in currency or collateral. The marriage ceremony
of these people is one grand display of barbaric ritualism. Among
the very poor class of these so-called domesticated natives, where
the enormous fees demanded by the church are beyond their means,
the two sexes were accustomed to live together under mutual vows,
but since the American occupation marriages by the ecclesiastics is
not compulsory, and this practice of mutual assent is fast dying out.

Among some of the pagan tribes, especially the Igorrotes, the
marriage ceremony is a sort of a catch if you can affair, in which
the prospective groom is led a chase about the village by the bride-
to-be, and for a time feigns to catch her, finally he secures his prize

26 METZGER— THE FILIPINO. [March 17,

and upon bringing her before her parents, in very much the man-
ner one might lead a reluctant dog at the end of a chain, they bow
down and bring their heads together sniff in the air violently (the
native substitute for osculation) and receive, at the same time, the
parent's sanction which is demonstrated by the pouring of cocoanut
oil over their heads. No feast follows as among the Christian tribes
but the bridal pair flee to the mountain fastnesses, where they re-
main for a fortnight subsisting on wild berries and fruits and then
return to their native village or clan to take up their abode.

A curious custom which prevails among the more ignorant of the
domesticated class of natives, a relic of barbarism, is the practice
of closing all the windows and doors of the house and filling every
available inch of floor space with the presence of neighbors, during
the birth of a child, while the male members of the family thrash
about the room, flourishing large knives or bolos, like so many
mad-men, in their attempt to drive out the evil spirit, as they are
wont to believe influences the destiny of the mother. This bar-
barous practice is carried still farther in some cases by making long
gashes through the skin of the encienle in the hope that the devil
may have an easier exit.

The Filipino funeral is yet another exemplification of the pecu-
liar customs of this type of Oriental. It is a display of fantastic
barbarism and blase sensualism. There is the ghastly bier with its
harsh and crude ornaments of wood and metal, a relic of his un-
tutored and savage ancestry. A native band precedes the funeral
cortege to the grave, playing some wierd uncanny air followed by
a group of professional mourners and the members of the deceased
family, exhibiting no signs of grief or regret but an air of stoical
indifference, in fact it is not uncommon to see them follow along
chatting and smoking as if they were but repairing to some place of
jollification. In the large cities, this burial custom is somewhat
modified by the introduction of a tawdry bier on wheels and drawn
by four or more horses, with footmen and runners dressed in the
garb of the sixteenth century courtier, the whole rendering this
solemn procession ludicrous and insensate. Upon reaching the
burial ground the corpse is stripped and wrapped in a piece of
sdula or matting, and without further ceremony deposited in a vault
or grave until such time as the rental expires, when, if not renewed,
it is resurrected and thrown to bleach under a tropical sun, with

i9°5-] METZGER— THE FILIPINO. 27

hundreds of others, unclaimed and forgotten, upon the so-called
bone-pile. Every necropolis has its bone-pile. It is an institution
of the church and like the potters-field is the final resting place for
many a departed being destitute of friends or kin.

Sunday in the Philippines, as in all Spanish countries, is the
great theater day and all the large towns of the islands have their
various play-houses. The dramatic composition is always in the
native dialect and usually melo- dramatic in character. To the
European the plays are highly ludicrous and extremely tiresome, as
the several parts are not memorized by the actors but are repeated
after a prompter, who is seated in front of the stage and' not infre-
quently in full view of the audience. The plot is always some sup-
posed conflict of times past between the Mohammedans of the south
and the early Christians. There is much palavering with painful
attempts at oratory and brandishing of knives. Then comes the
bloody conflict, the wild beast of the forest puts in its appearance,
the ghost walks and the curtain is finally drawn amidst the loud
applause of the audience. These plays, like those of the Chinese,
not infrequently run for days before the climax is reached and the
plot unraveled.

The matter of bathing practiced by this beople is worthy of men-
tion. Notwithstanding the filth of the average native's house and
the unsanitary surroundings these same people may be seen each
morning bathing in the waters of a nearby stream. If this is not
accessible they will find a pool in which to bathe, even should this
pool be nothing more or less than a composite of all manner of
filth. He must take his daily bath no matter in what or with what
and not infrequently this latter resolves itself into nothing more
than a bowl of water and a gourd for a dipper. His bath like his
siesta is, I believe, more a habit than a sanitary necessity in the
eyes of this people. Men and women bathe together and with little
or no respect for modesty.

The Filipino, as a people, wash their linen as do the East
Indians by beating them with a pamalo upon the rocks. Needless
to say the clothing suffers no little in consequence of this treat-
ment.

Being naturally prone to superstitious beliefs the early native
accepted all the fantastic tales of the early missionaries, and the
modified heathen rites adopted by the Church were received will-

28 METZGER— THE FILIPINO. [March 17,

ingly by them. He loved the pompous ritual, the gaudy and
elaborate robes, the glitter of gold and silver and the images of
saints. All these appealed to his savage nature and his ancient
tribal legends, and this ocular demonstration seems to have im-
pressed him with the sanctity of the system and the infallibility of
its believers. The result is, that to-day, shrines are to be found in
almost every semblance of a town throughout the islands where the
faithful Filipino comes at least once a year to worship.

There seems to be no limit to the number of saints, there is the
patron saint of the islands, Santa Rosario, and the innumerable
local sainte whose images are revered and worshipped for some
wonderful mystic power of healing the sick and halt or some
marvelous act they are supposed to have performed in the dim past.
The victory over the Chinese, during the invasion of Li-ma-hong,
is accredited by the natives to the appearance of Saint Francis on
the walls of the city. The legend of the celestial protector of
Manila is not less interesting. It is related that in Dilo, near
Manila, a wooden image of St. Francis de Assisi was seen to weep
so copiously that many cloths were moistened by its tears, and
again this same image with its hands uplifted and opened during
three hours asked God's blessing on the city of Manila, then
closing its hands it grasped a cross and skull so firmly that these
appeared to be one and the same thing. Vows were made to this
saint, who was then declared protector of the capital. Others of
equal significance might be mentioned but this will suffice to show
the innate mysticism of these people. Many of these images are
most tawdry and elaborately ornamented. I believe the most
elaborate I have seen, outside of the metropolis, was in the town
of Quingua, province of Bulican. The image was that of a man
astride a horse and attired in a gorgeous robe. In his uplifted
hand he held a dangerous looking knife and under his pranc-
ing steed lay the prostrated form of a Mussulman, bleeding and
wounded unto death. This was mounted on an elevated carriage,
and strung about the platform were the heads made of carved wood,
mutilated in appearance, representing the many victims of this
venerated saint. The whole was a barbarous display of cruelty and
superstition.

The roguery of the Filipino is not infrequently manifested
through the agency of these saintly images, and it was only within

9o5.] METZGER— THE FILIPINO. 29

•

the past few years this was brought to the American public's notice
through an unjust attack upon the army in permitting the supposed
looting of one of the churches of the Colony and bringing into the
states one of these sacred images. Upon investigation this " Black
Christ," over which the stir was raised, proved to be a private in-
stitution of some scheming natives working upon the superstition
of their people to extort money for personal gain. The image was
an exquisitely carved piece of wood, waxed and stained to a deep
brown, while the eyes were of glass and framed with eyelids most
human, and the whole enveloped in rich drapery. With the aid of
a ventriloquist (Jose Zaide) the natives were led to believe that
this "Black Christ" was the new Messiah through whom their
sole redemption from the torments of hell could only be obtained
by the making of large donations of money.

Other superstitious beliefs might be mentioned, such as the dia-
bolical influence supposed to be possessed by certain persons which
preserves them from all harm even refractory to the effect of bullets,
called the anting-anting. Then again the belief held by many,
that a crime escapes punishment if committed in Easte week, be-
cause the thief on the cross was pardoned of his sins, and many
more might be enumerated if but time permitted.

Before taking up' the third great classification of these domesti-
cated natives, I wish to make mere mention of the sport of hunt-
ing the wild-buffalo and boar much engaged in by these people
and the bull fights, which until 1885 obtained throughout the
principal cities of this dependency. Likewise a brief description
of this freemasonry that exists on the islands, the so-called Kati-
punan. This is a Tagalog word the meaning of which is league.
The organization was originally perfected with the object of retri-
bution and was the result of the confederation of the various dis-
satisfied islanders under the leadership of one Andres Bonifacio, a
native half caste, who drew up its constitution and devised its
mystic rites, which were of a dread and impressive character,
breathing vengeance upon Spain and more especially the monas-
tics. Since the end of Spanish rule in the archipelago the Kati-
punan has been felt not a little by the American forces operating
in the islands, and it must be admitted that it is a powerful agent
in the political prosperity of this Colony. In 1896 there was
known to have been at least fifty thousand leaguers and by 1900
this number was trebled.

30

METZGER— THE FILIPINO.

[March 17,

As regards this third division of the domesticated Filipino, the
so-called Moros, I can say but little and that principally from the
observation of others, as it has never been my lot to have been
thrown in contact with these people as a collective body. These
people occupy the islands of Mindanao, Palawan and the Sulu

Fig. 5. Mussulman Girl, "so-called Moro."

sultanate. Their early history is vague and dissipated. It is
generally conceded, however, that these people are the descen-
dants of the Mussulman Dyaks of Borneo, their ancestors having
been a great chief and his retinue, who early in the sixteenth
century fled his native land and settled on these islands, bringing
with them the Mohammedan faith. This strange people never
yielded to either Spanish arms or Spanish monastics, but continued

i9°s] METZGER— THE FILIPINO. 31

throughout Spain's regime to rule by tribal custom under the direc-
tion of a Datto or chief and recognizing only the spiritual supremacy
of the Sultan, whose position is hereditary under the Salic law and
who annually makes his trip to Mecca.

The Mussulmans are a valiant and merciless people and for cen-
turies they controlled the high seas in that part of the world, ravish-
ing the coasts in their piratical workings. It was not until the
introduction of steam vessels that Spain was able to cope with these
robbers of the seas. The Moro is very much averse to work, con-'
sequently he is not an agriculturist. His whole ambition in life
seemingly is to strut about in gaudy attire, and encased in a veri-
table arsenal of knives, etc.

Slavery exists in an occult sense among these people. There
are slaves by birth and slaves by conquest, such as insolvent debtors
and prisoners of war. Unlike the other tribes of the islands the
veracity of these people is not to be questioned, for to lie with
them is a heinous crime and deserving of severe punishment, the
penalty of which is usually the severing of the tongue or splitting
the mouth.

Until 1902 these people gave the United States authorities no
trouble but the moment their ancient rights, customs and religion
were supposed to have been interfered with, it was the stirring up
of a hornet's nest.

Many other characteristics might be enumerated if but time per-
mitted, however, this will suffice to show the character of these
people as we see them to-day.

32 MATHEWS— ABORIGINES OF AUSTRALIA. [March i7

SOCIOLOGY OF THE ABORIGINES OF WESTERN
AUSTRALIA.

R. H. MATHEWS, L.S.
( Read March jj, igoj. )

Five years ago I communicated an article to the Society,1 deal-
ing with some of the customs of the natives of Western Australia.
On that occasion I described the organization of a number of tribes
possessing four divisions in their social structure. In the present
paper it is proposed to give a short explanation of a different
organization, found among some tribes occupying the northeastern
corner of Western Australia, comprising the country drained by the
sources of the Ord, Denham, King and other rivers, Stirling Creek,
Sturt Creek, Margaret River and the Upper Fitzroy. Some of the
best known of the aboriginal tribes within the immense geographic
limits mentioned, are the Lunga, Kityu, Charrau and Nining.

All the details given in this article have been gathered by me
through the kind assistance of correspondents who reside in the
Kimberly district of Western Australia, in the region inhabited by
the tribes treated of. I sent them categorical lists of all the points
upon which I wanted information and gave them directions how to
proceed with the investigations. From the reliable character of
my correspondents, and my own general knowledge of the subject,
I feel sure that their work can be depended upon. It is unneces-
sary to add that I am under no obligations to any other authors.

A whole tribe, or it may be a community of several tribes, is
nominally divided into two portions, which may be called phratries,
groups, or any other distinguishing title. Next there is a reparti-
tion of each phratry into four parts, which for purposes of refer-
ence, may be called sections or classes. A name is given to each
section, by means of which the members of the different divisions
are readily distinguished ; and identification is further facilitated
by a masculine and feminine form of every one of the eight
names.

A phratry therefore contains four given sections of men, who

'"Native Tribes of Western Australia," Proc. Amer. PHILOS. Soc, Vol.
XXXIX, pp. 123-125.

*9°5-]

MATHEWS— ABORIGINES OF AUSTRALIA.

33

marry certain four sections of women. In other words, the men
of one phratry marry the daughters of the men of the other, in a
certain fixed rotation. The constitution of the phratries, the
nomenclature of the sections, with the order of intermarriage and
the designation of the children, will be readily understood by an
examination of the following tabular synopsis ;

Table I.

Phratry.

Father.

Mother.

Son.

Daughter.

C Changura

Nungulla

Chabuldyi

Nabicherri

Chanima
j

.Nulima

Chungarin

Nabungarti

1

| Chungulla

Nangilli

Chambin

Nambin

I Chulima

Nabana

Chakara

Nakara

f Chakara

Nabicherri

Chulima

Nulima

I Chambin
j

Nabungarti

Chungulla

Nungulla

| Chungarin

Nambin

Chanima

Nabana

[ Chabuldyi

Nakara

Changura

Nangilli

The above table gives the phratry, father, mother, son and
daughter on the same line across the page. For example, Chan-
gura takes a Nungulla as his wife, which is the ordinary or normal
rule of marriage and may be called " No. i." He could instead
marry a Nulima, which I shall designate as "No. 2." Or he could
mate with a Nabana woman as "No. 3." And lastly, he may
espouse a Nangilli, who can be distinguished as " No. 4." Mar-
riages of the "No. 1 " type, which are those given in the table,
are the most usual ; " No. 2 " is the next most in favor; whilst
"No. 3" and "No. 4" are more or less uncommon, although
quite lawful.

In the tribes we are now discussing the section to which the
children belong, and consequently the phratry also, is invariably
determined through the women. Taking an example from phratry
A in Table I : If Changura wed a Nungulla, as in the table, his
children will be Chabuldyi and Nabicherri ; if he take a Nulima
spouse, they will be Chungarin and Nabungarti ; if he choose a
Nabana, the offspring will be Chakara and Nakara ; and if his wife
be a Nangilli, then his family will be Chambin and Nambin.

We will now show the wives eligible to Chanima, the next name
in Table I. He marries Nulima as his tabular wife or "No. 1 ";
he takes Nungulla as his alternative spouse, or " No. 2 "; he mates
with Nangilli as "No. 3," and he can marry a Nabana woman as

34 MATHEWS— ABORIGINES OF AUSTRALIA. [March i7>

" No. 4." Similarly Chungulla and Chulima can marry either of
the women opposite their names in the table as "No. 1" and
"No. 2" wives; or they can take Nulima or Nungulla as their
" No. 3 " and " No. 4."

It appears, then, that any specific man in Phratry A can marry
any one of the four women opposite to him in the column headed
"mother" in the table. Everything which has been said respect-
ing the marriages in Phratry A applies equally to the marriages of
the men and women in Phratry B.

All the people have totems, consisting of animals, plants, the ele-
ments, and so on, but there is no well established descent of any
given totem from the parents to their offspring. Indeed, there
could not be any regular succession of the totems in a tribe where
the intermarrying laws are as stated in Table I. For example, if
descent were through the males, and Changura's totem were a
bandicoot, it would not only be liable to be disseminated through
the children of any or all the sections in Phratry A, but in the next
generation it would be similarly distributed to the children of all
the men in Phratry B. Hence, in a tribe where the sociology is so
constituted, we discover that in some cases the totems follow the
father, in others the mother, and again in other instances the chil-
dren inherit the totem of neither parent. The totem of the off-
spring is determined by the old men in accordance with customary
laws, which need not now be entered upon.

Space willnot permit of a genealogical tree, but the reader can
easily construct one for himself from the following explanation. A
study of Table I discloses that Chabuldyi, the first name in the
"son" column, has a normal or "tabular" father, Changura.
But he may have what we shall distinguish as an "alternative"
father. Of these "alternative" fathers Chanima is the most
general, whilst Chungulla and Chulima are not so frequent.

Looking at Table I, we see that Changura's father is Chabuldyi,
and the latter' s father is Changura. That is, Changura's paternal
grandfather is Changura, the same as himself. Then Changura
marries his father's " tabular " father's sister's son's daughter Nun-
gulla, as " No. 1" wife already described. Or he marries his father's
" tabular " father's sister's daughter's daughter Nabana as " No. 3."
Again, Changura may espouse his father's "alternative" father's
sister's son's daughter Nulima for "No. 2." Or he can take his
father's "alternative" father's sister's daughter's daughter Nan-
gilli as his " No. 4 " wife.

xgos.] MINUTES. 35

The genealogy of Changura's wives could likewise be traced
through his mother's father. A woman ascertains who are her
potential husbands by going back to her mother's "tabular"
father, or her mothers " alternative " father, as well as her father's
father, from which point the pedigree is the same in principle as
that of last paragraph. It is manifest, therefore, that whichever
one of the four specific women which a man is allowed to take as a
wife, possesses practically the same relationship to him, although
through different channels. The lineage from which a man obtains
his wife is decided by the elders of the tribe.

Parramatta, New South Wales.
February, 1905.

Stated Meeting, April 7, 190J.

President Smith in the Chair.

A letter was read from Mr. James Douglas describing a so-
called shower of toads which he saw in the Sulphur Spring
Valley, Arizona, and confirming the observations of Dr. C. C.
Abbott (see these Proceedings, Vol. XLIII, p. 163).
The decease was announced of the following members :
Henri Louis Frederic de Saussure, at Geneva, Switzerland,

on February 20, 1905, stt. 75.
F. A. Randall, M.D., at Warren, Pa., on January 23, 1905.
Dr. Alexander C. Abbott read a paper on " Epidemic
Cerebro-Spinal Meningitis."

General Meeting, April 12, zj, and 14, ipoj.
April 12,

Afternoon Session.

President Smith in the Chair.

The President opened the meeting with a brief address of
welcome.

An invitation was received from the Naturwissenschaftliche
Verein fur Schleswig-Holstein to be represented at the cele-

36 MINUTES. March 17,

bration of the Fiftieth Anniversary of its foundation to be
held at Kiel on June 17 and 18 next. On motion, the Presi-
dent was authorized to appoint delegates.

The following papers were read :

"The Weal-Relation," by Prof. Lindley M. Keasbey, of
Bryn Mawr, Pa.

"A Plea for Governmental Supervision of Posts Necessitat-
ing Normal Perception of Color," by Dr. Charles A. Oliver,
of Philadelphia.

"The Present Status of the International Catalogue of Sci-
entific Literature," by Dr. Cyrus Adler, of Washington.

" The Composite Character of the Babylonian Creation
Story," by Prof. Morris Jastrow, Jr., of Philadelphia.

"The English Masque," by Prof. Felix E. Schelling, of
Philadelphia.

" The Emancipation of the Waterways," by Prof. Lewis M.
Haupt, of Philadelphia.

"The Beginnings of Lumbering as an Industry in the New
World," by Mr. John E. Hobbs, of North Berwick, Maine.

April 13,

Morning Session.

President Smith in the Chair.

The following papers were read :

"The Structure of the Lignified Cell Wall," by Prof. John
M. Macfarlane, of Lansdowne, Pa.

"New Species of Genus Nepenthes," by Prof. John M.
Macfarlane, of Lansdowne, Pa.

" On Thought Transference Among Animals by Touch
and Scent," by Mr. Alden Sampson, of Haverford.

" Mosaic Development in Ascidian Eggs," by Prof. Edwin
G. Conklin, of Philadelphia.

"The Oligodynamic Action of Copper on Some Intestinal
Organisms," by Prof. Henry Kraemer, of Philadelphia.

lew.] MINUTES. 37

"Observations on Columbium and Tantalum," by Dr.
Edgar F. Smith, of Philadelphia.

"The Effects Upon Metabolism of Preservatives Added to
Foods," by Dr. Harvey W. Wiley, of Washington.

"The Use of the Rotating Anode and Mercury Cathode in
Electro-Analysis," by Lily G. Kollock and Dr. Edgar F.
Smith, of Philadelphia.

Afternoon Session.
Vice-President Scott in the Chair.

The following papers were read :

" The Rounded Sands of Palaeozoic Formations," by Mr.
Gilbert van Ingen, of Princeton.

" A Review of Lacroix's Work on the Montagne Pelee," by
Prof. Angelo Heilprin, of Philadelphia.

" Notes on the Genus Sinopa," by Dr. W. D. Matthew,
of Princeton.

"The Mammalian Fauna of the Fort Union Beds," by Mr.
M. S. Farr, of Princeton.

"The Marsupial Fauna of the Santa Cruz Beds," by Mr.
W. J. Sinclair, of Princeton.

"The Mutual Affinities of the Species of the Genus Cam-
barus," by Dr. A. E. Ortmann, of Pittsburgh.

"The Faunal Relations of the Ryu-kyu (Loo Choo)
Islands," by Dr. Henry A. Pilsbry, of Philadelphia.

Evening Session.
Prof. Russell H. Chittenden, of New Haven, read a paper
on " Reason and Intelligence vs. Custom and Habit in the
Nutrition of the Body."

April 14,
Morning Session.
Vice-President Newcomb in the Chair.
The following papers were read :

"The Secular Perturbations of the Earth," by Mr. Eric
Doolittle, of Upper Darby, Pa.

38 MINUTES. [March i7f

"On the Problem of Four Bodies," by Prof. Edgar Odell
Lovett, of Princeton.

" Radio-Activity in Solar Phenomena," by Prof. Monroe B.
Snyder, of Philadelphia.

" Evidence Relating to Latitude Variations of Short Periods.
From Observations at the Flower Observatory During the Year
1904," by Prof. C. L. Doolittle, of Philadelphia.

" Enquiry Into the Pressure and Rainfall Conditions of the
Trades-Monsoon Area, by W. L. Dallas, of the Meteorological
Office, India.

" On the Construction of Isobaric Charts for Upper Levels
and Their Dynamic Importance in Dynamic Meteorology," by
Dr. J. W. Sandstrom, of Stockholm.

"The Straight Line Concept," by Prof. P. A. Lambert, of
Bethlehem, Pa.

Executive Session.

President Smith in the Chair.

The pending nominations for membership were read and
the Society proceeded to an election.

Afternoon Session.

President Smith in the Chair.

The Tellers of Election reported that the following candi-
dates had been elected to membership :
Residents of the United States:
Joseph S. Ames, Ph.D., Baltimore ;
Thomas Chrowder Chamberlin, Ph.D., LL.D., Chicago ;
William Gilson Farlow, Cambridge ;
Charles H. Frazier, M.D., Philadelphia ;
David Starr Jordan, Stanford University, Cal. ;
George Lyman Kittredge, LL.D., Cambridge;
Robert G. Le Conte, M.D., Philadelphia ;
Eliakim Hastings Moore, Chicago ;
George T. Moore, Ph.D., Washington ;
Richard A. F. Penrose, Jr., Ph.D., Philadelphia ;

MINUTES. 39

Francis P. Venable, Ph.D., LL.D., Chapel Hill, N. C;

J. Edward Whitfield, Philadelphia ;

Bailey Willis, E.M., C.E., Washington.

Foreign Residents :

Yves Delage, Paris ;

Otto Nordenskjold, Stockholm ;

William Matthew Flinders-Petrie, D.C.L., LL.D., F.R.S.,
London ;

Edward Sievers, Leipzig ;

Sir William Turner Thiselton-Dyer, LL.D., Ph.D., F.R.S.,
Kew, England.

The following papers were read :

" On the Theory of the Double Suspension Pendulum," by
Prof. Robert S. Woodward, of New York.

"The Relation between the Economic Depth of a Bridge
Truss and the Depth Which Gives Greatest Stiffness," by Prof.
Mansfield Merriman, of South Bethlehem, Pa.

" On tne Dispersion, Absorption, Fluorescence and Mag-
netic Rotation of Sodium Vapor," by Prof. Robert Williams
Wood, of Baltimore.

" On a Possible Case of Scattering of the Ultra- Violet
Light by Gas Molecules," by Prof. Robert Williams Wood, of
Baltimore.

" The Use of the Falling Plate Oscillograph as a Phase
Meter," Dr. William McClellan, of Philadelphia.

" On the Brains of Scymnus, Mitsukurina and Chlamydo-
selachus, with Remarks Upon Selachian Brains from Stand-
points Morphic, Ontogenic, Taxonomic, Phylogenic and Peda-
gogic," by Prof. Burt G. Wilder, of Ithaca.

40 OLIVER— GOVERNMENT SUPERVISION OF POSTS. [April 12.

A PLEA FOR GOVERNMENTAL SUPERVISION OF
POSTS NECESSITATING NORMAL PER-
CEPTION OF COLOR.

BY CHARLES A. OLIVER, A.M., M.D.
( Read April 12, /goj. )

When it is realized how important becomes normal perception
of color in situations in which accurate color-vision is one of the
main requisites or the sole determining factor for the safety of
lives and the protection of property, it will be at once understood
that definite rules for the obtainance of color-material, the con-
struction of test and governing objects, and the choice of standards
of necessary color-sense, should all be placed under the supervision
of a controlling body from whom all requisite laws shall proceed,
all regulations exercised, and all appeals of enforcement made.

Arbitrary selection of color-material, even though scientifically
and properly obtained primarily ; voluntary employment of neces-
sarily many empirical — and hence ofttimes, imperfect — methods ;
and the existence of multitudinous controlling corporate bodies for
the adjudication of uncertainties, neglect, and intentional wilful
acts — must all exist — as they practically now do — just as long
as no steps are taken to place the entire question under the super-
vision of a national governing board.

Railway service, no matter what the form of motor may be or in
what manner the necessary duties are performed, is mainly governed
during actual work by the proper and ready recognition of color-
signals which are placed sufficiently distant for safety to those for
whom the signalling is intended ; naval and marine transport
throughout the world is mostly accomplished amid its many vicis-
situdes of atmospheric and hydrostatic change, by quick and certain
detections of chosen peculiarities of color situated at safe points of
definite signification ; and army signalling and geodetic survey work
in their every varying degrees of necessity of occasion, are largely
dependent for success upon both aided and unaided color vision.
These conditions granted, it will be at once seen how vast the field
of color employment is, how necessary that proper material shall
be correctly used, and how important it becomes that the perfor-

igos-] OLIVER— GOVERNMENT SUPERVISION OF POSTS. 41

mance of the actual work shall be limited to those who possess
normal color-vision.

The facts set forth in this brief communication once recognized
and systematically applied, thousands of lives must be annually
protected and millions of property yearly saved ; a result for which
this plea — a most urgent one — is offered ; a plea which demands
that in this country — these United States of America — there
shall be established a board of authority composed of those who
are best suited for the establishment and the continued furtherance
of the required work performed, meeting in association with repre-
sentatives from the variously affected departments of state.

42 HAUPT— EMANCIPATION OF THE WATERWAYS. [April

EMANCIPATION OF THE WATERWAYS.

EY LEWIS M. HAUPT.

{Read April 12, iQOj. )

Probably no expenditure made by the Government produces a
larger return than that for the development of our waterways.
They are the lines of least resistance but in a state of nature they are
not always available. Their economic possibilities are inestimable,
when not obstructed by bars or tolls. By the improvement of the
channels connecting the Great Lakes to a depth of 20 feet, not
only has the cost of the transportation been greatly reduced but the
enormous stimulus given to manufacturers has added largely to the
population and wealth of the cities encircling these waters. Thus
the rate on a bushel of wheat from Chicago to New York by the
Lakes and Erie Canal, in 1866, when the Sault Canal depth was
limited to 12 feet, was 29.62 cents, whereas the rate on a 20-foot
draught in 1904, was only 4.71 cents, or only about one sixth,
thus effecting a saving of 84 per cent. The rail charges between
the same terminals were, for the year 1866, 32.79 cents, and in
1904, 1 1. 1 1 cents showing a reduction of about two thirds in the
charges by rail. From these significant figures it appears that
while the charges by rail and water had both been greatly reduced,
in 1866 the water charge was 90 per cent, of that by rail while in
1904 it was only 42 per cent.

But a still more impressive illustration as to the beneficial effects
of this improvement is set forth by the statement made in 1892 by
Senator Wm. P. Frye, in presenting his committee report, wherein
he said that for the year 1890 "The total expenditure for water
improvements of the lakec has amounted to about $30,000,000, or
approximately one fifth of the annual saving effected in transporta-
tion. . . . Our waterways have acted as the most powerful regu-
lators of rates. . . . When it is considered that a diminution of one
mill per ton on the railroads of the country effects a saving of
nearly $100,000,000 to the shippers in transportation, the value
of this restrictive power cannot be overestimated." Had the dis-
tinguished Senator added as a recognized fact that such regulation
by water does not reduce, but greatly increases the revenues of the

»9°5] HAUPT— EMANCIPATION OF THE WATERWAYS. 43

railroads he would but have emphasized the Commercial Paradox,
which comparatively few persons appear to recognize.

In 1890 the unit rate by rail was about 9 mills and by the lakes
alone was 1.2 so that the computed saving on the tonnage moved
by water that year was $147,027,514. Applying the same method
to the tonnage and rates of 1903 it is found that the water rate is
about 6.7 mills less that that by rail while the total ton-mileage of
the lakes is 28,974,660,408 so that the economy effected for the
year 1903 is about $194,139,206. Attention is directed to another
impressive result of the deepening and enlargement of the capacity
of the channels, in the greatly increased size, tonnage and economy
of operation of the vessels engaged in this traffic. Thus from 1855
to 1883, or during the 26 years when the draught was limited to
12 feet the traffic increased from 106,296 to 2,042,259 registered
tons, giving an average increment of 56,918 tons per annum.
From 1883 to 1896, when the Weitzel-Lock was in operation, with
its 16 feet depth, the annual increment was 935,211 tons and after
the opening of the great Poe-Lock in 1896 it immediately expanded
to 2,750,000 tons so that the registered tonnage in 1902 reached
the unprecedented total of 31,955,582, in the seven months of
open navigation. Again the value of land is effected by its earn-
ing capacity as measured by the price of its products on the spot
and this in turn is a function of the cost reaching the ultimate
consumer.

Thus the effects of the cheaper water routes manifests itself most
remarkably, as will be seen by reference to the average values of
the farm products of the several states as furnished by the Depart-
ment of Agriculture. From the statistics covering a decade, it is
found that the lowest average price for wheat is in Nebraska and
that it increases in value as the seaboard is approached. The dif-
ference in price between the 50.9 cents per bushel in Nebraska and
the 7S cents at New York, 1,214 miles distant, is 27.1 cents per
bushel or $8.94 per ton which gives 7.2 mills for the ton-mile rate
which is just the average for the United States, so that the price on
the farm in Nebraska is regulated by that at the port of export, less
the freight charges.

In Missouri where wheat maybe shipped by the Mississippi river
to New Orleans from St. Louis, 1,162 miles for 4.88 cents per
bushel or $1.61 per ton the rate becomes only 1.4 mills by water.

44 HAUPT— EMANCIPATION OF THE WATERWAYS. LApni 12,

If sent to New York by rail, 946 miles, it is 6 mills per ton-mile.
In consequence of this possible competition therefore the average
price paid to the Missouri farmer is ten cents a bushel higher than
that paid in Nebraska and this at 12 bushels to the acre means a
net return of $1.20 per acre on his crop.

Extending this analysis to the cereals of the two adjacent states
of Kansas and Nebraska the former having the advantage of greater
proximity to the waterways, for the year 1901 it was found that the
five cents higher price realized on the wheat crop, gave to Kansas
$4> 953>965 greater revenue than her neighbor, while at nine cents
more per bushel on corn her advantage was $5,535,543 and for
oats at six cents, it was $1,039,944, makinga total of $11,530,000
on these three cereals. In the same manner it is found that if Ne-
braska could have marketed her grain at Kansas prices she would
have received $14,267,000 more, in one year. The total expendi-
tures on the rivers and harbors of the country up to September 19,
1900, is reported to have been $370,411,124.44, 4 per cent, of
which would represent the annual loss to one state due to the
absence of water competition.

THE POLICY OF OTHER COUNTRIES.

It is not surprising therefore that in the sagacious French Re-
public which has expended over $700,000,000 on her internal
waterways, which are free of tolls, her economists believe this
policy to be fully justified by the indirect returns and the thrift and
prosperity of the people incidental thereto.

So too the Dominion of Canada has not hesitated to provide the
munificent sum of $95,316,910.07 for her system of internal water-
ways, which have returned only about one eighth of this sum, yet
the Government recognizes "that waterways and roadways are
essential to the commercial life of the country."

Great Britain has learned from a sad experience that the purchase
of 1,138 miles out of a total of 3,906, by the railroads, up to
1883, has so retarded her trade that she is no longer able to com-
pete successfully with her foreign rivals and Parliament had pro-
hibited the further control of the waterways by hostile interests
and is returning to the policy of rehabilitating them under cor-
porate management. Moreover it is shown that the 2,768 miles
under independent control, in 1898, earned a net profit of $1,080

i9°5] HAUPT— EMANCIPATION OF THE WATERWAYS. 45

per mile while the returns from the 1,138 miles, managed by the
railroads, only averaged $200 per mile. To secure the rights and
privileges of an open port the Manchester district contested for
enabling legislation for five years at a cost of $750,000 against the
allied interests of the railroads and the port of Liverpool but now
that the great work is completed, at a cost of about $75,000,000
the 13,000 vacant dwellings and factories are filled and as many
more have been added to the district, while the interests formerly
opposed, on principle, are all doing a much larger business than
before.

Belgium, but little larger than Vermont, has 1,300 miles of
waterways of which the center is Antwerp. Notwithstanding the
fact that the State owns most of the railways it has encouraged the
construction of the canals so as to render the transportation "as
cheap as possible, that by this means the Belgian manufacturer may
be enabled to compete on most advantageous terms with his foreign
rivals." During the last 25 years about $90,000,000 have been
spent on ports and canals, so that goods can be carried in 300-ton
barges directly from the factory to the ship and by the economies
thus effected the manufacturer can underbid his foreign competitor.

Germany is building an extensive system of canals to connect the
Rhine with the Vistula, passing through her national capital.
Russia is urging a thousand-mile canal to unite the Baltic and Black
seas. France is proposing further extensions to her ample facilities
and intends making a sea-port of Paris. Austria and Italy are also
expending large sums for the benefit of their trade with foreign
countries and yet the astute American who is on the alert for the
best and most economic administration apparently fails to appreciate
the great utilities and possibilities lying almost in a state of nature,
at his very doors.

POLICY OF THE UNITED STATES.

As evidence it is necessary to refer to the condition of the canals
of this country to-day as compared with those of the past century.
Massachusetts claims the honor of building the first canal around
the falls of the Connecticut in 1792-3 and the first railroad at
Quincy in 1827, 34 years later. During this period a large num-
ber of canals were incorporated to connect navigable waters, and
the discovery of hard coal in Eastern Pennsylvania in 1792 also

46 HAUPT— EMANCIPATION OF THE WATERWAYS. [April 12.

stimulated the opening of canal routes to the great cities of the
seaboard and for its transportation to the manufactories. Thus the
Delaware and Hudson, the Morris and Essex, the Schuylkill Navi-
gation, the Chesapeake and Ohio, the Delaware and Raritan, as
well as the James River and Kanawa, the Pennsylvania, the Schuyl-
kill & Susquehanna and the Erie were well under way or completed
prior to the advent of railroads ; but it soon after became apparent
that a railroad constructed by private capital could not conduct a
profitable business as a competitor of a free waterway built and
operated by public funds, so that a war of extermination began
between these interests and it became necessary to purchase or lease
the canals to control their tonnage. Instead of enlarging and
modernizing them for the interest of the lessees and the public
they have in some cases been abandoned and in others only suffi-
cient traffic is carried to maintain the charters. The result of this
policy is well illustrated in the history of the State works of Penn-
sylvania where between 1865 and 1874 some 701 miles of canals,
which had cost over $33,000,000 to build, were abandoned. In
a similar way 656 miles of the Ohio canals were obliterated having
cost nearly $11,000,000. New York has been more fortunate in
having lost only about 269 miles which cost something over $10,-
000,000, but the determined effort now making to prevent the
enlargement of the Erie Canal to even 1 2 feet depth indicates that
the active opponents to our waterways are not yet convinced that
their best interests are conserved by these great arteries of cheap
transportation. The beneficial effects of the cheapest water com-
petion in the country upon railroad interests may be seen along the
Great Lakes where the roads skirting their banks are amongst the
best revenue producers in the United States. If it were possible to
purchase the 90,000 square miles of non-productive water-surface
and convert it into arable land the railroad interests would not per-
mit it to be done as it would exterminate the prosperous cities and
industries which these waters have created, and ruin the tonnage
incidental thereto, yet they persist in obstructing deep water legis-
lation. By the end of 1835 there were about 2,700 miles of
canals open and in use and only about 1,000 miles of railroad ; in
1889 the canal mileage had fallen to 2,305.2 while the railroad
mileage had increased to 157,976 miles and to-day it is not less
than 212,000. Of the canal mileage only 40.6 is under the con-

igos] HAUPT— EMANCIPATION OF THE WATERWAYS. 47

trol of the general Government and 2,264.6 is under State or cor-
porate control. This does not include the 1,078 miles of slack-
water river improvement, making in all only about 3,400 miles for
the entire internal water-borne commerce of the United States.

What this indifference to the earning capacity of canals means
in the cost of wear and tear, for maintenance, may be well exem-
plified by a comparison of the reports of the United Railroads of
New Jersey for the best year of the canal traffic before it was
leased by the railroad, and when its traffic reached nearly 4,000,-
000 tons per annum.

In the reports of the company for the year 1866 it is stated :

The cost of the Camden and Amboy R. R. and its equipment $10,099,000

The cost of the canal and appurtenances 4,381,251

The cost of operating the railroad for the year was 3,801,732

The cost of operating the canal for the same period 360,513

The net revenues from the railroad were 511,162

The net revenues from the canal were 933,642

So that the railroad returned a little more than five per cent, while
the canal earned nearly twenty-three and the operating expenses
were less than one tenth of the former. This financial statement is
independent of the general benefit conferred upon the public at
large by the lower charge for freight carried.

From the above statements as to the great economic advantages
of canals, the neglected condition of our own and the activity shown
in foreign countries which are thoroughly alive to their importance,
it would seem incredible that this government has failed so fre-
quently to act upon or authorize others to engage in most laudable
projects, which call for no appropriations from the general treasury
for construction, and that petitions of influential communities and
large industrial centers are set aside on the score of economy or for
other pretexts so that these most important economies in interstate
traffic are prevented from securing legislation for periods varying
from ten to twenty or more years. Some of the most worthy
projects have been before Congress for nearly a half century and do
not appear to be much nearer fruition than when they were first
proposed.

THE OHIO RIVER.

The largest manufacturing district in the world, that at Pitts-
burgh, has been praying Congress for a charter to construct a ship

48 HAUPT— EMANCIPATION OF THE WATERWAYS. [April 12.

canal to connect the Ohio river at Beaver with Lake Erie at
Ashtabula, so that the congestion of the trade in coal, iron and
steel may be raised and the price of these commodities be reduced,
but in vain. In this district the annual tonnage now exceeds 86,-
000,000 which is greater than that of any port in the world, and
the great rivers leading to the sea are not yet navigable for boats of
even six feet draught. They must wait for floods to float them to
the markets. What this means may be best shown by the experi-
ence of the season of 1895 when the coal which had been accumu-
lating from April 18 until November 28, seven months, and which
amounted to 1,200,000 tons was providentially released by a flood
only in time to prevent it being frozen in and a large part of it
lost. As it was, the cost of keeping the barges afloat amounted to
$2,000 a day. The value of the plant thus tied up was estimated
at $6,500,000.

Although the improvement of this river has been discussed, sur-
veyed and frequently reported upon, the first dam, that at Davis
Island, was not opened until 1885 and since then another, at
Beaver, has been completed, twenty-eight miles below. Four
above and five below Beaver are under contract, but it is estimated
that between Pittsburgh and Cincinnati thirty-seven locks and
dams will be required and fourteen more below Cincinnati ; all for
a six-foot navigation, but already it is found insufficient and nine
feet are now required. At this rate it may well be asked when will
the 1,000 miles be available? This is all down grade and amongst
the cheapest systems in the world — on a six-foot draught the esti-
mated cost is .675 mills per ton-mile and on a nine-foot, .39 mill.
The lowest rail movement is believed to be that across the Lake
Divide, on the Bessemer and L. E. R. R. where the charge was
1.87 mills in 1901, and 2.10 mills in 1904 — or three times the
river rate.

THE COASTWISE CANALS AND PRIVATE ENTERPRISE.

Again for more than twenty years urgent demands have been
made for the creation of harbors of refuge along the New Jersey
coast, where there have been 368 wrecks in ten years, which is
recognized as one of the most dangerous on the great bay between
Cape Cod and Cape Hatteras, but while several estimates have
been submitted for projects costing from three to four millions

ig°5] HAUPT— EMANCIPATION OF THE WATERWAYS. 49

each they have been rejected as unworthy of improvement because
of the absence of sufficient local commerce, caused by the existing
bars which it is desired to remove. The interior coastwise canals
have been recommended for about a century, but as yet only a
few links have been built and those mainly by private and State
aid. Massachusetts, has authorized private companies to open a
canal across Cape Cod ; New Jersey, across its girdle ; Delaware
and Maryland through their peninsula ; Virginia from the Chesa-
peake to Albemarle Sound ; South Carolina from the Santee to
the Cooper rivers twenty-two miles, opened in 1802 ; and many
others. The State of Illinois has authorized the levying of a
special tax which has been expended in cutting the Chicago Drain-
age Canal through the Sag to the Illinois river. Thus past history
and present experience point conclusively to the greater efficiency
of the policy of constructing local works under local legislation and
and supervision rather than to attempt to legislate for the entire
country, by general appropriations made in Congress where so many
other matters of a political nature consume time and prevent action,
or where sectional jealousies have operated to restrain important
measures. Even at this date there are said to be works recom-
mended for approval aggregating nearly $500,000,000, in rivers
and harbors alone, to meet immediate demands, yet it is extremely
difficult to pass a bill for even the most urgent improvements. So
that it has recently been deemed necessary to authorize private
parties, corporations or municipalities to make their own improve-
ments at their own cost subject to the approval of the plans by the
Government, but without authority to charge tolls, or to collect
revenues. As this is not a practical, commercial proposition, it
has been further amended, in the last act, by giving authority in
several special instances to private individuals to open channels and
charge tolls, the Government reserving the right to recover control
after a period of years.

Thus the pressure for commercial channels which it is beyond
the power of the general Government to furnish in a reasonable
time, is leading back to the original policy of local control and de-
velopment of the lines of least resistance for our internal commerce
which has done so much to open up the country prior to the de-
struction of our merchant marine in 1867 when it was the pride of
the nation, and mistress of the seas.

50 HAUPT— EMANCIPATION, OF THE WATERWAYS. [Aprils,

If the Government desires to adhere to the policy of expending
seventy-five per cent, of its revenues for the war, navy and pension
establishments it would seem to be wise to surrender its jurisdiction
over the secondary rivers and harbors, to local or State authorities
that there may be opportunities afforded for the creation of chan-
nels of ample capacity not only for commerce but also for the use
of the military and naval arms of the service in case of necessity
that they may be operated on safe strategic bases between naval
depots. Thus may the waters of the country be emancipated from
the shackles which have so seriously retarded their development.

i9°S-J

ON INTESTINAL ORGANISMS. 51

THE OLIGODYNAMIC ACTION OF COPPER FOIL ON
CERTAIN INTESTINAL ORGANISMS.

BY HENRY KRAEMER.

( Read April /j, 1903. )

Carl von Nageli, probably the greatest botanist of the last cen-
tury, being both a philosopher and a true scientist, passed away on
May 10, 1 89 1. Among his papers was found the manuscript of a
paper entitled " Ueber oligodynamische Erscheinungen in lebenden
Zellen," which, together with an added note by Cramer, was pub-
lished by Schwendener several years after his death in Neue Denk-
schriften der schweizerischen naturforschenden Gesellschaft.1

This really remarkable paper, while it has attracted considerable
attention, does not seem to have been given the credit in some
quarters that its merits deserve. In the light of more recent bio-
logical studies it has proved to be one of the most important papers
that was written by Nageli, and illustrates both the fertility of his
resources and the incisiveness of his genius.

In this paper Nageli showed how exceedingly sensitive certain
living plants are to very minute quantities of various metals. For
forty years he had been studying the algae, but it was not until some
time in the '8o's during an illness that he observed that if alga?
were placed in distilled water they were killed. This he at first
attributed to various causes, but found upon analysis of the water
that it contained traces of copper, and later experiments showed
that the copper, which had been dissolved by the water in its pas-
sage through the copper still, was the toxic agent. He then carried"
on a large number of experiments placing copper coins in distilled
water, and even went so far as to calculate approximately the amount
of copper which was dissolved.

1 For example, in the English translation of Pfeffer' s Physiology of Plants, Vol.
II, page 260, it is stated that copper is poisonous to Spirogyra in the proportion
of one part of copper to 1,000 million parts of water, an observation made by
Nageli in the paper referred to above, and yet no mention of this paper is made in
the citation of literature, which would lead the reader to believe that one of the
other investigators quoted deserved the credit for the discovery.

52 KRAEMER— ACTION OF COPPER FOIL

[April 13,

In these experiments Nageli used 2 -pfennig pieces, consisting of
95 parts of copper, 4 of tin and 1 of zinc. These were cleaned
with sand, and twelve of them were placed in 12 liters of distilled
water and allowed to remain for several days. The solution was
evaporated, the residue dissolved in hydrochloric acid, and the
copper precipitated as sulphide. This precipitate was dissolved in
nitric acid and an excess of ammonia added, producing a blue solu-
tion. On comparing the intensity of color of this solution with
that of other solutions containing known quantities of copper sul-
phate, Nageli estimated that it contained 1.3 parts of copper to 100
million parts of water. He found that this solution was toxic to
various species of Spirogyra, and a further experiment showed that
if the solution were diluted ten times, that is, so that it contained
1.3 part of copper to 1,000 million parts of water, it would still kill
Spirogyra.

Inasmuch as solutions containing such extremely minute quanti-
ties of copper were toxic to Spirogyra, Nageli was inclined to believe
that the toxic action was different from ordinary chemical poison-
ing. This view appeared to him to be strengthened by the fact
that the effects produced in the cells were different from those pro-
duced by ordinary poisons or those resulting from the natural death
of the organism.

It has been supposed by some later investigators1 that Nageli did
not regard the copper as being in a state of solution, yet the
experiments just described clearly show what his information was
on this point, and in another part of his paper he distinctly states
that he so regards it. The marvellous thing to him, as to us, was
that such minute quantities of copper exerted toxic action, and at
first he was inclined to believe that the effect produced was due to
a new force " Isagitat," and in his original manuscript he used the
word " isagische " in describing it. But this term was later replaced

1 On page 23 of his paper Nageli says, " Die oligodynamischen Eigenschaften
des Wassers lassen sich also in alien Fallen auf Stoffe, die ini demselben gelost
sind, zuruckfiihren. Nun weicht aber das durch Metalle oligodynamische
gewordene Wasser in seinem Verhalten wesentlich ab von anderen Losungen.
Eine Salz or Zuckerlosung verliert ihre eigenschaften nicht, wenn unloslicbe
Korper in dieselbe gelegt werden und sie erteilt den Wandungen des Gefasses
nicht die Fahigkeit, reines Wasser wieder salsig or siiss zu machen, wahrende
analoge Erscheinungen bei den Kupferlosungen eintreten."

I9oSj ' ON INTESTINAL ORGANISMS. 53

by that of " oligodynamische " derived apparently from two Greek
words meaning the force within a small quantity of substance.1

There seems to be some confusion among recent writers as to
the condition of the copper produced by placing copper foil in
water, and it is customary to speak of the solution as being a solu-
tion of colloidal copper. While it has been customary since the
classical experiments of Graham to apply the name colloid to those
substances which in solution or suspension will not pass through
animal membranes, still recent researches have shown as pointed
out by Noyes 3 that there are two subclasses of colloidal mixtures, —
the one having the characteristic properties of true solutions, that
is, possessing osmotic pressure, diffusibility and usually a limited
solubility at some temperature ; the other being without these prop-
erties and being in the nature of macroscopic and microscopic sus-
pensions. Considering the origin of the copper in solution it would
properly belong to the class of colloidal suspensions, but it has none
of the properties of this class of substances ; and it differs funda-
mentally from the so-called colloidal solutions not only in origin
but in that it possesses the property of permeating colloids, as the
cell wall and the organized contents of the cell, thereby producing
marked disturbances in the cell and thus resembling the crystalloids.

It is well known as stated by Copeland .and Kahlenberg2 (page
455) that "every metal in contact with water and air is subject to
some change. It reacts with oxygen and carbonic acid dissolved
in water, or with the water itself, to form oxides, hydroxides, car-
bonates, basic carbonates, or acids, which in greater or less degree
pass into solution. When this chemical action is sufficient for the
effect to become visible, the metal is tarnished or corroded ; and
even gold and platinum lose their lustre." Nageli in his paper
(page 24) says that a solution of copper manifesting oligodynamic
properties results only when copper is placed in water containing
oxygen and carbon dioxide ; but so far no one has determined
which compound, or compounds, of copper is formed under these
conditions.

1 As further indicating the meaning that Nageli had in mind we quote from his
paper (page 8) as follows : " Ich will nun, um eine bestimmte und feste Bezeich-
nung zu haben, die specifische Wirkungen des Giftes die chemischen, die jenigen
der noch unbekannten Ursache, in dem ich dem Endresultat vorgreife, die oligo-
dynamischen nennen."

54 KRAEMER— ACTION OF COPPER FOIL [April i3>

While Niigeli's paper was incomplete he nevertheless had carried
on a sufficient number of experiments to show that there is a
marked difference in the sensitiveness of various species of
Spirogyra to the oligodynamic action of copper. He found, for
instance, that Spirogyra orthospera and other Spirogyra s with lense-
shaped nuclei were more resistant than the remaining species.1
He also showed that there was a marked difference in the sensitive-
ness in some of these plants (S. m'tida) at different times of the
day.

While Niigeli confined his attention to studies on Spirogyra
using copper coins2 to produce the oligodynamic effects, other in-
vestigators since his time have carried on experiments with other
organisms both plant and animal and have employed metallic cop-
per and salts of copper as well. One of the most important of
these researches is that by Israel and Klingmann.4 These investi-
gators studied the effects of copper on certain bacteria (as Bacil-
lus typhi, B. coli and Microspira comma'), as well as certain animal
organisms (as Amoeba, Difflugia oblonga, Hematococcus pluvialis,
Paramecium Bursaria, Spirostomum ambiguum, Vorticella micro-
stoma and Sty lony chia mytilus), and also on Spirogvra. They used
copper foil and found that it had a marked toxic effect on all of
the bacteria that they worked with, B. typhi being the most
sensitive. They also found that by placing the solutions of
copper containing the organisms in an incubator at a temperature
of 35°-4o° C. the toxic effects were manifested in i hour, whereas
at the ordinary temperature similar effects were produced in two
hours. In the case of the animal organisms, while the toxic effects
were visible in most instances in but a few minutes, in Vorticella it
required several hours for any toxic effects to be observed and it
was found that Sty lony chia might resist the action for 24 hours.
These authors further found that water in which copper foil had
been placed for 24 hours, could be diluted 100 times and still
manifest oligodynamic effects on Spirogyra. In the latter instance

1 Cramer made a similar observation with S. setiformis (?).

Israel and Klingmann (page 307) found that S. Crassa was killed in 15 min-
utes, S. majuscula in 30 minutes and S. laxa in 75 minutes.

2 During his investigation Nageli also discovered that minute quantities of other
metals, as silver, lead, tin, iron and mercury manifested oligodynamic properties
similar to copper.

i9o5]

ON INTESTINAL ORGANISMS. 55

the time required was 24 hours as against 8 minutes in the first
instance.

While we have seen that solutions containing minute quantities
of copper are exceedingly toxic to certain organisms, other investi-
gators have shown that various plants not only withstand the influ-
ence of relatively large quantities of copper sulphate, but under
certain conditions even appear to be benefited by its presence.
With these various data before us we may say that while copper
has a specific toxic action even in very minute quantities on certain
organisms, it should be borne in mind that these same organisms
manifest a specific sensitiveness towards copper and various other
metals.1

These data are not only of great interest from a scientific point
of view but in their practical application are of very great impor-
tance, and it was to be expected that pharmacologists would appre-
ciate the important bearing of this line of investigation on their
work. Cushny 5 among pharmacologists early recognized the value
of these researches and the possibilities in their application in the
prevention and treatment of disease. He states that while copper
is comparatively harmless to man, yet it is exceedingly toxic to
certain microorganisms and intestinal parasites. He says :

" Small quantities of copper may be taken for indefinite periods
without any symptoms being induced, so that so far as man is con-
cerned the general action of copper is unknown. ... On the
other hand, copper is a deadly poison to several of the lower plants.
Thus, traces of copper added to the water in which they live,
destroy some of the simpler algae, and Nageli asserts that 1 part

1 While various explanations might be offered to show why such extremely
minute quantities of copper in solution are sufficient to kill unicellular and fila-
mentous algre, bacteria, and unicellular animal organisms, and yet not affect
multicellular plants and animals, whose cells are as delicate in structure as those
of the unicellular organisms, it seems that this is in a measure due to the fact that
in the latter the entire individual is comprised in a single cell, which performs all
the vegetative as well as reproductive functions, and being entirely surrounded by
the copper solution, all the life process are affected, there being no way for the
organism to distribute the solution to other cells, and thus by a dilution minimize
the toxic action of the copper. Or if some of the cells in the multicellular organism
are destroyed or injured by exposure to the solution, others are formed to take their
place from the more or less deep-seated meristematic cells. It is true that the
idiosyncrasies in these organisms should also be borne in mind, some of them
being more resistant than others.

56 KRAEMER— ACTION OF COPPER FOIL [April 13,

of copper in 1,000,000,000 parts of water is sufficient to kill these
plants. . . . Locke found that the traces of copper contained in
water distilled in copper vessels were sufficient to destroy tubifex
(one of the annelid worms) and tadpoles, while Bucholtz states
that the development of bacteria is stopped by a solution of copper
sulphate under 1 per cent, in strength. Copper thus seems to have
a very powerful poisonous action on certain living forms and to be
harmless to others, and the subject deserves further investigation.
It is possible that it may prove to act prejudicially to some human
parasites, and it is certainly less dangerous to man than many other
remedies used as parasiticides and disinfectants."

It was not, however, until the publication of the bulletin on "A
Method of Destroying or Preventing the Growth of Algae and Cer-
tain Pathogenic Bacteria in Water Supplies" by Moore and Keller-
man,6 nearly a year ago, that the very great practical significance
of work along these lines became apparent and general interest was
aroused in the subject.

Since last fall we have been carrying on a series of experiments
in the Microscopical Laboratory of the Philadelphia College of
Pharmacy ' with the view of testing the efficiency of metallic copper
for destroying typhoid and colon bacilli in water. Some of the
results obtained have already been published.7

In presenting the results of our experiments sufficient of the
details will be given to show the manner in which the work was
conducted.

In the first series of experiments here recorded water under three
different conditions was employed : ' («) Distilled water which was
prepared from tap water by first treating it with potassium perman-
ganate and then distilling it two or three times by means of
apparatus constructed entirely of glass ; (b) filtered tap water,
prepared by means of a Berkefeld filter attached to a copper spigot ;
(V) tap water, collected after being allowed to run for five minutes,
the spigot being the usual copper one. All of these were sterilized
in an autoclave at 1 io° for 30 minutes.2

The cultures of typhoid and colon which were used were pure
cultures developed in bouillon for 18 to 24 hours.

1 1 acknowledge my indebtedness to Mr. John R. Rippetoe for valuable assis-
tance in carrying on the experiments recorded in this paper.

2 In all of our work we found in the blank experiments that water which had
been sterilized in an autoclave remained sterile.

ON INTESTINAL ORGANISMS.

57

To 200 c.c. of samples of water prepared as stated, and contained
in sterile Erlenmeyer flasks, were added two 3-mm. loops of the
fresh bouillon cultures of typhoid and colon bacilli respectively.
Counting the duplicate experiments provided for, we thus had a
series of 1 2 flasks, 6 of them containing typhoid bacilli, and 6 colon
bacilli.

For determining the number of organisms, 1 c.c. of the respective
solutions was transferred directly to a Petri dish by means of a
sterile i-c.c. pipet, and to this was added 10 c.c. of Heyden's
nutrient agar, which had been kept at a temperature of 40 ° C. for
some time. Three separate plates of the water in each of the 12
flasks was made immediately upon the addition of the cultures, and
both the plates and the flasks were kept at a temperature of 35 °

Table I. — Experiments with Bacillus coll.

Water Without Copper Foil.

Triple
Distilled
Water

At time of add-
ing culture.
At end of 4

hours.
At end of 8

hours.
At end of 24

hours.
At end of 48

hours.
At end of 6

days.
At end of 14

days.
At end of 21

days.
At end of 28

days.
At end of 53

days.
At end of 60

days.
At end of 83

days.
At end of 90

days.
At end of 1 30'

days.

Filtered

Tap
Water.

7,746 11,246

7,655 '■ 5,075

7,735 ! 3,"5
1,000,000 1,000,000
1,200,000 1,600,000
1,200,000 1,000,000

Tap
Water.

8,283

7,665

7,000

1,500,000

2,000,000

1,200,000

Water With Copper Foil.

Triple
Distilled
Water.

,866

No or-
ganisms.

1,060,000 910,000 2,245,000

700,000 462,000 650,000

700,600 462,446

602,000 456,000

! 583,200 421,000

215,600 128,766

208,133
289,333

48,433
146,543

649,666
693,000

687,333
206,950
147,000
225,400

4,410

No or-
ganisms.

Tap

6,790

No or-
ganisms.

1 The nutrient medium used in the plates made at the end of 130 days was agar
having an acidity of 0.5 percent.

58

KRAEMER— ACTION OF COPPER FOIL

[April

C. to 37° C. To six of the flasks were then added strips of copper
foil about 15 mm. wide and 18 cm. long, these being corrugated
in such a manner that the entire surface was exposed to the water.
Plates were made from all the 12 flasks at the end of 4 hours and
8 hours, and 1 day, 2 days, and 6 days, even in the cases where no
organisms remained, and in the cases in which they continued to
develop, also at the end of 14, 21, 28, 53, 60, 83, 90, 120, 130
and 134 days. The results are given in the accompanying tables.

Table II. — Experiments with Bacillus typhosus.1

Water Without Copper Foil.

Triple Filtered
Distilled Tap
Water. Water.

Tap
Water.

Water With Copper Coil.

Dis'tilled Filtered Tap

Water. Water. Water.

At time
At end

of adding culture,
of 4 hours.

24 "
4S «
6 days
14 "
21 "
28 "
602
90

days

134

3.740
2,835

3,85o

3,75o

3,8i5

1,850

16,380

39,690

153,600

295,866

239,400

78,75o

34,440

4,75o
No or-
ganisms.

3,675
3,8i5

1,995
i,435

i,540

3,986 127

No or- No or
ganisms. ganisms. ganism

1,400
No or-

3,920

65,500

221,867

961,800

346,500

9,156

7,875

1 Bouillon cultures of the different samples of water, at the end of 60 days,
gave with Widal's test the characteristic behavior of typhoid organisms. After
60 days the organisms were found to be very long and more or less filamentous
and did not respond to Widal's test. I am indebted to Dr. Herman B. Allvn,
Philadelphia, for specimens of typhoid blood.

It is seen in the foregoing tables that in all the flasks to which
copper foil had been added all of the organisms were destroyed in
less than four hours, and furthermore the solutions remained
sterile as shown by plates made for a number days thereafter.

I may say that every single experiment which we have conducted,
not only those given in the foregoing tables, but all others, shows
that copper foil is exceedingly toxic to colon and typhoid bacilli,
particularly the latter.

It will be seen further that in the filtered water, to which no
copper foil had been added, the typhoid organisms did not develop
as was the case with the tap water and distilled water, although

ON INTESTINAL ORGANISMS.

59

there was a larger number of organisms to begin with ; while the
colon bacilli multiplied considerably in the filtered water still there
was a very marked inhibiting action. At first I was inclined to
attribute this diminution in the number of the organisms to minute
traces of copper in the flasks, but subsequent experiments showed
that this was not the case. I was, then, inclined to attribute
these rather anomalous results to the presence of extremely small
quantities of copper dissolved by the water in its necessarily slow
passage through the copper spigot to which the filter was attached.
In order to test further the validity of this assumption another
series of experiments was conducted using (a) tap water, (/;) water
filtered through a stone filter,1 and (V) water filtered through a
Berkefeld filter. The water in each case was sterilized in an auto-
clave at a temperature of no° C. for 30 minutes, and iS- to 24-
hour cultures of typhoid and colon bacilli were respectively added
to the samples of water at the ordinary temperature. The results
are summarized as follows :

Table III. — Experiments with Bacillus
Water.

coli and B. typhi in Filtered

Stone
Filtered
Water.

Berkefeld
Filtered
Water.

Bacillus typhi.

^ Stone ! Berkefeld

Tap Filtered Filtered

Water. Water. Water.

At time of adding

culture.
At end of 2 hours.

" " 4 «

" " 6 "

" " 8 "

« "24 "

« "48 "

" "7 days.

" " 14 «

" " 21 "

" "3° "
« "60 "

5,040

6,426
8,505

6,930

16,065

315,000

630,000

10,611

18,270

24,570

28,350

77,175

630,000

1,000,000

7,875

10,269

6,600

5,500 j 2,714

5,040

2,646

3,654

150,000

200,000

,512 1,764

2,520 No or-
ganisms.

250
150
38
39

1,289,333

1,505,700

599,333 , 80,770

900,000
730,800

1,260,000
945,000

94,500 j No or-
ganisms.
t49,33l I "

2,930
3,829
1,820
9,000

43
No or-
ganisms.

1 In the preliminary experiments with samples of water that had been filtered
through a stone filter or a Jewett filter, it was found that there was a similar in-
hibiting action on the organisms to that of water from the Berkefeld filter.
This action was supposed to be due to the influence of the copper in the spigot
attached to the receiver of the filter, and was overcome by removing the spigot
and using a rubber stopper fitted with a glass tube.

60 KRAEMER— ACTION OF COPPER FOIL [April 13,

It is seen from the foregoing table that while we began with
approximately 5,000 organisms of colon bacilli to the cubic centi-
meter in the case of the tap water, there were over 700,000 at the
end of sixty days ; and that in the case of the stone filtered water
where the initial number of organisms was about 10,000 they in-
creased on an average similar to those in the tap water. In the
case of the water from the Berkefeld filter, however, beginning with
10,000 organisms to the cubic centimeter, there was a rapid
diminution of the organisms, so that but about 2,500, or about 25
per cent, of the organisms persisted at the end of six hours, and
while they continued to multiply after this still the number was
considerably less than in either the tap water or stone filtered water,
showing that with Berkefeld filtered water there is some agency
which inhibits the growth of the colon bacilli. This we concluded
to be due to the copper dissolved from the spigot to which the
filter was attached, as already suggested.

In the experiments with the typhoid organisms it was found that
they multiplied in number in both the tap water and stone filtered
water persisting for fourteen days, after which they disappeared, as
was also the case in some other experiments ; but in the case of
Berkefeld filtered water they entirely disappeared within four hours,
which was also the case in three other experiments not here re-
corded. It may also be stated that it was not unusual to observe in
the case of both tap and stone-filtered water, where cultures of the
typhoid bacillus were used, that if the organisms persisted until the
fourteenth day, they would multiply enormously after that as shown
for tap water and distilled water in Table II.

In an investigation of this kind many lines of experiment are
suggested, and it was thought desirable to carry on another series
of experiments with a view of testing the toxicity of solutions in
which metallic copper had been allowed to remain for varying
lengths of time. In these experiments sterilized distilled water
and stone filtered water were used. To 600 cc. of water in a
graduate 8 strips of copper foil 15 x 130 mm. were added. The
graduate was agitated continuously and 100 cc. of the solution
were removed at the end of 1, 5, 10, 20 and 30 minutes. The
respective solutions were placed in Erlenmeyer flasks and sterilized
in an autoclave at no°C. for 30 minutes. To these were added
18- to 24-hour cultures of typhoid bacilli, and plates made with
results as indicated in the two following tables :

^OS-]

ON INTESTINAL ORGANISMS.

61

Table IV. — Experiments with Bacillus typhi in Distilled Water in
Contact with Copper Foil for Varying Lengths of Time.

Water

Water in Contact with Copper for

without

1.5

. IO

20

3°

minute. minutes.

minutes.

minutes

minutes.

At time of adding culture

3-45»

7,119 7,420

6,791

8,631

I2J26

At end of 2 hours.

5.292

No or- No or-
ganisms, ganisms.

2,139

3,150

6,l88

4 "

6,489

No or-
ganisms.

25

420

6 "

5,95°

"

"

" "

3

35

8 •'

4,410

No or-
ganisms.

12

" 24 "

6,4S9

"

"

' '

"

No or-
ganisms.

« 48 "

8,410

"

"

"

"

"

3 days.

11,466

"

"

"

"

"

4 "

7,560

"

"

"

"

"

7 "

2,SqS

"

"

"

"

Table V. — Experiments with Bacillus typhi in Stone Filtered Water
in Contact with Copper Foil for Varying Lengths of Time.

Water

Water in Contact with

Copper fo

without

1

5

10

20

3°

minute.

minutes.

minutes.

minutes.

minutes.

At time of adding culture.

5,050

4,725

7,221

6,111

13,482

",403

At end of 2 hours.

4,599

4,977

6,615

7,056

16,000

II,o88

4 "

6,300

5,859

3,906

6,339

14,000

15,482

6 "

7,"9

6,300

4,250

5,4i8

8,946

5,574

8 "

4,914

8,064

5-48I

5,645

7,951

5,624

10,710
10,700

3,213
i,i55

2,205
142

1,925
104

4,4IO
790

" 48 »

3

" 3 days.

11,277

152

No or

ganisms.

No or-
ganisms.

123

No or-
ganisms.

« 4 «

io,395

No or-
ganisms.

No or-
ganisms.

" 7 "

7,899

"

"

"

Table IV shows that in the experiments made with distilled water,
the mere contact of the copper foil with the water for from 1 to 5
minutes imparted to it sufficient toxicity, or oligodynamic property,
to kill the typhoid organisms placed in the solution within two
hours, when the organisms did not exceed approximately 7,000 to
the cubic centimeter, or 700,000 to the entire solution. Where
the number of organisms in the solution exceeded this number ap-
proximately three-tenths of 1 per cent, persisted four to eight
hours longer.

62

KRAE.MER— ACTION OF COPPER FOIL

[April 13,

In the case of stone filtered water (Table V) a longer time was
required to affect the organisms. This is probably accounted for
by the fact that the water contained other substances which modi-
fied the action of the copper either precipitating it, absorbing it,
or even adsorbing it, and thus weakening the solution.1

As showing the influence of a material which would be in the
nature of a food to the organisms and which at the same time
would have a tendency to inhibit the oligodynamic action of the
copper solution, the following experiments were conducted using
filtered water : (a) Berkefeld filtered water ; (&) stone filtered water.
In both series of experiments 1 cc. of nutrient bouillon was added
to 200 cc. of water, which was then sterilized in the autoclave,
and the typhoid organisms added after cooling.

Table VI. — Experiments with Bacillus typhi in Filtered Water Con-
taining Bouillon.

Berkefeld Filtered Water.

Stone Filtered Water.

Without
Bouillon.

With
Bouillon.

Without

Bouillon.

With
Bouillon.

At time of adding culture.
At end of 4 hours.
8 "

" 24 "

« 48 "

" 7 days.

" 14 «

" 30 »

" 60 «

7-245
550
No organisms.

1,296

9
5
17

2,500,000
7,000,000

14,044

11,907

7,560

No organisms.

2,151
1.323

4,820
3,000,000
4,500,000
2,255,000

11,109
6,466

In the case of the Berkefeld-filtered water it is seen that there
was no growth in the flasks to which bouillon had not been added,
after four hours ; and while there was a diminution of the number

'Nageli found (p. 13 of his paper) that the oligodynamic action of a copper so-
lution could be lessened by the introduction of the following substances : Sulphur
(either roll or flowers), carbon (either graphite or soot), coke, coal, peat, black
oxide of manganese, starch, cellulose (either as Swedish filter paper, or cotton,
linen or wood fiber), silk, wool, stearic acid, paraffin, gum, dextrin, egg albumin
and glue.

True and Oglevee s have studied the influence of insoluble substances on the
toxic action of poisons and have confirmed several of Nageli's observations.

Moore and Kellerman have shown in their recent bulletin the relative decrease
of toxicity of copper sulphate solutions depending on the amount of organic
matter present, the amount of carbon dioxide in solution or the temporary hard-
ness of the water.

IO0S.]

ON INTESTINAL ORGANISMS.

63

of organisms in those solutions containing bouillon between the
first 4 and 24 hours, there was after this a marked increase in
growth. This increase in development would appear to begin after
the last inhibiting traces of copper are removed, either by precipi-
tation in the organisms or by the bouillon.

Other experiments which we conducted showed that there was a
difference in the persistence of the typhoid organisms depending
upon whether the cultures added to the water were 2 4 -hour or 14-
day bouillon cultures, as seen in the following table.

Table VII. — Experiments with Cultures of Bacillus typhi of
Different Ages.

TapW

ater.

Berkefeld Filtered Water.

24-Hour

14-Day

24-Hour 14-Day

Cultures.

Cultures.

Cultures. Cultures.

At time of adding cultures.

3,058

1,050

1,983 952

At end of 4 hours.

682

1,105

40 574

8 "

440

604

No organisms. 215

" 24 "

137

217

" 106

" 48 "

63

179

150

7 days.

No organisms.

49

35

The figures in Table VII, show that the older cultures of the
typhoid organisms were most resistant in the tap water, and that
they survive over 7 days in Berkefeld-filtered water, the 24-hour
cultures usually being destroyed in about 4 hours.

Table VIII.-

-experiments on tap water with copper foil and copper
Sulphate.

Tap Water without

Copper Foil or

Copper Sulphate.

Berkefeld
Filtered
Water.

Tap
Water

with

Copper

Foil.

Tap Water with
Copper Sulphate.

1 Part to 1 Part
100,000 1,000,000

At time of adding copper
foil or copper sulphate.

On drawing tap water or
before filtering.

After filtering.

At end of 2 hours.

days.

39,000

32,666
21,300
40,900
41,000
68,933

39,ooo

8,233

46,800
666

27,133

I 35,666

■ 29,266

20,516

! 9,866

61,466 609,900

87,100(500,200

n,ooo 395,300 33,600
'343,700

9,500
7,766
10,200

13,333

102,200

1 ?3j

3°o

66

200

300

3,633
185,000
211,760

8,233

1,833
1,300
2,233
1,166
112,300

97,l5o
134,000

64 KRAKMER— ACTION OF COPPER FOIL [April 13,

At the beginning of our investigation a number of experiments
were made with a view of testing the comparative efficiency of both
copper foil and copper sulphate in destroying the organisms in
tap water, and it is thought that the results obtained are of suffi-
cient interest to present at this time.

It is observed that in the case of the Berkefeld filtered water, 99
per cent, of the original number of organisms were removed by
the process of filtration. When copper foil was introduced into
the water about 75 per cent, of the organisms were destroyed in
8 hours, although in other experiments where larger quantities of
water were used from 85 to 97 per cent, of the organisms were
destroyed. When copper sulphate was added to the tap water, so
that there was 1 part to 100,000 of water, 97 per cent, of the
organisms was destroyed in 8 hours. When the strength was reduced
so that there was 1 part of copper sulphate to 1,000,000 parts of
water, there was a reduction of 86 per cent.

Owing to the sensitiveness of typhoid and colon bacilli to the
influence of copper, as previously shown, it may be inferred that
they would have been included in the 75 to 97 per cent, of the
organisms destroyed.

CONCLUSIONS.

From the experiments thus far conducted as well as the results
obtained by other writers, the following conclusions may be drawn :

1. Certain intestinal bacteria like colon and typhoid are com-
pletely destroyed by placing clean copper foil in water containing
them, or by adding the organisms to water previously in contact
with copper foil.

2. The toxicity of water in which either copper coins or copper
foil has been added is probably due to a solution of some salt of
copper, as first suggested by Nageli.

3. The copper is probably in the form of a crystalloid rather
than that of a colloid, as it has the property of permeating the cell
walls and organized cell contents of both animals and plants,
thereby producing the toxic effects.

4. While the effects produced by the oligodynamic action of
copper are apparently different from those of true chemical poisons,
the difference is probably in degree only and not in kind.

5. Certain lower organisms including both plants and animals
possess a specific sensitiveness to minute quantities of copper, and

I90S.] ON INTESTINAL ORGANISMS. 65

it has been shown that they are not restored on transferring them
to water free from oligodynamic properties.

6. Oligodynamic solutions of copper are obtained by adding
either copper coins, copper foil or salts of copper to water ; when
copper foil is used, sufficient copper is dissolved by the distilled
water in i to 5 minutes to kill the typhoid organisms within two
hours.

7. A solution of copper may lose its toxicity by the precipita-
tion of the copper as an insoluble salt or compound ; by its ab-
sorption by organic substances ; or by adsorption by insoluble sub-
stances.

8. The oligodynamic action of the copper is dependent upon
temperature as first pointed out by Israel and Klingmann.

9. The effects of oligodynamic copper in the purification of
drinking water are in a quantitative sense much like those of filtra-
tion, only the organisms removed, like B. typhi and B. coli are
completely destroyed.

BIBLIOGRAPHY.

1. Nageli.

Ueber oligodynamische Erscheinungen in lebenden Zellen. Neue Denk-
schriften der schweizerischen naturforschenden G*esellschaft (33-34),
1S93-1S95, pp. 1-5 1.

2. Copeland and Kahlenberg.

Trans. Wisconsin Academy of Sciences, Arts and Letters, 1898 and 1899,
pp. 454-474-

3. Noyes.

The Preparation and Properties of Colloidal Mixtures. Jour. Avier.
Chem. Soc, Vol. 27, 1905, pp. 85-104.

4. Israel and Klingmann.

Oligodynamische Erscheinungen (v. Nageli) an pflanzlichen und thier-
ischen Zellen. Virchoto 's Arc/iiv, 147, 1S97, pp. 293-340.

5. Cushny.

Pharmacology and Therapeutics, 1 899, p. 159.

6. Moore and Kellerman.

U. S. Department of Agriculture, Bureau of Plant Industry, Bulletins 64,
1904 ; and 76, 1905.

7. Kraemer.

Amer. Jour. Phariii., 76, I904, pp. 574-581 ; Amer. Medicine, 9, 1905,
pp. 275-277.

8. True and Oglevee.

The Effect of the Presence of Insoluble Substances on the Toxic Action of
Poisons. Bot. Gaz., 39, 1905, pp. 1-21.

6U WILEY— EFFECT OF PRESERVATIVES. [April 13,

THE EFFECT OF PRESERVATIVES ON METABOLISM.

BY H. W. WILEY, M.D.

( Read April /j, /goj. )

The question of the use of preservatives in food products has of
late assumed an importance greater even than in previous years.
A tendency to legislation of a prohibitory character has developed
in all civilized countries. Many preservatives are now forbidden
by law in Germany, France, Italy, Spain, Austria, and many of
the States of the United States. It seems that it is scarcely just to
legislate against preservatives individually rather than as a class.
Universally excepted from prohibitory or restricted legislation are
the preservatives in common use, namely, sugar, salt, vinegar and
wood smoke. The basis of all prohibitory legislation, at least, the
alleged basis lies in the fact that the preservatives restricted or for-
bidden are injurious to health. If literature on the subject is con-
sulted some conflicting statements are found emanating from scien-
tific sources apparently of equal reliability. It is evident, therefore,
that there is a very widespread difference of opinion among
physiological chemists and hygienists respecting the effect of pre-
servatives added to foods upon the public health. The data of
research are very extensive in experiments in vitro, with the lower
animals and with man. It cannot be denied that there are many
grounds for the prohibitive and restrictive legislation referred to.
There are other questions which must be considered in connection
with this, namely, the dangers which attend the use of nonpreserved
foods and the effects which the prohibition of preservatives might
have upon the price of foods. The latter is a purely economic
subject and does not enter into the present discussion. It is evi-
dent that if a preservative is injurious to health it will in some way
affect the metabolic process. It will either derange digestion or
interfere with assimilation and excretion.

There are many apparently almost insurmountable difficulties in
the experimental determination of this problem with man himself.
A merely negative result is not sufficient to secure a verdict of
acquittal. The reason of this is apparent, namely, the fact that in-

igo5] WILEY— EFFECT OF PRESERVATIVES. 67

dividuals present such marked differences in their powers of resist-
ance. One person may be affected with great facility while an-
other person, subjected to the same treatment shows no sign of
injury. The object of restrictive laws is, of course, the protection
of the weakest and not of the strongest. Hence, I think it may be
laid down as a direct principle of legislation that the addition of
any substances to foods whatever not necessary in their preparation
which affect the health of the most susceptible should be prohibited
or so regulated that danger of injury even of the weakest may be
eliminated.

I have now to briefly record the results of experimental work on
strong and healthy young men. I can do no more than merely
state the principal points which were noticed. First, the action of
borax and boric acid on nitrogen metabolism was extremely slight.
There was, however, a very slight tendency manifested in the ex-
periments which extended over a period of nearly eight months to
inhibit the excretion of nitrogen. The general effect, however, on
nitrogen metabolism is not of sufficient magnitude to warrant the
drawing of any definite conclusions. The effect of the borax and
boric acid upon the metabolism of phosphoric acid is very marked.
A very much larger quantity of phosphoric acid is excreted under
the influence of these preservatives than without them. Borax and
boric acid appear to increase the digestibility of the fats in food.
In other words there is slightly less fat in the feces during the ad-
ministration of these bodies than without them. These preserva-
tives have a slight tendency to diminish the utilization of calories
of foods. In other words, there is a great number of non-metabo-
lized calories in the feces during the administration of the preserva-
tives than without them. Both boric acid and borax have a slight
tendency to increase the traces of free albumen in the urine. Boric
acid has a decided tendency to increase the acidity of the urine.
Borax has a decided tendency to diminish the acidity of the urine,
establishing often the amphoteric reaction and occasionally an
alkaline reaction. Both these bodies when exhibited over long
periods in small quantities tend to disturb the digestion by dimin-
ishing the appetite and inducing a feeling of heaviness in the head
or often headache of a persistent character. The results of these
influences are seen in a gradual diminution of weight.

In large doses, from one to three grams per day, both boric acid

68 WILEY— EFFECT OF PRESERVATIVES. [APrili3,

and borax, when their use is continued for a short time, tend to
produce a feeling of distress and even nausea. There is, however,
no tendency to produce diarrhea. The general effect produced by
borax and boric acid upon health and digestion is decidedly un-
favorable, whether by large doses over a short period of time or in
the case of small doses, namely seven and one half grains or half a
gram per day, over a period of fifty days.
Bureau of Chemistry, U. S. Department of
Agriculture, Washington, D. C.

r9°5-J

MATTHEW— THE OSTEOLOGY OF SINOPA. 69

NOTES ON THE OSTEOLOGY OF SINOPA, A PRIMI-
TIVE MEMBER OF THE HY.LNODONTIDtE.

BY W. D. MATTHEW.

( Read April /j, igoj. )

The following observations are based upon a nearly complete
skeleton of a Middle Eocene creodont discovered by Mr. Walter
Granger near Fort Bridger, Wyo., in 1902. The specimen is the
property of the National Museum and the full description will be
published under the auspices of that institution. I am indebted
to the Secretary of the Smithsonian Institution for permission to
publish this abstract in advance.

The skeleton is unusually well preserved, and practically complete
except for the feet. Most of one fore and one hind foot are pre-
served, the others are missing. It is believed to be one of the most
perfect skeletons ever found in this formation and is of interest as a
typical generalized Creodont. The points of especial interest in its
study were : (1) the relations of the Creodonta to marsupials and
Insectivora, and (2) the relations of Sinopa to Hycenodon and to the
Oxysenidae.

Sinopa was the first fossil carnivore described from the Eocene
of North America and is a characteristic genus of the Lower and
Middle Eocene found in Europe as well as in this country. The
dentition of this or allied genera has been well known from the
descriptions of Cope and Scott, and Wortman in 1902 described a
skull and some parts of the skeleton which he referred to Sinopa.
The complete knowledge of the skeleton enables us to determine
its relationships with certainty, and for the most part confirms the
views hitherto generally accepted.

The animal was a little smaller than a coyote, but in proportions
much more like the Tasmanian wolf, the lower limbs and feet being
much shorter and less compact than in any of the Canidse, and the
tail long and heavy. The skull is long both in cranial and facial
regions, the long basicranial region being very characteristic of
carnivora, while in marsupials and insectivores the basicranial region
is short. The mastoid has a small exposure on the side of the

70 MATTHEW— THE OSTEOLOGY OF SIXOPA.

[April 13.

skull, as in carnivora, while in marsupials and insectivores it has a
large exposure on the back of the skull. The brain is very small
and of inferior type, as in marsupials and all primitive mammals.
The occipital and sagittal crests are high, as in the carnivorous
marsupials. The tympanic bullae are not preserved and probably
were incompletely if at all ossified, and loosely attached to the
skull as in marsupials and insectivores. In modern carnivora they
are completely ossified and fast to the skull. But there is no trace
in Sinopa of the supporting plates from the alisphenoid and basi-
sphenoid bones around the margin of the bulla, the so-called
" false bulla," which is more or less developed in most insectivora
and marsupials. In Hyanodo'n the bullae are ossified to a varying
degree in the different species, in some apparently not at all, in
others a loosely attached bony ring, in others again a complete
osseous bulla ; but there is no trace of false bulla.

The teeth resemble those of many carnivorous marsupials, the
molars being triangular with transverse and oblique shearing edges ;
but the dental formula is that of eutherian mammals, three incisors,
a canine, four premolars and three true molars, while the marsupials
have four or five incisors, canine, three premolars and four true
molars. The angle of the lower jaw is like that of typical carniv-
ora, and shows no trace of the marsupial inflection. This inflected
angle is seen quite as clearly in Cretaceous as in modern marsupials
and is evidently a distinction of very ancient origin.

The details of construction of the skull, especially the basicranial
bones and foramina, agree entirely with the true carnivora, and
show that the marsupial resemblance is a superficial one.

The vertebrae agree with carnivora in all important points. The
vertebral artery perforates the atlas and does not perforate the
seventh cervical. This condition prevails in carnivora and most
eutherians ; in marsupials as far as I have examined, the reverse is
the case.

There are 13 dorsals and 7 lumbars, making a dorsolumbar
formula of twenty as in carnivora instead of nineteen as in marsupials.
The dorsolumbar formula is known in only a few creodonts. In
Oxyecna, and probably in Patriofelis and Hyeenodon, it was twenty
as in Sinopa; in Dromocyon nineteen according to Wortman. It is
probable that in all Oxyaenidae and Hyaenodontidae it was twenty
and in the Mesonychidae nineteen, this family approaching the

i9°5-]

MATTHEW— THE OSTEOLOGY OF SINOPA. 71

marsupials in two or three other important characters, and differing
rather widely from the remaining creodonta. The lumbar region is
long and the vertebrae large with long transverse processes, indi-
cating a flexible body with great leaping powers, as in primitive
mammals generally. Among modern carnivora the cats, viverrines
and mustelines retain more of this character than the other groups.

The limbs show a considerable degree of cursorial adaptation for
an Eocene carnivore, the bones being longer and the feet more
compact than in the majority of creodonts. The scapula is nearly
as long and narrow as in the dogs (the anterior border is incom-
plete and is restored too wide in the mount) ; the humerus com-
pares with that of the cat ; the femur retains a vestigial third
trochanter, but its distal end is deep and narrow, almost ungulate
in type ; the ulna is somewhat more robust than the radius, as in
creodonts generally, and in most insectivora and marsupials ; in
modern carnivora the shaft of the ulna is reduced to a varying
degree.

There are five well developed toes on each foot and the axis of
symmetry in both fore and hind foot passes through the middle
digit (mesaxonic) as in Hyanodon. In all modern carnivora and
in the Mesonychidas among creodonts, the axis of symmetry lies
between the third and fourth digit (paraxonic). In the Oxyaenidas
the weight is distributed over comparatively short spreading digits
so that the axis of symmetry is not well defined (amphaxonic).
The scaphoid, lunar and centrale bones of the wrist are separate as
in creodonta, instead of united as in true carnivora ; the arrange-
ment of the carpals resembles that in Hycznodon, but their vertical
diameters are greater. The fibula is large and has a considerable
facet for the calcaneum, and the contact between astragalus and
cuboid is slight as in Hymnodon.

The skeleton represents an undescribed species nearly allied to
S. rapax Leidy. The skull described by Wortman as Sinopa agilis
differs considerably in dentition, etc., and should be distinguished
generically ; the generic name Prototomus Cope, is probably avail-
able for this form. The most important distinctions from Sinopa
in the teeth are the closely connate paracone and metacone on
M1---!, absence of metacone on M&, reduced heels of the lower
molars, and much compressed premolars.

In all respects Sinopa appears as a primitive member of the

72 MATTHEW— THE OSTEOLOGY OF SINOPA. [April 13,

Hygenodont phylum. The genera Sinopa, Prototomus, Cynohycen-
odon, Pterodon and Hyccnodon show a series of stages in the develop-
ment of a highly specialized sectorial dentition, and with some
exceptions, in the specialization of the skull and skeleton so far as
they are known. The geological occurrence of the known species
of these genera precludes their being regarded as in the direct line
of phyletic descent. Sinopa and Prototomus are found together in
the Lower and Middle Eocene, while Cynohyoznodon, Pterodon and
Hycenodon occur together in the Oligocene. But without doubt
the genera represent very closely the stages through which the
phylum passed in its evolution, and that is about as much as it is
safe to assert of most phylogenetic series.

The relationship of Sinopa to the Oxyaenidae, especially to
Limnocyon, is not yet clear. There is a great deal of resemblance
in skeletal characters, a marked diversity in the more significant
features of the skull. Most of the resemblance, perhaps all, is to
be explained as due to retention of primitive creodont characters,
but some may indicate a nearer relationship of Hyasnodonts to
Oxyaanids than to any other creodont family.

1905.]

SINCLAIR— FAUNA OF SANTA CRUZ BEDS.

THE MARSUPIAL FAUNA OF THE SANTA CRUZ BEDS.

(Plates I and II.)

BY WM. J. SINCLAIR.

[Read April 13, fcpoj. )

The Patagonian marsupials of the Santa Cruz epoch are of pecu-
liar interest from the relationship which they bear to certain Aus-
tralian and Tasmanian forms. This relationship establishes the
reality of a former land connection between the Australian region
and South America, so plainly indicated by the distribution of the
Tertiary marine mollusks, fishes, land shells, decapod Crustacea and
plants.1

These marsupials are referable to three families, remnants of
which survive in widely separated parts of the world. The Thyla-
cynidas are represented by at least four genera in the Santa Cruz
fauna, where they occupy the place of the placental carnivora. The
Didelphyidae include the genus Microbioiherium and several other
imperfectly known forms, comparable in size to some of the smaller
South American opossums. The Santa Cruz diprotodonts belong
to a third family which may be called the Casnolestidae. A single
representative of this family, dznolestes, survives in Ecuador and
Colombia.

The Thylacynid.e.

This family is sharply separated from the Dasyuridse and all other
existing carnivorous marsupials by the absence of the metaconid in
the lower molars and by the great reduction of the outer cingulum

'Ortmann, A. E., Reports of the Princeton University Expeditions to Pata-
gonia, iSqb-iSqq, Vol. IV, pp. 299-302, 1902.

Ortmann, A. E., "The Geographical Distribution of Freshwater Decapods
and its Bearing upon Ancient Geography," Proc. Amer. Phil. Soc, Vol. XLI,
pp. 267-400, 1902.

Pilsbry, H. A., "Distribution of Helices in Time and Space," Manual of
Conchotomy, Series 2, Vol. IX, pp. xxxviii et set/., 1894.

Lydekker, R., "A Geographical History of Mammals."

Hedley, C, " Considerations on the surviving refugees in Austral lands of
ancient Antarctic life," Proc. Roy. Soc. Ar. S. Wales, August, 1895, p. 3, foot-
note 1.

74 SINCLAIR— FAUNA OF SANTA CRUZ BEDS. [April 13.

and styloid cusps in the upper teeth. In the Dasyuridre, these
styles are almost as high as the outer cusps of the trigon. The
family name, based on the Tasmanian marsupial wolf Thylacynus,
was proposed by Buonaparte in 1838, and may very properly be
extended to include the related South American forms.

The Santa Cruz thylacynes were predatory carnivores. An indi-
cation of their pugnacious habits is afforded by the traces of wounds
received in fighting, which are found occasionally on the skull and
mandible.

These carnivores have been placed by Ameghino in a sub-order
named by him the Sparassodonta, a group which he regards as refer-
able neither to the creodonts, the placental carnivores, nor the car-
nivorous marsupials. That the so-called Sparassodonta are true
marsupials, and not worthy of sub-ordinal rank, is fully apparent
from the following characters, which they possess in common with
existing marsupial carnivores :

1. A typical marsupial dental formula, ^, \, f, \.

2. The number of successional teeth is reduced below that char-
acteristic of the placentals.

3. The nasals are broad posteriorly, excluding from contact the
frontals and maxillas. There is usually a small contact in existing
carnivorous marsupials. A similar broadening of the nasals is ob-
servable in Mesonyx, Harpagolestes and Dromocyon among the
Creodonta.

4. Anteroposterior shortening of basis cranii.

5. Lachrymal spreading out on the face ; lachrymal duct within
the orbit. An internal opening of the lachrymal duct is observable
in Tliylacynus.

6. Inflected mandibular angle.

7. Excavation of the premaxillse for reception of the tips of the
lower canines as in the dasyures, Thylacynus and the opossums.

8. Basisphenoid and alisphenoid ridged as in existing marsupial
carnivores and unlike the structure of this region in the placentals.

9. Posterior extension of the malar bar to form the pre-glenoid
process.

10. Posterior border of palate thickened. This structure is
observable also in certain creodonts.

11. Posterior border of palate perforated by a large foramen on
either side of the posterior nares.

i9°5-]

SINCLAIR— FAUNA OF SANTA CRUZ BEDS.

12. An alisphenoid bulla present in some genera, absent in
others. Tympanic annular and unfused with the adjacent elements
in the former, unknown in the latter.

13. Basisphenoid perforated by internal carotid artery.

14. Presence of avascular foramen (the post-zygomatic of Cope)
perforating anteriorly the base of the zygoma below or within the
lip of the post-glenoid foramen.

15. Presence of a large vascular foramen (the sub-squamosal of
Cope) perforating the squamosal on or above the crest which con-
nects the base of the zygoma with the inion. This is absent in the
placental carnivores.

16. Sutures of the skull distinct. Not strictly a marsupial char-
acter, but indicative of marsupial affinities when considered in con-
nection with the other characters presented.

The four best known genera may be arranged as follows :

A. Skull brachycephalic. Alisphenoid not dilated to form an auditory bulla.

I. Dental formula |, i, |, |- Protocone on upper molars reduced. Mi
biscuspidate with paracone and antero-external style. Posterior pre-
molars greatly enlarged. Talonid of MT with single conical cusp.
Terminal phalanges round, blunt, and broadly fissured at the tips.

B. Skull dolichocephalic. Borhycena.

(a) Alisphenoid bulla absent.

1. Dental formula |?, i, f,|- Protocone well developed on M1 and M-,
reduced on 1VP-. Mi with vestigial protocone and metacone. Poste-
rior premolar not greatly enlarged ; in the inferior series not exceeding
the median premolar in size. Talonid of MT small and basin-shaped.
Terminal phalanges laterally compressed, sharply pointed, and slightly
cleft at tips. Prothylacynus .

(b) An alisphenoid bulla.

1. Dental formula |, i, |, |. Protocone well developed on Mi-^-. Mi

with small conical protocone, large paracone and antero-external style ;
metacone reduced to the merest vestige or absent. Premolars increas-
ing regularly in size posteriorly in both upper and lower series.
Talonid of MT enclosing a small basin-shaped area, unicuspidate.
Terminal phalanges uncleft, laterally compressed and pointed.

Cladosiclis.

2. Dental formula f, \, f, |- Protocone well developed on all the upper

molars. Mi with protocone enclosing a basin-shaped area ; paracone
and antero-external style large ; metacone vestigial or absent. Upper
premolars increasing regularly in size posteriorly ; median and posterior
lower premolars subequal. Talonid in MT large and strongly bicuspi-
date. Terminal phalanges laterally compressed and pointed without
clefts. Amphiproznverra.

76 SINCLAIR— FAUNA OF SANTA CRUZ BEDS. [April i3j

The Santa Cruz thylacynes are short-legged animals with large
heads, long necks and heavy tails. These characters are well
shown in the accompanying restorations of Prothylacynus patagoni-
cus and Cladosictis In stratus (Plates land II) reproduced from the
forthcoming Volume IV of the Reports of the Princeton Univer-
sity Expeditions to Patagonia. In addition to the characters
already mentioned, the following are worthy of notice :

i. The facial region of the skull is short in proportion to the
length of the cranium. The brain case is small and greatly con-
stricted postorbitally. The orbits are placed much further forward
than in the Dasyuridre, opossums, or Thylacynus. The jugal arches
are robust and broadly expanded, and the sagittal and lambdoidal
crests well marked but not very high. The palate lacks the vacui-
ties present in all existing carnivorous marsupials, but is perforated
by a number of accessory palatine foramina. Between the molars,
the margin of the palate is depressed into deep hemispherical fossa;
for reception of the tips of the lower teeth when the mouth is
closed. The occiput is semicircular in outline in contrast with its
triangular shape in the dasyures, Sarcophtfus and Thylacynus. The
lachrymal canal opens well within the orbital rim. In the majority
of living marsupials, the opening of the lachrymal duct is placed
either on or external to the orbital rim. Thylacynus is transitional
between these two types of structure in that it possesses a double
lachrymal perforation, one branch of the canal opening without
and the other within the orbit. Borhycena and Prothylacynus,
resemble Sarcophilus in the fusion of the mandibular symphysis.
In the remaining genera the symphysial union is ligamentous.

2. The molars are of the same type as in Thylacynus, differing
principally in the greater reduction of M4-, the loss of all the styloid
cusps except the antero-external, and the character of the heel of
the last lower molar, which may be either small and conical,
basin-shaped or bicuspidate. The premolars are unreduced in
number, and usually increase in size posteriorly in both series.
The canines are long, sharply pointed and slightly curved in the
smaller genera. In Borhycena the fang is swollen and the point
short and blunt. The incisors in Borhycena are reduced to f , an
exceptional formula among marsupials in that the number above
and below is the same. In Amphiproviverra the median pair are
conical and approximated at the tips as in Dasyurus and DiJcl-

I9oS.] SINCLAIR— FAUNA OF SANTA CRUZ BEDS. 77

phys. The posterior premolar is preceded, in Cladosictis, by a
deciduous tooth resembling the first molar. x\ccording to Ame-
ghino, the median premolar and canine in this genus also have
deciduous predecessors, and in Borhycena the canine is said to dis-
place a deciduous tooth.

3. The atlantal intercentrum is unfused with the base of the
neural arch in Borhycena and Amphiproviverra, as it is also in
Thylacynus. In Prothylacynus and Cladosictis complete fusion has
taken place with obliteration of the sutures. An atlantal fora-
men for the transmission of the spinal nerve and vertebral artery is
present in all the genera except Borhycena, which resembles Phasco-
lomys in transmitting the nerve and artery through a groove in the
anterior margin of the neural arch. The axis carries a large hatchet-
shaped neural spine. The bases of the transverse processes of the
second to the seventh cervicals are perforated for the transmission
of the vertebral artery. The dorso-lumbar vertebral formula was
probably nineteen as in Tiiylacynus : thirteen dorsals and six lum-
bars. As in that genus, the anticlinal vertebra is the tenth dorsal.
Two vertebras are coossified in the sacrum. The tail was undoubt-
edly long, very heavy and greatly thickened at the base.

4. The limbs are short in proportion to the length of the body
and the feet small with spreading toes. The trochlear surface of the
astragalus is short and flat with feebly differentiated facets for the
tibia and fibula which latter articulates with the calcaneum. In
Prothylacynus, the hallux is reduced to a deformed metatarsal,
which carries no phalanges and terminates distally in a blunt rounded
knob. In Cladosictis, the hallux is small, judging from the size of
its articulation on the entocuneiform. It may have supported
phalanges. The hallux in Amphiproviverra is large and opposable
indicating that this genus was probably arboreal. The pollex is
known in Amphiproviverra and Cladosictis. In these genera, the
phalanges of the pollex are deflected toward the inner side of the
foot as a result of the enlargement of the outer condyle of the
metacarpal of the thumb. In gait, the Santa Cruz thylacynes
were probably plantigrade. In striking contrast with these extinct
genera the pes of Thylacynus shows a peculiar cursorial modification.
Not only is the gait of this animal digitigrade, and the hallux
entirely obliterated, but the ectocuneiform has shifted t6\vard the
outer side of the foot until it is supported almost entirely by the

78 SINCLAIR— FAUNA OF SANTA CRUZ BEDS.

L April 13.

cuboid. In the Santa Cruz forms, this shifting has progressed to
about the same extent as in Sarcophilus. There is no trace of syn-
dactyly. The manus and pes are pentadactyl in Amphiproviverra
and Cladosictis. The manus is pentadactyl in Borhycena and prob-
ably also in Frothy lacynus. The hallux is reduced to a vestige in
the latter genus. Its condition in Borhycena is unknown.

5. The pelvis is without trace of epipubic ossifications in Clado-
sictis. The pubes are not preserved in the only specimen of Pro-
thylacynus in the Princeton collection, and the pelvis of Bor-
hyana and Amphiproviverra is unknown. The patella is ossified in
Amphiproviverra and Prothylacynus. The radius and ulna are
capable of some degree of pronation and supination. The tibia and
fibula are unfused. The inner humeral epicondyle is perforated by
a large foramen in Prothylacynus and Cladosictis ; imperforate in
Amphiproviverra. The supinator ridge terminates in a hook-shaped
extremity in Prothylacynus. This is wanting in Amphiproviverra
and Cladosictis.

THE DIDELPHYID.E.

The Didelphyidas are represented in the Santa Cruz fauna by
several genera of which Microbiotherium is the best known. In
dental formula and the structure of the lower molars Microbio-
therium agrees with Didelphys, differing from all the opossums in
the greater reduction of the outer cingulum, styloid cusps, and
metacone spur in the upper molars. The posterior premolar is en-
larged in both the upper and lower dental series. The premolars
are double-rooted in the majority of the species and decrease in
size anteriorly. The molars in both series decrease in size pos-
teriorly as in the existing didelphyd genus Caluromys.

THE CVENOLESTID.E.

This family, better known as the Epanorthidae, includes all the
Santa Cruz diprotodont marsupials. As the genus Palceothentes,
defined by Ameghino in 1887, has priority over Epanorthus, pro-
posed by him two years later, necessitating the rejection of the
latter, the family has been renamed after its best known representa-
tive, Ccenolestes. All the members of this family are small animals
and are very incompletely known.

Three subfamilies may be recognized. The more primitive
members of the first of these, the Caenolestinae, form a connecting

igos.] SINCLAIR— FAUNA OE SANTA CRUZ BEDS. 79

link between the polyprotodont and diprotodont marsupial sub-
orders in possessing, in the lower jaw, the tuberculo-sectorial type
of molar characteristic of the polyprotodonts combined with a
diprotodont modification of the median incisors. One of the
minute Santa Cruz forms has the same inferior dental formula as the
opossums. Unfortunately, nothing is known of the upper denti-
tion, skull and feet of this important transitional form.

The second subfamily, the Palseothentinae, contains the largest
of the Santa Cruz diprotodonts. The upper molars of the Palaeo-
thentinae resemble closely those of certain bunodont phalangers.
The first is fully quadritubercular. The second has a rudimentary
hypocone. The third and fourth are tritubercular. The lower
molars are lophodont. The posterior upper premolar and first
lower molar are modified as sectorial teeth. The dental formula
varies in the different genera but there are always four molars above
and below. The members of this subfamily form a regular pro-
gressive series in the shortening of the anterior portion of the
mandible and the reduction of the posterior lower premolar from a
double-rooted fully functional tooth to a single-rooted more or less
vestigial condition.

The Abderitinse, the third subfamily, are the most specialized of
the Santa Cruz diprotodonts. The first lower molar is greatly en-
larged, vertically grooved, and notched along the cutting edge of
the crown, resembling in some respects the peculiar sectorial teeth
of the multituberculate Plagiaulacidae. The sectorial in Abderites,
however, possesses a large bicuspidate heel, which is lacking in the
Plagiaulacidae, and the remaining molars are quadritubercular.

The Csenolestidae are examples of the restrictive influence of
competition on adaptive radiation. During the Santa Cruz epoch
they were crowded into obscurity by a horde of placentals, sloths,
rodents, and ungulates, and had no opportunity to attain the high
degree of adaptive specialization shown by the Australian dipro-
todonts, although so far as can be judged, they possessed as much
latent capacity toward variation as do their nearest living allies, the
phalangers.

RELATIONSHIPS OF THE SANTA CRUZ MARSUPIALS.

The Patagonian thylacynes do not represent the main line of
descent which ended in Thylacynus. In all the Santa Cruz genera

80 SINCLAIR— FAUNA OF SANTA CRUZ BEDS. [April 13,

the last upper molar has undergone greater reduction and the sty-
loid cusps have decreased in number, the antero-external alone being
represented. Apart from these advanced characters in the denti-
tion, the Santa Cruz thylacynes are of a distinctly more primitive
type than their surviving Tasmanian relative, which has progressed
in the lengthening of the face and posterior shifting of the orbit,
the increased brain capacity, the acquisition of palatal vacuities,
the prenatal shedding of the deciduous teeth, the external shifting
of the outer cuneiform, and the loss of the hallux. With the ex-
ception of the reduced hallux in Prothylacynus, transitions to these
advanced types of structure do not appear in the Santa Cruz mem-
bers of the family.

The marsupial faunas of those formations in Patagonia older
than the Santa Cruz are still too imperfectly known to afford a
secure basis for phylogenetic speculation, but it may confidently
be expected that the common ancestor of Thy lacy mis and the ex-
tinct Santa Cruz types will be found among them. In fact, certain
large carnivorous marsupials from the Pyrotherium beds named by
Ameghino, Proborhyama and Pharsophorus retain the metaconid in
the lower molars as in the Dasyuridai, while the premolar formula
is unreduced as in the Thylacynidre.

The affinities of Microbiotheriimi are unquestionably didelphyd.
The genus can not be regarded as ancestral to any of the existing
South American opossums as the degree of reduction of the ex-
ternal cingulum and styloid cusps in the upper molars is greater.

The most primitive of the Csenolestidae, the genus Halmarliiphus,
is transitional to the Polyprotodontia and represents, with little or
no modification, a type which is not only ancestral to the Palreo-
thentinae but agrees perfectly with the "minute insectivorous
forms which, apart from the diprotodont modification of the ante-
molar teeth, possessed a full antemolar formula," indicated by
Bensley's1 studies as the ancestors of the Phalangerinae. Un-
fortunately this interesting transitional genus is known only from
the lower jaw. The Palaeothentinae are important in retaining
constructive stages in the evolution of the bunodont type of molar
characteristic of the more primitive of the existing phalangers.
The Abderitinas are highly specialized diprotodonts which appear

1 Bensley, B. A., "The evolution of the Australian marsupials, etc.," Trans.
Linn. Sec, London, ser. 2 (Zool.), vol. 9, p. 1 39, 1 903.

rgos.] SINCLAIR— FAUNA OF SANTA CRUZ BEDS. 81

to have become extinct with the Palaeothentinae at the close of the
Santa Cruz epoch, while the less specialized Caenolestinae were able
to persist to the present day.

The Csenolestidae resemble the primitive phalangers in so many
respects that it is impossible to escape the conclusion that the two
families are related and not merely convergent groups. With the
exception of Halmarhiphus, a persistent ancestral type, the Santa
Cruz diprotodonts possess specializations in dental structure which
prevent their being regarded as direct ancestors of the phalangers,
but favor the idea that both groups are descended from a common
ancestry.

Considerable evidence is now available to show that a land con-
nection between Patagonia and the Australian region existed noj
later than the close of the Cretaceous or beginning of the Tertiary *
and it is possible that at this time the interchange of marsupials
between the two continents was effected. Whether the marsupials
originated in South America and migrated thence to Australia, or
the reverse, can not at present be demonstrated, but a South Amer-
ican origin for at least some of the existing Australian and Tas-
manian types appears probable in view of their unmistakable rela-
tionships with Santa Cruz forms.

Princeton University, April, 1905.

EXPLANATION OF PLATES.

Plate I. Cladosictis luslratus. Restoration based upon two specimens in the
collection of Princeton University. The skeleton measures 3 feet
over all. Restored parts are indicated by a cross.

Plate II. Prothylacynus patagonicus. Restored from a single specimen in the
Princeton collection. The skeleton measures 4 feet 8^ inches
over all. The restored parts are indicated by a cross.

1 For a summary of the evidence see Ortmann, Reports of the Princeton Uni-
versity Expeditions to Patagonia, Vol. IV, pp. 310-324.

82 LAMBERT— THE STRAIGHT LINE CONCEPT. [April i4,

THE STRAIGHT LINE CONCEPT.
BY P. A. LAMBERT.

{Read April 14, /<?PJ.)
INTRODUCTION.

The foundation of a science is the system of assumptions which
gives precision to the concepts with which the science deals. It is
essential that the system of assumptions together with the results
obtained by applying the processes of logic to the concepts shall be
free from contradiction. This freedom from contradiction is gen-
erally established by showing that the system of assumptions gives
precision to some complete number system of arithmetic.

It is an important problem in any science to reduce the system
of assumptions to a minimum. This problem is solved by exclud-
ing all assumptions which are logical consequences of other assump-
tions. When the system of assumptions of a science is reduced to
a minimum the omission of any one assumption or the change of
any one assumption will either lead to a contradiction or change
the concepts of the science.

In order that a science shall not become a mere exercise in men-
tal gymnastics the results obtained by applying the processes of
logic to the concepts of the science must agree with observed results
in the processes of the physical world.

The assumptions of a science are also called the axioms of a
science. The assumptions of geometry are called axioms by Hil-
bert in the Grundlagen der Geometric

THE STRAIGHT LINE.

The logical entities with which rational geometry deals are the
concepts named the poirrt, the straight line and the plane. Pre-
cision is given to these concepts by axioms which have been ar-
ranged by Hilbert in five groups, called axioms of relation, axioms
of order, axioms of congruence, axioms of parallels and axioms of
continuity.

The importance of the straight line, whether by straight line we
understand the intuitive entity of experience or the logical entity

I905j LAMBERT— THE STRAIGHT LINE CONCEPT. 83

of rational geometry, depends primarily on the fact that a straight
line is determined by any two of its points and can be indefinitely
extended between any two of its points. Right here arises a ques-
tion that can not be answered by experience or experiment. If the
straight line is indefinitely extended beyond any two of its points,
will there be found on the straight line two points at infinity, one
point at infinity, or no point at infinity ? This question, of course,
can not be answered until precision has been given to the term
distance.

In the plane determined by a given point and a given straight
line, draw a straight line through the given point intersecting the
given straight line and revolve the straight line about the given
point. When the point of intersection of the revolving line with
the given line moves to an infinite distance from the foot of [the
perpendicular from the given point to the given line, the revolving
line is said to become parallel to the given line. In how many
positions does the revolving line become parallel to the given line ?
This, again, is a question that can not be answered by experience
or experiment. It can be answered only by the axiom of parallels.

If the axiom of parallels is made to read : Through a given point
without a given line one and only one parallel to the line can be
drawn — we have a geometry in which the straight line has only
one point at infinity. This is the geometry of Euclid. Since the
parabola meets the straight line at infinity in only one point, this
geometry is also called the parabolic geometry.

If the axiom of parallels is made to read : Through a given point
without a given line two and only two parallels to the line can be
drawn — we have a geometry in which the straight line has two
and only two points at infinity. Since the hyperbola intersects the
straight line at infinity in two points, this geometry is called the
hyperbolic geometry. The hyperbolic geometry was developed by
Euclidean methods by Lobachevski and Bolyai.

If the axiom of parallels is made to read : Through a given point
without a given line no parallel to the line can be drawn — we have
a geometry in which the straight line has no point at infinity.
Since the ellipse does not intersect the straight line at infinity in a
real point, this geometry is called the elliptic geometry. The
elliptic geometry has been discussed by Riemann, Clifford and
Newcomb.

84 LAMBERT— THE STRAIGHT LINE CONCEPT. [April i4,

THE EXPRESSION FOR DISTANCE.

Much of the apparent mystery of hyperbolic and elliptic geom-
etry vanishes when precision is given to the term distance. Dis-
tance is the result of measurement, and the measurement of a straight
line requires that any part of the straight line may be applied any-
where along the straight line. If A, B, Care any three points in
a straight line, and B is between A and C, the expression for dis-
tance must satisfy the equation

distance AB + distance BC= distance AC.

In a system of measurement introduced by Cayley in the Sixth
Memoir on Quantics and developed by Klein, the expression for
the distance between two points on a straight line is a function ot
the cross-ratio of these two points and two fixed points on the
straight line. Let the fixed points be X, YandA, B, Cany three
points taken in order on the straight line. By definition the
cross-ratio of the four points A, B, X, Fis

(AX)/(AY)+(BX)/(BY).

It follows from this definition that :

cross-ratio ABXY X cross-ratio B CX Y= cross-ratio ACXY.
Applying logarithms to this equation

log cross-ratio ABXY ' -\- log cross-ratio BCXY

= log-cross-ratio A CYX.
The expression

k log cross -ratio ABXY

where k denotes any constant may therefore be taken as the expres-
sion for the distance between the points A, B}

The pair of fixed points X, Fis called the absolute of linear
measurement. When one of the points A, B coincides with a
point of the absolute the distance AB becomes infinite. Hence
when the absolute consists of two distinct real points, the straight
line has two points at infinity ; when the absolute consists of two
coincident points, the straight line has one point at infinity ; when

1 The constant k must be so determined that the expression for distance has a
real value. Since the logarithm is a many-valued function for which the series
of values differ by multiples of 2k\/— I, when k is imaginary the expression for
distance is a many-valued function for which the series of values differs by multi-
ples of some real constant.

i9°5-]

LAMBERT— THE STRAIGHT LINE CONCEPT. 85

the absolute consists of a pair of imaginary points, the straight
line has no point at infinity.

In the geometry of two dimensions the absolute must be the locus
of the point pairs which are the absolute of all the lines in the
plane. It follows that the points on the absolute are the points at
infinity in the plane. By substituting in the equation of the abso-
lute f(x, y) = o for x and y respectively (xx + Xx2)/(i + '•) and
0'i + l)\)/(l + '•) there will be found two values of X, say A, and
/.,, to which correspond the points of intersection of the straight
line through the points (.\\, j\) and (x2, v2) with the absolute, and
the cross-ratio of the points (.v,,/,), (xv y3) and the points of
intersection with the absolute is Xjk%. This cross-ratio is there-
fore readily calculated whether the points of intersection are real or
imaginary.

If the equation of the absolute in homogeneous coordinates is
v2 + v2 — 4a2/" = o, in order that the distance between two points
within the absolute shall be real the constant k must be real.
Every straight line determined by two points within the absolute
has two points at infinity and we have the hyperbolic geometry of
two dimensions. Points without the absolute are non-existent in
this geometry.

If the equation of the absolute in homogeneous coordinates is
x2 +y + ^z2/2 = o the constant k must be assumed imaginary in
order that the distance between two points of the plane shall be
real. The straight line has no point at infinity and we have the
elliptic geometry. In this geometry the straight line has a finite
length and must return into itself. The distance between two
points has a series of values differing by multiples of the length of
the entire straight line.

The points whose homogeneous coordinates are x = 1, y =
y/ — 1, t= o) x = 1, y = — s/ — 1, /= o satisfy the equations of
the absolute in both the hyberbolic and elliptic geometries. These
two points, named the imaginary circular points at infinity, consti-
tute the absolute of plane parabolic geometry. The parabolic
geometry is therefore a common limiting case of the hyperbolic
and elliptic geometries. By a suitable choice of the constant k
the parabolic geometry becomes the geometry of Euclid.

In the geometry of three dimensions the absolute of hyperbolic
geometry may be written v2 4- r2 + z1 — 4a2/2 = o ; the absolute

86 LAMBERT— THE STRAIGHT LINE CONCEPT. [April i4)

of elliptic geometry x'2 + y2 + z'2 -f 4a2/- = o ; the absolute of
parabolic geometry, x2 -f- y1 -f- r = o, t — o, again a common
limiting case of the absolute of hyperbolic and elliptic geometry.

By taking a in the equation of the absolute sufficiently large the
hyperbolic and elliptic geometries approach identity with the para-
bolic geometry in finite regions of space, so that experience or
experiment can never determine that the space of experience is
hyperbolic, elliptic or parabolic.

The expression for the distance between two points must satisfy
the requirement that the distance between two points shall be the
same for all positions of the straight line on which the two points
are located. A collinear motion of space into itself is represented
analytically by a linear transformation which transforms the abso-
lute into itself. The cross-ratio is an invariant of linear transforma-
tions. Hence the definition of distance k x log cross-ratio satisfies
also this requirement of the expression for distance.

By the calculus of variations it is proved that in the elliptic,
hyperbolic and parabolic geometries the straight line is the shortest
distance between two points. Hilbert, by taking for absolute a
triangle, has proved that the sum of two sides of a triangle may be
equal to or less than the third side.

ANGLE MEASUREMENT.

In Cayley's system of measurement the measure of an angle is
defined as a constant times the logarithm of the cross-ratio of the
pencil of four rays formed by the sides of the angle and the
tangents to the absolute from the vertex of the angle. If the
equation of the absolute in line coordinates is /(//, v) = o, the
measurement of angles about the point of intersection of the lines
(«j, i\) and (u2, 7'2) is analytically identical with the measurement
of distance on a line through two points.

It follows from the definition that a right angle is an angle
whose sides are harmonic conjugates with respect to the tangents
from the angle vertex to the absolute. In the hyperbolic geometry
any line through the pole of a given line with respect to the abso-
lute intersecting the given line is perpendicular to it ; the angle
between lines intersecting on the absolute is zero, hence the two
lines drawn from a given point to the intersections of a given line
with the absolute are parallel to the given line ; the sum of the
angles of a triangle is less than 1800.

i9°5-]

LAMBERT— THE STRAIGHT LINE CONCEPT.

The chief attraction of hyperbolic geometry lies in the fact that
one has the power to see the whole of hyperbolic space and to
direct geometric constructions from a vantage point outside of this
space. For example, to draw a common perpendicular to two straight
lines, not intersecting and not parallel, connect by a straight line the
poles of the given straight lines with respect to the absolute. This
problem has been solved by Hilbert by methods such as a being
living in hyperbolic space would be obliged to use. It is a simple
matter to determine directly from the expression for distance that
the locus of points in the hyperbolic plane equidistant from a given
straight line is an ellipse tangent to the absolute where the given
line meets the absolute.

GEOMETRY ON SURFACES OF CONSTANT TOTAL CURVATURE.

If Rl and R2 are the maximum and minimum radii of curvature
of the normal sections of a curved surface at any point of the sur-
face, the reciprocal of the product of Rx and R2 is called the Gaus-
sian or total curvature of the surface at this point. The geometry
of geodesies on surfaces whose total curvature is constant has strik-
ing analogies to plane Euclidean geometry. Euclid's definition
of a straight line as a line which lies in the same manner with
respect to all the points in the line, and his definition of a plane
as a surface which lies in the same manner with respect to all
straight lines in the plane, when taken in connection with Euclid's
"Common Notions" implies the congruent displacement of a
straight line into itself, that is the displacement of the straight line
into itself such that any two points of the line may be made to coin-
cide with any other two points of the line provided the distance
between the first pair of points equals the distance between the
second pair ; and the congruent displacement of the plane into
itself, that is the displacement of the plane into itself such that any
portion of the plane bounded by straight lines may be made to
coincide with any other portion provided the two portions are
bounded by straight lines of equal length and the corresponding
angles are equal. Now surfaces whose total curvature is constant
and geodesies on these surfaces also possess this property of con-
gruent displacement, provided displacement is suitably defined.

If the constant total curvature of the curved surface is — a'1, the
geometry of geodesies on the curved surface is identical with the

88 LAMBERT— THE STRAIGHT LINE CONCEPT. [April 14,

geometry of the plane with hyperbolic measurement.1 The type
of surfaces with constant negative total curvature is the pseudo-
sphere of revolution, generated by revolving the tractrix about its
asymptote.

If the constant total curvature of the curved surface is -f- a2, the
geometry of geodesies on the curved surface is identical with the
geometry of the plane with elliptic measurement.1 The type of
curved surfaces with constant positive total curvature is the sphere.
It is important to note that the entire elliptic plane is represented
on the hemisphere.

These statements show the reasonableness of using as equivalent
the terms elliptic space and space of positive curvature ; hyperbolic
space and space of negative curvature ; parabolic space and space
of zero curvature.

CONTINUITY OF THE STRAIGHT LINE.

It remains to examine the elemental structure of the straight line.
Adopting as definition of continuity the totality of all real num-
bers, is the totality of distances from a fixed point of the line to all
other points of the line continuous? This question must be an-
swered by establishing a correspondence between sets of numbers
and points and lines, that is by a system of analytic geometry.

Let any pair of numbers (x, y) correspond to a point, any pair
of numbers (//, v) correspond to a straight line, and let the equa-
tion ux + vy -+- i = o denote that the point (x, y) is on the line
(u, 7>). The straight line is now determined by any two points
and two straight lines intersect in only one point, that is, the
straight line is the straight line of Euclid. If x, y and it, v are
any numbers of the totality of numbers obtained from unity by
applying a finite number of times the operations addition, subtrac-
tion, multiplication, division and taking the positive square root of
unity plus the square of any number previously determined, Hilbert
has proved that all the constructions of Euclid are possible. The
straight line, however, is clearly not continuous, for no transcen-
dental numbers occur in the totality of numbers represented on the
straight line.

The continuity of the straight line is not a necessity of Euclid's

1 Except for certain self-evident limitations due to the peculiarities of the
surface.

I9o5.] LAMBERT— THE STRAIGHT LINE CONCEPT. 89

geometry, it is not an intuitive property of the straight line and it
cannot be proved by experiment. The continuity of the straight
line can be established only by means of axioms, whether these
axioms take the form given by Dedekind in his Essays on Number
or by Hilbert in his Foundations of Geometry. When the contin-
uity of the straight line has been established the Cartesian geometry
at once follows.

Lehigh University,
April S, 1905.

MAGELLANIC PREMIUM

Founded in 17S6 by John Hyacinth de Magellan, of London
I905

THE AMERICAL PHILOSOPHICAL SOCIETY

Held at Philadelphia, for Promoting Useful Knowledge

ANNOUNCES THAT IN

DECEMBER, 1905

IT WILL AWARD ITS

MAGELLANIC GOLD MEDAL

to the author of the best discovery, or most useful invention,
relating to navigation, astronomy, or natural philosophy
(mere natural history only excepted) under the following

- CONDITIONS :

1. The candidate shall, on or before November 1, 1905, deliver free of
postage or other charges, his discovery, invention or improvement, ad-
dressed to the President of the American Philosophical Society, No. 104
South Fifth Street, Philadelphia, U. S. A., and shall distinguish his per-
formance by some motto, device, or other signature. With his discovery,
invention, or improvement, he shall also send a sealed letter containing
the same motto, device, or other signature, and subscribed with the real
name and place of residence of the author.

2. Persons of any nation, sect or denomination whatever, shall be
admitted as candidates for this premium.

3. No discovery, invention or improvement shall be entitled to this
premium, which hath been already published, or for which the author
hath been publicly rewarded elsewhere.

4. The candidate shall communicate his discovery, invention or im-
provement, either in the English, French, German, or Latin language.

5. A full account of the crowned subject shall be published by the
Society, as soon as may be after the adjudication, either in a separate
publication, or in the next succeeding volume of their Transactions, or in
both.

6. The premium shall consist of an oval plate of solid standard gold of
the value of ten guineas, suitably inscribed, with the seal of the Society
annexed to the medal by a ribbon.

All correspondence in relation hereto sholild be addressed

To the Secretaries of the

AMERICAN PHILOSOPHICAL SOCIETY

No. 104 South Fifth Street

PHILADELPHIA, U. S. A.

TRANSACTIONS

OF THE

American Philosophical Society

HELD AT PHILADELPHIA

For Promoting Useful Knowledge

Vol. XXI, New Scries, Part I. Just Published

The Morphology of the Skull of the Pelycosaurian Genus
Dimetrodon. By E. C. Case

Subscription— Five Dollars per Volume

Separate parts are not sold

Address The Librarian of the

AMERICAN PHILOSOPHICAL SOCIETY
No. 104 South Fifth Street

PHILADELPHIA, U. S. A.

H^\^~ PROCEEDINGS

AMERICAN PHILOSOPHICAL SOCIETY

HELD A I PHILADELPHIA

FOR PROMOTING USEFUL KNOWLEDGE

Vol. XLIV. May-July, 1905. No. 180.

CONTENTS.

The Mutual Affinities of the Species of the Genus Cambarus,
and their Dispersal over the United States. By Dr. A.
E. Ortmann 91

The LTse of the Rotating Anode and Mercury Cathode in
Electro- Analysis. By Lily G. Kollock and Edgar F.
Smith 137

Observations on Columbium and Tantalum. By Edgar F.

Smith 151

Enquiry into the Pressure and Rainfall Conditions of the Trades-
monsoon Area. By W. L. Dallas 159

The Relation Between the Economic Depth of a Bridge Truss
and the Depth that Gives Greatest Stiffness. By Mans-
field Merriman 164

On the Use of the Falling Plate Oscillograph as a Phase Meter.

By William McClellan 166

Some Observations on Columbium. By Roy D. Hall and

Edgar F. Smith 177

PHILADELPHIA

THE AMERICAN PHILOSOPHICAL SOCIETY

104 South Fifth Street

1905

GENERAL MEE i 1XG— 1906

The next General Meeting of the Society will be held on April 17-20.
1906; beginning on the evening of Tuesday. April 17.

Wednesday, April iS, will be devoted to the presentation and discussion
of scientific papers, and Thursday, 19 and Friday, 20, to the ceremonies
connected with the celebration of the 200th Anniversary of the Birth of
Benjamin Franklin.

Members desiring to present papers on subjects of science at the Gen-
eral Meeting are requested to communicate with the Secretaries at the
earliest possible date.

CORRIGENDUM

In No. 179 the volume number should be corrected to read Vol. XLIV.
A new title page, containing the correct volume number, to be substituted for
the one in No. 179, will appear in the concluding number of this volume.

Members who have not as yet sent their photographs to the Society will
confer a favor by so doing: cabinet size preferred.

It is requested that all correspondence be addressed
To the Secretaries of the

AMERICAN PHILOSOPHICAL SOCIETY
104 South Fifth Street

Philadelphia, IT. S \

AUG 14 1905

PROCEEDINGS

OF THE

AMERICAN PHILOSOPHICAL SOCIETY

HELD AT PHILADELPHIA
FOR PROMOTING USEFUL KNOWLEDGE

Vol. XLIV. April-May, 1905. No. 180.

THE MUTUAL AFFINITIES OF THE SPECIES OF THE

GENUS CAMBARUS, AND THEIR DISPERSAL

OVER THE UNITED STATES.

(Plate III.)

BY DR. A. E. ORTMANN.

{Read April ij, fcpoj.)

In a previous study of the geographical distribution of the cray-
fishes of the United States (see Proc. Amer. Phil. Soc, xli, 1902,
pp. 277-285), the present writer took it for granted that the
division of the genus Cambarus, into five groups, as introduced by
Faxon {Mem. Mus. Harvard, 10, 1885), and the arrangement of
the species within each group adopted by him, would correspond,
as far as one might expect, to the natural affinities.

This, however, is true only to a certain degree. There is no
doubt that Faxon correctly recognized the chief systematic groups
within the genus, and that he also had, in most cases, appropriate
views as to the relationship of smaller groups of species. But accept-
ing his system as a whole, and trying to correlate it with the peculi-
arities shown by the geographical distribution, a number of cases
are revealed, where such a correlation is not very apparent, and
attempts to give a reasonable theoretical explanation prove to be
more or less unsatisfactory. I shall mention here a few instances.

1. The fifth group of the genus follows, in Faxon's system, after

PROC AMER. PHILOS. SOC. XLIV. l8o. G. PRINTED JULY 28, I905.

92 ORTMANN— AFFINITIES OF CAMBARUS.

[April

the fourth, and he apparently believes, that it is connected genet-
ically with the latter. The geographical distribution, however, is
opposed to this assumption, and a closer study has led me to think
that there is no such affinity between these two groups, and that the
fifth is more closely allied to the first and second. (Compare Ort-
mann, I. c, 1902, p. 283.)

2. Faxon believes (/. c. , 1885, p. 19) that the first group con-
tains the most primitive forms. This is not probable when we con-
sider the very highly specialized character of most of the species.
Indeed, there are rather primitive forms among them, but they are
clearly not as primitive as certain species of the second group (Ort-
mann, 1903, p. 283), and further, the main range of the first group
occupies a territory that is, geologically, comparatively young,
namely, the lowlands of the southern states (Mississippi, Alabama,
Georgia, Florida), of which we know that they became land by
degrees during the Tertiary period, the more southern parts in very
recent times. It is not very likely that this recent land is occupied
by an ancient group of animals.

3. I strongly object to placing Cambarus pellucidus, the blind
cave-species of Kentucky and Indiana, with the first group, where
it stands entirely isolated, morphologically as well as geographically.
If we place this species at the beginning of the fourth group, it
comes into an assemblage, from which it is not so strongly differ-
ent. It will always remain a remarkable, and, as Faxon believes,
a primitive type, but it is not the most primitive type of the genus
in all respects. In the shape of the male organs it certainly points
rather to the fourth group than to the first.

4. Faxon places C. blandingi at the head of the genus: this is
apparently due to the desire to let the type-species of the genus
stand first. This, however, may convey the wrong impression,
that C. blandingi is the lowest (or else the highest) form of the
whole genus. But I do not think that it is either, and regard it as
a highly specialized (but not the most highly specialized) form of
a branch of the genus that is rather ancient. The distribution of
C. blandingi has all the characters of a comparatively modern
encroachment upon foreign territory.

5. I believe that the second group of Faxon contains the most
primitive types of the genus. But this is to be understood "cum
grano salis." There are, in this group also some very highly

1905.] ORTMANN — AFFINITIES OF CAMBARUS. 93

specialized types (C. gracilis and allies), and even some of the
primitive forms (C. cubensis) possess some peculiar and apparently
advanced characters. I think we can express it this way : among
the second group, there are species that approach most closely the
original, stock of the genus, but they themselves are modified to a
degree. If I am to single out a species that possibly is the most
primitive, I should name C. digueti Bouvier (Bull. Mus. Paris,
1897, p. 224), which is identical with C. carinatus Faxon (Pr.
U. S. Mies., 20, 1898, p. 648). It is remarkable that the most
primitive forms are found in Mexico (and Cuba), which agrees
well with the theory of the origin of the genus in these parts (see
Ortmann, /. c. , 1902, p. 283).

CHARACTERS OF THE GENUS CAMBARUS THAT SERVE TO
DISTINGUISH GROUPS AND SPECIES.

Sexual Organs. — Already the earlier writers (Girard, Hagen)
have pointed out the importance of the anterior pair of abdominal
appendages of the male (copulatory organs) for systematic pur-
poses. Faxon made large use of them in defining his five groups,
but within the groups he rarely tried to avail himself of these limbs
to reveal the mutual affinities of the different species. It is now
generally known that these organs furnish not only the best specific
characters, but that their similarity in certain assemblages of species
clearly indicates genetic relationship. In close connection with
the shape of this organ is that of the female " annulus ventralis,"
as we now know, the receptaculum seminis. There is, however,
not so much variety in the shape of the latter organ, it is not so
polymorphous, and the main types occur in different groups, which
is apparently due to the more simple structure of this organ.
Nevertheless, the annulus — in connection with the male organ —
is rather important. We may add that in a number of species
the annulus is not very well known.

As regards the male organs, their shape is very complex and
much varied. Several main types may be distinguished, and these
again show much diversity. It is hardly likely that in so complex
an organ the identical form has developed several times, that is to
say that there are cases of parallelism ; where there is identity or
similarity of these organs in different species, this is generally and
surely a sign of close affinity. Only one or two exceptions in the

94 ORTMANN — AFFINITIES OF CAMBARUS. [April i3,

first and second group are known to me, where the question is ad-
missible, whether convergency plays a part. There are rarely two
species known, where the shape of the male organs is absolutely
identical (except in the third group of Faxon, where they are re-
markably uniform), and since they have developed from the begin-
ning in three or four different main lines, it is easily understood
why they furnish the best specific characters as well as the best
criteria for judging the affinities. Thus the danger of being misled
by convergency of structure, which is the chief impediment of
properly recognizing natural affinities in any group of animals, is
here reduced to a minimum. We shall see below that by the ac-
tual use of this principle we arrive at conclusions that render the
investigation of the development of the genus Cambarus a com-
paratively easy task, furnishing a clue to the explanation of the
geographical distribution ; further, the study of the male organs
gives us a standard by which to judge the other characters that are
of systematic value, and as we shall presently see, there is hardly
another structure that has the same value for revealing the affinities
within the genus, that is to say, the same characters generally de-
velop independently in different groups, being clearly subject to
parallelism, presumably under the influence of similar external con-
ditions. In a few cases the latter is very evident.

Copulatory Hooks of the Male. — Faxon lays much stress upon the
number of hooks present in the male on the ischiopodite of the sec-
ond, third, or fourth pereiopods, which are used to take hold of the
female in copulation. The third pereiopods always possess these
hooks, and in many cases only this pair is present. But sometimes
there is an additional pair on the second, or on the fourth pereio-
pods. The number of pairs of hooks is very constant in the single
species (except for occasional abnormities;, and it is remarkable
that certain types of male sexual organs are generally connected
with certain type of hooks ; this is chiefly the case in the third,
fourth and fifth group of Faxon, while it is not in the first or sec-
ond, where similar types of sexual organs may be connected with
different types of hooks.

If we consider that the presence of two pairs of hooks is cer-
tainly a more highly advanced stage than that of only one pair, that
is to say, that the difference of the number of hooks is only a differ-
ence in the degree of development of one and the same feature, it

i9°5-]

ORTMANN— AFFINITIES OF CAMBARUS. 9i

is easily conceivable that the increase of the number of hooks may
have taken place independently in different groups, and we shall
see below that there is at least one case (C. pellucidus), where we
are to assume an independent origin of an additional pair of hooks
on the fourth pereiopods ; this is also rendered probable by the
variability of this character seen in this species. That Faxon's
fifth group has developed an additional pair of hooks independently
is clearly shown by the fact that here it is the second pair of pereio-
pods that carries the additional hooks.

Indications of a more general tendency to increase the number
of hooks are found in occasional freaks in other groups ( C. propin-
gi/us, C. virilis).

General Shape of Carapace. — The primitive type of the cara-
pace seems to be more or less ovate, generally depressed. It
assumes, however, sometimes a more cylindrical form (in some
cave-species), and in some cases it is rather compressed. The
latter character is most remarkable in all burrowing species, and
has developed independently at least in two groups (second and
third of Faxon, gracilis- and diogenes-groups).

The Areola seems to be originally rather broad and short. But
there is a general tendency of it to become narrower, and at the
same time to increase its relative length. This is evidenced in
almost all groups, and a narrow, sometimes partly obliterated
areola is found in species that have nothing whatever to do with
each other. In fact, it is only the fifth group of Faxon where this
tendency is not manifested. Generally, the length of the areola is
correlated to the width, but there are exceptions.

Rostrum. — The shape of the rostrum is characteristic for most
species, but it is available only as a specific character. The
original type seems to be a rather long rostrum, with more or less
parallel margins, with a marginal spine on each side, and a rather
long acumen. The chief tendency in further development is for
the marginal spines to disappear, and for the whole rostrum to
become shorter. This, however, is found in all five groups of
Faxon in species which are not at all allied to one another. Even
certain peculiar types of rostrum may reappear in a widely diver-
gent group. Thus the d/andingt'-type is imitated, if the expression
is permitted, by C. immunis, and the burrowing species possess all
a rostrum of similar shape.

96 ORTMANN — AFFINITIES OF CAMBARUS. [April i3>

[It is very remarkable, that burrowing crayfishes of the southern
hemisphere (Parastacus defossus Fax. from southern Brazil is at
hand) resemble the North American burrowing species in a remark-
able degree externally, chiefly so in the shape of carapace, rostrum,
chel?e, length of abdomen, etc.]

The Chela are very variable in shape : they are fully developed
only in old males, but generally quite characteristic for the species.
One and the same type is often common to large groups of species,
and thus they are often a good help in the investigation of the
natural affinities. But in other cases a similar form of chelae is
found in different groups which is most striking again in the bur-
rowing species.

Among the more primitive species the shape of the chelae seems
to be more or less subcylindrical, and rather elongated. This
shape is found in Faxon's first, second, third, and fifth groups, but
only in the first, second and fifth it is frequent. Removing C.
pellucidus to the fourth group, also in the latter this type of chela?
is represented. In part of the second group, and in the third and
fourth, a more or less ovate, broad, and depressed chela becomes
common, but there is not much uniformity in detail, each group
generally developing its own type.

The above are the more important characters. We see that all
of them must be dealt with cautiously, if they are to be used for
the investigation of affinities of species. Indeed, in many cases,
they support the conclusions arrived at by the examination of the
the sexual organs, but very frequently similarities of the above
characters are due to convergency. The same is true of all other
characters, such as armature of carapace, chelae, shape of "epistoma,
antennal scales, of abdomen, telson, etc.

SUBGENERA OF CAMBARUS, ACCORDING TO THE CHIEF TYPES OF THE
SEXUAL ORGANS OF MALE.

There are three chief types of the male sexual organs (first pair
of abdominal appendages), the last of which is easily divided into
two subtypes. According to these, I should like to distinguish
four subgenera, as follows :
i. Subgenus: Cambarus (sens, strict.).

Sexual organs of male stout, more or less straight, and compara-
tively short, truncated or blunt at the tip, the outer part ending in i-j

i9oS.] ORTM ANN— AFFINITIES OF CAMBARUS. 97

horny teeth, which are sometimes recurved, or compressed, or plate-
like, and are always sharply distinguishable from the blunt end.
Inner part terminated by a shorter or longer, acute spine, which is
sometimes distinct from the tip of this part, so that it appears two-
pointed. In the male the third or the third and fourth pereiopods
have hooks.

2. Subgenus : Cambarellus nov. subgen.

Sexual organs of male stout, straight, or slightly curved at the
tips. Outer part ending in two horny teeth, which are rather long,
taper rapidly, and are not sharply distinguishable from the end,
which is not truncated. Inner part terminated by a rather long,
acute spine. In the male, the second and third pereiopods have
hooks.

[The two following subgenera represent the third type of male
sexual organs, in which both parts, outer and inner, each terminate
in only one tooth, which is rather slender, and not sharply distin-
guishable from the end, which is never truncated.]

3. Subgenus : Faxonius nov. subgen.

Sexual organs of male shorter or longer, not very stout, generally
slender, or with slightly curved tips. Tips never truncated, ending
always in two more or less elongated spines, the one formed by the
outer part, and horny, the other formed by the inner part and
softer. There is never more than one tip to the outer part, and there
is no terminal tooth distinguishable, but the tip tapers gradually,
or the whole outer part is setiform. In the male generally the third
pereiopods only have hooks, very rarely (in C. pellucidus') hooks
are found on third and fourth pereipods.

4. Subgenus : Bartonius nov. subgen.

Sexual organs of male very uniform throughout the subgenus.
They are short and thick, inner and outer part each terminating in
only one short and thick spine, tapering to a point. Both terminal
spines are strongly recurved, forming with the basal part about a
right angle. In the male, only the third pereiopods possess hooks.

Subgenus : Cambarus. (Type : C. blandingi. )

This subgenus comprises Faxon's first and second group, exclud-
ing the species C. pellucidus. Both groups are rather heterogeneous,
and so is this subgenus, and there are considerable variations in the
male sexual organs. The chief feature of the latter is their blunt

98 ORTMANN — AFFINITIES OF CAMBARUS. [April 13,

ending, a character that possibly points to the condition seen in
the genus Potamobius ; for the rest, the terminal teeth are quite
variable, but always very characteristic for the species.

It is advisable to distinguish groups within this subgenus, not
only with reference to the sexual organs, but also with reference
to the hooks of the male, for the presence of one or two pairs of
hooks seems to constitute, as already Faxon recognized, important
differences, the presence of two pairs, on third and fourth pereio-
pods, being evidently a more advanced stage. Using in addition
some other differences of the areola and the chelae, we obtain the
following three sections.

1. Section : C. digit eti.

Sexual organs of male with one to two teeth at the tip of the outer
part. Male with hooks on third pereiopods. Areola wide or nar-
row, but never obliterated, about half as long as the anterior section
of the carapace {incl. rostrum), or shorter. Chela elongated and
subcylindrical.

2. Section : C. gracilis.

Sexual organs of male with one to tiuo teeth at the tip of the outer
part. Male with hooks on third pereiopods. Areola obliterated in
the middle, considerably longer than half of the anterior section of
the carapace. Chela short, broad, ovate.
J. Section : C. blandingi.

Sexual organs of male with one to three teeth at the tip of the outer
part. Male with hooks on third and fourth pereiopods. Areola
wide or narrow, rarely obliterated in the middle, shorter or longer.
Chela generally elongated, narrow, and subcylindrical.

The most primitive sexual organs are found in species of the first
section, where there is only one tooth at the end of the outer part.
Similar sexual organs are found in the second (C. advena) and in
the third section (C evermanni} : in the latter cases, however, I
think we have to deal with parallelism, the single tooth in both
cases being due possibly to reduction. Since these two species are
very rare and poorly known, and since C. evermanni belongs to a
group that offers other difficulties, further investigations are needed.

A closer examination may reveal the fact, that the sexual organs
of the digueti-group are more sharply distinguished from those of
C. advena and evermanni. Through the courtesy of Professor E.
Bouvier of Paris, I have received two cotypes (male and female)

i9°5 ]

ORTMAXN — AFFINITIES OF CAMBARUS. 99

of his C. digueti, which show, on the one hand, that C. carinaius
Fax. is a synonym of this species, and on the other hand, that the
sexual organs have a rather peculiar shape. The figures of these
organs, given by Faxon (/V. U. S. Mus., 20, 1898, pi. 65, f. 2
and 3) are absolutely correct, but the description (p. 648) is rather
short and unsatisfactory. Faxon says: "Inner and outer parts
ending in a small horny tooth, anterior margin furnished with a
small tooth near the tip." Fig. 2 represents this organ of the
right side, seen from the outside : the outer part ends bluntly,
without a distinct tooth, while the inner part ends in a rather
pointed tooth, outside of which is a sharp spine that is longer than
the outer part. Faxon's Fig. 3 represents the identical part seen
from the inside : only the two tips of the inner part are seen here,
and the tip of the outer part is hidden behind the end of the inner ;
the inner part is flattened and hairy on the inside, and the " shoul-
der " ( " small tooth near tip " ) is distinctly developed. My male
specimen of C. digueti 'agrees in every detail with the figures of Faxon.

A very similar structure is seen in C. cubensis (Faxon, 1885, pi.
7, f. 5), only here the flattened face of the inside is dilated, and
the shoulder is more prominent. In both cases, there are practi-
cally three tips to this appendage, two of which belong to the
inner part.

The description of this organ in C. mexicanus (Faxon, /. c, p.
50) agrees closely, but possibly the "small, procurved spine" at-
tributed to the external part belongs to the internal, and then there
would be complete agreement.

The double tip to the inner part, and the shoulder, which has a
very peculiar position, possibly give to these three species a more
isolated position within this subgenus, and might possibly justify
the creation of a separate subgenus, which then should stand at the
head of the genus. This would also agree well with the geographi-
cal distribution.

Aside from these more primitive species (C. digueti, cubensis,
mexicanus), the first section contains two others (C. simulans and
gal Unas), which mark the transition to the third section, from
which they differ only by the number of hooks of the male ; the
third section contains more advanced forms of the simulans '-type.
The second section is a peculiar side branch going off from the first
section, which has acquired burrowing habits; this is known posi-

100 ORTMANN — AFFINITIES OF CAMBARUS. [April i3,

tively of C. gracilis and of C. advena (Hyeme vitam degit subter-
ranean!. Aestate in fossis invenitur. Leconte).

The large number of species known in the third section makes a
further division desirable, which is easily made according to the
following characters :
i. Group : {spicu lifer) .

Outer part of sexual organs with two or three recurved teeth, with-
out prominent angle {shoulder) on anterior margin. Rostrum with
marginal teeth, acumen rather long. Areola wide, rarely narrow,
distinctly shorter than half of the anterior section of the carapace.
2. Group: {blandingi).

Outer part of sexual organs with three {rarely two) recurved
teeth, inner part with terminal spine directed obliquely outward. No
shoulder on anterior margin. Rostrum with marginal teeth, acumem
rather short. Areola narrow, generally distinctly lunger than half
of the anterior section of the carapace.
j. Group: {clarki).

Outer part of sexual organs with two compressed tubercles, inner
part straight, directed forwards. Anterior margin with a distinct
shoulder. Rostrum with marginal teeth, acumen rather short.
Areola very narrow, often obliterated in the middle, about half as
long as anterior section of carapace.
4. Group : {alleni) .

Outer part of sexual organs with one or tiuo teeth, often peculiarly
formed {compressed and plate-like), inner part straight or oblique.
No shoulder on anterior margin. Rostrum without marginal teeth
{at least in the adult stage). Areola moderately wide, about half as
long as anterior section of carapace.

There is no doubt, that the spicu lifer-group is the most primitive
of these, and that the others represent special modifications, each
developed in a different direction.

The following key for the identification of the species of the sub-
genus Cambarus is submitted ; it is claimed that this key represents
— as far as is possible in a " key " — the natural affinities. If adult
males of the first form are at hand, it should be possible, in every
case, to correctly identify the species.
1. Section of C. digue ti (see p. 98).

flj Sexual organs of male with only one terminal tooth on outer part, inner part
with two tips ; anterior margin with an angular projection (shoulder ) near
the tip {digueti-group).

1905.] ORTM ANN — AFFINITIES OF CAMBARUS. 101

bx Rostrum with marginal teeth.

cx Sexual organs of male with inner part not broadly dilated on inner
side, curved forward at apex ; shoulder small. Rostrum carinated
above. Carapace with lateral spines.

C. (Cambarus) digneti Bouv.
c2 Sexual organs of male with inner part greatly dilated, forming a
broad, flat, setose plate on inner side ; shoulder strongly developed.
Rostrum not carinated. Carapace without lateral spines.

C. (Cambarus) cubensis Er.
b2 Rostrum without marginal teeth, subplane above. Carapace without lat-
eral teeth. Inner part of sexual organs flattened within, but not greatly
dilated. C. (Cambarus) mexicanus Er.

?2 Sexual organs of male with two terminal teeth on outer part, one of which is
flat and disk-shaped, inner part with one terminal spine; without shoul-
der on anterior margin. Rostrum without marginal teeth (simulans-
group).
bx Terminal teeth of sexual organs oblique, both of about the same length
Acumen of rostrum longer. C. {Cambarus) simulans Fax.

b2 Terminal teeth of sexual organs straight, one much longer than the other.
Acumen of rostrum shorter. C. (Cambarus) gallinas Cock, and Port.
2. Section of C. gracilis (see p. 98).

?! Rostrum suddenly contracted into a short acumen. Sexual organs with two
teeth at end of outer part. Terminal spine of inner part straight, longer
than outer part.
bx Anterior margin of carapace forming a blunt suborbital angle.

C. ( Cambarus) gracilis Bund.
b2 Anterior margin of carapace not forming a suborbital angle.

C. (Cambartis) hagenianusYzx.
i2 Rostrum triangular, margins not suddenly contracted to form an acumen. Sex-
ual organs with only one compressed, triangular tooth at the end of outer
part. Inner part straight, not longer than outer.

C. (Cambarus) advena (Lee).

3. Section of C. blandingi (see p. 98).
1. Group of C. spiculifer (see p. 100).
(j Areola wide. Chelre rather broad. Two lateral spines on each side of the
carapace.
bx Chela; with large, remote tubercles. Margin of rostrum converging.
Outer part of sexual organs with two terminal teeth.

C. (Cambarus) spiculifer (Lee).
b2 Chel?e with small, crowded tubercles. Margins of rostrum subparallel-
Outer part of sexual organs with three terminal teeth.

C. (Cambarus) versutus Hag.
i2 Areola wide or narrow. Chelae generally narrower. One lateral spine on
each side of the carapace.
<$j Rostrum subplane above, ciliated. Areola wide. Outer part of sexual
organs with two terminal teeth, the inner part with terminal spine di-
rected outward. C. (Cambarus) pubescens Fax.

102 ORTM ANN — AFFINITIES OF CAMBARUS.

[April 13,

b2 Rostrum concave above, smooth. Areola narrower.

Cj Margins of rostrum subparallel. Outer part of sexual organs with two
terminal teeth, the inner part straight.

C. (Cambarus) angustatus (Lee).
c2 Margins of rostrum convergent. Outer part of sexual organs with
three terminal teeth, the inner part directed outward.

C. (Cambarus) lecontei Hag.

2. Group of C. blandingi (see p. 100).

ax Eyes rudimentary. Outer part of sexual organs with two terminal teeth.

C. ( Cambarus) acherontis Loennb.
a2 Eyes well developed. Outer part of sexual organs with three terminal teeth.
bx Sexual organs straight, terminal teeth well developed.

cx Sexual organs not excavated on outer side near distal end.

C. ( Cambarus) blandingi ( Harl. )
c2 Sexual organs excavated on outer side near distal end.

C. (Cambarus) hayi Fax.
b2 Sexual organs curved back distally, terminal teeth minute.

C. (Cambarus) J "alia x Hag.

3. Group of C. clarki (see p. 100).

ax Rostrum concave above, acumen slightly longer. Shoulder of sexual organs
slightly developed. C. (Cambarus) clarki Gir.

a2 Rostrum plane above, acumen shorter. Shoulder of sexual organs very promi-
nent. C. (Cambarus) troglodytes (Lee).

4. Group of C. alleni (see p. 100).

«j Outer part of sexual organs with one or two terminal teeth ; inner part not
longer than the outer. Hooks of fourth pereiopods of male not bituber-
culate.
bx Rostrum concave above. Outer part of sexual organs with one recurved
terminal tooth ; inner part with the terminal spine placed obliquely.

C. (Cambarus) evermanni Fax.
b2 Rostrum plane above. Outer part of sexual organs with terminal part
plate-like, covering the inner part, and with two very small teeth.
* cx Chelae bearded on inner margin. C. (Cambarus) barbatus Fax.

c2 Chelae not bearded on inner margin.

C. (Cambarus) wiegmanniYx. (?).
a.. Outer part of sexual organs forming at apex a broad, flattened plate, whose an-
terior margin is furnished with hairs and one strong seta, the posterior
margin of the plate produced anteriorly into a blunt process. Inner part
produced into an erect spine, which is much longer than the outer part.
Hooks of fourth pereiopods of male bituberculate.

C. (Cambarus) alleni Fax.

Note : The position of C. wiegtnanni is very doubtful, since the
male sexual organs are unknown. It has been placed with C. bar-
batus by Hagen and Faxon, but only the external resemblance to

i9°5-]

ORTM ANN — AFFINITIES OF CAMBARUS. 103

this species speaks for its position here. The geographical distribu-
tion, however, is entirely opposed to it, and I very strongly suspect
that it belongs somewhere else.

GEOGRAPHICAL DISTRIBUTION OF THE SUBGENUS CAMBARUS.

Taken as a whole, the subgenus Cambarus occupies a rather con-
tinuous area, with a possible interruption in northern Mexico: this
gap, however, may be due only to the incompleteness of our
knowledge. It covers Mexico, and a large part of the southern,
central and eastern United States, but leaves unoccupied the
mountainous region of the East ; it is lacking in the larger part of
Tennessee, in Kentucky, West Virginia, Pennsylvania, and north-
ward. The largest number of species is found in the southeastern
states : Mississippi, Alabama, Georgia, and this region represents
at present the center of frequency of the subgenus. From here it
extends, gradually declining, westward into Texas, northward up
the Mississippi valley, becoming quite scarce north of the State of
Missouri (only two species), and further it has populated the
Atlantic coast plain as far north as New Jersey (only one species
north of South Carolina).

Regarding the single sections, the distribution shows rather
peculiar features. The digueti-section is characterized by a marked
discontinuity : two species are found in Mexico, one in Cuba, and
two in New Mexico, Texas and Kansas. Since I consider this
section the most primitive of the genus, this discontinuity is highly
interesting, and tends to confirm this view. And further, this
peculiar distribution probably indicates the direction of the immi-
gration into the United States. The most primitive forms {C.
digiteti and mexicanus) are still preserved in the original home of
the genus, in Mexico, while two other, somewhat more advanced
species ( C. simulans and gailinas) occupy the higher plains lying
to the east of the Rocky Mountains in the southwestern United
States. These parts are largely formed by Cretaceous deposits,
and represent the first land-connection between western and eastern
North America after the Upper-Cretaceous separation. It is very
significant, that just these parts contain the most primitive forms
of the United States, and thus the distribution of the digueti-section
clearly indicates this old condition prevailing at the end of the

104 ORTMANN — AFFINITIES OF CAMBARUS. [April .3,

Cretaceous and the beginning of the Tertiary time, and also gives
a clue as to the direction of the migration : it did not go over the
lowlands of Texas, which are geologically younger, but over the
higher plains of the interior. (See Ortmann, 1902, pp. 282-285,
p. 388.)

The gracilis-section, which is a specialized type, arising from the
more primitive forms of the subgenus, forms in the distribution of
the species C. gracilis a direct continuation of this southwestern
range of the digueti-section : C. gracilis is found from eastern Kansas
through Missouri, to Illinois, Iowa, and southern Wisconsin. This
is in the same line of the migration marked by the distribution of
the species of the digueti-section, and plainly its continuation in a
northeastern direction. However, the two other species of the
gracilis-section, C. hageniamts and advena, are entirely isolated,
being found only far in the east, in the lowlands of Georgia and
South Carolina. Here again we have discontinuity, indicating old
age. I have no doubt, that these separated localities once were
connected, namely from Kansas and northern Texas over Arkansas
and across the Mississippi valley into Mississippi, and the northern,
higher parts of Alabama and Georgia, including probably Tennessee.

Thus I think that the most primitive forms of Cambarus occu-
pied, in the United States, first the Cretaceous plains of the south-
west, necessarily reaching in very early times the Ozark Mountains,
following the Ozark uplift into Illinois and beyond, and, on the
other hand, crossing the present Mississippi valley, and reaching
the southern end of the Appalachian system, and finally the sea
coast in Georgia and South Carolina. Representatives of the
primitive sections of the subgenus have now disappeared in the
Appalachian region, and this is very likely due to the fact, that,
as we shall see below, just in this region some other very vigorous
groups developed, which apparently suppressed those earlier forms.
In the southwestern extremity, where these new groups are rather
scarce or entirely lacking, there was a chance for the old types to
survive, and this may account for the presence of C. simulans and
gallinas in this region, while C. gracilis, which is found right in
the chief domain of the subgenus Faxonius, survived possibly on
account of its different habits. For similar reasons C. hagenianus
and advena may have survived at the extreme eastern seashore.

The third section of the subgenus Cambarus represents typically

,9o5.j ORTM ANN— AFFINITIES OF CAMBARUS. 105

the distribution of the whole subgenus, with the exception that it
is not found in the extreme west and in Mexico. (I disregard C.
wiegmani, since I do not believe that its position with this section
is correct.)

Here again we have peculiar facts of distribution. The more
primitive forms (spiculifer-group) are restricted to the states Georgia,
Alabama, and northwestern Florida. Thus they come into close
contact with the hypothetical old range of the more ancient types
of the subgenus in the southern Appalachians, and I believe that
they originated from an original stock of the digueti-section, that
immigrated into the lowlands south of the mountains, which became
dry land by degrees during Tertiary times. Here in these low-
lands, chiefly in Alabama and Georgia, is the center of origin of
the blandingi-section, which represents a secondary center for the
subgenus. The more primitive forms {spiculifer-group) still stick
to this center, while the more advanced forms have spread out from
here as follows.

The blandingi-group invaded ( C. fallax) northern Florida, and
spread out northeastwardly along the Atlantic coast plain ( C.
blandingi-typicus), and also it migrated westward aud northward,
up the Mississippi valley (C. hayi and blandingi acutus). The
clarki-group extended chiefly westward from northern Florida far
into Texas (C. clarki), and slightly eastward into South Carolina
(C. troglodytes, in South Carolina and Georgia). Finally, the
alleni-group occupied Florida : C. alleni, the most aberrant form,
goes farthest south here (Caloosahatchee River, Lee Co.). (The
other species, C. evermanni and barbatus, are known from scattered
localities in Georgia, western Florida, and Mississippi, and their
distribution needs further investigation ; C. wiegmanni from Mexico
possibly does not belong here. )

Thus the distribution of the subgenus Cambarus illustrates the
early history of the immigration of the genus into the United
States, and it also illustrates the later population of the southern
parts of the United States during Tertiary times by forms of the
blandingi-section. The latter prevail here, and hardly ever had
any competitors, and thus the southern states are at the present
time the center of the frequency of the whole subgenus. They are,
however, the center of origin only for the blandingi-section, while
the center of origin of the subgenus is to be sought in Mexico.

106 ORTMANN — AFFINITIES OF CAMBARUS. [APriii3,

The more advanced forms of the subgenus Cambarus generally
seem to prefer the ponds, lakes, and sluggish streams of the lowlands.
Subgenus: Cambarellus (Type: C. montezuma).

This subgenus corresponds to the fifth group of Faxon.

Faxon compares the male sexual organs with those of his fourth
group (= Faxonius), but I rather think that they are more closely
allied to those of his first and second group (= subgenus Cambarus^.
This latter relation, with the more primitive forms of the subgenus
Cambarus, is confirmed by other characters : carapace and areola
which are rather primitive, at least not very highly advanced ; the
rostrum has lateral teeth, which show a tendency to disappear ; the
chelae are very simple, more or less elongated and subcylindrical,
which is distinctly a primitive feature. The annulus ventralis of
the female seems to be very remarkable in C. montezuma (movable,
fixed only at the posterior end), and also in C. shufeldti (a trans-
verse curved ridge, the hind side of the ridge concave).

The three species of the subgenus may be distinguished as follows :

ax Sexual organs of male with straight terminal teeth. Carapace with lateral
spines. Rostrum with distinct marginal spines.

C. {Cambarellus) shufeldti Fax.
a2 Sexual organs of male with curved terminal teeth. Carapace without lateral
spines. Rostrum with or without marginal spines.

61 Carapace slender and subcylindrical. Rostrum longer and narrower,
with sharp marginal spines, and long, spiniform acumen.

C. {Cambarellus) chapalanus Fax.
/'., Carapace ovate. Rostrum shorter and wider, with or without marginal
spines, in the first case, the acumen is much shorter.

C. [Cambarellus) montezuma Sauss.

C. shufeldti is apparently more primitive than the other two
species. I have no doubt that Cambarellus took its origin from the
most primitive species of the subgenus Cambarus (digueti-group) ,
but developed in a peculiar direction, which is chiefly characterized
by the male sexual organs, and by the presence of hooks on the
second pereiopods, a condition that is found nowhere else in the
genus.

The distribution of this subgenus also suggests its antiquity, for it
is characterized by a strong discontinuity, C. shufeldti being found
in Louisiana, the other two species in Mexico. This geographical
discontinuity is accompanied by morphological discontinuity, the
former species differing very strongly from the two latter. While

I905-]

ORTMANN — AFFINITIES OF CAMBARUS. 107

C. chapalanus and montezumce still remain in the original home of
the genus, although they have changed a good deal, C. shufeldti
seems to be an early emigrant, which, however, has not much
changed. Further investigations in this subgenus are much needed.

Subgenus: Faxonius (Type: C. limosus)}

This subgenus corresponds to Faxon's fourth group, with the
addition of C. pellucidus. As regards the latter species, which
Faxon places with his first group, apparently chiefly on account of
the presence of hooks on the third and fourth pereiopods in the
male, it is easy to see that the sexual organs do not agree with the
blandingi-type. Faxon himself says (1885, p. 42), that they are
very simple, and generally admits that this species unites characters
of different groups. Looking at the figures of the sexual organs
given by Hagen (111. Cat. Mus. Harvard, 3, 1870, pi. 1, f. 68-71),
and Hay {P. U. S. Mus., 16, 1893, pi. 45, f. 11-14), I fail to
see any similarity to any of the species of the subgenus Cambarus,
but their shape approaches rather closely that of some species of
Faxon's fourth group, namely : C. limosus, indianansis and sloanei.
Indeed, in C. pellucidus this organ is different from any one of
these, but it agrees with them in the more or less straight and
simple form, with the outer and inner parts separated at the tips
for a short distance ; there is also no trace of a terminal truncation.
The rostrum and the chelae are rather primitive in C. pellucidus,
while carapace and areola are peculiar, which is possibly a char-
acter due to the subterranean life (see Faxon).

If we place C. pellucidus with the species of the fourth group
named above, it loses its isolated position also with reference to
the geographical distribution: it is found in a region (Kentucky
and southern Indiana), where at least two of the above species are
also found : C. indianansis and sloanei.

I think, that C. pellucidus is a rather primitive form, connecting
the subgenus Faxonius with the more primitive forms of Cambarus

1 Astacus limosus of Rafinesque has been considered by all authors (Girard,
Hagen, Faxon) as very probably identical with A. ajfinis of Say. Although
Rafinesque' s description is very poor, the locality given (" muddy banks of the
Delaware near Philadelphia") renders it absolutely certain that C. affinis was
intended. There is no other species on the banks of the Delaware but this, and it
is so abundant there, that it even attracts the attention of the casual observer.
Thus I do not see why the older name of Rafinesque should not be restored.

PROC. AMER. PHII.OS. SOC. XLIV. iSo. H. PRINTED JULY 28, I905.

108 ORTMANN — AFFINITIES OF CAMBARUS. [April i3,

(digueti-group) , and that the development of an additional pair of
hooks on the fourth pereiopods is a parallelism to the similar ten-
dency in the more highly advanced forms of the subgenus Cambarus
{blandingi group) : to the latter, C. pellucidus has no direct relation
at all.

With regard to all the rest of the species of this subgenus, I
agree with Faxon in thinking them to form a natural, genetically
connected group. Nevertheless there is much diversity within
this subgenus, and is chiefly indicated by the shape of the male
sexual organs. Faxon did not use the latter in arranging the
species of his fourth group, and thus his key (1885, p. 86) is, as
he admits himself, artificial to a degree. But I shall show here,
that according to the sexual organs we can divide the subgenus in
groups, which seem to be quite natural.

1 . Section : C. limosus.

Sexual organs short, rather thick up to near the tips, reaching to
the base of the third pair of pereiopods. Tips split for a short
distance, each tapering to a point. Hooks on third, or on third and
fourth pereiopods.

This is the most primitive section of the subgenus, and it is also
in other characters quite indifferent, and not highly specialized ;
and further, it appears a little heterogeneous. The rostrum is
quite uniform in shape, generally with marginal spines (except in
certain varieties of C. pellucidus) , with a rather long or a moderate
acumen. The areola is wide and of medium length (except C.
pellucidus) ; the chelae are comparatively narrow and without
remarkable features (except in C. harrisoui).

The annulus of the female shows the tendency to develop tuber-
cles upon its face ; these tubercles have a more or less central
position (limosus, indianensis), or a posterior (sloanei), or have
the shape of a transverse ridge (harrisoni), or form a "median
keel" (pellucidus).

2. Section : C. propinquus.

Sexual organs shorter or longer, not thick, deeply split at the tips,
tips slender, more or less straight, sometimes the onto- one slightly
curved, but never both tips curved in the same direction. Always
only third pereiopods 7vith hooks (barring freaks) .

The other characters are very uniform in this section. The
rostrum possesses with one exception ( C. tnedius), marginal spines,

1905.]

ORTMANN — AFFINITIES OF CAMBARUS. 109

and a rather long or moderate acumen. The carapace is of normal
shape, oval and depressed ; the areola uniformly rather wide, and
there is no tendency to become narrow. There !are, however,
some differences in length : generally, the areola is about half as
long as the anterior section of the carapace (inch rostrum); but in
certain species (erichsoni and forceps) it is slightly, and in one
species (spinosus) decidedly shorter, and in two others (nistieus
and medius) it is decidedly longer. The chelse in this section'are
also rather uniform, but not very primitive : they are more or less
broad and ovate. The fingers (in old males) generally are gaping
at the base, and in contact distally, and the movable finger pos-
sesses a peculiar S-shaped curve. The immovable finger is generally
not bearded at the base (a slight indication of a beard is seen in :
C. propinquus, obscurus, neglectus). In C. forceps, the fingers are
unusually and widely gaping, up to the tips. In C. medius the chelae
are unusually broadly ovate, and the movable finger has no S-curve.

The annulus of the female is flat, with a median depression and
raised margins. Very often the anterior margin is elevated into
tubercles, and in C. hylas the posterior margin is very prominent,
which is rather unusual in this section, and ought to be confirmed
by additional investigations.

This section contains ten species, which may be divided into two
groups.

1. Group: {propinquus).

Tips of sexual organs comparatively short, reachitig only to the
third rarely {in erichsonianus) to the second pereiopods, without or
with {obscurus) a shoulder on the anterior margin. Outer tip regu-
larly tapering from base to end.

2. Group.- {rusticus).

Tips of sexual organs long, reaching rarely only to the second, gen-
erally to. the first pereiopods, mostly with a shoulder on the anterior
margin. Outer tip not regularly tapering, but thin {setiform) from
base to end.

C. erichsonianus forms a transition between the two groups : the
sexual organs are rather long, but they lack a shoulder, and in shape
they resemble those of C. propinquus.

3. Section : C. virilis.

Sexual organs generally quite long {rarely rather stout), reaching
about to the second pereiopods, deeply split at the tips, tips slender

110 ORTMANN— AFFINITIES OF CAMBARUS.

[April i3)

{rarely shorter} and more or less strongly curved backward, both in
the same direction. Always only the third pereiopods with hooks
{barring freaks').

The shape of the sexual organs is quite uniform in this section,
and they do not vary much in the different species, with one excep-
tion : C. difficilis. Here they are remarkably short and stout,
reaching only to the third pereiopods. But we cannot separate
this species on this account from the section, since in other charac-
ters it is closely allied to C. palmeri.

This section closely approaches the propinquus type, especially
that represented by C. rusticus, in fact, the curvature of the tips of
the sexual organs is the only important differential character. Be-
sides, however, there is in no case a shoulder developed here,
which is so frequently seen in the propinquus-section.

In other characters this section is more variable than the propin-
quus-section, and this is most evident in the width and length of
the areola. The chelae are built according to the type of the
propinquus-section, but a remarkable character is the presence of a
dense tuft of hairs {beard) at the base of the immovable finger.
This beard is absent in C. compressus only. In two species, C.
alabamensis and compressus, the chelae are very broad, and excep-
tionally smooth.

The annulus of the female is depressed in the middle, with
raised margins, similar to that of some species of the propinquus-
section (virilis, longidigitus) . In other cases it is elevated posteri-
orly, and the anterior part is depressed ; it is never elevated ante-
riorly, as is generally the case in the propinquus-section. (In some
species, alabamensis and mississippiensis, the description of the
annulus is inadequate).

The eleven species of this section are easily arranged into three
groups according to the areola.
i. Group: {alabamensis).

Areola wide a fid short.

2. Group : (virilis').

Areola narrow, of medium length.

3. Group.- {palmeri).

Areola obliterated in the middle, of medium length.

4. Section : C. lancifer.

Sexual organs very peculiar ; short, and with slightly curved tips,
the outer tip remarkably compressed.

>9°5-l

ORTMANN — AFFINITIES OF CAMBARUS. Ill

This section is formed to receive an isolated species, the position
of which seems quite uncertain. There is a remote resemblance of
the sexual organs to those of C. difficilis of the third section of this
subgenus, and in other characters there are resemblances to C. mis-
sissippiensis, namely in the lack of marginal spines of the rostrum,
and in the obliteration of the areola. The annulus of the female
agrees with C. palmeri in being depressed in front, and prominent
and tuberculated behind : but a similar shape is found in the sub-
genus Bartonius. On the other hand, also the male sexual organs
can be compared with Bartonius, although they are by no means
identical with the very uniform type seen in the latter subgenus.
The chelae, according to the description, are very peculiar, namely
long and subcylindrical, the palm with subparallel margins : this is
entirely unlike anything that is seen in the virilis-section of the
present subgenus, and rather stamps this species a primitive one.
Then, again, this species presents in the elongate rostrum and
antennal scale very unusual features.

Thus it is hard to form a positive opinion about its position. I
should not hesitate to place it with the palmeri-group of the virilis-
section, if it was not for the primitive character of the chelae. Ac-
cording to the latter, and possibly also according to the sexual
organs, we might place it at the beginning of the subgenus, as a
peculiarly developed primitive form, but It also may be the most
highly specialized form of the subgenus. The distribution (Mis-
sissippi and northeastern Arkansas) would fit either assumption.

KEY TO THE SPECIES OF THE SUBGENUS FAXONIUS.

i. Section of C. limosus (see p. 108).

«j Generally third and fourth pereiopods with hooks in the male. Carapace sub-
cylindrical. Areola wide and long. Chela; subcylindrical. Eyes rudi-
mentary. C. (Faxonhts) pellucidvs (Tellk. ).
rtj Only third pereiopods with hooks in the male. Carapace ovate, depressed.
Areola rather wide, of medium length (about half as long as anterior
section of carapace). Chelae not subcylindrical, compressed, and more
or less ovate. Eyes well developed.
bx Sexual organs thick, swollen in the middle, tips short and stout, both
slightly curved in the same direction. C. (Faxonius) harrisoni Fax.
b2 Sexual organs short, thick, but not swollen, straight. Tips divergent.
i\ Sides of carapace with one spine behind the cervical groove.

dx Sexual organs with tips not crossed, the outer directed outward,
the inner inward. C. (Faxonzus) sloanei Bund.

112 ORTMANN — AFFINITIES OF CAMBARUS. [April 13,

d2 Sexual organs with tips crossed, the outer directed inward, the
inner outward. C. (Faxonius) indianensis Hay.

c2 Sides of carapace spinose, several spines behind cervical groove, and
spines on the hepatical region. Tips of sexual organs crossed.

C. (Faxonius) limoszts (Raf. ).

2. Section of C. propinquus (see p. 108).

1. Group of C. propinquus (see p. 109).

<?j Sexual organs reaching to the third pereiopods, with or without shoulder.

bx Rostrum with or without median keel. Sexual organs without shoulder

on anterior margin. C. {Faxonius) propinquus Gir.

b2 Rostrum without median keel. Sexual organs with shoulder on anterior

margin. C. (Faxonius) obscurus Hag.

a2 Sexual organs reaching to the second pereiopods, without shoulder.

C. (Faxonius) erichsonianus Fax.

2. Group of C. rusticus (see p. 109).

ax Rostrum with marginal spines. Carapace with a lateral spine.

bx Margins of rostrum concave. Sexual organs reaching to the second perei-
opods
cx Tip and marginal spines of rostrum bent upward. Fingers of chela
gaping only at base. C. (Faxonius) rusticus Gir.

c2 Tip and marginal spines of rostrum not bent upward. Fingers of
chela gaping to the tips. C. (Faxonius) forceps Fax.

b2 Margins of rostrum straight, generally subparallel. Sexual organs reach-
ing to the first pereiopods.
cx Rostrum with distinct median keel. Sexual organs without shoulder.

C. ( Faxonius) neglectus Fax.
c2 Rostrum without median keel. Sexual organs with more or less dis-
tinct shoulder.
d1 Areola shorter than half of the anterior section of carapace.

' C. (Faxonius) spinosus Bund.
d2 Areola half as long as the anterior section of carapace.
<?j Margins of rostrum almost parallel,

C. (Faxonius) putnami Fax.
e2 Margins of rostrum distinctly convergent.

C. (Faxonius) /ty /as Fax.
a2 Rostrum without marginal spines. Carapace without lateral spines.

C. (Faxonius) medius Fax.

3. Section of C. virilis (see p. 109).

1. Group of C. alabanicnsis (seep. no).
ax Areola very short. Carapace not compressed.

C. (Faxonius) alabamensis Fax.
a2 Areola a little longer. Carapace compressed.

C. (Faxonius) compressus Fax.

2. Group of C. virilis (see p. no).

tfj Margins of rostrum concave, acumen moderately long, together with marginal
spines bent upward. G. (Faxonius) meeki Fax.

igo3] ORTM ANN — AFFINITIES OF CAMBARUS. 113

a2 Margins of rostrum straight, parallel or convergent. Marginal spines and
acumen not bent upward.
bx Acumen of rostrum long, marginal spines sharp, margins parallel. Fingers
of chela long. C. (Faxonius) longi digitus Fax.

b„ Acumen of rostrum short, marginal spines small or absent, margins more
. or less convergent,
fj Acumen of rostrum not considerably shorter than width of rostrum at
base ; marginal spines small, but present ; margins slightly
convergent ; upper surface slightly concave. lingers of chelre
not remarkably long and not emarginate at base.
dx Sexual organs longer, slightly curved.

C. (Faxonius) virilis Hag.
d2 Sexual organs shorter, more strongly curved.

e1 Immovable finger bearded at base, chela for the rest with-
out hairs. C. [Faxonius) uais Fax.
e2 Immovable finger bearded at base, chela pilose.

C. (Faxonius) pilosus Hay.

if2 Acumen of rostrum considerably shorter than width of rostrum at base ;

marginal spines generally wanting (rarely present and small ) ; upper

surface deeply concave ; margins strongly convergent. Movable

finger of chela with a deep emargination at base of inner margin.

C. (Faxoniits) immunis Hag.

3. Group of C. palmeri (see p. no).
«j Rostrum with marginal spines.

bx Sexual organs long. C. (Faxonius) palmeri Fax.

b2 Sexual organs remarkably short. C. (Faxouius) dijjicilis Fax.

a2 Rostrum without marginal spines. C. (Faxonius) mississippiensis Fax.

4. Section of C. lane if er (see p. no).

Rostrum very long, without marginal spines. Antennal scale very long. Areola
obliterated in the middle. Chelae long, subcylindrical.

C. (Faxonius) lancifer Hag.

GEOGRAPHICAL DISTRIBUTION OF THE SUBGENUS FAXONIUS.

The area occupied by this subgenus is almost entirely continuous ;
it extends over all of the central parts of the United States, from
northern Texas to Lake Winnipeg in Canada, and from Kansas to
the Appalachian Mountains. To the south, it hardly encroaches
upon the domain of the subgenus Cambarus, being found only in
the northern parts of Alabama and Georgia. To the North, it
reaches the Great Lakes, and follows down the St. Lawrence valley.
Eastward, the Allegheny Mountains apparently form a boundary,
but at two places it has crossed these mountains, namely in the
north, where C. limosus is found in the lowlands and rivers of

114 ORTMANN — AFFINITIES OF CAMBARUS. [April i3j

Virginia, Maryland, Pennsylvania and New Jersey ; and in the
south, where C. spinosus and erichsonianus cross over from the
Tennessee River drainage into that of the gulf and the Atlantic
Ocean in Alabama, Georgia, South and North Carolina. These
latter cases are continuous, the same species being found in both
drainages, while in the former case discontinuity is implied, C.
limosus being cut off and isolated from the rest of the range of the
subgenus.

Generally speaking, this subgenus seems to belong to the
great rivers of the interior basin, its center lying about in the
region where the rivers Missouri, Mississippi, and Ohio come
together, that is to say, in the states of Mississippi, Kentucky,
southern Illinois, and southern Indiana. From this center it
spreads out in the directions of these rivers and tributaries, chiefly
toward the North and Northeast. However, the area remained not
restricted to the Mississippi drainage, but crossed the divides into
other systems in the following cases : From the Tennessee River
two species {spinosus, erichsonianus} have crossed over into the
Gulf and Atlantic drainages, and from the upper Ohio drainage
another species (limosus) has crossed over into the Chesapeake and
Delaware Bay drainage. Another species (mississippiensis) is found
in the Gulf drainage (outside of that of the Mississippi River) in
the state of Mississippi. In the North the area largely extends
into the drainages of the great lakes, and even into that of Hudson
Bay (through the Red River of the North and Winnipeg Lake).

Studying the distribution of the single sections, the following is
to be remarked. The most primitive section (that of C. /imosus)
is marked by discontinuity : C. limosus being found on the Atlantic
coast plain, C. pellucidus, indianensis, sloanei in Kentucky and
southern Indiana, C. harrisoni in Missouri. This discontinuity,
chiefly the isolation of C. limosus, is accompanied by morpholog-
ical isolation, the latter species possessing in its spinosity a charac-
ter, that only recurs in the allied, but otherwise peculiar species,
C. pellucidus. This latter species, as well as C. sloanei, indi-
anensis and harrisoni, undoubtedly are the last remnants of
the primitive stock of the subgenus in its original home, i. c,
in the central basin formed by the three great rivers. Thus the
geographical distribution of the limosus-section confirms the char-
acter of antiquity : most of the species remain in the original

I9oS.] ORTMANN — AFFINITIES OF CAMBARUS. 115

home, while C. tiniosus apparently is an early emigrant that has
crossed over into the Atlantic drainage, and has been entirely
cut off from the connection with the original stock. At present,
I am not prepared to say which was the way by which C. limosus
reached its present habitat.

The section of C. propinquus contains quite a number of species :
studying their distribution, we see that the distributional areas of
the two groups into which this section is divided correspond to the
main ranges of two species, while the other species seem to be
rather local forms of these. The typical form of the propinquus-
group, C. propinquus, occupies a continuous range that belongs in
part to the Mississippi drainage (Iowa, Illinois, Minnesota), in an-
other part to the Ohio drainage (in Indiana), and for the rest to
the Lakes and St. Lawrence drainage (in Michigan, Ohio, Penn-
sylvania, New York and Canada). Compared with C. rusticus,
this range is more northern and northeastern, and it is remarkable,
that there is hardly a locality known for the typical C. propinquus,
that lies south of the Terminal Moraine of the Wisconsin ice
sheet. C. obscurus is found at the eastern edge of the range of C.
propinquus, namely in the upper Ohio drainage in western Pennsyl-
vania and western New York (See Ortmann, Ann. Carnegie A/us.,
v. 3, 1905, p. 387-406), and seems to be the representative form
of C. propinquus, in this region.

C. rusticus, the typical species of the other group of this section,
has a wide range over the central basin, from Ohio, Indiana, and
Kentucky to Iowa, Missouri, and Tennessee. With reference to
C. propinquus it is more southern and western, although it extends,
in Ohio, far northward, and is found in the lake drainage in Michi-
gan and Wisconsin. (The investigation of the distribution 01
these two species, rusticus and propinquus, in Ohio, Indiana, Illi-
nois, Michigan, and Wisconsin will certainly be very interesting.)
Associated with C. rusticus in the same group are six other species :
all of these are rather local, and all are found at or near the edge
of the range of C. rusticus. C. forceps, spinosus, and putnami are
found at the southeastern edge, namely in the Cumberland and
Tennessee river drainages in Kentucky, Tennessee, and northern
Alabama. One of these species (spinosus) has crossed over into
the Gulf and Atlantic drainages in northern Georgia, South and
North Carolina. (This is an additional case throwing light upon

116 QRTMANN — AFFINITIES OF CAMBARUS. [April i3,

the changes of the drainage systems in the southern Appalachians,
see: Simpson, Science, 12, 1890, p. 133, and chiefly Adams,
Americ. Natural., 35, 1901, p. 844 ff. ; where on p. 849 three
species of Cambarus are mentioned (C. spinosus, extraneus, and
erichsonianus) that belong into this category). The species C.
neglectus, hylas, and medius belong to the southwestern and western
edge of the range of rusticus, and are found in Missouri, Arkansas,
Texas, Kansas, and Iowa. Thus it is evident, that the six species
morphologically allied with C. rusticus in the same group, express
this relation also in their distribution, being apparently locally
modified forms of the rusticus-type, and being naturally found
just where we ought to expect them, namely at the edge of the
range of this rusticus-type.

C. erichsonianus seems to be abnormal : morphologically we have
placed it with C. propinquus, but its range is far remote from it in
eastern Tennessee and central Alabama (in both the Tennessee
and Alabama river drainages). But, as we have seen above, its
position is a little uncertain, it resembling C. rusticus and its allies
to a degree, and the distribution suggests the same : it clearly
agrees better in this respect with C. forceps and spinosus, and it
would thus become another local form of the rusticus-type. Fur-
ther investigations on this question should be made.

The third section, that of C. virilis, has been divided into three
groups. The virilis-group agrees somewhat with the rusticus-group
in its range, belonging to the central basin, only being a little
more western, and considerably more northern : it is hardly found
in the drainage of the Ohio, but it is very abundant in that of the
Mississippi and Missouri, and crosses over not only into the lake
drainage, but also into that of Hudson Bay (Winnipeg Lake).
The typical species of the group {virilis) occupies almost all of
this range, while four other species associated with it (nieeki, longi-
digitus, nais, f>ilosus) apparently are local forms of it, being found
at or near the southwestern extremity of the range of C. virilis
in Arkansas and Kansas. C. i/nmunis is a peculiar type of the
virilis-group, and its range coincides with the southern part of the
range of C. virilis (Kansas, Missouri, Iowa, Illinois, Indiana,
Ohio) : this is interesting in so far as this occupation of the same
territory by two closely allied species is rendered possible as it
seems in this case, by the different habits : as far as we know, C.

i9o5j ORTMANN— AFFINITIES OF CAMBARUS. 117

immunis inhabits the (often temporary) shallow, stagnant ponds
and roadside ditches of the western prairies, and is a burrower,
while C. virilis prefers rocky places in running streams. (See
Harris, Americ. Natural., 35, 1901, f. 187 ff., and Kansas Univ.
Quart., 9, 1900, pp. 268 and 270).

Of the other two groups of the third section, that of C. alaba-
mensis contains only two species, which are very local, being found
only in northern Alabama. Both are rather primitive, and appar-
ently are the last remnants in the Tennessee drainage of a once
more widely distributed stock. The difficilis -group seems to rep-
resent a southern extension of the subgenus Faxonius : the species
are found in western Tennessee, Missouri, Arkansas, Indian Ter-
ritory, northeastern Texas and Mississippi, all in the drainage of
the lower Mississippi (below Cairo), only C. mississippiensis be-
longs to the Tombigbee river drainage.

C. lancifer would agree in its range (Mississippi and Arkansas)
with this latter group.

The species of this subgenus, generally, are river-species, and
prefer the large rivers of the great central basin. Some species
have become lake-forms ( C. propinquus, for instance), and others
ascend the rivers into the smaller streams (chiefly so in the Tennes-
see and upper Ohio drainages), but they rarely inhabit true moun-
tain streams.

Further investigation of the distribution of this subgenus should
pay particular attention to the ways by which several species have
crossed the divides of the Hudson Bay, Great Lakes, and Atlantic
coast plain drainage systems. It is very likely that wandering of
the divides has played here an important part.
Subgenus: Bartonius (Type : C. bartoni).

This subgenus, which corresponds to the third group of Faxon,
is a very natural one, and, in my opinion, contains the most mod-
ern and most highly specialized forms in those that have acquired
burrowing habits (diogenes-section) . There are, however, other
species, which are rather primitive, as indicated by certain char-
acters.

The length of the areola, in this subgenus, is rather variable : in
the extraneus-section it is shortest, about half as long as the anterior
section of the carapace, and it is even shorter than that in C. acu-
minatus. In all other species it is considerably longer. The an-

118 ORTMANN— AFFINITIES OF CAMBARUS.

[April 13,

nulus of the female is, corresponding to the uniformity of the male
organs, also very uniform, and is characterized by its posterior ele-
vation. Aside from the length and width of the areola, the shape
of the chelae, the presence or absence of marginal spines of the
rostrum, and the shape of the carapace serve to distinguish the more
primitive forms from the more highly developed, and furnish a
division of the subgenus into sections as follows :

1. Section : C. hamitlatus.

Carapace subcylindrical. Rostrum with or without marginal
spines. Chelce long, subcylindrical. Areola rather long. Eyes
rudimentary.

Only two species, C. hamulatus and setosus, belong here, both
blind cave-forms. They do not seem to be closely related to one
another, since they differ in very important characters. The sub-
cylindrical shape of the chelae, however, indicates, that both are
rather primitive, and have become separated from the primitive
stock of this subgenus very early, and probably independently.
The shape of the carapace, the long areola, and the rudimentary
eyes are very likely due to parallel development, brought about by
the similar conditions under which these species are found. (See
Faxon, Pr. U. S. Mus., v. 12, 1890, p. 628).

2. Section : C. extrancus.

Carapace more or less ovate, depressed, with lateral spines behind
cervical groove. Chela not very elongated, depressed, and rather
broad, but a little more elongated than in the following sections.
Areola more or less wide, of medium length, about half as long as
anterior section of carapace, sometimes slightly shorter, rarely, in C.
cornutus, the areola is rather long. Eyes well developed.

Two of the species belonging here ( C. extraneus and jordam)
are typical, and are unquestionably the most primitive forms of the
subgenus, as is shown by the shape of the carapace, the rostrum,
and chelae, at least as compared with the following sections. The
third species, C. cornutus, stands by itself, and is a rather aberrant
form, peculiar on account of its antennae, which have a large, com-
pressed flagellum, ciliated on inner margin. Also the spines of the
rostrum (upturned) are peculiar. In the long areola, it is rather
advanced. It seems to be a peculiar local form, developed out of
the primitive stock now represented by C. extraneus and jordani,
and we may safely leave it with this section, since the only alter-
native would be to create for it a separate section.

i9oS.] ORTM ANN — AFFINITIES OF CAMBARUS. 119

3. Section : C. bartoni.

Carapace ovate, depressed, with or mostly without lateral spines.
Rostrum without marginal spines. Chela comparatively short and
broad, depressed, ovate. Areola wide or narrow, generally distinctly
longer than half of the anterior section of the carapace, only in one
case ( C. acuminatus) slightly shorter than half of the anterior section.
Eyes well developed.

The four species belonging here are all closely allied to one
another. Their chief differences are furnished by the shape of the
rostrum, width and length of areola, and shape of chelae : but all
are built according to the same plan.

4. Section; C. diogenes.

Carapace ovate, compressed, without lateral spines. Rostrum
without marginal spines. Cheloz short and broad, depressed, ovate.
Areola very narrow or obliterated in the middle, always distinctly
longer than half of the anterior section of the carapace. Eyes well
developed.

The five species belonging into this section also form a very
natural group. They are connected with the bartoni-section through
C. latimanus (chiefly its var. striatus Hay). The peculiar, com-
pressed shape of the carapace (and possibly other characters, as
shape of rostrum, narrow areola, shape of chelae) seems to be
closely connected with the habits : all these species (it has not
been reported for C. uhleri but it is likely also the case with this
one) are burrowing species and so-called chimney-builders. This
habit begins to appear in the bartoni-section : C. bartoni often, but
not always, makes burrows and chimneys, apparently forced to do
so, when the water supply of the small mountain streams, in which
it lives, begins to run short in dry seasons. With the species of
this group, this habit becomes firmly established, and they never
live without making burrows, having abandoned the streams and
brooks, and taken to swampy and springy places, generally to the
groundwater, where it is found at a short distance below the surface.

The species of this section are distinguishable by the width of
the areola, shape of rostrum, shape of the chelae, and in some cases
by peculiar colors. I believe that it is the most highly specialized
group of the whole genus, as is indicated partly by the burrowing
habits, no doubt an extreme adaptation, and, in one species (C.
uhleri), by the adaptation to brackish and salt-water, which is
found in no other case in the genus.

120 ORTMANN — AFFINITIES OF CAMBARUS. [April 13,

KEY TO THE SPECIES OF THE SUBGENUS BARTONIUS.

i. Section of C. hamulatus (see p. 118).

ax Rostrum with marginal spines. Areola wide.

C. (Bartonius) hamulatus (Cope and Pack.).

a2 Rostrum without marginal spines (rarely with spines in the young). Areola

narrow. C. (Bartonius) setosus Fax.

2. Section of C. extraneus (see p. 118).

tfj Antennae with normal flagellum.

bx Rostrum concave above. Areola rather wide.

C. (Bartonius) extranius Hag.

b2 Rostrum flat above. Areola narrower. C. (Bartonius') jordani Fax.

a2 Antennae with very long, compressed flagellum, which is ciliated on the inner

side. C. (Bartonius) cornutus Fax.

3. Section of C. barton i (see p. 119).

flj Rostrum long, tapering from base to tip. Areola very wide and short, a little
shorter than half of the anterior section of carapace. Carapace with lateral
spines. C. (Bartonius) acuminatus Fax.

a2 Rostrum shorter, suddenly contracted to a short acumen. Areola moderately
wide or narrow, distinctly longer than half of the anterior section of cara-
pace. Carapace with or without lateral spines.

bx Areola rather wide. Chelae smooth, punctate, inner margin of palm with
one or two rows of tubercles.
cx Fingers of chelae broad, slightly gaping at base, not bearded.

C. (Bartonius) bartoni (¥.).
c2 Fingers of chelae subcylindrical, widely gaping at base, the outer one
bearded at base. C. {Bartonius) longulus Gir.

b2 Areola narrower. Chelae rough or tuberculated.

C. (Bartonius) latimanus (Lee).

4. Section of C. diogenes (see p. 119).

«! Areola very narrow, but not obliterated. Color very striking.
bl Rostrum broad. Outer margin of hand serrate. Color red.

C. (Bartonius) carolinus Er.
b7 Rostrum narrower. Outer margin of hand not serrate. Color blue.

C. (Bartonius) monongalensis Ortm.
a2 Areola obliterated in the middle. Color dull, greenish or brownish.
bx Rostrum concave above.

cx Fingers of chelae not remarkably flattened, the inner one without dis-
tinct excision at base, the outer one not bearded,

C. (Bartonius) diogenes Gir.

cs Fingers of chelae flattened, the inner one with distinct excision at

base, the outer one bearded. C. (Bartonius) argillicola Fax.

/>., Rostrum flat above. C. (Bartonius) uhleri Fax.

i9o5.j ORTMANN — AFFINITIES OF CAMBARUS. 121

GEOGRAPHICAL DISTRIBUTION OF THE SUBGENUS BARTONIUS.

This subgenus is characteristic for the mountainous regions of
the east of the United States, that is to say, for the Appalachian
mountains, but the more highly developed, burrowing species have
in part descended from the mountains, and spread largely over the
central portions of this country. The greatest number of species
is found in the southern extremity of the Appalachian system, and
there is no question that we have to regard this as the center of
origin of the subgenus.

The two cave forms of the first section are widely separated from
each other. This indicates, on the one hand, that they are not
very closely allied, and, on the other hand, the discontinuity thus
displayed again indicates antiquity. The one, C. hamulatus, is
found in a cave in eastern Tennessee, that is to say, right in the
center of origin of the subgenus, while the other one, C. setosus,
comes from a cave in Jasper Co., Missouri (in the Ozark region).
This is very remarkable, and very likely indicates, that the center
of origin of the subgenus possibly includes the Ozark Mountains,
west of the Mississippi : this is further suggested by the reported
presence of C. carolinus in the northeastern part of Indian Terri-
tory, not far from the locality of C. setosus (see below). Conse-
quently, we are to regard C. setosus as the last remnant of the
primitive forms of the subgenus surviving in the western extremity
of the original home.

We have regarded, morphologically, the second section of the
subgenus as the most primitive group of it : this view is supported
by the geographical distribution. C. extraneus is known from
northern Alabama, northern Georgia, Tennessee, and Kentucky
(see below, p. 134) ; C. jordani is found in northern Georgia;
and C. cornutus in Kentucky (locally, only in Edmonson Co.).
Thus all the localities are in or near the old center of origin of the
subgenus. The presence of C. extraneus in the Cumberland and
Tennessee river drainages, as well as in the Alabama river drain-
age indicates an old drainage feature, namely the Appalachian
river (see above, p. 116).

The third section presents very interesting conditions, such as we
have noticed in several groups of the subgenus Faxonius. Here
we have apparently one widely distributed, typical form, C. bartoni :
this is found all along the Appalachian mountains and extends very

122 ORTMANN — AFFINITIES OF CAMBARUS.

[April 13,

far to the northeast. This species has followed, in its dispersal,
chiefly the direction of the strike of this mountain chain, and
reaches now from Tennessee to Maine and New Brunswick. East-
ward, it hardly decends to the Atlantic plain, at any rate it does
not spread over it, and westward it goes as far as Indiana, always
preferring smaller streams in mountainous or hilly regions.

C. bartoni possesses several marked varieties, chiefly at the south-
ern and southwestern extremity of its range, in Kentucky, Ten-
nessee and northern Georgia ; one variety (robustus) seems to
follow the northwestern edge of the range of the main species,
from Ohio through northwestern Pennsylvania to western New
York (and in Canada). This variety has also been reported from
Maryland and Virginia, but I doubt that this is actually the same
thing (see below, p. 135).

Besides, there are three other species in this section, which are
closely allied to C. bartoni. One of them, C. acuniinatus, is found
in North and South Carolina, at the southeastern edge of the range
of C. bartoni ; the second, C. latimanus, fringes the southern and
southwestern extremity of the area of C. bartoni in South Carolina,
northern Georgia, northern Alabama, and central Tennessee ; and
the third, C. longulus, is apparently a form belonging to the high
mountains, being found in the middle of the southern part of the
main range of C. bartoni along the highest mountain chains of
North Carolina, Tennessee, Virginia, and West Virginia. Thus it
is beyond question, that we can regard these three species as local
forms of C. bartoni, the one belonging to the high mountains,
another being its southeastern, the third its southern and south-
western representative.

While the first and second sections characterize the earlier stage
of the distribution of the subgenus, the third section expresses its
advance and dispersal over the eastern mountain system of the
United States.

Finally, the fourth section (of C. diogenes) offers remarkable
conditions. Two of the species, belonging here ( C. carolinus and
monongalensis') are evidently a little more primitive than the rest.
C. carolinus seems to possess a wide range within the Appalachian
system. It is a true mountain form, and is found from northern
South Carolina to southern Pennsylvania, thus representing the
same direction of migration as C. bartoni. from southwest to north-

iyoS.] ORTM ANN — AFFINITIES < >F CAMBARUS. 123

east, parallel to the strike of the mountains. This species, how-
ever, has also been reported from Indian Territory (Ozark region).
This locality is very strange, and at present is not connected with
the main range, no localities being known in Missouri, Arkansas
or the larger part of Tennessee (except the eastern extremity).
But it is possible that a connection exists here, and if this should
be so, this would indicate, as has been said above (p. 121) that the
Ozark region is to be included in the original home of the sub-
genus. C. monongalensis apparently is a representative form of C.
caro Units in southwestern Pennsylvania.

The most puzzling distribution is offered by the remaining three
species, of which C. diogenes is the most widely distributed. This
species has an eastern and a western range on both sides of the
Allegheny Mountains. Apparently it has descended from the
mountains, that is to say, represents a more highly specialized
branch of the original mountain-loving chimney-builders. It has
descended into the Atlantic coast plain on the one side, and is
found from New Jersey to North Carolina (Cape Fear). On the
other side, it has descended westward, and is found from south-
western Pennsylvania over all the states north of the Ohio (also in
Kentucky) as far north as Minnesota and Wisconsin, westward to
Iowa (also reported from southwestern Wyoming and Colorado),
Kansas, and southward to Louisiana. This immense distribution
represents possibly the widest known range of any of the species
of crayfishes of the United States. The question remains open,
whether the eastern and western range of C. diogenes is actually
connected across the mountains.

Of the other two species, C. ulilcri clearly is a local form of C.
diogenes, inhabiting the sea coast (brackish and salt marshes) in
Maryland. C. argillicola is morphologically very closely allied to
C. diogenes, and might be regarded, at least in Ohio, Michigan and
Canada, as a local form developed at the northern edge of the range
of C. diogenes. But the fact that C. argillicola is also found in cen-
tral and southern Indiana, in southern Illinois, and that it has been
reported from Mississippi and southern Texas (Victoria and Bra-
zoria), does not render this assumption probable : further investi-
gations of the range of these two species {diogenes and argillicola)
in the south and west are desirable, before their mutual geographic
relation can be ascertained.

PROC AMER. PH1LOS. SOC. XLIV. iSo. I. PRINTED JULY 29, I905.

124 ORTMANN — AFFINITIES OF CAMBARUS. [April t3j

Thus the burrowing species of the diogenes-section of the subgenus,
while conforming in part to the original habit of living in the moun-
tains, have in another part abandoned their original mountain home,
and largely spread over the plains. That this was possible is no
doubt due to their peculiar way of living. Aside from C. gracilis of
the subgenus Cambarus, which is also a burrowing form, and occu-
pies a certain part of the western plains, there are no other forms
in the central basin that have acquired this habit, and thus C. diog-
^nes did not find any competition, and was able to occupy a large
territory. That C. diogenes is a very vigorous and flourishing form
is also demonstrated by the fact that it attains, chiefly on the wes-
tern plains, a considerable size.

GENERAL CONSIDERATIONS AND CONCLUSIONS.

We have divided the genus Cambarus into four subgenera :
Cambarus, Cambarellus, Faxonius, Bartonius. Cambarus orig-
inated in Mexico, and immigrated, probably at the beginning of
the Tertiary, into the southwestern and southern United States,
originally occupying only the southwestern Cretaceous plain, the
Ozark Mountains, and the southern extremity of the Appalachian
System. A side branch, Cambarellus, has also its center in Mex-
ico, and spread, possibly along the sea coast, to Louisiana. In the
central and southeastern parts of the United States three new cen-
ters developed. The one is a secondary center for the subgenus
Cambarus, and lies at the foot of the Appalachian Mountains in the
lowlands of Alabama and Georgia. Here the more advanced forms
of this subgenus took their origin, and spread all over the Atlantic
and Gulf coast plain, and further up the Mississippi valley. These
are species inhabiting chiefly ponds, lakes, and sluggish streams of
the lowlands. Another subgenus, Faxonius, developed in the
central basin of the three great rivers, spreading over almost all of
the Mississippi drainage, and crossing over into the Hudson Bay,
Great Lakes, and even into the Atlantic drainages, probably by the
aid of shifting divides. The species belonging to this subgenus
are chiefly true river species. Finally, a fourth subgenus, Barton-
ius, developed in the mountainous region of the southern Appalach-
ians, probably including also the Ozark region, and from here it
spread chiefly over the Appalachian chain in a northeasterly direc-
tion as far as New Brunswick. Most of the species belonging

i9°5-]

ORTMANN — AFFINITIES OF CAMBARUS. 125

here are inhabitants of smaller mountain streams and brooks. A
peculiar group separated from these, the section of C. diogenes,
which acquired burrowing habits, and is originally also a mountain
loving group, but began to descend into the lowlands. Finding no
competition here, on account of its peculiar mode of life, it had
a chance to spread over a large area.

The centers for the more highly advanced forms of the subgenus
Cambarus, and for the subgenera Faxonius and Bartonius, appa-
rently form physiographically differentiated parts of one larger cen-
ter, situated in the southeast of the United States, clearly corre-
sponding to the southeastern center of dispersal of Adams (Bio-
logical Bulletin, 3, 1902, p. 115 ff. )l Adams discusses this center
chiefly with reference to the glacial and postglacial time, but it ex-
isted, no doubt, also during the Tertiary, and the development of
the different branches of Cambarus falls, in my opinion, chiefly
into the preglacial time. As Adams maintains, this center is quite
distinct from the southwestern center on the arid plateau of Mexico
and the adjoining parts of the United States. This latter does not
seem to be very important for the later development of the genus,
arid regions being generally unfavorable for crayfishes. In older
Tertiary times, however, also the southwestern center played a part,
in fact it is the original center of the whole genus Cambarus.

The different "outlets or highways of dispersal," as Adams
(/. c, p. 123) has characterized them, are rather well represented
in the distribution of Cambarus, and here again I believe, that they
were efficient in preglacial times as well as in postglacial times.
The Mississippi valley route is represented in the dispersal of the
subgenus Faxonius, and also by that of the blandingi-group of the

1 Adams' southeastern center does not include the central basin, and he thinks
that the Mississippi river (although it undoubtedly possessed a fauna of its own)
was largely populated by way of the Tennessee River, which, after having cap-
tured the upper course of the old Appalachian River, opened an outlet to its fauna
toward the Mississippi. This is no doubt quite correct with reference to the
freshwater shells, and, as has been pointed out already by Adams, finds some sup-
port in the distribution of certain crayfishes (/. c. , p. 849). But as we have seen
in the above pages, the center of Faxonius in the central Mississippi valley is
very marked, and apparently distinct from the other two centers. It is, however,
easy to unite all three of them, and regard them as parts of one larger center of
older (old Tertiary?) age, including parts that are differentiated physiographi-
cally, as indicated above.

126 ORTMANN — AFFINITIES OF CAMBARUS. [April i3j

subgenus Cambarus (C blandingi acutus). The route along the
coastal plain seems to be the least frequented, only C. blandingi
typicus characterizing it. But then again the Appalachian plateau
formed a third outlet to the north ; this is clearly indicated by the
dispersal of the subgenus Bartonius. Adams says very pertinently
(p. 129): "dispersal is both forward and backward along these
highways," and thus we see that in special cases the direction of
the migration may become the opposite. In one case (C clarki}
we have a reversed current of migration from the southeastern
United States toward the southwest, going in a direction opposite
to the general direction of immigration of the whole genus. A
direction downward the Mississippi valley (southward) is probable
in the palmeri-group of Faxonius, and C. diogenes seems to repre-
sent the identical reversed direction, descending the Ohio valley
from the Allegheny Mountains. The same species shows indica-
tions of a reversed migration on the Atlantic coast plain, from
Maryland to Virginia.

That the different centers of origin assumed above are very likely
correct, is shown by a two-fold consideration. First, the largest
number of species of each subgenus is generally found in or near
these centers (Adams, /. c, 1902, p. 128: first criterium), and
then the more primitive forms of each subgenus are found there
(third criterium of Adams). For the subgenus Cambarus, this is
not entirely true, Mexico possessing only two species, while Kansas
possesses three of the more primitive forms, but this may be due to
deficiency of our knowledge, or else it is due to interruption and
breaking up of the old southwestern range of the genus ; it is ap-
parently not so flourishing any more in these parts as it used to be.
Of the more highly advanced forms of the subgenus Cambarus
{blandingi-section) , the largest number of species is recorded for
Georgia (7), Florida (6), and Alabama and Mississippi (4 each).
The most primitive forms {spiculifer-group) are found in Georgia,
Florida and Alabama.

The subgenus Cambarettus also makes an exception, two species
being found in Mexico, and only one, but this a more primitive
one in Louisiana.

The subgenus Faxonius possesses the largest number of species
in Arkansas (8), and in Missouri and Indiana (7 each). Illinois
has only 4, but this may be due to defective knowledge. The more

I9o5.] ORTMANN — AFFINITIES OF CAMBARUS. 12"t

primitive forms of the limosus-section (aside from C. limosus itself)
are found in Indiana, Kentucky and Missouri, that is to say, in the
same general region.

The subgenus Bartonius has the largest number of species in
Tennessee (6); then follow: Georgia, North Carolina, Virginia
and Pennsylvania (with 4 each). The more primitive forms of the
extraneus-section are found in Georgia, Alabama, Tennessee and
Kentucky. In Tennessee is also found one of the blind cave forms
(C. hamulatus). Thus also here is apparently a mutual relation
between center of origin, location of most primitive forms, and
center of frequency. This rule, consequently holds good in the
section of C. blandingi of the subgenus Cambarus, and in the sub-
genera Faxoniits and Bartonius, while it is not very evident in the
more primitive forms of .the subgenus Cambarus, and in the sub-
genus Cambarellus}

A few peculiar and striking facts ought to be mentioned especi-
ally.

Discontinuity of distribution proof of antiquity. — We have found
this rule substantiated in the following cases: (1) In the distribu-
tion of the more primitive forms of the subgenus Cambarus (sec-
tions of C. digueti and gracilis) ; (2) in the subgenus Cambarellus ;
(3) in the limosus-section of the subgenus Faxonius; (4) in the
hamulatus-section of the subgenus Bartonius. The discontinuity
offered by C. wiegmanni in the alleni-group of the subgenus Cam-
barus needs further investigation, and cannot be regarded as estab-
lished before the systematic position of this species has been posi-
tively ascertained.

Morphologically isolated species occupy isolated stations. — This is
illustrated by : (1) C. cubensis in Cuba ; (2) C. shufeldti in Louisi-
ana ; (3) C. limosus on the Atlantic coast plain from New Jersey
to Virginia; (4) C. harrisoni in Missouri; (5) C. alabamensis
and compressus in northern Alabama: (6) C. setosus in Missouri
(cave-form).

Closely allied species occupy neighboring areas. — This is most evi-

1 Addition to our knowledge may change this considerably. I only call atten-
tion to the fact, that up to very shortly ago only two species of Bartonius were
known from the state of Pennsylvania. Investigations during the last four years
have revealed the presence of two more species, thus doubling the number. This
may happen in any other state.

128 ORTMAXX — AFFINITIES OF CAMBARUS.

[April 13.

dent in the following cases, where groups of species occupy a cer-
tain range, but represent each other in the different parts of this
range : ( 1 ) in the spicalifer-group of the subgenus Cambarus : spic-
ulifer in northern and central Georgia, versutus in central and
southern Alabama, and in northwestern Florida, pubescens in eastern
Georgia, angustatus in southeastern Georgia. (2) In the clarki-
group : clarki, parallel to the Gulf coast from Texas to Florida,
troglodytes in corresponding localities in Georgia and South Caro-
lina. (3) Limosus-section of Faxonias : indianensis in southwestern
Indiana, sloanei in southeastern Indiana and Kentucky. (4) Pro-
pinquus-group : propinquus has a western and northern distribution ;
it is represented in western Pennsylvania by obscurus. (Between
both possibly is C. prppinquus sanborni, occupying an intermediate
range.) (5) Rusticus-group : spinosus is southern and eastern
(North and South Carolina, northern Georgia, northern Alabama
and eastern Tennessee), while putnami is more northern (Ken-
tucky). (6) In the palmeri-group the different species occupy dif-
ferent parts of a range that includes Mississippi, western Tennessee,
Arkansas, Indian Territory and northeastern Texas.

Groups of allied species are oj ten formed by a typical species, which
shows a wide range, while the allied species form a fringe on the edge
of this range thus representing local forms. This is shown beauti-
fully in the following natural groups : (1) Rusticus-group: the typ-
ical form is rusticus, the local forms at the edge of its range are :
forceps (southeast), neglectus (west and southwest), spinosus and
putnami (southeast), hylas (south), medius (south); probably also
erichsonian us (southeast). (2) Virilis-group : the typical form is
virilis, the local forms are : mceki, longidigitus, nais, pilosus, all in
the southwest. (3) Barto?ii-section: bartoni is the typical form, the
local forms of it are : acuminatus (southeast), latimanus (south and
southwest); in this section also a mountain form has developed
within the range (longulus), and varieties are found in the southern
section of the range, as well as at its northwestern edge. (4) C. mon-
ongalensis is a local form developed at the northwestern edge of the
range of C. carolinus. (5) In the diogenes-section, at least one
species, C. uhleri, seems to be a local form of the widely distributed
C. diogenes, developed at the eastern extremity of its range.

More or less closely allied species, occupying the same or nearly the
same territory, generally possess different habits. In most of the

I9o5.] ORTMAMN — AFFINITIES OF CAMBARUS. 129

species, we do not know much about their habits, but a few remark-
able cases may be mentioned, (i) C. virilis and C. immuuis,
although sharply separated, are rather closely allied, and occupy
large identical tracts of the central states. We know that C. virilis
prefers running water with stony bottom, while C. immunis is a
pond and ditch form (see above, p. 117). (2)6". monongalensis
inhabits, in western Pennsylvania, almost the same territory that
is occupied by C. diogenes. The first, however, belongs to the
hills, the second to the lowlands (see Ortmann, Ann. Carnegie
Mus., v. 3, p. 400).

The various drainage systems have a different effect upon the
species of the different subgenera, which is apparently due to funda-
mental differences in their habits. (1) Bartonius is preeminently a
mountain-stream group. It goes up into the smallest streams, up
to their very sources. In this region, changes of drainage, due to
piracy, are common, and rather the rule than the exception, and
thus the species quite generally occupy the headwaters of streams
running in different directions from the divides. This is exampled
by the distribution of the following species : extraneus, bartoni,
longulus, latimanus, carolinus, and probably also by diogenes. (See
Adams, " Migration of Divides," in Americ. Natural., 35, 1901,
p. 844). (2) The blandingi-section belongs originally to the low-
lands of ths Gulf and Atlantic plain. Here removal of barriers
largely has taken place, and thus the species of this group belong
to the drainages of different coast rivers, for instance : lecontei
blandingi, clarki, troglodytes, alleni. (See Adams, ibid., p. 842 :
"In a country approaching base-level a wide distribution of the
fauna will be facilitated.") (3) The subgenus Faxonius belongs
to the great rivers of the interior basin, and does not ascend far into
the headwaters, at least in the mountainous regions, and also does
not descend far toward the coastal plain. Consequently, the drain-
age systems being more permanent, the distribution of these species
is more closely connected with the latter. We may, perhaps, com-
pare this — in a very general way — with the period of maximum
roughness of Adams (/. c), although this does not hold good for
all of this immense region. Indeed, there are important excep-
tions, and the subgenus has crossed over into the lake-drainage
(C propinquus, obscurus, rusticus, virilis, immunis}, and evert into
the Hudson Bay drainage (C virilis). This has been brought

130 ORTMANN— AFFINITIES OF CAMBARUS. [April 13,

about, apparently, by extensive shifting of divides, and we know
positively, that this has taken place in great style during and after
glacial times. The eastern mountains (Appalachian system) have
formed a sharper barrier, but also here certain species have been
able to cross : in ancient times C. limosus, in more recent times
C. obscurus (see Ortmann, Ann. Cam. Mus., v. 3, p. 406). The
most interesting region is at the southern extremity of the Appa-
lachian system, as we shall presently see.

Very important drainage changes, that have taken place in the
southern Appalachian system, are clearly indicated by the distribution
of crayfishes, and tend to confirm the results obtained by Simpson and
Adams for the freshwater mollusks (see above p. 116). In the
region of the Alabama River drainage and that of the Tennessee
River, we had at a certain time, a large river running to the South,
the Appalachian River, the upper course of which was deflected
toward the Northwest, forming the present Tennessee River. The
former unity of the drainage system is indicated by identical or
closely allied species found now in both systems. The following
species illustrate this : C. erichsonianus, exiraneus, Jordan/', lati-
vianus, and possibly others. Further investigations of the condi-
tions present in these regions are very desirable.

This is, I think, a rather satisfactory outline-sketch of the dis-
tribution of the genus Cambarus over the United States. But it is
only a sketch, and more detailed investigations are much needed.
We see that the migrations of the different groups are very com-
plex, the directions of the migrations crossing at various angles,
often being directly opposed to each other. (See map, plate III.)

Further, we are to emphasize, that our knowledge is by no
means complete with regard to the distributional facts. There is
hardly a single case, where the actual boundaries of a species are
known. We have a large number of locality-records, and by plot-
ting them on a map, we obtained a general idea of the range of the
different species, but rarely we know the exact limits, and nobody
has ever tried to ascertain these, except the present writer in a
very limited region, in western Pennsylvania (see Ann. Carnegie
Mus., v. 3, 1905). But this ought to be done -by all means, and
there is no doubt, that very interesting results will be obtained.

It may be remarked in conclusion, that I do not think that a
number of reported localities for certain species are trustworthy.

I9o5] ORTMANN— AFFINITIES OF CAMBARUS. 131

It is astonishing how easy records and museums specimens become
mixed up, and a number of localities which are given bona fide by
various authors are very questionable. In the following, I put
together those records, that appear — at least to me — doubtful or
in need of confirmation. At the same time, a number of new
records is given which have been made use of in the above pages.

C. blandingi (Harl.).

New Localities. — ■ Millpond at Plainsboro, Middlesex Co., New
Jersey, coll. by the writer (Cam. Mus. ). — This species is further
abundant in the millpond of Grover's Mills, Princeton Junction,
Mercer Co., N. J. (seen by the writer), and is rare in the Dela-
ware-Raritan Canal, at Aquaeduct near P/inceton, Mercer Co., N.
J. (seen by the writer).

C. clarki Gir.

New locality. — Devils River, Val Verde Co., Texas, coll. by
H. A. Pilsbry, 1903 (specimens in Philadelphia Acad, and Cam.
Mus.).

C. timosus (Raf. ).

New localities. — Stony Brook, Princeton, Mercer Co., N. J.,
coll. by the writer, May 30 and Sept., 1898 (Cam. Mus.). —
Delaware-Raritan Canal, at Aquaeduct near Princeton, Mercer Co.,
N. J., coll. by the writer, Jan., 1899 (Cam. Mus.). — Delaware
River, North Cramer Hill, Camden Co., N. J., coll. by the writer,
Sept. 18, 1904 (Cam. Mus.). — Collected by the writer at the
following new places in Eastern Pennsylvania in September, 1904 :
Delaware River, Torresdale Fish Hatchery, Torresdale, Philadel-
phia Co.; Marcus Hook Creek, Marcus Hook, Delaware Co.:
Little Neshaminy Creek, Grenoble, Bucks Co.; Delaware River,
New Hope, Bucks Co.; Schuylkill River, West Manayunk,
Montgomery Co. (Cam. Mus. ). — Further: Tributary of Brandy -
wine Creek, Chadds Ford Junction, Chester Co., Pa. (Acad.
Philad. ). — Delaware River at Holmesburg, Philadelphia Co., Pa.
(Acad. Philad. and Cam. Mus.). — Gettysburg, Adams, Co., Pa.,
coll. by H. A. Pilsbry (Acad. Philad.). — Potomac River,
Cherry Run, Morgan Co., W. Va. , coll. by the writer, Sept. 23,
1904 (Cam. Mus. ).

Doubtful and spurious older records. — Hagen gives, in 1870,
Niagara (L. Agassiz) ; Lake Erie; New York (Mr. Pike) ; and
Pittsburg. Faxon (1885) drops New York and Pittsburg, but

132 ORTMANN — AFFINITIES OF CAMBARUS.

[April 13 ,

again gives Niagara ( " there is no doubt of the correctness of the
determination" ), and Lake Erie (Peabody Ac. Sci.). In 1890,
Faxon says of the latter specimens, that they "are too small to
determine with certainty." He further gives, in 1885, Lake Su-
perior (Boston Soc. Nat. Hist.). I do not entertain the slightest
doubt that all these localities are wrong. As to Niagara, which is
founded upon the authority of L. Agassiz, we only have to con-
sider that the same locality upon the same authority is given also
for C. propinquus, and it is quite probable, that specimens of C.
limosus were put by mistake into a jar containing C. propinquus.
As to Lake Erie and Lake Superior, some other species may be in-
tended, or a similar mistake has been made : I do not believe,
most emphatically, that this species is found in the lake-region.
With regard to the absence of C. limosus in the state of New York,
we possess the testimony of De Kay (Zool. N. Y. , 6, 1844, p.
23) : "I have searched for it {Astacus affinis') without success in
the tributaries of that stream (Delaware) within the limits of this
State."

C. propinquus Gir.

New Localities. — Lake Erie, Lorain Co., Ohio, Lorain gill nets.
May 1, 1892, coll. by H. Warden (Mus. Oberlin). These speci-
mens from the lake are the true C. propinquus, while all other
specimens from the tributaries of the lake in Lorain Co., Ohio,
belong to propinquus sanborni, see below. Crooked Lake, Oden
near Petoskey, Emmet Co., Mich., coll. by E. B. Williamson, Sept.
1, 1904 (Cam. Mus.). This is the northernmost exact locality
known, and is very near to a locality recorded by Ward (Bull.
Mich. Fish Comm., 6, 1896, p. 15), but not recorded by Faxon,
namely: Lake Michigan and Pine Lake at Charlevoix, Charlevoix
Co., Mich.

Doubtful Locality. — The latter localities in northern Michigan
render it possible that the old records of Lake Superior, given by
Hagen on the authority of L. Agassiz, may be correct. But since
to L. Agassiz also the record of C. rusticus and virilis for Lake
Superior are attributed, we have again several species mixed up,
and it is better to wait for a confirmation.

C. propinquus sanborni Fax.

New Localities. — Oberlin, Lorain Co., Ohio, is the type-locality
(Faxon) for this form. I have seen it (Mus. Oberlin) from the

igos.] ORTMANN — AFFINITIES OF CAMBARUS. 133

following localities in this region and the state of Ohio : Water-
works reservoir, Oberlin, and Plum Creek, Oberlin ; further : Ver-
million River, Beaver Creek, French Creek, all in Lorain Co. ;
Killbuck Creek, Creston, Wayne Co.; Tuscarawas River, Gnaden-
hutten, Tuscarawas Co. The latter two localities belong to the
Ohio drainage, while the rest is lake drainage. This variety forms
a morphological link between C. propinquus typicus and C. obscurus,
and seems to be intermediate also in its range.

C. rusticus Gir.

The locality Lake Superior (L. Agassiz) given by Hagen (1870)
needs confirmation. As I have shown elsewhere (Ann. Car. Mus.,
v. 3, 1905, p. 387), the locality Pittsburgh is wrong.

C. neglect us Fax.

New Locality. — Rogers, Benton Co., Arkansas, coll. by H. A.
Pilsbry, March 25, 1903 (Acad. Philad. and Cam. Mus.).

C. putnami Fax.

New Locality. — Rockcastle River, Livingston, Rockcastle Co.,
Ky., coll. by E. B. Williamson, June 21, 1904 (Cam. Mus.).
(See Williamson, Ohio Natural., 5, 1905, p. 311.)

C. virilis Hag.

New Locality. — Sandy Lake, Ontario, Canada, coll. by G. H.
Clapp (Cam. Mus.). This species has been reported by Ward
(Bull. Mich. Fish Coram., 6, 1896, p. 15) from Lake Michigan
and Pine Lake, Charlevoix Co., Mich.

The locality Lake Superior, given on the authority of L. Agassiz by
Hagen (1870), has been confirmed by Faxon (1885) on the authority
of C. L. Herrick, and falls within the known range of the species.

Doubtful Records. — Lake George, N. Y. (L. Agassiz) has been
recorded by Faxon (1885) with a ? . It surely is very doubtful.

Faxon also mentions this species from Laramie City in Wyoming ;
this may be correct, but needs confirmation. He records it further
from near Bridgeport, Jackson Co., in northern Alabama, in the
Tennessee drainage (U. S. Mus.); I seriously doubt the correct
ness of this locality, since it is the only one east of the line formed
by the Mississippi and Ohio rivers, and is far remote from the rest
of the range.

C. immunis Hag.

New Localties. — Lamoni, Decatur Co., Iowa, coll. by J. B.
Hatcher (Cam. Mus.). This species is also found in northern

134 ORTMANN — AFFINITIES < >F CAMBARUS. [April 13,

Ohio, as first indicated by Osburn and Williamson (6 Ann. Rep.
Ohio Ac. Sci., 1898, p. 21), in Sandusky, Erie, and Lorain Cos.,
and in Lake Erie. I have seen specimens (Mus. Oberlin) from
Huron River, Huron, Erie Co., and from Oberlin, Lorain Co.
(Waterworks Reservoir and Plum Creek),

Doubtful Records. — Hagen (1870) gives Huntsville, Madison
Co., northern Alabama. This is possibly not this species, at any rate
it is "not normal" (Faxon, 1885, p. 100). The locality is too far
separated from the rest of the range, to be accepted without hesitation.

Faxon (1885) gives: New York (L. A. Lee); Laramie, Wyo-
ming (U. S. Mus.) ; Orizaba, Mexico (U. S. Mus.), and further in
1898 he adds : small stream flowing into Oneida Lake, N. Y. The
locality in Wyoming may be correct, but we have to try to connect
it with the rest of the range, before accepting it. Orizaba, Mexico,
is no doubt wrong, and I do not hesitate for a moment to drop it.
Oneida Lake in New York seems very strange, since there are no
connecting localities with northwestern Ohio. I cannot accept this
locality unless verified by unequivocal evidence.

C. palmeri longimamis Fax.

New Locality. — Limestone Gap, Choctaw Mt., Indian Terr.,
coll. by H. A. Pilsbry (Ac. Philad. and Carn. Mus.).

C extraneus Hag.

New Locality. — Rockcastle River, Livingston, Rockcastle, Ky.,
coll. by E. B. Williamson, June 21, 1904 (Carn. Mus.). This is
in the Cumberland River drainage ; previously, this species was
known only from Tennessee, Alabama and Georgia. (See William-
son, Ohio Natural., 5, 1905, p. 310.)

C. bartoni (F.).

New Localities. — Small streams, Princeton, Mercer Co., N. J.,
coll. by the writer (Carn. Mus.); East Canada Creek, Herkimer
Co., N. Y., coll. by R. Ruedemann (Carn. Mus.); Selbysport,
Garret Co., Md., coll. by the writer (Carn. Mus.); Cherry Run,
Morgan Co., W. Va., coll. by the writer (Carn. Mus.); Green-
ville, New Castle Co., Del. (Ac. Philad.). The following locali-
ties in eastern and central Pennsylvania are represented in the
Carnegie Museum (coll. by the writer): Driftwood and Sinnama-
honing, Cameron Co.; Keating Summit, Potter Co.; Wills Creek,
Mance, Somerset Co.; Cush-Cushion Creek, Indiana Co.; Cresson,
Cambria Co.; Ashville, Cambria Co.; Hollidaysburg, Blair Co.;

I9o5]' URTMANN — AFFINITIES OF CAMBARUS. 135

Wissahickon, Philadelphia Co.; Shoemakersville, Berks Co.; Valley
Forge, Chester Co.; Grenoble, Bucks Co.; New Hope, Bucks Co.;
West Manayunk, Montgomery Co.; Wallingford, Delaware Co.
Other new localities in eastern Pennsylvania are : Headwaters of
Loyalsock Creek and Ganoga Lake, Sullivan Co. (Ac. Philad.);
Pinegrove, Cumberland Co. (Ac. Philad.).

Doubtful Record. — Lake Superior, given by Hagen on the au-
thority of L. Agassiz, is undoubtedly wrong. As to records of this
species from Ohio see C. bartoni robustus.

C. bartoni robustus ( Hag. ) .

New Localities. — Small stream tributary to Rockcastle River,
Livingston, Rockcastle Co., Ky., coll. by E. B. Williamson, June
21, 1904 (Carn. Mus. ) . These specimens agree well with young
individuals of this variety ; adult ones are not in the lot. (See
Williamson, Ohio Natural., 5, 1905, p. 310.) Oberlin, Lorain
Co., Ohio (Mus. Oberlin). This form was doubtfully reported
from Knox Co., Ohio, by Osburn and Williamson (1896). All
specimens from Oberlin seen by the writer belong to this variety.
The typical form seems to prevail in southern Ohio.

Doubtful Records. — Faxon (1885) gives Decatur, Macon Co.,
111., but this needs confirmation. Further it is doubtful, whether
the form called by this name in Maryland and Virginia is identical
with the true (northern) robustus.

C. bartoni longirostris (Fax.).

Doubtful Record. — Pollard, Escambia Co., Alabama, seems
doubtful, since it is close to the Gulf coast, and far away from the
original mountain home of this form.

C. latiinanus (Lee).

The locality, Ocean Springs, Miss., is doubtful for the same reason.

C. carolinus Er. (= dubius Fax.).

The reported occurrence of this species in Indian Territory
(Faxon, 1890) seems strange. It must be looked upon as doubtful
till the connection with the rest of the range is established.

C. diogenes Gir.

New localities. — Cooper, Greene Co., Iowa, coll. by J. B.
Hatcher (Carn. Mus.); Seaford, Sussex Co., Delaware, coll. by S.
N. Rhoads, June 18, 1903 (Ac. Philad. and Carn. Mus.). — Ober-
lin, Lorain Co., Ohio (Mus. Oberlin). — The specimens from this
locality have been mentioned by Osburn and Williamson (1898)
as C. dubius ?, but they are typical C. diogenes.

136 ORTMANN — AFFINITIES OF CAMBARUS. [April i3>

Doubtful records. — Faxon, 1885, gives Deer Park, Garrett Co.,
Md. This should be confirmed ; according to the. writer's experi-
ence, C. carolinus ought to be expected there. If confirmed, this
locality will be highly interesting.

Faxon further gives : Cheyenne, Wyoming, and Clear Lake,
Colorado ; in both cases the most western extremity of the range
of the genus is reached. Harris (Kansas Univ. Quart., 9, 1900, p.
267) gives: Boulder, Colorado. This serves to establish the cor-
rectness of the above records, but the connection with the rest of
the range must be found (I have not been able to locate Clear Lake
in Colorado). The southern localities for C. diogenes recorded by
Faxon, Monticello, Lawrence Co., Miss., and New Orleans,
Louisiana, certainly need further support.

C. argillicola Fax.

New locality. — Oberlin, Lorain Co., Ohio (Mus. Oberlin). — I
have seen three specimens from Oberlin (adult and young male,
adult female), two of which bear the label: Hovey's Ice house,
northeast of Oberlin, coll. by Leuthi, Sept. 29, 1892.

Doubtful records. — The localities, Kinston, N. Carolina, and New

Orleans, Louisiana, given by Faxon in 1885 are doubtful, as

admitted by himself. The localities given in 1898, Victoria and

Brazoria, Texas (U. S. Mus.), most emphatically need confirmation.

Carnegie Museum,

Pittsburgh, April 7, 1905.

explanation of plate

v,„.

The plate is introduced to illustrate the centers of origin, and the chief directions
of migration of the different subdivisions of the genus Cambarus. Circles or ellip-
ses indicate centers of origin, the lines radiating from these, and ending in an
arrow-point, indicate the migration. The different colors mark the different sub-
genera : Red, Cambarus ; brozvn, Cambarellus ; green, Faxonius ; blue, Bartonius.

It will be remarked that two centers are given for the subgenus Cambarus ; the
one in Mexico marks that of the more primitive forms, the other in Alabama and
Georgia, that of the more highly advanced forms (bla?idingi -section). This latter
one, as well as the subgenera Faxonius and Bartonius, took their origin probably
from a primitive stock of the subgenus Cambarus, immigrated into the southern
United States along the broken red line running from Kansas to Alabama.

For further particulars see text, pp. 103, 106, 113, 121, and 124 ff.

igos-] KOLLOCK-SMITH — ELECTRO-ANALYSIS. 13^

[Contribution from the John Harrison Laboratory of Chemistry.]

THE USE OF THE ROTATING ANODE AND MERCURY
CATHODE IN ELECTRO-ANALYSIS.

BY LILY G. KOLLOCK AND EDGAR F. SMITH.

( Read April 13, iqos- )

First Paper.

Several investigations made in this laboratory have shown that
when in electro-analysis the anode is rotated high currents can be
used and metals be precipitated completely in very short periods
of time ; further, by the use of mercury cathodes most interesting
determinations and separations of metals are possible.1 In the
latter case, however, the anode has been stationary, and the elec-
trolyte consequently not agitated. Then, of course, the precipi-
tation of the metal has been comparatively slow. Observing
the splendid results got with the rotating anode, when platinum
was the cathode, we determined to use a combination of rotating
anode and mercury cathode. This was accordingly done, and in
some preliminary trials made last August (1904), the results of
which were briefly alluded to in a communication published in
the Jour. Am. Chem. Soc, 26, 16 14, mention was made that
0.4810 gram of copper could be precipitated in twenty-five
minutes, and that this success could be had with other metals.
Since then we have made additional experiments which we desire
to record here. Not only is the time factor reduced for the metals
studied, but the plan of combining a mercury cathode with the
rotating anode gives an inexpensive form of apparatus which will
eliminate the platinum dish, cone or cylinder from electro-analysis
and thus remove an expensive factor.

Apparatus. — The decomposition cell is a tube 3.5 cm. in
diameter and 7.5 cm. in height, made from a test tube. Soften
the bottom of the tube in a blast lamp flame, then push through it
a platinum wire two centimeters in length, so that its end projects
0.5 cm. into the tube. Flatten the bottom of the tube on an
asbestos plate and anneal it in the ordinary way.

xJour. Am. Chem. Soc, 25, 884: 26, 1 124.

138 KOLLOCK-SM1TH — ELECTRO-ANALYSIS. [April 13

The anode, 7.5 cm. in length, is made from platinum wire 1
mm. in diameter, coiled into a flat spiral 1.5 cm. in diameter. It
is inserted in a chuck carried by the rotator which is also provided
with three pulleys varying from 2 to 5 cm. in diameter. These
pulleys are connected by a belt to two pulleys on the motor.
With this arrangement the rotation of the anode could be varied
from 100 to 1800 revolutions per minute. During the decom-
position an amperemeter, a voltmeter and a rheostat, allowing of
resistance from .1 to 100 ohms, were kept in the circuit.

The precautions indicated by Myers in his paper with regard to
the decomposition cell were observed. If care be taken to have the
cell as clean as possible there will be no trouble experienced with
the amalgam subsequently adhering to its sides. The mercury, be-
fore using, should be washed with alcohol and ether and after the
odor of the latter has disappeared, be placed in the desiccator until
it is weighed. It was generally allowed to remain for about five
minutes on the balance pan before taking the final weight. In prac-
tice a beaker containing a large quantity of mercury, so prepared,
should be kept in the desiccator ready for use. The mass of the
mercury taken in a single experiment varied from forty to fifty grams.
This was frequently used for two or three determinations, except
in the case of chromium, where it was found advisable to use it
but once. The cathode surface in the first experiments upon zinc
was 3.5 sq. cm., but throughout the rest of the work it was about
9 sq. cm. After weighing the decomposition cell and mercury,
the solution to be electrolyzed should be introduced. The volume
of the electrolyte is always recorded in the accompanying tables.
The cell should then be placed upon the copper plate and the
anode lowered into the solution. The distance between the
cathode and anode depended upon the volume of the electrolyte.
When the volume was five cubic centimeters the electrodes were .5
cm. apart and in other instances 1 cm. was their distance apart.
The difference did not appear to materially affect the rate of
deposition. The tube should be covered. The anode should next
be rotated and the connection made with the required number of
chloride accumulator cells. The speed of the anode was varied
either by using less current for the motor or by changing the com-
bination of pulleys. With the higher currents recorded, the solu-
tion was frequently heated to boiling. When this occurred the

i9°5-] KOLLOCK-SMITH — ELECTRO-ANALYSIS. 139

current invariably dropped sometimes as much as one ampere. But
upon washing down the cover glasses with cold water it rose to its
former strength. The dropping of the current is probably due to
the accumulation of steam bubbles upon the electrodes. During
the electrolysis some of the solution will of course be carried to the
sides of the containing vessel and to the cover glasses by the escap-
ing gases or by the agitation of the liquid. After many trials it
was found that it is unnecessary to wash down this portion when
the higher currents are used. The condensed steam continually
frees the sides from the solution. The cover glasses may now and
then be tilted against the sides of the tube in order to run off the
water which collects in large drops.

It has been repeatedly observed in the present work that the
greater the concentration of the electrolyte, the greater the rapidity
of deposition, but the last traces of metal were always difficult
to remove. For this reason, after a solution had become colorless,
the electrolytic action was continued several minutes in order to
precipitate the minute amount remaining unprecipitated. It is,
therefore, also important to have the volume small toward the end
of the decomposition.

When the metal has been completely deposited, the anode
should be stopped, the cover glasses removed and the decomposi-
tion cell filled with distilled water. This should then be siphoned
off to the level of the spiral and the liquid replaced by distilled water
until the current drops to zero. This wash water should always be
put aside and tested in order to ascertain that the metal has been
completely deposited. The current should next be interrupted and
the tube removed and washed again with distilled water, inclining
and twirling the cell in order to more completely wash the amalgam.
As much of the water as possible should be poured from the cell
and the amalgam then be washed twice with absolute alcohol and
twice with ether. It should be wiped dry on the outside and after
the volatilization of the ether be placed in the desiccator and
weighed as previously described.

Experimental Part.
Zinc.
The first experiments made after those described in the Jour.
Amer. Chem. Society 26, 16 14, were upon zinc sulphate. They

PROC. AMER. PHILOS. SOC. XLIV. l8o. J. PRINTED JULY 31, I905.

140

KOLLOCK-SM1TH — ELECTRO-ANALYSIS.

|April 13,

were conducted in order to ascertain the rate of deposition with
varying concentration, current strength, electromotive force, speed
of anode' and how the quantity of metal in the mercury affected the
subsequent rate of deposition. The solution for the first experi-
ments contained 0.2025 gram of metallic zinc in 10 cc. This was
determined by the electrolytic method, depositing it upon a plati-
num dish from an ammonium acetate electrolyte. The speed of
the anode was 400 revolutions per minute. The current strength

was one ampere and the
E.M.F. was 5 volts. The
volume of the zinc sulphate
solution equaled 15 cc, the
current acted thirty minutes.
The solution siphoned from
the tube showed no trace of
zinc Consecutive experi-
ments so conducted gave the
following results in 25 min-
utes: .2027, .2030, .2025,
.2025, .2021, .2027, .2025
grams. Two trials were made
with the same conditions but
using a volume of 10 cc in-
stead of 15c c. It was found
that the zinc was completely
separated in twenty minutes.
Experiments were then
made to determine the rate
of deposition in successive
periods of time and the curve constructed from the data thus obtained,
using periods of time for abscissas and masses for ordinates. The
conditions employed were those given above. The results were as
follows :

In 5 minutes o.ll96gram.

" 10 " 0.1774 "

" 15 " 0.1897 "

"20 " 0.2002 "

" 25 " 0.2027 "

Upon employing a current of 2 amperes, adding sulphuric acid
to increase the conductivity, the entire amount was deposited in

p

&

/

/

/

/

/

/

/

/

/

/

0

/

—

0

5

0

5

0 ■

g-

Curve i. Zinc — 1 Ampere, 5 volts.

1905.]

KOLLOCK-SMITH — ELECTRO-ANALYSIS.

141

15 minutes. The following conditions were employed : Total vol-
ume 15 c.c, sulphuric acid 0.4 c.c, current strength 2 amperes,
pressure 7 volts, speed of anode, 500 revolutions per minute.

In 5 minutes 0.1860 gram of zinc was deposited.
" 10 " 0.1998 " " " i( "

" 15 " 0.2020 " " " " "

Double the quantity of zinc mentioned above was dissolved in
15 c.c. To this was added .25 c.c. of concentrated sulphuric acid,
the anode was rotated at the rate of 800 revolutions per minute and
the solution electrolyzed. In thirty minutes the zinc was com-
pletely deposited, using a current of 1.5 amperes and 10 volts.

1

t

y

/

17

/

1

1

1

£11

5

0

5

0 a;

nvtc

i

/

/

/

1

n

Curve 2. Zinc— 2 Amperes, 7 volts.

Curve 3. Zinc-2 Amperes, 6 volts

In 10 minutes 0.3701 gram was deposited.
" 15 " 0.3997 " "
" 20 " 0.4011 " " "

" 30 " 0.4058 "

Curve 3 was drawn from these results.

The same mass of zinc in twenty cubic centimeters was electro-
lyzed with a current of 2 amperes and 6 volts, other conditions
being identical.

142

KOLLOCK-SMITH — ELECTRO-ANALYSTS.

[April 13,

Curve 4. Zinc-

In 10 minutes 0.3352 gram was deposited.
" 15 " 0.4010 " " "

"20 " 0.4030 " " "

" 30 " 0.4050 "

Curve 4 was drawn from these re-
sults. A comparison of the third
and fourth curves shows the effect of
greater dilution upon the quantity of
zinc deposited in the first ten minutes.
Two experiments were made to learn
the effect of different speeds of the
anode upon the rate of precipitation.
It was found that the amount of zinc
deposited under a rotation of 440 revo-
lutions per minute, and 1,000 revolu-
tions per minute was only .0004, which
is within experimental error, showing
that between these limits there is no
apparent effect. It was also discovered
that when more than 1 gram of zinc was
present in the mercury, the latter should
not be further used if it is desired to
-1.5 Amperes, 10 volts, obtain results in the shortest period.

Zinc.

I

4

/

/

/

/

/

'

I

/

/

/

/

/

/

11

b

0

.5 '•/

a
1
1

Zinc

Present
in Gram.

Sulphuric

Acid

Present

Volume
in c.c.

Current.
Am-
peres.

Volts.

Revolu-
tions of
Anode
Per

Time

in

Minutes.

Zinc

Found

in Gram.

Error in
Gram.

W

Minute.

j

0.2025

O

15

I

7

75°

30

O.2027

+ 0.0002

2

"

0

15

I

7

75°

25

O.203O

+ O.OOO5

3

"

O

15

I

7

750

25

O.2OI5

— O.OO I

4

"

O

15

I

7

750

25

0.2020

— O.OOO5

5

"

O

15

I

7

750

25

O.2025

6

"

0

IO

2

7

750

25

O.2024

— O.OOO I

7

"

25

IO

2

7

75°

3°

O.2O27

+ 0.O0O2

8

O.4050

25

20

1-5

b

750

45

O.2054

— O.OOO4

9

O.2025

25

IO

I

5

750

25

O.2025

10

"

25

IO

I

5

750

25

O.2029

+ O.OOO4

11

"

25

15

I

5

750

25

0.2025

12

"

25

IS

I

5

75°

20

0.2027

-|- 0.0002

13

"

25

15

2

b

750

15

0. 2030

+ O.OOO5

H

"

25

15

2

b

750

i5

0.2020

— O.OOO5

15

"

25

15

2

b

750

15

O. 202 1

— O.OOO4

16

O.4050

2S

i5

5

8

I4OO

6

O.4057

+ O.OOO7

17

"

25

15

5

8

480

6

O.4045

— O.OOO5

18

"

2S

15

S-b

7-5

4S0

8

O.4042

— O.CO08

19

"

25

IO

5

7

b40

5

O.4050

i9o5 ]

KOLLOCK-SMITH — ELECTRO-ANALYSIS.

143

To 10 c.c. of the zinc sulphate solution 0.4 c.c. of concentrated
sulphuric acid was added, after which it was electrolyzed by a cur-
rent of 5 amperes and 7 volts ; the speed of the anode being 640
revolutions per minute. Under these conditions 0.405 gram of zinc
was precipitated in five minutes.

Copper.

Having found that .405 gram of zinc e
could be deposited in from five to eight
minutes it was decided to try other
conditions upon copper than those rec-
orded in the previous paper, in order
to reduce the time factor. By using
higher currents and greater concentra-
tion of the electrolyte this was accom-
plished.

A solution of copper sulphate con-
taining 0.3945 gram of metallic copper
in five cubic centimeters was used for
these experiments. This quantity of
metal was precipitated finally in five
minutes. The solution became color-
less in three minutes. Twice this quan-
tity (.789 gram) was deposited in ten
minutes, although the solution had be-
come colorless at the expiration of Curve 5.
seven minutes. The volume in this Peres> 6 volts-
case being ten cubic centimeters it appeared the last traces of cop-
per required more time for precipitation. A current of 5 amperes
and 6 volts was used, sulphuric acid being introduced to increase
the conductivity.

The current strength recorded in the following table was main-
tained during the greater part of the electrolysis. When it showed
a tendency to rise, on the liberation of the acid, additional resis-
tance was thrown into the circuit. The following rates of depo-
sition of copper were determined under the preceding conditions.
The anode made 640 revolutions per minute.

In I minute 0. 1800 gram of copper was deposited.
"2 " 0.3400 " " " " "

" 3 " 0.3664 " " "
" 4 " 0.3945 " "

" 5 " 0.3945 " " «

/

/

1

/

3 4

hmuUi

144

KOLLOCK-SMITH — ELECTRO-ANALYSIS.

[April 13.

Copper.

0.

Copper
Present
in Gram.

Sulphuric
Acid
Present

Volume
in c.c.

Current.
Am-
peres.

Volts.

Revolu-
tions of
Anode
per

Time

in

Minutes

Copper

Found

in Gram.

Error in
Gram.

W

Minute.

I

O.789O

•25

12

3-5

6

I200

IO

O.79CO

— O.OOI

2

0.3945

•15

12

4

6

I080

5

O.394I

— O.OOO4

3

0.3945

•25

12

3-5

6

I200

6

0.3942

O.OOO3

4

0-3945

•15

12

5

6.5

I200

5

0-3944

— O.OOO I

5

0-3945

O

IO

2-4

9-7

I200

6

O.3946

— O.OOOI

6

0-3945

•17

IO

3-5

8.5

I200

4

0.3944

O.OOOI

7

0.3945

•17

IO

4

6

1080

5

O.3946

— O.OOOI

Nickel.
A nickel sulphate solution containing 0.4802 gram of metal in ten
cubic centimeters was used in the following experiments*, and after
finding that this quantity was completely deposited in the mercury
in twenty minutes with a current of 2 amperes and 7 volts, the rate
of deposition in succeeding periods of time was determined with a
current of 2.5 amperes and 6 volts.

In 2.5 minutes 0.2017 gram of nickel was deposited.

7-5 '

0.4095

10 '

0.4651

12. s '

o.4774

15

0.4802

s

a

Nickel
Present

Sulphuric
Acid

Volume

Current.
Am-

Volts.

tions of
Anode

Time
in

Nickel
Found

Error in

p,

in Gram.

in c.c.

peres.

per

Minutes

in Gram.

«

Minute.

O.4802

1

O.4802

■25

18

2

7

600

18

2

O.4802

•25

12

3-5

7

600

16

0.4799

— O.OOO3

3

O.4802

•25

12

2-4

6-5

600

IO

O.4S06

-j- O.OOO4

4

O.4802

•25

12

6

5

500-

7

O.4804

— 0.0C02

5

O.4802

•25

12

5

6.5

600

10

O.4796

— O.OO06

6

O.9604

•25

IO-3O

4

6

I IOO

10

0-9597

— O.OO07

7

O.4802

•25

12

3

7-S

I IOO

10

O.4806

—O.OOO4

8

O.4802

•25

12

3

7

I IOO

10

O.4796

— O.OO06

9

O.9604

•25

12

3-5

7

I IOO

16

0. 9604

10

O.4802

•25

12

5

7

640

12

O.4S09

— O.OOO7

11

O.4802

•25

12

5

6

S80

8

O.4806

— O.OOO4

12

O.4802

•25

7

6

S

I2CO

9

O.480I

— O.OOOI

*3

O.4802

•25

7

6

6

I200

7

O.480I

O.OOOI

• On employing a current of 6 amperes and a pressure of 5 volts
the solution became colorless in four minutes. Not a trace of

KOLLOCK-SMITH — ELECTRO-ANALYSIS.

145

nickel was found in the liquid after seven minutes. The amalgam
was very bright and of the consistency of soft dough, when one
gram of nickel was combined with the usual quantity of mercury
(40 grams).

qrams

48
4b

44
4^

4

38

,,

s

/

-

1

3

.28
26
24

18

,

12

.06
06

'

02

\

r

>

5
it

/

/

21

.16

1/

1

a

i

.08

/ ' '

/

Vr

1

\_r

t>

~1

1

S

^

5

Curve 6. Nickel — 2. 5 Amperes,
6 volts.

Curve 7.

volts.

Cobalt — 5 Amperes, 5

Cobalt.
This metal does not appear to enter the mercury with the same
rapidity as nickel under similar conditions. The last minute traces
are more difficult to remove. Various conditions were used. When
no sulphuric acid was added the current was at first low, but it
rapidly rose as the decomposition proceeded. The conditions,
giving the total cobalt in the least time, were the following : 10
c.c. of solution, containing 0.3535 gram of cobalt ; 0.25 c.c. of
sulphuric acid and a current of 5 amperes with a pressure of 6
volts. The speed of the anode was 1,200 revolutions per minute.
The solution became colorless in seven minutes, but ten minutes
appeared to be necessary for the removal of the last traces of the

14G

KOLLOCK-SMITH — ELECTRO-ANALYSIS.

[April 13,

metal. On using the same amount of cobalt in a volume of five
cubic centimeters, other conditions remaining unchanged, all of the
metal separated in seven minutes, thus :

In I minute 0. 1197 gram, of cobalt was deposited.
3 " 0.2930 " " "

5 " 0.3300 " " "

6 " 0.3520 " " "

7 " 0.3535 " » "
10 " 0.3530 " " "

The curve (7) was constructed from these results.
Cobalt.

i
e

V

0.

Cobalt
Present
in Gram.

Sulphuric
Acid
Present

Volume
in c.c.

Current.
Am-
peres.

Volts.

Revolu-
tions of
Anode
Per

Time

in

Minutes.

Cobalt

Found

in Gram.

Error in
Grams.

W

Minute.

I

0.3525

•35

■5

5

7

1250

15

0.3522

— O. COO3

2

0.3525

•25

15

3

5

980

18

0.3524

— O.OOO I

3

0.3525

•25

'5

4

6

GOO

14

0.3523

— 0.0002

4

0.3525

•25

10

4

6

860

16

0-3530

+O.OOO5

5

0.3525

•5

10

4

6

IOOO

15

0-3530

+ O.OOO5

6

0-3525

0

10

4

6

I24O

16

0.3528

+ O.OOO3

7

0.3525

•25

10

3

6

1200

IO

0.352I

— O.OOO4

8

0.3525

•5

10

6

6

I200

IO

0.3530

— O.OO05

9

0.3525

.25

10

5

8

SOO

IO

0.3522

— O.OOO3

10

0 3525

•25

10

3

8

I4OO

12

0-3523

— 0.0002

11

0.3525

•5

10

6

5

800

II

0.3530

— O.OO05

12

O.7050

•5

15

6

7

I200

3°

O.7052

+ 0.0C02

13

O.I762

•35

! 10

1 4

8

560

7

O.I762

Chromium.
A solution of chromium sulphate was electrolyzed with currents
varying from 1 to 4 amperes and 7 to 12 volts and with a varying
quantity of sulphuric acid. It was found by Myers that the addi-
tion of the acid was necessary, otherwise, there was a separation of
the oxide of chromium throughout the liquid ; but too much acid
retards or entirely prevents the decomposition. When 10 drops
{40 drops = 1 c.c.) were added and a current of 2.5 amperes and
6 volts applied one half hour was necessary to deposit o. 23 gram of
chromium. With 0.5 c.c. of acid and a current of 5 amperes and
4 volts the solution at the end of 60 minutes did not appear
to have lost its color. Experiments were then made to learn how

KOLLOCK-SMITH — ELECTRO-ANALYSIS

147

much chromium, if any, was deposited when the acid was present
in large quantity. Thus, with a current of 4 amperes and 7 volts,
solution containing 1 c.c. of acid, 0.05 gram of metal was precipi-
tated in forty-five minutes ; while with two cubic centimeters of
acid and a current of 1 ampere and 4 volts the mercury showed no
increase in weight after thirty minutes. The following results,
obtained in the use of smaller amounts of acid, confirm this. By
adding 10 drops of acid (= .25 c.c.) and employing a current of
4 amperes and 7 volts, the liquid became colorless in thirty minutes,
but forty minutes were necessary for the complete removal of the
metal. With the same quantity of acid, and a current of 5 amperes
and 8 volts, the chromium was completely precipitated in thirty
minutes. With five drops of acid and a current of 3 to 4.5 am-
peres and 8 volts, the solution became colorless in eleven minutes.
It, therefore, seems that more than three drops of acid are sufficient
to materially affect the rate of precipitation. More than two drops
of acid must be present to prevent the separation of chromic oxide

f

/

/

/

/

/

/

f

/

f

1

-/

1

1

Curve 8. Chromium — 3.5 Amperes, 11-10 volts.

which always took place with less than that amount of acid. The
following conditions gave the most rapid determination : A volume
of the solution, containing 0.1180 gram of chromium and three
drops of sulphuric acid (40 drops = 1 c.c), was electrolyzed with
a current of 4 to 5 amperes and six volts, the speed of the anode
being 400 revolutions per minute. In four minutes the solution
was colorless and in six minutes the chromium was found to be
completely deposited. The solution was siphoned off in the man-

148

KOLLOCK-SMITH — ELECTRO-ANALYSIS.

[Apri!

ner previously described, but after the cell was removed anhydrous
alcohol was poured in as quickly as possible and the operation re-
peated twice and followed by two washings with ether in order to
prevent, if possible, oxidation of the chromium. Oxidation, if it
occurred, was but slight, for the error never exceeded 0.0007 gram.
Curve 8 was constructed from the results given below :

In 2 minutes 0.048 gram of chromium was deposited.

' 4

' 0.085

' 6

' 0. 1000

' 8

' 0.1105

' 9

' 0.1 185

' 10

0.1185

Chromium.

s
1

0.

w

Chromium
Present
in Gram.

Sulphuric

Acid.
40 Drops
= 1 c.c.

Volume
in c.c.

Current
Am-
peres.

Volts.

Revolu-
tions of
Anode

per
Minute.

— u
lg

Chromium

Found
in Gram.

Error in
Gram.

1

0.1 1 So

5

IO-15

3-4

7

28o

J5

0.1 186

-f O.OOO6

2

0.1 180

3

IO-I5

2-4

1 1-9

280

i5

0.1 187

-f O.OOO7

3

0.1 180

3

IO-15

i-3

9

640

20 0. 1 185

- O.OOO5

4

0.1 180

3

8-15

1-5-3

10-8

220

15 0. 1 186

-f O. OO06

5

0.1 1 80

3

IO-15

i-3

1 1-9

520

20 0.1 186

-f O.OOO6

6

0.1 180

3

5-15

1-2

1 1-9

640

17 0.1175

— O.OOO5

7

0.1 180

3

5-15

2-4

9-8

480

15 j 0.1 180

8

0.2360

3

5-15

2-5

10

520

150 0.2355

— O.OOO5

9

0.1 180

5

5-15

3

7-5

400

15

0.1 179

— O.OOO I

10

0.1 180

3

7-i5 4-5

8

640

i 6

O.II75

— O.OOO5

11

0.1 180

3

7-15 3-4

10-9

640

10

0.1 180

12

0.1 180

7

7-15 ' 3-4

10-8

200

13

0.1 187

-[-O.OOO7

*3

0.1 180

3

5-15 3-5

8

640

1 II

0.1177

— O.OOO3

14

0.2360

4

5-i5 1 3

12

640

35

0.2359

— O.OOO I

15

0. 1 180

3

5-15 i 3-4

10-8

320

11

0.1179

— O.OOO I

16

0.1 180

3

5-15 3-4

10

540

0.1 182

^0.0002

Iron.
In experimenting with salts of this metal it was soon discovered
that sulphuric acid in large amount retarded its precipitation. It
was also noticed when higher currents were used that the solution
became very hot and assumed a decidedly pink color,1 which dis-
appeared on the addition of cold water or when the cover glasses
were removed, allowing the steam to escape rapidly and thus decreas-
ing the pressure and consequently the temperature of the boiling
solution. The color reappeared a few seconds after the cover glasses
were replaced.

1 Due to traces of manganese.

KOLLOCK-SMITH — ELECTRO-ANALYSIS.

149

The conditions which gave the most satisfactory results were as
follows :

Volume of solution 5 c.c. containing 0.275 gram of metallic iron,
3 drops of concentrated sulphuric
acid and a current of 3 to 4 am-
peres and 7 volts. The rotation of
the anode varied from 520-920
revolutions per minute. The iron
was completely deposited in seven
minutes. The following observa-
tions on rate of deposition were
made under the conditions just
given :

In 2 minutes o. 1760 gram of iron deposited.
"4 " 0.2000 " " "

(i 5 << 0.2050 " " "

" 8 " 0.2075 " " "

In addition to the results just
described a solution of ferrous sul-
phate containing 0.1945 gram of
iron in 10 c.c. was used to get
further working conditions. By
using a current of 3.5 A. and
10-9 volts, with about 900 revolu-
tions per minute of the anode,
the total iron content was deposited in fifteen minutes. The
residue from the decompositon cell when oxidized by nitric acid
and tested with potassium sulphocyanide gave no color. The

1

2

/r

IS

/

/

It

.13
12
.11

05

.02

fl

i

Curve 9.
volts.

Iron — 3.5 Amperes, 7

s

Revolu-

a

Iron
Present

Acid
in Drops.

Volume

Current.
Am-

Volts.

tions of
Anode

if

Iron
Found

Error in

&

in Gram.

40 Drops

peres.

„,per

HS

in Gram.

a

=1 c.c.

Minute.

1

O.2075

7

5

4-5

8-7

520

14

O.2072

— O.OOO3

2

O.2075

4

5-i5

5-4

b.5-5

680

14

0. 2078

+ 0.0003

3

O.2075

5

5-10

3-2-4

6.5

680

15

O.2077

— O.OOO3

4

O.2075

3

5

2-2.5

7-b

680

is

0.2073

O.OC02

5

O.2075

3

5

4

6-5

6SO

10

O.2080

-+-O.OOO5

6

O.2075

3

5

3-4-5

7-b

920

7

O.2078

+ O.OOO3

7

O.2075

3

5

2-3

6

740

9

O.2076

+O.OOOI

8

O.2075

3

5

2-4

6-5-5-5

700

9

O.2076

+ O.OOOI

150

KOLLOCK-SMITH — ELECTRO-ANALYSIS.

[April i3)

experiments made to determine the rate of deposition with a lower
current (1-2.5 amperes and 10-9 volts) while the other conditions

ls

.2

.18

--'

.16

/

14

I

.12

1

.01

1

/

i

06

02

1

15 Minutes.
Curve 10. Iron — 1 to 2.5 Amperes, 10-9 volts.

remained as above, gave the following results which appear in
Curve 10.

In 2.5 minutes o. 1141 gram of iron was deposited.

" 5 " 0.1787 " " » "

" 7.5 " 0.1945 " " " "

" 10 " 0.1950 " " " " "

When the residue was tested with potassium sulphocyanide no
iron was detected. By the addition of 3 drops (40 drops = 1 c.c.)
of sulphuric acid and using a higher current (3.5 amperes) in 10
minutes a faint reaction for iron was observed indicating that the
acid has some retarding influence. In fifteen minutes under these
conditions the iron had completely separated. By using a higher
current 3 amperes and 9 volts under the same condition all the
iron was deposited in ten minutes.

University of Pennsylvania.

SMITH— COLUMBIUM AND TANTALUM. 151

[Contribution from the John Harrison Laboratory of Chemistry.]

OBSERVATIONS ON COLUMBIUM AND TANTALUM.
BY EDGAR F. SMITH.

{Read April 13, 1905. )

In 1 So 1 Hatchett, while studying minerals in the British Mu-
seum came upon a specimen from Haddam, Conn., which attracted
his attention because of its rather high specific gravity and its bril-
liant black color. A portion of this material was given him for ex-
amination, with the result that he discovered in it a new metallic
acid, to the metal of which he applied the name columbium. It
was his earnest hope that he might obtain larger quantities of the
American mineral in order to exhaustively study the new element,
and it is of interest to remark that Hatchett fondly expected this
material assistance from Thomas Peters Smith, a member of this So-
ciety and an enthusiast in chemical science, who on his return from
England met an untimely death on shipboard.

In 1802 Ekeberg, of Sweden, while examining an unknown min-
eral, found that it contained a new metallic acid, to the metal of
which acid he assigned the name tantalum, because "when placed
in the midst of acids it is incapable of taking any of them up and
saturating itself with them." Later, Wollaston (1809) strove to
prove that columbium and tantalum were identical. In this he
failed. The few reactions known even at that early day differ-
entiated the new elements. About 1840, Heinrich Rose, in study-
ing similar minerals, from other localities, came to the conclusion
that the American mineral contained an element absolutely differ-
ent from tantalum, and called it niobium. Subsequently, owing
to his inability to account for the peculiar products which he got
by chlorinating a mixture of the oxide of the new element and car-
bon, he asserted that, in addition to niobium, there was present
pelopium (1846). Later (1853), however, he seems to have ar-
rived at the opinion that niobic acid and pelopic acid were differ-
ent oxides of niobium. The first he called niobic acid and the
second hyponiobic acid. Hermann, also, contributed to the un-
certainty which surrounded the two elements (columbium of Hat-

152 SMITH— COLUMBIUM AND TANTALUM. [April i3)

chett and tantalum of Ekeberg) in that he announced the existence
of ilmenium ; but it was not generally accepted by chemists. Her-
mann, however, persisted in his declaration that it occurred along
with the other two elements to which reference has been made. v.
Kobell believed that he had detected dianium in allied minerals.
This unfortunate state of affairs prevailed until the early sixties,
when Marignac, after a careful study of a number of minerals from
various localities, announced the existence in them of but two ele-
ments — the columbium of Hatchett and the tantalum of Ekeberg
— and added that the confusing reactions which had perhaps led
Heinrich Rose — but most certainly, Hermann and von Kobell —
astray were to be explained by the presence of titanium in all tan-
talites and columbites. It is only fair to say that Marignac never
succeeded in obtaining titanium from any one of the minerals in
which, according to him, it occurred, associated together with
columbium and tantalum. Indeed, in his concluding paper on
columbium he frankly acknowledged that the columbium com-
pounds, which he used for the determination of the atomic weight of
the metal, contained titanium, and that he knew of no method by
which the latter could be separated from columbium. Marignac's
conclusion was accepted by the chemists of the world as final.

When we come to examine the evidence which Marignac gives
for the presence of titanium, we find that it is, practically : that
the recrystallization of a double fluoride of columbium and potas-
sium, supposed to contain titanium, gave rise, gradually, to a frac-
tion which became more insoluble in water and the molecular
weight of its acid oxide approached, within ten or more units,
that required by titanic oxide. It is well to bear in mind that
at no time did Marignac, whose ability and keen insight one would
not for a moment question, give any tests which are ordinarily
regarded as indicating titanium. He assumed it to be present on
the evidence mentioned above, viz : the greater insolubility of
the double fluoride and an approximate molecular weight corre-
sponding to that required by titanic oxide. Ever since Marignac's
day chemists the world over have tacitly accepted titanium as asso-
ciated with columbium in columbites. They have also estimated
its quantity by methods suggested from time to time. One of these
methods is based on the color reaction which titanium salts give
with hydrogen peroxide. Its intensity, compared with that shown

i9°s ] SMITH— COLUMBIUM AND TANTALUM. 153

by a known amount of titanium, has been regarded by most
analysts as entirely satisfactory. For the determination of the
amount of titanium in columbium many methods have been pro-
posed, but this is not the place to discuss them or their value.

As late as 1877, in the last paper published by Hermann, he an-
nounced the element neptunium and claimed to have obtained it
from the acid mother liquors remaining after tantalum potassium
fluoride, ilmenium potassium fluoride and columbium potassium
fluoride had been crystallized out. The particular mineral in
which he observed it was a columbite from Haddam, Conn.; in
other words, the same mineral in which columbium had been orig-
inally discovered by Hatchett.

The existence of neptunium has never been contradicted. This
is probably because the majority of chemists thought that the ver-
dict in regard to the constitution of tantalites and columbites had
been given by Marignac, and that the numerous, unusual reactions
of the metallic acids contained in those minerals, noted and com-
mented upon at various times by Heinrich Rose, von Kobell,
Blomstrand and Hermann, were all due to the contaminating
influence of titanium.

The two elements, columbium and tantalum, in their derivatives,
have received comparatively little attention within the last quarter
of a century, although at intervals attempts have been made to
clear up the mystery which, in a certain sense, surrounds them.
In this laboratory, several investigations upon derivatives of them
have been made. These being not wholly satisfactory, about three
years ago, 50 lbs. of columbite from South Dakota and 25 lbs.
from Haddam, Conn., were worked up, with a view of getting an
abundant supply of starting-out material ; with the view, also, of
studying anew the various derivatives of both columbium and tan-
talum. In the year 1903-1904, Dr. R. D. Hall devoted, in this
laboratory, much time and labor to the double fluorides of tantalum
and columbium. He made a comparative study of the reactions of
the same with the reactions of titanium. His results have been
published, but from an examination of them it will be observed
that he was not able to find any tests, while using the double fluor-
ides, which differentiated titanium from columbium so thoroughly
that he could expect to obtain a complete separation of these two
most interesting elements. It is true that he was able, by precipi-

154 SMITH— COLUMBIUM AND TANTALUM. [April 13,

tating potassium columbium fluoride incompletely with ammonia,
to get a columbium oxide, which apparently gave no response to the
hydrogen peroxide test upon its application, or to the reagent called
chromotropic acid. Hence he inferred that he had eliminated the
titanium from the columbium. More recent work with large quan-
tities of material has demonstrated that in the latter cases it was
impossible to entirely remove the metallic acid which gave the
color tests. It was further found that by the action of sulphur
monochloride upon the oxides of columbium and titanium, cor-
responding chlorides were produced ; but again, it proved im-
possible to wholly expel the titanium from the columbium by
this process, notwithstanding titanium chloride is an exceed-
ingly volatile liquid and columbium chloride a solid, crystalline
body.

In the present year, we have, in this laboratory, prepared large
quantities of the double fluoride of tantalum and potassium, and
found no difficulty whatsoever in eliminating from it every trace of
what was supposed to be the titanium double fluoride.

Having thus, at our disposal, such generous amounts of pure
tantalic oxide, free from columbic oxide, in short, really pure tan-
talic oxide, it was determined to make a new study of the double
fluorides of tantalum with the alkali metals and organic bases.
This was undertaken in order to discover, if possible, why Dr.
Pennington, when working in this laboratory in 1895, obtained
double fluorides of tantalum and columbium with caesium, which
showed these unusual formulas: 15CSF. TaF5 and yCsF. CbF5,
which varied so widely from those generally followed by the
double fluorides of tantalum and columbium, and were not in ac-
cord with the law proposed for double halides (Amer. Chem. Jour.,
v. 291).

At the outstart it was thought that this re-investigation of the
double fluorides would prove to be an easy and simple matter.
But it was not long until it was seen that the discordant results of
Dr. Pennington were probably due to the fact that there was more
than one csesium tantalum fluoride. Indeed, the latest work done
in this laboratory, by Mr. C. W. Balke, proves that there are two
coasium tantalum fluorides, two rubidium tantalum fluorides, two
sodium tantalum fluorides, two ammonium tantalum fluorides, and
so forth, of these ratios :

i9°s-] SMITH— COLUMBIUM AND TANTALUM. 155

TaF5.CsF 2TaF5.3RbF

TaF5.2CsF TaF5.2NaF

TaF5.2NH4F TaF5.3NaF

TaF5.3NH4F TaFs.2KF

The existence of several such double fluorides with each of the
alkali metals naturally raises the question whether these salts ought
to be used for the determination of the atomic weight of tantalum,
inasmuch as each salt is likely to be contaminated with smaller or
larger quantities of the other, depending upon the condition or the
care with which they are prepared. Marignac used potassium
tantalum fluoride and ammonium tantalum fluoride in his re-deter-
mination of the atomic weight of tantalum. It would seem, from
the study of the salts just mentioned, that even this skilled and
careful analyst could not have been sure that he had a definite,
homogeneous body in the determinations which he made. Of
course, if there was even a slight amount of a second salt in the salt
used for the atomic weight work, it would naturally vitiate the final
result. Of all the double- fluorides of the alkali metals and bases
with tantalum which have thus far been studied by Mr. Balke, that of
sodium and tantalum, of the ratio 3 to 2, seems to be the one having
some definite and most stable characteristics. We hope to re-deter-
mine the atomic weight of tantalum, but it is not probable that we
shall use any one of the double fluorides, of which mention has been
made, although they appeal strongly because of the ease with
which they can be crystallized. The uncertainty, however, as to
whether they are really absolutely of one definite ratio every time
that they are crystallized is uncertain. Hence they had better be
abandoned in atomic weight determinations.

The question may also be asked, may not the double fluorides of
colurnbium with the alkali metals, which have been used for atomic
weight purposes, been contaminated with salts of varying ratios ?
This point will receive attention.

Turning again to colurnbium, it seems proper to record that
having eliminated the tantalum completely from a mixture of oxides
obtained from Haddam columbite, the remaining potassium colurn-
bium oxy-fluoride was crystallized a number of times from water
and also from solutions containing much hydrofluoric acid. This
procedure finally gave a mother liquor that was decidedly acid. A
metallic acid remained in this mother liquor. According to Her-

PROC AMER. PHILOS. SOC XLIV. l8o. K. PRINTED JULY 31, I905.

156 SMITH— COLUMBIUM AND TANTALUM. LApm .3.

maun, in his communication of 1877, this acid should be neptunic
acid. Therefore, the acid mother liquor was treated as directed
by Hermann; namely, it was evaporated, the- residue was dissolved
in water and the boiling solution precipitated with an excess of
caustic soda. The precipitate, after the liquid had become cold,
was filtered out, pressed thoroughly from adherent water and then
boiled with 25 times its own weight of pure water. Everything
dissolved. The solution was perfectly clear. On cooling, there
separated from it the beautiful needle-like crystals of sodium colum-
bate. According to Hermann, the precipitate which was collected,
pressed out and then boiled with water, should, if neptunium were
present, have left a slimy mass, insoluble in water. This, Her-
mann said, was sodium neptunate. It should be observed that our
experiments were made with the final acid liquors obtained from
the double fluorides present in columbite from Haddam ; further,
that we proceeded in strict accordance with the directions of Her-
mann and having done all this, did not obtain a gelatinous mass
which might have been sodium neptunate. In Hermann's com-
munication, to which reference has been made so frequently, he
lays great stress on the fact that the distinguishing reaction of
neptunium is the beautiful golden yellow color which sodium nep-
tunate imparts to a salt of phosphorus bead in the reducing flame.
It is needless to add that we tried on different occasions to find
neptunium, according to the directions of Hermann ; but our
search was fruitless. On one occasion, however, we obtained a
mass, not great in amount, which, in the inner blow-pipe flame,
did impart a yellow color to the salt of phosphorus bead, but more
careful examination of this residue demonstrated that it contained
tantalum, iron and some columbium. The intense golden yellow
color, which was so strongly emphasized by Herman, we could not
get ; so that it is very probable that neptunium, like ilmenium and
the other metals announced from time to time as present with
columbium and tantalum must really be placed in the list of defunct
elements. It has not been our wish to bury this candidate for
elemental honors. Indeed, we would have been only too glad to
have found the evidences of its existence and to have confirmed the
observation of that earnest and sincere student of chemical science,
who, in his tireless labors, frequently felt confident that he had
fallen upon the cause of the varying results observed with colum-
bium and tantalum.

I905-] SMITH— COLUMBIUM AND TANTALUM. 157

In this connection it may be added that, having freed the tanta-
lum and columbium oxides as thoroughly as possible from ordinary
contaminations, the problem of removing tungsten and tin con-
fronted us. After much experimentation, we found that the cer-
tainty of the removal of these impurities could only be had by
fusing the tantalum and columbium oxides with sodium carbonate
and sulphur. It is true that small quantities of tantalum and
columbium will be lost, being carried along with the tungsten and
tin, but as we were seeking a method of purification and not a sepa-
ration, we adopted this course. It is the one which was pursued
by Heinrich Rose. Our own experience leads us to say that the
removal of tungsten and tin from columbium and tantalum oxides
cannot be realized by digestion with ammonium sulphide. Indeed,
not only did we find the fusion with the sodium carbonate and sul-
phur necessary, but that working in large quantities of material, as
in our case, two and three refusions with these reagents were found
necessary. Another point of interest in connection with the puri-
fication of the tantalum and columbium oxides may be mentioned.
It has frequently been said that in crystallizing out the double
fluorides of these metals, if titanium be present with them, it will
be found in the potassium tantalum fluoride. We have encountered
no difficulty in getting potassium tantalum fluoride perfectly free
from what is supposed to be titanium by one or two crystallizations.
It has been assumed that as potassium titanium fluoride is rather
insoluble in water, it would naturally go with potassium tantalum
fluoride. This, however, seems to be an incorrect observation. It
masses with the columbium potassium fluoride ; at least the element
which gives the yellow color with hydrogen peroxide, or a rose red
with chromotropic acid is always found associated with the colum-
bium. How to free the columbium from titanic acid we do not
know. We are in precisely the same position as that of Marignac,
notwithstanding we have probably made greater efforts than he to
remove it from the columbium. It is this point in our investigation
upon which we have been continuously at work for the last year.
We have tried fractional precipitation with ammonia water, frac-
tional crystallization of the double fluorides, fractional chlorination
of the oxides in the presence of carbon and the action of numerous
organic bases, without finding any way of effecting a separation.
Indeed, the separation of columbium and titanium is a problem

158 SMITH— COLUMBIUM AND TANTALUM. |Apnii3)

which the analyst has not solved up to the present. As remarked,
it has received and is receiving our daily attention. Once having
achieved this result and having definitely determined the character
of the color-giving metallic acid, or proved it to be titanium, with-
out any further doubt, we then hope to subject the purified coiumbic
oxide to a searching review in all its derivatives, just as we are now
doing with the compounds of tantalum. In anticipation, it may
be said that there are some most interesting complexes of tungstic
acid with tantalic oxide and also of tungstic acid and coiumbic
acid. These are under study at present. These complexes, also,
have brought analytical problems that are most puzzling. Yet our
progress with them leads us to hope for a separation and a satisfac-
tory solution of the same.

Some attention has likewise been given to per-tantalates and per-
columbates. It would not be the least surprising to find these de-
rivatives answering admirably for atomic weight work.

University of Pennsylvania.

CONDITIONS OF THE TRADES-MONSOON AREA. 159

ENQUIRY INTO THE PRESSURE AND RAINFALL CON-
DITIONS OF THE TRADES-MONSOON AREA.

BY W. L. DALLAS.

{Read April 14, igoj.)

In 1900 the writer undertook the discussion of the seven mon-
soon seasons 1893 to 1899 and showed that during those seven
years there occurred a series of oscillations of pressure and that be-
tween these oscillations and the monsoon rainfall over India there
existed a very distinct and marked relationship. The data used in
this discussion consisted of the mean monthly and seasonal varia-
tions of pressure over India, derived from all the stations employed
in the Daily Weather Report of the Meteorological Department,
and the mean monthly and seasonal variations of pressure over the
Equatorial Belt and the Arabian Sea as given by the pressure obser-
vations recorded (1) at the Seychelles, Zanzibar and Mauritius and
(2) on board ships traversing the Arabian Sea and the South-east
Trades Region.

The relationship as established for those seven monsoon seasons
was as follows: (1) The Indian monsoon rainfall was in defect
during the rising portions of these pressure oscillations and in
excess during the falling portions while the amount of the rainfall
variation agreed directly with the rapidity of the pressure changes.
(2) The pressure oscillations exhibited a periodicity of about four
years.

It was carefully pointed out at the time that the discussion dealt
solely with the seven years under review so that, though the agree-
ment there disclosed was exact and clear, it was obvious that a
much longer series of observations would be required before it
would be safe to assert that the period of the oscillations and
the relationship between the pressure oscillations and the rainfall,
as disclosed in the discussions, could be accepted as having a gen-
eral application. As a matter of fact, before the publication of the
paper, it had already become apparent that the relationship had
not been maintained, while a simple examination of the existing
rainfall data of India showed that there does not obtain any simple

160 DALLAS— PRESSURE AND RAINFALL [April 13,

four-year cycle in the Indian rainfall. The author believes that
these four-year oscillations form the basis of the weather changes
over the Indian monsoon area, though there occur at times violent
or spasmodic interruptions, the cause of which is not as yet appar-
ent, and that these interruptions are the cause of the great irregu-
larities in the course of the pressure cycle and in the occurrence of
the variations of rainfall. Since the history of these seven mon-
soon seasons was written, Professor Bigelow's " Contributions toCos-
mical Meteorology " ] has appeared. In it the following paragraph
occurs :

" The increase of solar magnetic intensity is synchronous with a
diminution of temperature but with an increase of pressure and this
function persists throughout every phase of the research. In spite
of some irregularity there is a distinct conformity in the general
sweep of these curves and also in the tendency to describe crests
during the same years. Indeed the occurrence of four subordinate
crests in the n -year periods suggests strongly that a 2 f -year period
is superposed upon the long sweep of that period curve. Appar-
ently this minor period is the basis of the seasonal variations of the
weather conditions of the U. S. A. more than anything else, so
that in long range forecasting this period must be very carefully
considered. ' '

It will be noticed that the period of these minor oscillations as
then determined by Professor Bigelow was 2f years for the United
States. Subsequently Professor Bigelow produced his "Report on
the Barometry of the United States" and from the complete data
there employed he obtained an eight-year cycle of pressure which
is a simple multiple of the four year cycle determined for the Indian
Monsoon Area. Professor Bigelow's researches terminate with the
year 1899 but where they overlap the Indian series of observations
the principal characteristics of the two series agree. Thus Pro-
fessor Bigelow obtains a maximum in 1896 and a minimum in 1898
with pressure rising again to a maximum through 1899. In the
Indian equatorial area the period is approximately four years, and
the writer agrees in believing that these minor oscillations of
pressure are mainly influential in determining the seasonal varia-
tions of weather. On this point it appears probable that the ex-

1 See Monthly Weather Review, July, 1902, and especially, Weather Bureau
Bulletin, No. 21, pp. 125-6, Washington, 1898.

I905-] CONDITIONS OF THE TRADES-MONSOON AREA. 161

perience of the Indian area will be found to correspond with that
of the United States. It must however be born in mind that the
investigation is one of extreme complexity and that superposed on
the four-year or minor oscillation there are great irregularities
which cannot now be explained but which at times completely upset
the regular course of the cycle. Notwithstanding these irregulari-
ties and interruptions it appeared to the writer that in face of the
remarkable agreement between the pressure oscillations and the
rainfall during the years 1893 to 1899 'li was worthwhile to con-
tinue the discussion in a more exact and detailed manner so as to
determine (1) over what area the pressure oscillations extended
(2) how far they agreed in amplitude and in time throughout the
affected area and (3) what relation the rainfall of the whole mon-
soon area bore to the pressure oscillations.

The author has collected and discussed a large amount of ma-
terial and has arrived at certain conclusions which he regards as
tentative and far from satisfying. He feels doubtful if the obser-
vations would fulfil the requirements which Professor Schuster laid
down as a means of estimating the reality of the periodicity, but the
investigation has brought out certain relationships which appear at
least worthy of record.

The tentative conclusions arrived at are as follows :
(1) That over the trades monsoon area — and most markedly
so over the equatorial belt — there occur four-year oscillations of
pressure; (2) that during the rising portions of these oscillations
the general rainfall of the trades monsoon area is below, and dur-
ing the falling portions is above the average, with a well-marked
minimum of rainfall in the first year of the cycle and a well-marked
maximum of rainfall in the third year ; (3) that from the Antarctic
or extreme southern regions there emanate at irregular intervals
rays or streamers of varying extent and intensity which occasion
increased atmospheric pressure over the affected area; (4) these
rays or streamers are apparently not in the least in the nature of
waves, as they affect large areas practically simultaneously and con-
tinue for considerable periods; (5) when these rays or streamers
are frequent and extensive, as in portions of the years 1899 and
1900, pressure ranges largely above the normal, but exhibits large
oscillations or fluctuations ; when on the contrary they are absent
as in portions of the years 1898-1899 pressure is low and the oscil-

162 DALLAS— PRESSURE AND RAINFALL [April 14.

lations small; (6) these variations are superposed on the four-
year cycle of the tropical belt, and are spasmodic, occurring at
irregular intervals over irregular areas so that their influence oc-
casions irregular variations of rainfall and irregularities in the pres-
sure cycles.

There appears to be no satisfactory explanation either of the four-
year cycle of pressure over the trades monsoon area or of the
irregular spasmodic disturbances of pressure referred to above.
With regard to the cycles it is possible that compensatory actions
are at work, so that when atmospheric pressure increases in one
part of the world it decreases in another, though the evidence of
the barometry of the United States is opposed to this and rather
suggests that the principal secular variations of pressure are of a
uniform character over the whole globe. It is impossible to believe
that the variations of pressure are a result of variations of rainfall.
For one thing, the variations are as marked in a dry area like Aden
as in a wet area like Bombay, and for another, the evidence, so far
as it can be sifted, shows that the variations of pressure precede the
variations of rainfall. Thus the increase of pressure which culmi-
nated in the large excess of pressure in the months of July, August
and September, 1899, commenced in February of that year, thus
preceding by some months and not succeeding the scanty rainfall of
that season.

The memoir contains all the figures and data on which the
enquiry is founded. Some of the observed changes are at present
quite inexplicable, but the observations are given as recorded so
that though the author has not succeeded in obtaining any con-
clusive results, it may be possible for other students of meteorology
with more available leisure to work them into a more harmonious
scheme.

In order to undertake this detailed examination the employment
of pressure or rainfall means of large areas has been abandoned,
and instead the actual monthly pressures and their variations for
certain selected stations, which it is believed represent fairly ade-
quately the whole monsoon area, have been used. The list of
stations includes : Batavia, Calcutta, Bombay, Aden, Cairo, Mad-
ras, Colombo, Seychelles, Zanzibar, Mauritius, Durban and Perth,
while in addition the marine observations of the Arabian sea and
the Equatorial belt have been utilized to obtain averages for those

igos ] CONDITIONS OF THE TRADES-MONSOON AREA. 163

areas, mainly with the object of determining whether, in the case
of these pressure oscillations, there occurs any horizontal transla-
tion in the pressure changes or whether they occur simultaneously
throughout the whole area.

Numerous tables and figures are given in the memoir to assist in
the discussion of the observations.
Meteorological Office,
India.

Ki4 MERRIMAN— DEPTH OF BRIDGE TRUSS. [April

THE RELATION BETWEEN THE ECONOMIC DEPTH

OF A BRIDGE TRUSS AND THE DEPTH

THAT GIVES GREATEST STIFFNESS.

BY MANSFIELD MERRIMAN.

( Read April 14, 1Q05. )

The fact that there is a certain depth for a bridge truss which
renders the quantity of material a minimum has long been known,
and the marked increase in the depth of bridge trusses which has
occurred during the past quarter of a century is due to the efforts
of manufacturers to use the least possible amount of material. It
has generally been supposed that the vertical deflection of a bridge
under a moving load decreases with the depth, and this is true for
plate girders. For a truss, however, investigations made by the
author show that the least deflection and hence the greatest stiffness
increases up to a certain limit, as the depth increases, and then
decreases, so that there is a depth which gives the truss its greatest
vertical stiffness.

The following are the results obtained by the author for the type
known as the deck Pratt truss. Let / be the span, d the depth, p
the panel length, and n the number of panels, so that /= np.
The economic depth was obtained by forming an algebraic expres-
sion for the amount of material in the truss in terms of its dimen-
sions, given loads and allowable unit-stresses, and then finding the
value of d/p which renders that expression a minimum. There
were found,

for n

= 4

8

12

d\P

= 1.29

i-73

2.08

d\l

= 0.32

0.22

0.17

2.65 3.21

0.13 O.I I

which shows that djp increases with length of span while ^///decreases
with length of span. To determine the depth that gives greatest
stiffness, an algebraic expression for the stored energy in the truss
due to the deformation of its members was formed and this equated
to the deflection due to the given loads. Then the values of djp
that render this expression a minimum were deduced for different
values of n, as follows :

igos] MERRIMAN— DEPTH OF A BRIDGE TRUSS. 165

30

n =. 4

8

12

20

d\p - 1.29

1.63

1.92

2.

4/ = 0.32

0.20

0 16

0.

0.09

which give laws similar to those of the economic depth, and which
show that the depth which gives the greatest stiffness is slightly
less than the economic depth. It hence appears that no additional
stiffness can be imparted to a bridge by giving to the truss a depth
greater than the economic depth.
April, 1905.

166 McCLELLAN— USE OF OSCILLOGRAPH. [April i4.

ON THE USE OF THE FALLING PLATE OSCILLO-
GRAPH AS A PHASE METER.

BY WILLIAM McCLELLAN,

RANDAL MORGAN LABORATORY OF PHYSICS, UNIVERSITY OF PENNSYLVANIA.
(Read April 14, /QOJ.)

The wave form of a periodic quantity is the curve which shows
the magnitude of the quantity for each instant of time. It is
always interesting and careful examination reveals relations that
could hardly be discovered in any other way. In alternating cur-
rent calculations, however, little can be done until the wave form
is known accurately. There are two general methods in use, by
which it may be determined — the point to point method and the
oscillograph method. In the first, the quantity is measured by a
meter, through which the circuit is closed, by a revolving contact -
maker, for an instant at any part of the wave for which it may be
set. The meter then indicates the value of the quantity at that
particular point only. By taking such readings at various points
in the cycle, the whole wave may be plotted. As this process is
somewhat laborious, various instruments, called wave tracers, have
been designed to facilitate the operation. In the Rosa curve tracer
a double potentiometer is used. The operator fixes his eye on the
galvanometer, and produces balance by means of a small crank,
which turns the cylinder carrying the potentiometer wire. When
this occurs, a second lever is pulled, which automatically prints a
point of the curve on a paper fixed in the proper position, and also
turns the contact-maker to the next position.

Either of the foregoing methods -requires considerable time to
plot a whole curve. The successive points are obtained from dif-
ferent waves. For example, a good operator can get a curve in
five minutes if the instrument is in order. If he is working on a
sixty-cycle circuit, he has obtained his curve from 18,000 succes-
sive waves. It will be a true curve, therefore, if he has kept his
conditions absolutely constant in the interval. This is always
troublesome to do, but particularly so in commercial work where
the operator seldom has control of the generator. To avoid this

igo5.]

McCLELLAN— USE OF- OSCILLOGRAPH.

167

difficulty, the oscillograph has been devised, by which it is possible
to obtain the form of a single wave, or a number of successive
waves.

The oscillograph is essentially a galvanometer of very short
period. The one used in this work as shown in Fig. i is of the
moving coil type, made under the Duddell patents. The field is

Fig.

supplied by an electromagnet, the coils of which are wound in sev-
eral sections, so that different voltages may be used for the exciting
current. The normal current nearly saturates the core, so that
slight changes in the value of the current do not cause appreciable
changes in the strength of the field. The coil consists of an in-
verted U with the ends rigidly fastened at the bottom by a rubber
block, and connections made to the binding posts. The upper
loop is threaded over a small pulley, to which is fastened the spring
by means of which the tension is applied to the strips. There are
two distinct loops, thus permitting the taking of two curves simul-

168

McCLELLAN— USE OF OSCILLOGRAPH.

[April

taneously. The free period is approximately one ten-thousandth
of a second, undamped. There are three mirrors, one fastened on
each coil, and one fixed in the center, to give a zero line. The
maximum current used is about one tenth ampere.

To obtain a curve, it is necessary to provide uniform motion
perpendicular to the motion of the mirrors. In this instrument it
is accomplished by means of a falling photographic plate. This

motion is uniformly accelerated, but the error in the length of the
plate is very slight, though measurable. The error amounts some-
times to about a half per cent, of the wave length. The arrange-
ment can be understood from Fig. 2. The galvanometer M is
placed in a camera as shown. This is provided with a slit in the
end, through which parallel light is sent. The ribbon of light

i9°5.] McCLELLAN— USE OF OSCILLOGRAPH. 169

falls on the three mirrors of the galvanometer, and is reflected to
the cylindrical lens CL. This renders the parallel ribbon of light
a point. The lens is focussed so that this point is in the plane of
the photographic plate. The chute through which the plane is
dropped is about ninety-five centimeters long, giving the plate a
speed which allows the record of one twenty-five cycle wave on a
four by five plate. Light enters the chute through a slit which is
provided with the hand shutter S. The shutter is open when the
plate is dropping, but is closed before the plate is pulled back to
the top so that the slide may be inserted. This prevents any possi-
bility of fogging. The plate is carried in an ordinary wooden four
by five plate holder. This in turn is held by a light wooden car-
riage H, which is provided with springs on the sides and back.
These are adjusted so that they just bear on the surfaces of the
chute, thus providing a very steady motion of the plate during the
fall. The springs also serve the purpose of holding the front of
the plate holder tight against the chute. The plate is started by
means of a bulb release R, and is stopped by an airdash pot D.P.
The bottom of the carriage is provided with a leather packed brass
piston which fits the cup. Light is provided by a powerful 25-
ampere arc light, which has the usual condensers. In addition, to
get a proper parallel beam, the concave lens L is provided. The
whole camera is provided with leveling screws, in addition to those
for the galvanometer. This is necessary, since the chute must be
vertical. The low potential currents used in this work were brought
to the galvanometer by lamp cords, which passed through corks in
the side of the box. For convenience in lifting the plate holder
and carriage the sliding bracket F is provided. This is raised by
the knob and string 7\ lifting the carriage until it catches in the
release apparatus. The bracket then drops to the bottom of the
box. It is held in place by a rod, on which it slides. For access
to the camera, the whole top of the box is arranged to slide in a
light tight groove.

To adjust the apparatus for a curve the chute is first made per-
pendicular by the outside levelling screws. Then the galvan-
ometer is levelled with its own screws. The arc light is then
adjusted so that a strong beam of closely approximate parallel light
falls on the mirrors. The galvanometer is then adjusted, if need
be, so that the images pass through the center of the slit. Since

170 McCLELLAX— USE OF OSCILLOGRAPH. [April 14,

the mirrors have practically the same horizontal axis, though they
are not in the same vertical plane, necessarily, the spots when
focussed on the plate will be in the same horizontal line. They
must be adjusted, however, until they have the same vertical axis
when not vibrating. This adjustment, as well as the focussing, is
done by means of a small glass cylinder with a ground glass end.
This is entered through the back of the chute, and is of such a
length that the ground glass' is in the plane of the falling plate.
The focussing is finished and the cylinder removed. After a time,
that is with some experience, the focussing can be done from the
front of the box. There is no adjustment for the verticality of the
mirrors. The coils may be twisted, however, so that they may be
brought to various horizontal positions by adjusting screws on the
side of the standard.

The double oscillograph permits the simultaneous taking of two
independent curves. Since the loops are so fine, and the area
of the field so large, comparatively, the loops move in a constant
field. The amplitude of the wave is therefore proportional to the
maximum value of the current passing through the loops. This
will be so only when the damping is critical, that is, sufficient to
prevent running past the static position for the same current, and
not too much to prevent the loop reaching its static position. A
proper adjustment of a non-inductive resistance in series with the
loops would make the deflection a definite fraction of an ampere
per millimeter. One of the suggested uses to which the double in-
strument may be put is to obtain the current and potential differ-
ence curves for the same piece of apparatus. That this cannot be
done, at least exactly, will be apparent from a little study of the
conditions. The problem is similar to the wattmeter problem, in
which there is always a slight error, due to either the current or the
E.M.F. required for one of the coils. The inference is also fre-
quently suggested that the difference in phase, as indicated by the
record of the two curves, is the true difference of phase between
the quantities. That this is never true, and seldom approximately
so, will be apparent from the following discussion :

The simplest method of putting the oscillograph in circuit is
shown in Fig. 3. L is the apparatus for which the current and
E. M. F. curves are to be determined. Ox and 02 are the two loops of
the oscillograph joined in circuit as shown. Usually it would be

1905.]

McCLELLAN— USE OF OSCILLOGRAPH.

171

necessary to have a non-inductive resistance in series with each
loop to cut down the current to a proper value. A non-inductive
shunt is also frequently used with the current loop. Fig. 4 pro-

u

vides an analysis of the quantities involved. Let E be the poten-
tial difference between A and B. Lagging behind this at an angle
a is the current in L(i)3. Also lagging behind E at an angle d
is the current in 02(i2). But the current through Ox is the vector
sum of i and / or i . Now the angle, or rather the space on the

plate equivalent to the angle, is of course the angle of phase
between t\ and i2 and not between E and /3, as is frequently in-
ferred. It may approach in certain cases, but it is never the true
value. The angle desired, that is a, is given by the relation

a=8+(pjr0. (1)

PROC AMER. PHILOS. SOC XLIV. l8o. L. PRINTED AUGUST I, I905.

17:

McCLELLAN— USE OF OSCILLOGRAPH.

[April 14,

To obtain it we have the following derivation :

i\ _ _ sin ,5 sin (<p -f- 0)
t\ sin 0 sin 0

r = sin (f cot 0 — cos <p,

r

cot 0 = . cot cr,

sin y

. *. a = cot-1 I cot 0 I + a> + d.

\ sin vr ' /

Before we can determine a, therefore, we have to determine 8

(2)

Fig. 5.

The latter may be obtained by putting a non-inductive resistance
in place of L. This is shown in Fig. 5, from which as before

3 = cot-1 I - ; - — cot if' I 4- a'
\ sin e' /

(3)

where

r' =

5MM

To illustrate, and provide a test, arrangements were made as in
Fig. 6. An alternator which was under the control of the oper-

McCLELLAN— USE OF OSCILLOGRAPH.

173

ator, so that its speed could be maintained constant, was connected
through a lamp resistance, a variable standard of inductance, and
the loop Ov in series. The other loop was connected across the
inductance. The resistance of the standard of inductance was
9.89 ohms in all positions. The angle of lag for any given posi-
tion was therefore easily calculated. Moreover, as the resistance
of an oscillograph loop with its fuse, is also about ten ohms, quite
a difference between a and <p could be expected. Now we have to
determine r, <p, and 8, in order to determine a. The procedure
was as follows. With the inductance set to some definite value,
Z, a plate was taken. Such a plate is shown by Fig. 7. With

Fig. 9.

the standard set to zero another plate was taken. This is shown
in Fig. 8. Now to obtain the ratio r, we have to measure the
ratio of the amplitudes of the two waves. As these waves are
obtained from two different loops, it will be necessary to obtain
the ratio of the galvanometer constants of the two loops. A third
plate was taken with the two loops connected in series, and one
current passing through both. For accuracy in measurement, the
loops were connected oppositely, so that an apparent phase differ-
ence of 180 degrees results. This is shown in Fig. 9. The fol-
lowing measurements are then made. C = the ratio of galvanom-
eter constants of the two obtained from ratio of amplitudes in
Fig. 9.
r= iji2 obtained from ratio of amplitudes in Fig. 7.

174

McCLELLAN— USE OF OSCILLOGRAPH.

[April

A 4 and l^, wave-lengths obtained from Figs. 7 and 8 respectively.

/^ and l^, phase displacement, obtained from Figs. 7 and 8
respectively r obtained from ratio of amplitudes in Fig. 7.

During the taking of each plate the value of the frequency must
be observed.

R = resistance of the standard of inductance. We then have the
following :

= /36o,

9 = ; 36°-
V

By substituting <p' and r' in (3) we get 6, By substituting <s, d
and r in (2) we get «. The true value of a is given by the relation

tan"

Values for a as obtained from five different sets of plates are
given. A calculation on the first plate showed that 0 was prac-
tically equal to <p' , so that (3) was not used.

No.

H

'*

V

v

&

*

e

a

L

Tan"1 -g

degrees.

degrees.

degrees.

henrys.

degrees.

I

1. 14

1.74

.11

1.36

1.76

.025

5-i°

22.8

32-3

59

.040

56.8

2

M5ro

1.79

.10

i-5S

i-73

.025

5-2

20.0

23-5

47

.030

48.8

3

1. 121

1.98 .125

1.36

2.02

.02Q

3-0 °vo

22.3

31.8

59

.040

56.8

4

M3H

2.17-13

1.36

2. 1 1

.025

4-3 4

21.6

30.5

57

.040

56*. 8

5

1.12

2.08 1.13

i-37

2.12

.025

4.2

22.3

3i-8

59

.040

56.8

R = 9.

377

It is evident, that in the simplest application of the oscillograph,
a mere inspection of the plate is not sufficient to obtain the true
difference of phase. The general case is much more complex.
This is represented by the diagram of connections in Fig. 10, and
the vector diagram in Fig. 11. It is necessary to introduce a series
non-inductive resistence in circuit with 02, and a shunt on Ov in
order to bring the current to a proper size. The letters have the
same significance as before, and in addition s is the resistance of the
shunt, R' is the series resistance, i, is the total current, and i, is
the current in the shunt. As before, the angle {Etz) is the quan-
tity desired. Now a plate obtained with the connections ar-
ranged as in Fig. 10, would give us /,, i\, and the angle (*,*',),

i9°5] McCLELLAN— USE OF OSCILLOGRAPH. 175

granting that the loops have been calibrated so that current values
may be measured by them. It is evident that if the multiplying
power of the shunt for direct currents is known, it can be obtained
from z\ without serious error. Also since s is non-inductive, the
angle (i/3) can be known from a plate taken similarly to Fig. 8.
Then the angle (/,/,) can be calculated, and subtracted from the
angle (i\i.,) leaving the angle (/>",). Having the latter angle, with
it and /.,, (/>"3) can be calculated, and added to (/>',), and (Ef),
which is known as before, thus giving Ei3, which is the desired
angle.

The writer has not worked the latter discussion out for two
reasons. First, the errors of measurement on a photographic plate

U

Fig. io.

would not permit of any accurate results since in any case the
angles are small. Second, in the large majority of cases, the mul-
tiplying power of the shunt is so large, that the currents /, and i
are in the same phase, so far as could be measured. Also the re-
sistance in the 02 circuit so large that E and i are in the same
phase. Therefore the angle desired can be obtained exactly as 8
was in the preceding discussion, with the introduction of the multi-
plying power of the shunt. The use of ammeters in some of the
circuits would greatly facilitate matters in many cases.

The above discussion shows that, with the oscillograph, the phase
angle can be calculated. In no case can it be obtained by a simple
measurement on a single plate. Owing to the thickness of the lines
traced by the moving spot, measurements cannot be made closer
than a quarter of a millimeter, so that a long curve must be obtained
if any accuracy is to be attained, especially with small differences
of phase. While the double oscillograph could not be recom-
mended in any way as a phase meter, yet it does permit an ap-

176

McCLELLAN— USE OF OSCILLOGRAPH.

April

proximate value of this quantity being obtained. In most of the
cases to which the instrument has been applied, the process would
be quite simple, owing to the magnitude of the quantities involved.
The fact that the angle can be obtained is valuable in those cases
where curves are taken under fleeting or unknown conditions, in
which other methods could not be applied.

i9o5.] HALL AND SMITH— COLUMBIUM. 177

[Contribution from the John Harrison Laboratory of Chemistry. J

SOME OBSERVATIONS ON COLUMBIUM.

BY ROY D. HALL AND EDGAR F. SMITH.

( Read May ig, igoj. )

The starting-out material in this study was columbite from Law-
rence County, South Dakota. Its specific gravity equaled 5.86.
It contained 81 per cent, of the mixed oxides of columbium and
tantalum. The total quantity of substance decomposed by fusion
with acid potassium sulphate (58.5 kilograms) was 21.3 kilograms.
Each fusion was made in a platinum dish, using 100 grams of
mineral and 275 grams of acid potassium sulphate. While still
liquid the mass was poured into a porcelain dish. When cold the
fusion separated readily from the dish and was broken into small
pieces, which were boiled with water in large No. 11 evaporating
dishes until thoroughly disintegrated ; when they were transferred
to precipitating jars and the hydrates washed by decantation until
the wash water gave no precipitate or only a slight precipitate with
ammonium hydroxide. The solution and the washings were evapo-
rated to dryness. The residue was designated part I. The moist
hydrates of columbium and tantalum were covered with ammonium
sulphide and allowed to stand for several days. This ammonium
sulphide solution was decanted and designated part II. The re-
maining oxides were finally treated with very dilute sulphuric acid
and then thoroughly washed with water. The washings and the
diluted sulphuric acid solution were also evaporated to dryness and
marked part III. The residue labeled part I contained potassium
and the bases from the mineral in the form of sulphates. It was
dissolved in water, poured into five-gallon jars and there precipi-
tated with a slight excess of ammonium hydroxide. Having de-
canted the supernatant liquid the precipitate was washed once with
water, after which the hydrates were covered with a solution of
ammonium carbonate and allowed to stand for several days. The
ammonium carbonate solution was then siphoned off, acidulated
with dilute hydrochloric acid, and any metal present precipitated

178 HALL AND SMITH— COLUMBIUM. [May i9j

with a slight excess of ammonium hydroxide. The hydrate ob-
tained in this way was dissolved in dilute hydrochloric acid, and
an excess of ammonium carbonate, together with ammonium sul-
phide, added to its solution. The iron separated in the form of
sulphide and with it there was a small amount of titanium. The
filtrate from this precipitate was boiled with hydrochloric acid and
ammonium hydroxide added. The hydrate which was precipitated
was ignited and tested as to its photographic power. It gave a
picture after an exposure of five days. It contained uranium. It
showed the greenish color characteristic of U308. After this it was
fused with acid potassium sulphate, taken up in water and the solu-
tion boiled. A precipitate formed on cooling but disappeared on
warming, thus indicating the presence of members of the cerium
group or of zirconium. On the addition of ammonium hydrate
there separated a hydrate which, after filtration and washing, dis-
solved completely in a solution of oxalic acid. This pointed to
the presence of zirconium. The hydrates were again precipitated,
dissolved in sulphuric acid, and the solution neutralized with
potassium carbonate and saturated in the cold with potassium sul-
phate. By this precipitation the zirconium was obtained in the form
of insoluble double sulphate, while the uranium remained in solu-
tion. After filtering out the zirconium potassium sulphate it was
dissolved in hydrochloric acid and zirconium hydrate precipitated
with ammonium hydroxide. It was well washed with water and
dissolved in hydrofluoric acid. An equivalent amount of potassium
carbonate was added and, on evaporation, potassium zirconium
fluoride crystallized out. Three grams of this salt were obtained.
Analysis. — 0.5043 gram of the salt ignited with sulphuric acid
gave 0.5272 gram of K2S04 + Zr02, and contained 0.2212 gram
of Zr02, leaving 0.3060 gram of K,S04.

Calculated
K2ZrF6. Found.

2KF 4I05 40.47

ZrF4 58-95 59-52

100.00 99-99

The uranium in the filtrate from the zirconium was precipitated
with ammonium hydroxide, dissolved in hydrochloric acid, and an
excess of sodium hydroxide added to this to obtain the glucinum.
This was not found. The uranium was changed to nitrate and the

igos.]

HALL AND SMITH— COLUMBIUM. 179

solution allowed to evaporate, when large characteristic crystals of
uranium nitrate separated. The solution decanted from the original
ammonium hydroxide precipitate, which contained manganese and
allied elements, was treated with an excess of sodium carbonate.
The precipitate produced was allowed to settle, the supernatant
liquid decanted into other jars, and hydrogen sulphide passed
through it. A small amount of a black sulphide was obtained. It
consisted of zinc, iron, copper and nickel (from the crucible tongs).
The precipitate produced by sodium carbonate contained manga-
nese, zinc and iron. Ten grams of it were dissolved in hydrochloric
acid and the iron removed by the basic acetate method. The zinc
was then precipitated as sulphide. The latter was changed to
chloride and tested for gallium. Not a trace of the latter was
found. Zinc lines alone were shown in the spectrum.

Tin and tungsten were contained in part II. Part III was not
further examined. It may be concluded, therefore, that the
columbite from South Dakota contains as acids : tantalum, colum-
bium, titanium, silicon, zirconium, tin and tungsten ; as bases :
iron, maganese, zinc, uranium, copper (?) and nickel (?).

The moist metallic acids, after having been washed with dilute
sulphuric acid, were brought into a large platinum dish and dis-
solved in fairly concentrated hydrofluoric acid. This solution was
then filtered, through a hot water funnel, from undecomposed
mineral and from potassium silicofluoride (due to the presence of
some potassium sulphate in the moist oxides). The hydrofluoric
acid solutions were collected in large rubber dishes and sufficient
potassium hydroxide was introduced to convert the tantalum into
potassium tantalum fluoride, most of which separated out and was
removed by filtration. This precipitate was dried as far as pos-
sible by suction. It was washed once and then allowed to dry in
the air. It weighed n kilograms. The mother-liquor from the
potassium tantalum fluoride was evaporated in stages, potassium
hydrate being added. The columbium separated usually in hexag-
onal, hard, short crystals, such as separate from a strongly acid
solution containing an insufficient amount of potassium fluoride.
The total residue obtained in this way amounted to about 8 to 10
kilograms. These residues were decomposed by treating them with
twice their own weight of sulphuric acid, heating gently until the
bulk of the hydrofluoric acid was expelled, and then evaporating

180 HALL AND SMITH— COLUMBIUM. [May .9.

until the mass fumed strongly and maintaining the temperature
until the excess of sulphuric acid had been almost completely driven
out. Several hours were required for this. It is necessary in order
to get rid of the hydrofluoric acid. The residual mass was boiled
with water to extract the bases which dissolved as sulphates. The
insoluble hydroxides were thoroughly washed and dissolved in
hydrofluoric acid. The first crop of crystals, obtained by evapora-
tion with potassium hydroxide, was removed and the mother-liquor
then evaporated to dryness with sufficient potassium hydroxide to
change all of the metallic acids into double fluorides. A portion
of these crystals (first crop) was dissolved in water and the tan-
talum removed by adding dilute potassium hydroxide to the solu-
tion, which, after the formation of a permanent precipitate, was
boiled for some time. The precipitate consisted mainly of potas-
sium tantalum oxyfluoride. It was filtered out and the filtrate
evaporated to dryness. The residue was baked for some time at
2oo°. By this procedure some hydrofluoric acid was expelled and,
on taking up the residue with water and boiling, more potassium
tantalum oxyfluoride separated. By repetition of this process all
of the tantalum was removed from the solution. The only test relied
upon for the detection of tantalum was the solution of this pre-
cipitate in a drop of hydrofluoric acid and evaporation to crystalli-
zation. If needles separated their solubility in water was used to
ascertain whether they were potassium tantalum fluoride or potas-
sium columbium oxyfluoride. It is true that this test consumes
considerable time, yet it is the only satisfactory means of deter-
mining with which of the metals the chemist is dealing. The for-
mation of a precipitate by protracted boiling of a dilute solution of
potassium tantalum fluoride is not conclusive, for Krliss and Nilson
(Ber. 1 88 1, 1676) have shown that potassium columbium oxy-
fluoride deposits under like conditions a small amount of a salt
containing less fluorine. Further, the double fluoride must be
recrystallized several times, so that it will be sufficiently free from
acid that tantalum, if it is present in small amounts, may be pre-
cipitated by boiling.

Having freed the double fluoride from tantalum it was dissolved
in water and hydrogen sulphide conducted through its solution.
A slight precipitate of platinum sulphide was obtained. The fil-
trate from it was evaporated to dryness and the residue baked.

ig05.] HALL AND SMITH— COLUMBIUM. 181

On dissolving in water more platinum sulphide was found, but
when hydrogen sulphide was conducted through the filtrate no
further precipitation took place. The first crop of crystals got by
the evaporation of this solution showed the usual form of potassium
columbium oxyfluoride. They were allowed to dry in the air and
labeled crystals No. i (A). The filtrate from them was reduced
to a small bulk. Strong hydrofluoric acid was added when -the
needles of potassium columbium fluoride (K2CbF.) separated.
These were dried between bibulous paper and labeled crystals No.
2 (B). Samples from these two crops of crystals were analyzed.
Analysts of No. i (A) :

0.53 gram of salt gave 0.2346 gram of oxide and

0.3 16 1 gram of potassium sulphate
0.2346 : 0.3 1 61 :: x/2 : 174 x = R.205 = 258.2.

Calculated.
K2CbOF5H„0. Found.

K.2S04 57.81 59.64

Oxide 4452 44.26

The crucible in which the ignition of oxide occurred was stained.
This was undoubtedly due to the presence of tin, which had not
been removed, although hydrogen sulphide had been conducted
through the solution of the double fluoride.

Analysis of No. 2 (2?) :

0.S428 gram of salt gave 0.3635 gram of oxide and

0.4803 gram of potassium sul-
phate.

0.481 1 gram of salt gave 0.2082 gram of oxide and

°- 2 7 751 gram of potassium sul-
phate.

0.3635 : 0.4803 :: .v/2 : 174 x = 263.4.

0.2082 : 0.2775 :: xJ2 : 174 x = 250.1.

Calculated
K2CbF7. Found. Found.

K2S04 57.05 56.99 57.68

Oxide 43.93 43.13 43.28

1 Probably too high because it was not heated enough to expel all of the sul-
phuric acid.

182 HALL AND SMITH— COLUMBIUM. [May 19,

Specific gravity of oxide (B) :

18 5924 pyknometer -\- water, t = 180

.2346 gram of oxide taken
18.8270
18.7760 pyknometer -f- water -j- oxide .2346

.0510 gram water displaced .0510 4.60 sp. g.

To determine the titanium content of the columbium oxide re-
course was had to a comparison of the color tint produced by
hydrogen peroxide in an oxalate solution against known amounts
of titanium hydrate dissolved in oxalic acid. Thus, 0.2262 gram
of columbium oxide was fused with acid potassium sulphate and the
fusion dissolved in oxalic acid and diluted to 50 cubic centimeters.
It gave a color equivalent to 0.4 cubic centimeter of the standard
titanium solution (1 c.c. contained 0.00106 gram of titanium
dioxide). In other words, by this test the columbium oxide was
thought to contain 0.000424 gram of titanium dioxide, or .18 °/c.

Solubility of Crystals A.
One part of the salt was found to be soluble in a little over 12
parts of water. This is the solubility of potassium columbium
oxyfluoride.

100 grams of the residues obtained by the evaporation of mother-
liquors (page 180) to dryness were dissolved in water and fraction-
ally crystallized. After having removed as much of the tantalum as
possible by introducing dilute potassium hydroxide into the boiling
solution until a rather considerable and permanent precipitate was
obtained, the solution was boiled for some time. The first fraction
of crystals (2) and the third fraction of crystals (3) were removed,
after which hydrofluoric acid was added to the mother-liquor, from
which there separated a crop of needles, which we shall designate
crystals 4. These last were recrystallized from hydrofluoric acid.
They probably contained silicon and tantalum. The acid mother-
liquors from these different crops of crystals were treated as de-
scribed by Hermann (J. pr. Chan., Series 2, vol. 15, 105, 1877).
That is, they were treated with 20 parts or two liters of water ami
150 grams of sodium hydroxide. A clear solution resulted, from
which a fine crystalline precipitate separated. The filtrate from

i9°5-]

HALL AND SMITH— COLUMBIUM. 183

this precipitate gave no reduction test when treated with acid and
zinc, nor was anything obtained from it after having added dilute
sulphuric acid and a slight excess of ammonium hydroxide. It was
free from earthy bases and metallic acids.

The crystals of the sodium salt (2.5 grams), obtained as outlined
in the last paragraph, dissolved almost completely in 20 parts of
boiling water and separated in well-defined forms from the cold
solution. A portion of this salt heated in a salt of phosphorus bead
imparted a blue color to the latter in the reducing flame.

The next step was to decompose the solution of this crystalline
sodium salt with dilute sulphuric acid. The solution was hot. The
precipitate which separated was thrown upon a filter and washed,
after which it was dissolved in hydrofluoric acid and an equivalent
amount of potassium carbonate added in order to form potassium
columbium oxyfluoride. The solution was evaporated to dryness
upon a water bath, the residue repeatedly moistened, and evapora-
tion to dryness repeated until the odor of hydrofluoric acid could
not be detected by the smell ; then the salt was baked, taken up in
water, and the solution boiled for some time. A trace of tantalum
oxyfluoride separated. It was filtered out. On evaporation to
crystallization the leafy, characteristic crystals of potassium colum-
bium oxyfluoride appeared. They were dried between bibulous
paper and then analyzed.

Analysis :

0.5502 gram of salt gave 0.2452 gram of oxide and

0.3224 gram of potassium sulphate
0.2452 : 0.3224 :: x/2 : 174 x = 264.6.

Calculated

K2CbOF5H,0. Found.

Oxide 44-52 44-56

K2S04 57.81 58.60

A portion of crystals No. 4 (page 182) was recrystallized and
analyzed.

0.5588 gram of sample gave 0.2407 gram of oxide and

0.3232 gram of potassium sulphate
0.2407 : 0.3232 :: x/2 : 174 #=259.2.

Calculated
K2tbF7. Found.

Oxide 43-93 43-°8

K2SO, 57.05 57-84

184 HALL AND SMITH— COLUMBIUM. [May 19,

Analysis of recrystallized portion of crystals (4), page 182.

.5588 gram of sample gave .2407 gram of oxide.

.3232 gram potassium sulphate.
.2407 : 3232 :: x/2 : 174 x= 259.2.

Calculated.
K2CbF7. Found.

Oxide 43-93 43-°8

K,S04 57-05 57-84

The results of analysis as well as the behavior points to the fact
that the oxide contained in these residues is mainly columbium
oxide, Cb205, with a small portion of another oxide causing the
equivalent weights obtained to be too low. These results may in
part be due to the presence of some potassium silicofluoride, but
more likely to potassium titanium fluoride. Yet these last fractions
of double fluoride in which the titanium should be concentrated
show only very small amounts. Starting with so much material
the last fractions should show more titanium if it is present in the
mineral in appreciable amounts.

The only test for small amounts of titanium which we have are
the hydrogen peroxide test of Schonn, Jahresberichte, 1893, 901,
and the chromotropic acid test used by Geisow (Dissertation, 1902).
Of these the former is the only one relied on, and it offers the only
direct evidence which we have for the presence of titanium in the
potassium columbium oxyfluoride obtained from columbite.

The methods applicable in the preparation of columbium oxide
relatively free from titanium are as follows :

1. Crystallization of potassium columbium fluoride (K2CbF7)
which is not isomorphous with potassium titanium fluoride. The
difficulty in this case is that the hydrofluoric acid increases the
solubility of the columbium body and decreases that of the titanium
double fluoride so that the titanium would have less tendency to
concentrate in the mother liquors.

2. Fractional precipitation with dilute ammonium hydroxide.
The columbium hydrate is precipitated first and the titanium con-
centrated in the last fractions. No fraction consists entirely of
titanium hydrate ; even the last fraction is largely columbium
hydrate.

3. The formation of the chloride or oxychloride of columbium
and the chloride of titanium and separating these by distillation.

i9°5-]

HALL AND SMITH— COLUMB1UM. 185

4. Treatment of the hydrates with cold, fairly concentrated sul-
phuric acid, in which the titanium hydrate should dissolve and
leave the columbium.

All of these methods were tried in the endeavor to obtain suf-
ficient titanium from the columbium to identify it and prepare some
of its derivatives, e. g., the potassium double fluoride. In order to
try out method " 1 " the remainder of the residues (page 180) —
2 to 3 kilos — was crystallized twice from hydrofluoric acid, thus
getting potassium columbium fluoride (K2CbF7). The mother
liquors were united. They equaled 600 c.c. They were neu-
tralized with dilute ammonium hydroxide ; the precipitated hydrate
filtered out and the filtrate made alkaline with a large excess of
ammonium hydroxide, which gave a further precipitate. This last
hydrate was filtered out and dissolved in hydrofluoric acid. Potas-
sium carbonate was added and the solution evaporated to dryness
on a water bath, the residue taken up in water and crystallized.
The crystals obtained were short stubby needles. They were evi-
dently not potassium columbium oxyfluoride (K2CbOF.). Their
quantity was too small to recrystallize, but they were analyzed :

0.4672 gram of sample gave 0.1740 gram of oxide = 37.6 percent.
" " " " " 0.3050 gram of K2S04 = 65.3 per cent.

The sp. gr. of the oxide was found to be 5.2, although too little of
it was at hand for accurate work. The determination of the
titanium in the oxide colorimetrically gave 0.0106 gram of
Ti02= 6.1 percent.

The salt originally taken showed 0.62 percent, of its oxide to be
Ti02 by the colorimetric determination, while the double fluoride
of potassium and columbium obtained showed a Ti02 content
equal to .25 per cent, of its oxide. Hence it would seem that by
method " 1 " the titanium or oxide with lower sp. gr. and molecu-
lar weight did concentrate in the mother liquors.

Crystals 2 and 3 (page 182) were combined. Their total weight
was about one kilo. This, in portions of 100 grams at a time, was
dissolved in about 2 liters of water and fractionally precipitated
with ammonium hydroxide, using dilute alkali and working in the
cold with constant stirring. The alkali was added until the solution
was barely acid to litmus, the precipitate formed was filtered off and
the filtrate made alkaline with an excess of the precipitant. This

ISO HALL AND SMITH— COLUMBIUM. [May i9l

second fraction contained about 3-4 grams of oxide from each 100
grams of double fluoride taken. These ten last fractions were com-
bined and dissolved in hydrofluoric acid, 15 grams of potassium
hydroxide added, then dilute ammonium hydroxide until slightly
acid, and the precipitate filtered out. It was marked "A." Am-
monium hydroxide was then added to the nitrate until litmus
showed the reaction to be just neutral. The precipitate was
designated "B." "C" was obtained in the filtrate from "B"
by adding a large excess of ammonium hydroxide. It (C) pre-
sumably should contain most of the titanium from the 1,000 grams
of double fluoride taken. The oxide actually present in it was
changed to double fluoride by dissolving in hydrofluoric acid and
adding potassium carbonate. The first crop of crystals was ob-
tained from strong hydrofluoric acid. It recrystallized from the
same in needles. These were analyzed :

0.6954 gram of sample gave 0.3018 gram oxide and

0.3900 gram potassium sulphate,
.3018 : .3900 :: x/2 : 174, x= 269.3.

Calculated
K,,CbF7. Found.

Oxide 57-05 56.09

K2S04 43-93 43-40

The sp. gr. of the oxide equaled 4.45. 0.3 gram of it showed
the presence of the equivalent of .0025 gram of Ti02, or 0.83
per cent.

The needles from "C," not taken for analysis, and the mother
liquor were combined and evaporated to dryness on a water bath
to expel the excess of hydrofluoric acid. This was repeated once.
The solution of the salt was then fractionally precipitated with
amonium hydroxide, the fractions up to the point where litmus
showed a slightly alkaline reaction being discarded, when the fil-
trate from them was made strongly alkaline with ammonium hy-
droxide. The precipitate obtained was changed to its potassium,
double fluoride. About 2.5 grams of the double salt were got.

Analysis :

0.7248 gram of sample gave 0.3014 gram of oxide and
0.4205 gram of K,S04
0.3014 : 0.4205 :: x/2 : 174 .v= 249.4.

i9°5 ]

HALL AND SMITH— COLUMBIUM. 187

Calculated
K2CbOF6H20. Found.

Oxide 57-81 58.02

K2S04 44-52 4I-58

The sp. gr. of the oxide was found to be 4.667. 0.2930 gram
of oxide gave a color with hydrogen peroxide equivalent to .0322
gram of Ti02 =11 per cent.

The crucible in which the oxide was ignited was deeply stained.
The analysis and the color test showed the presence of titanium,
while the stain on the crucible and the high sp. gr: of the oxide
could be due to tin in considerable amount, which is precipitated
in the last fractions on fractional precipitation with ammonium
hydroxide.

The crystals and the mother liquor remaining from this salt
were dissolved in boiling water and an excess of sodium hydrate
added, when a heavy flocculent precipitate separated. This was
filtered, boiled with water and the insoluble portion changed to
double fluoride. It was again taken up in boiling water and an
excess of sodium hydroxide added. The precipitation was not
complete. The portion precipitated was treated with boiling
water, filtered, and the filtrate found to contain a large amount of
titanium. That portion of the precipitate insoluble in water was.
changed to double fluoride.

Analysis :

0.5570 gram of sample gave 0.2248 gram of oxide and
0.3468 gram of K2S04
0.2248 : 0.3468 :: x/2 : 174 x = 225.6.

The sp. gr. of the oxide was found to be 4.2. The oxide gave a
color with hydrogen peroxide showing the presence of about 21
per cent. Ti02.

Crystals "A" (page 181) were changed to double fluoride and!
recrystallized.

Analysts :

0.7184 gram of sample gave 0.31 88 gram of oxide and

0.4176 gram of potassium sul-
phate.
0.3188 : 0.4176 :: x/2 : 174 x = 265.7.

PROC. AMER. PHILOS. SOC. XLIV. l8o. M. PRINTED AUGUST I, I905.

188 HALL AND SMITH— COLUMBIUM. [May 19,

Calculated Found.
KjCbOFsHjO.

Oxide 44.52 44.38

K2SQ4 57.81 58.13

The sp. gr. of the oxide was found to be 4.481. 0.3160 gram of
oxide gave a color with hydrogen peroxide equivalent to .0022
gram Ti02 = . 7 per cent.

Crystals "B" (page 181) analyzed as follows:

0.8098 gram of sample gave 0.3614 gram of oxide and
0.4708 gram of K2S04
0.3614 : 0.4708 :: x/2 : 174 x = 267.0.

Calculated Found.
KXbOF6H,0.

Oxide 44-52 44-63

K2SO, 57-Si 58-14

The sp. gr. of the oxide was found to be 4.864. 0.3600 gram of
oxide gave with hydrogen peroxide a color equivalent to .0013.
gram of Ti02 = .36 per cent.

Having failed to get any evidence of the existence of neptunium
in the last fractions of the double fluoride from the South Dakota
mineral, it was decided to test some of the mineral from Haddam,
Conn., the source of the material used by Hermann in his investi-
gation. The last fractions of the potassium double fluoride from
5.87 kilos of columbite from Haddam, Conn., amounting to 100
grams, were dissolved in boiling water and an excess of sodium hy-
droxide added. The precipitate obtained was partly crystalline
and partly flocculent, as described by Hermann. It was filtered
out, dried on a porous plate, and boiled with 25 parts of water.
The crystals dissolved leaving a yellowish residue evidently con-
taining much iron. This residue gave a yellow colored bead in
the reducing flame containing so much iron that it was impossible
with a small blowpipe to keep it all reduced. Is it not probable
that this is what Hermann supposed was neptunium ?

This salt was fused with acid potassium sulphate to remove the
iron, the oxide remaining after extracting the fusion with boiling
water was changed to double fluoride, dissolved in boiling water,
and an excess of sodium hydroxide added. The precipitate ob-
tained was crystalline and perfectly soluble in water, leaving an
inappreciable residue. From all of which it may be inferred that

igos.] HALL AND SMITH— COLUMBIUM. 189

the material from the Haddam locality showed no more evidence
of neptunium than did that from South Dakota.

Ten kilograms of the main bulk of the potassium columbium
oxyfluoride were crystallized from strong hydrofluoric acid. Five
hundred grams were taken at one time and the mother liquors from
the two fractions combined, evaporated one half, and another crop of
crystals removed. The mother liquors from these were united and
evaporated, the crystals obtained were recrystallized, the mother
liquors from the recrystallization combined and evaporated to dryness
with sulphuric acid. The oxide obtained from that portion evapo-
rated to dryness was 5 grams. The fraction of crystals immediately
preceding, 60 grams in weight, was decomposed with sulphuric acid
and the oxide obtained from it. The 5 grams of oxide and about one-
half of the oxide from the 60 grams of double fluoride were heated
in carbon tetrachloride vapors, taking about 2 grams of oxide at a
time. The more volatile portion, which should contain the most
of the TiCl4, was distilled away from the CbOCl3. The portions
of this liquid were combined, the oxides — 3 grams — obtained
from it, and these oxides again heated in CC14. Again only the
liquid portion and that of the solid which was carried over mechan-
ically was taken. The oxide from this portion was changed to
double fluoride.

Analysis :

0.567S gram of sample contained 0.2196 gram of oxide and

0.3390 gram of K.,S04.
0.2196 : 0.3390 :: x/2 : 174 x = 225.4.

Calculated

K„CbOF6H20. Found.

Oxide 44-52 38. 6

K2S04 57-8i 59.7

Amount of Ti02 in the oxide .0594 gram = 29.7 per cent.

The oxide from the salt taken for analysis was combined with the
oxide from the double fluoride not taken for analysis and this with
the remaining half of the oxide from the 60 grams of double fluoride
mentioned above was mixed with the oxide from all the previous
double fluorides which had been tested for titanium and shown its
presence in a fair degree. This mixture of oxides was heated in a
current of sulphur monochloride, the chloride formed collected

190 HALL AND SMITH— COLUMBIUM. [May 19,

in a receiver, and this was then heated until all of the sulphur
monochloride was distilled out. This sulphur monochloride and
any other chlorides which it might contain was again distilled to
remove the last of the columbium chloride which might have been
carried over mechanically. It was then treated with water and
oxalic acid, the sulphur filtered off, and any oxide dissolved in the
oxalic acid precipitated with ammonium hydroxide. This hydrate
was changed to potassium double fluoride.
Analysis :

0.5444 gram of sample gave o. 1952 gram of oxide and

0.3850 gram of potassium sulphate
0.1952 : 0.3850 :: x/2 : 174 x = 175.9.

Calculated.
K2CbOF6H20. Calculated. Found.

Oxide 44.52 33.33 35.85

K.SO^ 57.S1 72.50 70.72

0.0310 gram of the oxide was found to contain 0.244 gram Ti02,
or 78.7 per cent.

A solution of the oxide in hydrochloric acid reduced with zinc
to an amethyst or violet color ; the oxide also gave a violet titanium
bead in the reducing flame with salt of phosphorus.

It would seem that about 80 per cent, of the oxide from this
double fluoride was Ti02.

The remainder of the double fluoride, about .5 gram, was dis-
solved in the mother liquor from which it came by heating, and
an excess of sodium hydrate added. A fiocculent precipitate
formed. After cooling, this was filtered off and the filtrate acidi-
fied with hydrochloric acid and tested for metallic acids with am-
monium hydroxide. A precipitate was obtained, which was filtered
out and, after washing thoroughly, dissolved in hydrochloric acid.

Upon passing hydrogen sulphide through this solution a heavy
precipitate of yellow stannic sulphide was obtained, showing that
tin had been carried over with the titanium by the sulphur mono-
chloride.

The precipitate formed by the excess of sodium hydroxide was
drained thoroughly and boiled up with water. Nothing went into
solution, as would have happened had there been any sodium
columbate in the precipitate. This well washed precipitate was

i9°5-J

HALL AND SMITH— COLUMBIUM. 191

dissolved in hydrofluoric acid and potassium carbonate added to
change it to the potassium double fluoride.
Analysis of the crystals obtained :

0.1S93 gram of sample gave 0.0642 gram of Ti02 and
o. 1366 gram of K2S04

Calculated. Found.

Oxide 33.33 33.91

K2S04 72.50 72.16

The determination of the Ti02 colorimetrically gave .0636 gram.
The salt was undoubtedly potassium titanium fluoride, proving
conclusively the presence of titanium in columbite.

Behavior of Solutions of the Double Fluorides of Colum-
bium and of titanium with a variety of bases.

Excess of sodium hydroxide was found to precipitate titanium
completely from a solution of potassium titanium fluoride, while
with potassium columbium oxyfluoride it gave a precipitate soluble
in slight excess but again insoluble and separating in a crystalline
form from a large excess of the sodium hydroxide. The precipi-
tate formed in the case of the titanium was insoluble in water,
while in the case of columbium the crystalline deposit was com-
pletely soluble in hot water. It was hoped that this difference of
behavior might afford a means of separating these two elements.
To test this experiments were tried as follows :

1. 0.9600 gram of K2CbOF. -f H20, containing 0.4272 gram
ofCb20., and 1. 1600 gram K2TiF6, containing 0.3753 gramofTi02,
were dissolved in 200 c.c. water, brought to boiling and an excess
of sodium hydroxide added. The precipitate which formed was
partly crystalline and partly flocculent. The solution was allowed
to stand over night. The precipitate was filtered out, drained, and
washed back into a platinum dish. It was covered with 200 c.c.
of water, brought to boiling, filtered hot, and washed with hot
water. The filtrate which should contain most of the columbium
and none of the titanium was brought to boiling and sulphuric
acid and ammonium hydroxide added. The hydrate obtained was
ignited to oxide and weighed 0.1640 gram. It was found to con-
tain .0117 gram of Ti02. The titanium content was determined
colorimetrically by fusing with potassium acid sulphate, dissolving

192 HALL AND SMITH— COLUMBIUM. [May x9>

the fusion in oxalic acid and comparing the color developed with
hydrogen peroxide with that of a titanium solution, of known
strength, in oxalic acid.

Part insoluble — 0.2749 Cb205, 0.3636 Ti02

Ti024Cb205 soluble portion ; Cb2054Ti02 insoluble portion.

2. A mixture containing 0.1270 gram K2TiF6, or .0423 gram
Ti02 and 2.5280 grams K2CbOF5.H20, or 1. 1376 gram Cb20., was
treated as above. The precipitate was nearly all crystalline. That
part of it insoluble in water weighed .0470 gram and contained
.0074 gram of TiOa leaving .0349 gram of Ti02 in solution (by
far the greater quantity).

3. 0.2830 gram K2TiF6, containing .0943 gram Ti02, and 2.7040
grams K2CbOF.H20, containing 1.2168 grams Cb205, were treated
as before. The precipitate was chiefly crystalline. A small part
of it was flocculent. The part insoluble in water weighed .0740
gram and contained .0148 gram Ti02, showing that .0795 gram
Ti02 went into solution and would be found with the bulk of the
columbium.

4. 0.6790 gram K,TiFfi, containing .2263 gram Ti02, and 2.2530
grams K2CbOF.H20, containing 1.0139 grams Cb20., were treated
as before. The precipitate contained a considerable amount of
flocculent material. The part insoluble in water weighed 0.1720
gram and contained 0.0710 gram Ti02, leaving 0.1553 gram of
Ti02 in solution.

The action of potassium hydroxide was also tried. It gave a
precipitate with columbium, soluble in an excess, and reprecipi-
tated by greater excess. When the solution was evaporated pearly
plates of potassium columbate separated out. With a solution of
K2TiF6 a heavy precipitate was obtained, but the filtrate gave a
slight test for titanium.

1. 3 1 30 gram K2TiF6, containing .4377 gram of Ti02, and 1.0060
gram K2CbOF5H20, containing .4467 gram Cb205, were dissolved
in 200 c.c. of water and the solution brought to boiling, when an
excess of potassium hydroxide was added. The precipitate obtained
weighed 0.5900 gram after ignition. The filtrate gave a very pro-
nounced test for titanium.

Solutions of K2TiF6 and K2CbOF5 were studied with various
organic bases in the hope that differences of behavior might pre-

•9°S-]

HALL AND SMITH— COLUMBIUM.

193

sent themselves which would lead to a quantitative separation of
these two elements.

Reagent.

1. Mono-methylamine,

2. Di-methylamine,

3. Tri-methylamine,

4. Tetra-methylamine,

5. Mono-ethylamine,

6. Di-ethylamine,

7. Tri-ethylamine,

8. Di-propylamine,

9. Amylamine,

10. Iso-butylamine,

11. Allylamine,

12. Ethylenediamine,

13. Propylenediamine,

14. Butylenediamine (secondary),

15. Butylenediamine (normal)

16. Hexylamine,

17. Benzylamine,

18. Benzylmethylamine,

19. Piperidine,

20. Camphylamine,

21. Di-benzylamine,

22. Pyridine,

23. Di-isobutylamine,

24. Tri-propylamine,

25. Di-amylamine,

26. Heptylamine,

27. Toluylenediamine (meta),

28. Picoline,

29. Tri-isobutylamine,

30. Bornylamine,

31. Aniline,

32. Toluidine (m),

^^. Mono-methylaniline,

34. Mono-ethylaniline,

35. Isoquinoline,

36. Quinoline,

Solution of K2TiF6. Solution of K„CbOF5.
Precipit. complete, Precipit. soluble excess.

" by large excess.

insoluble excess.

dif. soluble in water.

" soluble excess.

slightly soluble excess.
Precipit. complete.

Partial precipitation, "

"

Slight precipitation, "

nearly but not

quite complete.

" " "

not complete.

Precipit. not complete, "

heavy but not

complete.

Slight precipitation, "

"

" " after "

" "

24 hours,

Slight prec. on standing, "

slow — incom-

plete.

" " " "

heavy — incom-

plete.

" " " '

' nearly com-

plete,

194

HALL AND SMITH— COLUMBIUM.

[May

Reagent.

37. I lexylmethylenetetramine,

3S. Bromaniline (ra),

39. Chloraniline (o),

40.

Di-chloraniline,

41.

Di-ethylaniline, "

42.

Chloraniline (p),

43.

Di-methylaniline, "

44.

Xylidine (p),

45-

Xylidine (0), "

46.

Xylidine (mj, "

47-

Tetra-hydroquinoline, "

48.

Benzylaniline,

49-

Di-phenylamine, "

50.

Tri-benzylamine,

5i-

Naphthylamine (,3), "

52.

Naphthylamine («), "

53-

Nitronaphthalene, "

54

Bromphenylhydrazine, "

55-

Nitrophenylh)drazine, "

5^>.

Benzidine, "

57-

Nitraniline (0),

58.

Nitraniline (p), "

59-

Nitraniline (m), "

60.

Diphenyl, "

61.

Diphenyl carbonate, "

62.

Methyl carbonate, "

63-

Ethyl carbonate, "

64.

Piperine, "

65.

Mono-chlorhydrin, "

66

Tri-chlorhydrin, "

67.

Di-bromhydrin (/?), "

68.

Nitroso-dipropylin, "

69.

Nitroso-diethylene, "

70.

Nitroso-dimethylene, "

7i-

Succinimide, "

72.

Methyl-diphenylamine, "

73-

Tetra-nitromethylaniline, "

74-

Bromamiline, "

Solution of K2TiF6. Solution of K2CbOFs.

Slight prec. on standing, Free, heavy — in-
complete.
No precipitation, Slight prec. on standing.

Frecipit. slow-
plete.

Slight prec. on standing.
No precipitation.

The behavior of the bases which react with the above solutions of
titanium and columbium may be divided into the following classes :

1. Those which precipitate the titanium completely, and while
they precipitate the columbium dissolve it upon the addition of an

I9oS.J HALL AND SMITH— C0LUMB1UM. 1(J5

excess of the reagent to form columbates. This class while show-
ing pronounced difference of behavior is useless as a means of
separation, for upon treating a solution containing both titanium
and columbium with one of these reagents the titanium was found
both in the soluble and in the insoluble portion. Columbium was
also found in the titanium precipitate. The probable explanation
for this is that a columbate was formed which dissolved the freshly
precipitated titanium hydrate to a salt of a complex titanoso-
columbic acid, which was soluble, and also a small amount of an
acid salt or a free acid containing an excess of titanium, which was
insoluble.

2. To the second class of reagents belong those which precipi-
tate the hydrates of the two elements and are not sufficiently basic
to dissolve the columbium hydrate and form columbates. Those
reagents which are sufficiently basic to completely precipitate the
columbium are strong enough to partially precipitate the titanium.
Quinoline seemed the most promising of all the reagents which were
tried.

3. Those reagents which gave only a partial precipitation with
columbium and no precipitation with titanium solutions did not
precipitate columbium hydrate free from titanium, from a solution
containing both titanium and columbium. The hydrate pre-
cipitated always gave a strong test for titanium after dissolving it
in oxalic acid. This may have been due in part to the extreme
difficulty encountered in washing the precipitate free from mother
liquor.

4. The remaining reagents precipitated neither titanium nor
columbium.

REACTIONS OF THE DOUBLE FLUORIDES OF COLUMBIUM, TITANIUM,

TANTALUM, TIN AND TUNGSTEN WITH VARIOUS REAGENTS IN

CONCENTRATED SULPHURIC ACID.

A small amount of the reagent was dissolved in eight to ten
drops of concentrated sulphuric acid on a glazed porcelain sur-
face and the crystalline double fluoride introduced into this acid
solution. In most cases the color was destroyed upon diluting
with water. No color was imparted to tin solutions by any of the
reagents which appear below.

196

HALL AND SMITH— COLUMBIUM.

[May ,9>

Reagent. Ta.

Codeine, No color,

Morphine, Faint yel-
low,
Resorcinol No color, No color

Naphthol (/?)

Naphthol («)

Pyrogallol,

Salicylic acid,

Cinchonidine,

Apomorphia,

Narceine,

Bebeerina,
Narcotina,

Cb. Ti. \V.

No color, Faint pink ; may Light brown ; on stand-

be due to mor- ing trace purple,
phine,
Red to brown ; Gray brown, becoming

very delicate, purple, ll.fi ppt.
Red brown ; No color,
fairly delicate,
Faint yellow Coffee brown ; Brown, becoming dark

brown very delicate, blue.

Faint brown, Green to dark Deep blue ; very deli-
greenish brown, cate.
Yellow to Dull dark red, Deep red to brown to

light brown, dirty blue.

Very faint Deep red, Reddish yellow.

yellow,
No color, No color, On standing a slight

purple.
Yellow brown, Light red brown, Purple to brown to green

• and blue.
Brownish yel- Brown, Dirty dark green.

low,
No color, Clear brown, Dark brown to green.

Yellow, Brown, Light brown to green.

Strychnia, quinidia, cinchonidine and atropia gave no color with
any of the elements tested. Narceine and bebeerina alone in sul-
phuric acid gave a considerable color, and with them the amount
of reagent used must be very small or it will obscure any change
produced by the addition of the double fluoride. In this connec-
tion it is of interest to note that Levy (C. R., 103, 1074 and
1 195) studied the colors produced by the phenol-like bodies, dis-
solved in concentrated sulphuric acid, when brought in contact
with the oxides of titanium, tin, tantalum, columbium and other
elements, with the following results. Columbium could be tested
for in the presence of all the others by using codeine, as it gave
a pink color, while titanium yielded no color and tantalum
but a faint green. Titanium could be tested for by using mor-
phine, with which it gave a carmine color, columbium no color
and tantalum a yellow color passing into brown. Tantalum with
resorcinol gave a dirty green color, changing to amethyst and rose,
while titanium yielded a flesh red color going to chocolate brown,
and columbium a yellowish tint. None of the results were dupli-

I9°5-]

HALL AND SMITH— COLUMBIUM. 197

cated save the morphine test for titanium, which proved exceedingly
delicate, yet to have the color show definitely in columbium the
latter must contain .5 per cent, of Ti02. Codeine gave no color
with columbium, nor did resorcinol with tantalum, therefore Levy
could not have had pure material for his tests.

In our use of these reagents we failed to find satisfactory tests
except in the case of morphine for titanium. None answered for
columbium in the presence of titanium or for tantalum in the pres-
ence of columbium. Resorcinol proved to be a fairly delicate test
for titanium. It gave no color with columbium, tantalum or
tungsten.

ACTION OF HYDROCHLORIC ACID GAS ON IGNITED
COLUMBIC OXIDE.

0.25 gram of ignited columbic oxide was completely volatilized
in three hours in a current of dry hydrochloric acid gas. It vola-
tilized as a white powder with no indication of reduction by change
of color. The compound formed adhered to the walls of the glass
tube, was insoluble in oxalic acid, and only very slowly soluble in
concentrated sulphuric acid, requiring long boiling to dissolve a thin
layer. It contained hydrochloric acid, as was shown by washing
with ammonium hydroxide and testing the washings with silver
nitrate and nitric acid. It would be very difficult to collect in a
form convenient for analysis, yet this should be done, as the body
evidently contains no water, as is given in the formula of a similar
body obtained by Smith and Maas {Zeit. atiorg. Chem., 7, 96) by
passing moist hydrochloric acid gas over the hydrated oxide. It is
undoubtedly analogous to the body obtained on heating molybdic
acid in hydrochloric acid gas, namely, Mo032HCl. It is likely
Cb20..xHCl.

ACTION OF SULPHURIC ACID ON THE HYDRATES OF COLUMBIUM

AND TITANIUM AFTER RECIPITATION BY AMMONIUM

HYDROXIDE FROM SOLUTIONS OF THEIR

DOUBLE FLUORIDES.

The method used was to precipitate the hydrates from a weighed
amount of the double fluorides, filter and wash as thoroughly as
possible, then transfer to a weighed platinum dish, reweigh, the
difference being water, after which a weighed amount of sulphuric

198 HALL AND SMITH— COLUMBIUM.

[May 19,

acid of definite specific gravity was added and allowed to stand in
contact with the hydrates for a definite time. The portion in-
soluble was filtered off, and the amount of oxide going into solu-
tion determined by precipitation with ammonium hydroxide,
igniting and weighing. The amount of titanium in the oxide
which went into solution was determined colorimetrically.

In making these trials columbium oxyfluoride was used in which
the titanium oxide as compared with the columbium oxide was
.00095 gram Ti02, or 0.41 per cent. The specific gravity of the
sulphuric acid used was 1.145.

EXPERIMENTS.

1. 1 gram ofK2TiF6, containing 0.3330 gram of Ti02, was used
to obtain the hydrate. The latter was treated with 40 c.c. of
water and 70 grams of sulphuric acid. It dissolved completely in
fifteen minutes.

2. 0.4450 gram of columbic oxide, in the form of hydrate, was
treated with 40 grams of water and 108 grams of sulphuric acid for
one hour. Only a slight precipitate was obtained with ammonium
hydroxide in the filtrate. It was not weighed but contained
.00032 gram of Ti02, or .07 per cent, of the total oxide taken, or
one-sixth of the total Ti02 present.

3. Columbic hydrate, containing 0.4450 gram of oxide, was
treated with 60 grams of water and 123 grams of sulphuric acid for
four hours. The portion which dissolved equaled 0.0060 gram =
1.33 per cent., and contained .000424 gram of Ti02, or 7 per cent,
of the oxide dissolved and about one-fourth of the total titanium
present.

4. The hydrate from three grams of columbium oxyfluoride, equiv-
alent to 1.35 grams of oxide, was allowed to stand in contact with
50 grams of water and 100 grams of sulphuric acid for seventeen
hours. The acid solution showed 0.0236 gram of oxide = 1.75
per cent., containing .000954 gram TiO, = 4 per cent, of oxide
dissolved or 17 per cent, of the total Ti02 present.

5. Hydrate containing 0.445 gram of oxide when treated with
57 grams of water and 85 grams of H2S04 (sp. gr. = 1.435) f°r 45
hours showed in solution 0.0500 gram =11.1 per cent., containing
0.0008 gram Ti02 =1.6 per cent, of the oxide dissolved or 43 per
cent, of the total Ti02 present.

igo5.]

HALL AND SMITH— COLUMBIUM. 199

Known amounts of the two hydrates were next treated together
as in the following experiments.

6. 0.2816 gram of K2TiF6, containing .0939 gram of Ti02, and
0.5078 gram of K2CbOF5H20, containing 0.2255 gram Cb205,
were treated with 50 grams of water and 15 grams of sulphuric acid
(sp. gr.= 1.435) f°r f°ur hours. The amount of oxfde remaining
insoluble was only .07 gram. It was not examined as to its titan-
ium content.

7. The hydrate from o. 2420 gram K2TiF6, containing .0807 gram
Ti02, and that from 0.3454 gram of K2CbOF.H20, containing
0.1537 gram of Cb205, were covered with 70 grams of water and 5
grams of sulphuric acid of sp. gr. 1.435. The amount of oxide in
solution after four hours was 0.0800 gram, corresponding well with
the weight of the oxide of titanium present, but the insoluble portion
was found to contain 0.0350 gram of Ti02, determined colorimet-
rically ; so that only about one-half of the titanium hydrate had
been dissolved out while nearly as much columbium hydrate had
gone into solution. The acid used would not have dissolved any
columbium hydrate had it been free from titanium hydrate ; further
it would have dissolved out all of the titanium hydrate had it not
been mixed with the columbium hydrate. It may, therefore, be
concluded that this method of separation is worthless. It remains
to be seen how haloid acids would act.

THE CHROMOTROPIC ACID TEST FOR TITANIUM.

Geisow (Dissertation, 1902) observed that the color developed
by chromotropic acid with titanium solutions offered a very deli-
cate test for that element. In concentrated solution it gives a deep
red, in dilute solutions, a pink color. The color-giving compound
was isolated by Geisow and found to have the following composi-
tion : one molecule of chromotropic acid to four of Ti02 and nine
of H20.

As it was most important to find some means of estimating the
amount of titanium in columbium we were induced to study this
reaction of Geisow, using solutions of titanic hydrate in oxalic,
sulphuric and hydrochloric acids.

Solutions used :

(A) 0.53 gram of Ti02 dissolved in 3.42 grams of oxalic acid
and diluted to 500 c.c. 1 c.c. = .00106 gram of TiO, and con-
tained .00684 gram of oxalic acid.

200 HALL AXJ) SMITH— COLUMBIUM. [May i9>

(B) 10 c.c. of (A) diluted to ioo c.c. i c.c. = .000106 Ti02.

( C) 10 c.c. of (B) diluted to 100 c.c. 1 c.c. = .0000106 Ti02.

(D) 10 grams of oxalic acid in 100 c.c. 1 c.c. =0.1 gram of
oxalic acid.

(is) 1 gram chromotropic acid in 100 c.c.

50 c.c. Nessler tubes, one inch in diameter, were used for all of
the tests.

.5 c.c. of (is) in 50 c.c. of water gave a mere trace of color,
for which reason the solution to be tested was always compared
with another tube containing the same amount of chromotropic
acid, thus making allowance for the slight color given by the solu-
tion of that reagent.

.15 c.c. of (C) gave a faint pink color when added to 50 c.c.
of water containing .5 c.c. of (is) ; .00000159 gram of Ti02 in
50 c.c. of water gave a change of color ; .3 c.c. of the same solu-
tion gave a very distinct coloration, or .00000318 gram of Ti02 in
50 c.c. More than 1 c.c. of the chromotropic acid gave so much
color'as to interfere with the delicacy of the test.

EFFECT OF OXALIC ACID.

i c.c. of (is).

i c.c. of (-D), or 1 gram of oxalic acid in 50 c.c, required 2.6
c.c. of titanium solution (C) to show the characteristic pink color,
or .0000275 gram of Ti02.

1 c.c. of reagent was adopted as the amount best suited to use and
was the amount taken in all of the following cases unless otherwise
stated.

With 0.2 gram of oxalic acid, .0000339 gram of Ti022 in 50 c.c.
gave a pink color.

With 0.5 gram of oxalic acid, .0000275 gram of Ti02 was re-
quired to give the test for titanium.

1.0 gram of oxalic acid required .0000244 gram of Ti02 to give
the test.

In the presence of 2.0 grams of oxalic acid, .0000244 gram of
Ti02 gave a distinct pink color in 50 c.c, while with 5.0 grams of
oxalic acid, .0000265 gram of Ti02 gave a color.

The amount of oxalic acid seems to have little effect although the
presence of the acid diminishes the delicacy of the test, but this is
independent of the amount of the acid present when more than o. 1
gram is used.

I9o5.i HALL AND SMITH— COLUMBIUM. 201

EFFECT OF THE PRESENCE OF HYDROCHLORIC ACID.

The solution of hydrochloric acid used was one part of concen-
trated acid to five parts of water.

With i c.c. of hydrochloric acid (1:5) and 1 c.c. of chromo-
tropic acid in 50 c.c, .0000795 gram of Ti02 gave a distinct pink
color.

With 2 c.c. of hydrochloric acid (1:5) .00017 gram of Ti02was
required to give the color.

When 5 c.c. of hydrochloric acid (1:5) was used .000424 gram
of Ti02 gave a pink color to 50 c.c.

10 c.c. of hydrochloric acid (1:5) required .000848 gram of
Ti02 for the color.

20 c.c. of hydrochloric acid (1:5) required .00169 gram of
Ti02 in 50 c.c. to give a definite test for titanium.

It may, therefore, be concluded that the destruction of the color
given by chromotropic acid is in proportion to the amount of acid
present, so if this test is used hydrochloric acid should be absent.

EFFECT OF SULPHURIC ACID.

With 1 c.c. of sulphuric acid, specific gravity, 1.435, 0.000318
gram of Ti02 was required to give a pink color in 50 c.c. of
solution.

With 2 c.c. of sulphuric acid 0.000742 gram of Ti02 was required
to give the titanium test.

While more dilute solutions of sulphuric acid were not tried it is
evident that the effect of the sulpuric acid is roughly proportional
to the amount of acid present, and that any appreciable amount of
this acid seriously interferes with the delicacy of the test. The
same is true of hydrofluoric acid.

In the presence of oxalic acid in any appreciable amount chromo-
tropic acid will show .000025 gram of Ti02 in 50 c.c. very dis-
tinctly. Half that amount could be detected but the color is very
faint and its similarity to the color possessed by the solution of
chromotropic acid itself renders the detection of this amount uncer-
tain. In making the test it is best to avoid the presence of free
mineral acids, as they interfere and generally in direct proportion
to the amount of acid present. The neutral chlorides and sulphates
are without effect, as Geisow has stated. It is probable that the
color developed in oxalate solution could be used to determine the

202 HALL AND SMITH— COLUMBIUM. [May i9.

amount of titanium present by comparison with the color developed
by known amounts of titanic acid, but this method would offer no
especial advantage over the hydrogen peroxide method.

THE ACTION OF CARBON TETRACHLORIDE ON THE OXIDES
OF TITANIUM, COLUMBIUM AND TANTALUM.

According to Demarcay (C. R., 104, 111) carbon tetrachloride
vapor passed over the ignited oxides of columbium, titanium and
tantalum changes them to chlorides — in the case of titanium with
the formation of an intermediate oxychloride.

Lothar Meyer (Ber., 20, 681) found no action on oxide of
titanium. He did not try the other two.

Delafontaine and Linebarger (Jr. Am. Ch. S., 18, 532) found
that oxide of columbium was changed to oxychloride, CbOCl3, with
the formation of a small amount of the chloride. In the case of
tantalum the oxide was not driven from the boat but remained be-
hind as a pasty mass, suffering no change to chloride. They sug-
gest this as a possible separation of the two elements columbium
and tantalum.

The vapor of carbon tetrachloride was found to act slowly on
ignited titanic oxide at a low red heat, some chloride of titaninm
being continuously formed. In time all of the oxide was converted
into chloride.

The oxide of columbium is readily acted upon by carbon tetra-
chloride even at a low red heat. The principal product is the
white oxychloride. Some of the yellow chloride is simultaneously
produced. It continues to be formed in small amounts as the oxy-
chloride is sublimed in the vapors of carbon tetrachloride. Colum-
bium oxide heated in a sealed tube with carbon tetrachloride, is
completely changed to chloride after several hours at 2oo°-2 2 5°.
The chloride dissolves in carbon tetrachloride and separates from
it in large, well-formed, needle-like crystals.

The action of the vapors of carbon tetrachloride on ignited oxide
of tantalum is rapid, contrary to Delafontaine and Linebarger, con-
verting it into chloride, which can be readily freed from the carbon
tetrachloride and thus obtained pure. If the carbon tetrachloride
used contains traces of moisture oxide will be produced by the de-
composition of the chloride. This oxide dissolves in the fused
chloride and remains as a glassy mass upon sublimation of the

igos.] HALL AND SMITH— COLUMBIUM. 203

chloride. Therefore, care should be taken in the dehydration of
the tetrachloride used ; otherwise the product will be contaminated
with oxide. This seems to be the best method for the preparation
of tantalum chloride in large quantities and in a high state of
purity. The chloride is an excellent starting-out material for a re-
determination of the atomic weight of tantalum, a number none
too definite, as a study of the series of results obtained by Marignac
(Zei/. anal. Chem., 5, 478) by the analysis of potassium tantalum
fluoride and ammonium tantalum fluoride will show.

The action of carbon tetrachoride on the oxide of columbium
also affords an excellent method for the preparation of the oxy-
chloride of that element. It is produced, however, in a very
voluminous state, and mats together to a tough felt, completely
stopping up any tube used in its preparation. When heated in a
sealed tube it condenses on a warm surface to very compact lus-
trous silky needles. It is very difficult to remove the last traces of
columbium pentachloride from this body. This may be done,
however, by subliming it in a current of chlorine over ignited
oxide, but as long as any carbon tetrachloride is present the colum-
bium chloride will continue to be formed. To make the chloride
of columbium it is necessary to have recourse to the action of sul-
phur monochloride on the oxide or to act on the oxide with carbon
tetrachloride in a sealed tube.

PROPERTIES OF COLUMBIUM CHLORIDE.

As already mentioned, columbium chloride is soluble in carbon
tetrachloride, forming a yellow colored solution. It is much more
soluble when hot than when cold and crystallizes out on cooling in
well defined crystals. It is also soluble in sulphur monochloride,
the solution saturated in the hot being red in color and depositing
yellow crystals of the chloride on cooling. It dissolves in ether
with a yellow color. On evaporating this solution on a water bath
a thick liquid remains, and an acid vapor is given off, but no crys-
tals separate. Upon ignition the mass chars, then burns and leaves
a residue of oxide. On passing dry ammonia gas into the ethereal
solution of the chloride a heavy precipitate is formed. This is am-
monium chloride and columbium nitride. On washing with water
the ammonium chloride is dissolved out, leaving a white residue
which reverts on ignition to oxide of columbium, and when boiled

PROC. AMER. PHILOS. SOC. XLIV. l8o. M. PRINTED AUGUST I, I905

204 HALL AND SMITH— COLUMBIUM. [May i9.

with sodium hydroxide gives off ammoniacal vapors, thus pointing
to nitride of columbium, likely Cb3N5.

Columbium chloride containing some sulphur monochloride was
treated with benzene. The sulphur monochloride dissolved out,
while the columbium chloride was decomposed, leaving an insol-
uble gummy mass. Chloroform dissolved the chloride readily, but
the solution seemed to undergo decomposition on warming and
evaporating, as the liquid became brown and a brown powder sep-
arated. No crystalline product could be procured.

The chloride is also soluble in alcohol. In the cold there is no
decomposition. On warming and concentrating the solution acid
vapors were given off, due perhaps, as H. Rose has suggested, to
the formation of ethyl columbate. That there is no decomposition
in dilute solution is shown by the formation of the compound
CbCl5(C5HnN)6, which was obtained on adding piperidine to the
alcoholic solution (Zeit. anorg. Chem., 36, 100). Other bases,
such as aniline, pyridine, etc., gave addition products which were
insoluble in the solvent.

The best solvent for columbium chloride is carbon tetrachloride.
In this solution reactions should take place, as they do with other
chlorides, in aqueous solution ; also double chlorides, analogous
to the double fluorides, should be formed by bringing together
solutions of the chlorides in carbon tetrachloride.

POTASSIUM FLUOXYPERCOLUMBATE.

When potassium columbium oxyfluoride is dissolved in three
per cent, hydrogen peroxide the solution acquires a yellow color.
When a saturated solution cools a pasty mass of crystals separates.
These are very hard to free from mother liquor. When dry they
have only a faint yellow tint. On dissolving in water, containing
hydrogen peroxide, the solution again becomes yellow in color.
The salt obtained in this way is potassium fluoxypercolumbate of
the following composition — K2Cb02F..H20.

METHOD OF ANALYSIS.

The potassium was determined as sulphate and the columbium
as oxide in the usual way, that is, by expelling the fluorine with
sulphuric acid, boiling with water, filtering out the insoluble

I9oS.] HALL AND SMITH— COLUMBIUM. 205

columbium hydrate and evaporating the filtrate to dryness and
weighing the potassium sulphate after ignition.

The oxygen and water were determined in another sample by
weighing a portion of the substance in a tube sealed at one end,
covering it with a plug of ignited asbestos, connecting with a gas
burette and igniting. The oxygen was collected and measured,
the tube was reweighed, the loss being water and oxygen. The
water was obtained by difference.

Analysis ;

0.4004 gram of the salt gave o. 1672 gram of oxide and
0.2196 gram of K2S04

0.4432 gram of the salt lost 0.0470 gram, which contained 17.9
c.c. of oxygen at 240 and under 742 mm. pressure, or 0.0229
gram, the difference — 0.0241 gram — being water.

Calculated. Found.

K2S04 54. S9 54.84

Oxide 42.28 41-84

O (active) 5.05 5.16

H20 5.68 5.44

This salt was obtained and the above composition ascribed to it
by Piccini {Zeit. anorg. Chan., 2, 21). He regarded it as a deriv-
ative of percolumbic acid and not an addition product of potassium
columbium fluoride and hydrogen peroxide, because the water was
lost on heating at ioo°, while the oxygen did not escape until the
temperature reached 1500.

On crystallizing this salt from concentrated hydrofluoric acid
and hydrogen peroxide in the hope of getting a perfluoride
large plates were obtained, which were quite yellow in color with
a green tint when dry. They did not seem to differ if little or
much hydrofluoric acid was used. The crystals taken for analysis
were obtained from a solution consisting of one-half hydrofluoric
acid, 48 per cent., and one-half hydrogen peroxide, 3 per cent.
They were dried between filter paper and promptly weighed out
for analysis.

Analysis :

0.7260 gram of salt gave 0.3274 gram of oxide and

0.3982 gram of potassium sulphate.

206 HALL AND SMITH— COLUMBIUM.

[May 19,

0.7482 gram of salt lost 0.0784 gram on ignition, i. e., 29.1
c.c. of oxygen at 240 and under 742 mm. pressure, or 0.0374
gram, the difference — 0.0410 gram — being water.

Calculated. Found.

K2SO, 54.89 54-85

Oxide 42.28 42.34

O (active) 5.05 5.00

H20 5.68 5. 48

Hence it may be concluded that the salt obtained from strong
hydrofluoric acid is the same as that got when hydrofluoric acid
is not used.

It would seem impossible to obtain a derivative of percolumbic
acid which does not contain oxygen.

The salt separates from solutions containing hydrofluoric acid in
large well- formed plates, which may be easily measured. They
are much easier to handle than when crystallized from a solution
free from acid. The crystals are always greenish yellow in color.

Piccini states that the salt obtained by him had a slight yellow
tint, but that this color was completely removed by two recrystal-
lizations from hydrogen peroxide. The salt obtained above was
recrystallized six times from hydrogen peroxide containing hydro-
fluoric acid. The crystals from the last crystallization were fully
as highly colored as those which had not been recrystallized. They
were then recrystallized twice from hydrogen peroxide containing
no acid. The resulting salt was practically colorless, but it dis-
solved in water and hydrogen peroxide with a yellow color, which
was intensified by the addition of hydrofluoric acid and on evapor-
ating again to crystallization the crystals were as highly colored as
any obtained previously.

The oxide from the double fluoride, originally used, gave a color
equivalent to 0.4 per cent. Ti02. It was supposed that this color
was due entirely to titanium and that the yellow color of the solu-
tion and of the crystals of potassium fluoxypercolumbate was also
due to this element. To test this supposition 100 grams of the
purest potassium columbium oxyfluoride was crystallized twice from
strong hydrofluoric acid. The crystals obtained were decomposed
with concentrated sulphuric acid, and the hydrate after extraction
with water ignited to oxide. The color which this oxide devel-
oped in oxalic acid solution with hydrogen peroxide was equivalent

19oS.] HALL AND SMITH— COLUMBIUM. 207

to 0.24 per cent, of its weight of titanic acid. It was now heated
in sulphur monochloride and converted into chloride. The latter,
together with the excess of monochloride, was collected in a re-
ceiver and the sulphur monochloride distilled out in the hope that
any titanium tetrachloride present would be expelled with it. The
chloride remaining after removing the sulphur monochloride was
converted into oxide. It contained titanic oxide equivalent to
o. 16 per cent. The oxide was again heated in sulphur monochlor-
ide and treated as before. After the second treatment the titanic
oxide equivalent was .12 per cent, and the color now developed
was different. It was greenish yellow instead of yellow inclining
towards red, which is characteristic of titanium. About five grams
of the oxide which had passed through this treatment were changed
to double fluoride and crystallized from hydrogen peroxide and
hydrofluoric acid. Its solution, in hydrogen peroxide, was yellow
and its color increased in intensity on adding hydrofluoric acid.
The crystals from it were canary yellow with a tint of green, dif-
fering in no respect from those previously obtained.

About ten grams of this yellow salt were next dissolved in water
and hydrogen peroxide. This solution was distinctly yellow in
color. It was divided into two portions. To one portion 0.5
gram of potassium titanium fluoride was added. The color in this
portion became considerably deeper, but the excess of color was
completely discharged upon adding hydrofluoric acid, the two
solutions becoming again identical in color.

Potassium titanium fluoride dissolved in hydrogen peroxide to a
deep yellow-colored solution. On cooling crystals were deposited,
which were not yellow but colorless when completely free from
mother liquor. The addition of hydrofluoric acid to the colored
solution completely destroys the color, and in the presence of
hydrofluoric acid the salt formed is white, resembling potassium
titanium fluoride. When air dried it gives off neither water nor
oxygen on ignition.

The only elements which give a distinctive color in acid solution
with hydrogen peroxide and which might occur here are titanium,
vanadium and molybdenum. Of these the first has been excluded
and the second also by reason of the color which it gives (red to
rose red). There still remains molybdenum. Its color in an
oxalic acid solution with hydrogen peroxide is identical with that

208 HALL AND SMITH— COLUMBIUM. [May 19,

observed in the case of the columbium as free from titanium as it
could be obtained. Although the columbium oxide used for these
tests had passed through several manipulations which should re-
move molybdenum, such as, fusion with sodium carbonate and sul-
phur, changing to chloride with sulphur monochloride and dis-
tilling off the more volatile portion, it was thought best to determine
how much molybdenum would be required to give a test equal to
that obtained from the purest oxide of the columbium at hand.
To this end weighed amounts of molybdenum were dissolved in
oxalic and sulphuric acids, and the color, developed with hydrogen
peroxide, compared with that obtained with a standard titanium
solution of hydrogen peroxide.

1. 0.2780 gram of molybdic acid developed a color equivalent
to 0.0048 gram Ti02, or 0.0058 gram of molybdic acid will give a
color equal to that given under similar conditions by 0.0001 gram
Ti02.

2. 0.0660 gram of molybdic acid gave a color equal to 0.0015
gram Ti02, or 0.0044 gram of molybdic acid is equal to 0.0001
gram Ti02.

The variations in these results is due to the difficulty in matching
the different shades as to intensity of color. The average is about
right, or 0.0050 gram of molybdic acid is equivalent to 0.0001
gram TiO.,. Calculating on this basis, the best oxide of columbium
obtained, which gave a color equivalent to . 1 2 per cent TiO.,,
would contain 6 per cent, of MoOs if the color was due to the
presence of molybdenum, which would be impossible after the
treatments through which the oxide has passed. It had been
crystallized twice as potassium oxyfluoride, fused with sodium car-
bonate and sulphur, the tantalum removed, again crystallized as
the oxyfluoride of potassium and twice from hydrofluoric as potas-
sium columbium fluoride, then changed to chloride in sulphur
monochloride, the sulphur monochloride and the more volatile
portions distilled off and rejected, again changed to oxide and this
treatment with sulphur monochloride repeated. The final oxide
was converted into potassium fluoxypercolumbate and crystallized
once from hydrogen peroxide and hydrofluoric acid. This salt
was yellow in color, and 0.3540 gram of oxide from it, dissolved
in oxalic acid, gave with hydrogen peroxide a color equivalent to
0.000424 gram TiC)2, or .12 per cent. At the most it could not
have contained more than a bare trace of oxide of molybdenum.

i9°5-]

MALL AND SMITH— COLUMBIUM. 209

0.3470 gram of the oxide from the yellow fluoxypercolumbate
was dissolved in oxalic acid and chromotropic acid and diluted to
50 c.c. It gave a very slight pink color, about equal in intensity
to the color developed in 50 c.c. by .000025 gram Ti02 in the same
amount of oxalic acid on treating with chromotropic acid. This
would correspond to less than .01 per cent, of TiO, and is likely
not very far wrong.

From these experiments it may safely be concluded that the
color produced in hydrofluoric acid solution of columbium with
hydrogen peroxide is not due to the presence of titanium. Also
it is likely that columbium itself gives a distinctive color with
hydrogen peroxide, equivalent to from .10 per cent, to .15 per
cent, of its weight of Ti02, yet yellow green instead of straw
yellow, as is given by titanium in dilute solutions. Possibly there
may still be present some other element. For this careful search
will be made.

PREPARATION AND ANALYSIS OF THE YELLOW OXIDE
OF COLUMBIUM.

Hydrated oxide of columbium, containing ten grams of oxide,
was prepared by decomposing the double fluoride with sulphuric
acid, evaporating off the excess of acid and extracting with boiling
water. This hydrate was washed repeatedly with boiling water
and air dried. It was covered with about 20 c.c. of concentrated
hydrochloric acid and brought to boiling for several minutes, until
all of the lumps had thoroughly disintegrated, when it was diluted
to about three times its original volume with water. All but a few
particles were dissolved. This solution was filtered and an equal
volume of three per cent, hydrogen peroxide added. It became
yellow and after a few minutes a yellow precipitate appeared. The
solution was allowed to stand over night. The precipitate was
then filtered out, washed with cold water, in which it was insoluble,
and air dried.

Under the above conditions about one quarter of the oxide in
solution was precipitated by the hydrogen peroxide. If the re-
mainder of the oxide in solution were recovered and dissolved in
hydrochloric acid, as before, a fresh portion of it could be pre-
cipitated on adding hydrogen peroxide. The air-dried precipitate
lost oxygen and water on ignition and regained its white color.

210 HALL AND SMITH— COLUMBIUM. [May i>;

As the precipitation of the columbium was only partial it was
best to be certain of the identity of the portion precipitated. To
this end 2.5 grams of the yellow oxide were obtained, ignited to
remove the excess of oxygen, and changed to the potassium double
fluoride. This analyzed as follows :

0.6822 gram of the salt gave 0.3026 gram of oxide and
0.3966 gram of K2SOt

Calculated. Found.

Oxide 44-52 44.36

K4S04 57.S1 5S.14

Hence the compound obtained is a derivative of columbium.

Columbium is not precipitated from the solution remaining after
the yellow precipitate has been filtered out, by an excess of ammo-
nium hydroxide, until the hydrogen peroxide in the solution has
been destroyed.

Analysis of the yellow precipitate :

0.1917 gram of sample gave 0.1298 gram of Cb20..
0.2556 gram of sample gave 7.8 c.c. of oxygen at 220 and under
741 mm. pressure, equal to 0.0101 gram.

Percentage. Ratio.

Cb2Q5 67.71 I. OOO

O (active) 3.95 0.984

H20 (difference) 28.34 6.240

100.00

This corresponds to Cb(OH)6, or to Cb205H2025H20.

Melikoff and Pissarjewsky (Zeit. anorg. Chem., 20, 340) obtained
a percolumbic acid of the formula HCb04 + «H20, by heating
columbium hydrate with 30 per cent, hydrogen peroxide on a
water bath. They also obtained it by adding sulphuric acid to a
solution of sodium percolumbate, dialyzing out the excess of sul-
phuric acid and the potassium sulphate, then evaporating the clear
yellow solution to dryness on a water bath. They describe it as a
yellow amorphous powder, insoluble in water.

The color of these higher oxides seems characteristic of colum-
bium and is certainly not due to the presence of titanium. The
hydrate obtained by Melikoff and Pissarjewsky contained twice as
much active oxygen, in proportion to the columbium, as did the
hydrate obtained during this investigation.

I9o5.j HALL AND SMITH— COLUMBIUM. 211

DIFFERENCE IN SOLUBILITY OF DOUBLE FLUORIDES.

It is of interest to note that the solubility of potassium titanium
fluoride is increased upon the addition of hydrogen peroxide, while
that of the potassium columbium oxyfluoride is decreased. In
hydrofluoric acid this order is reversed, the columbium salt becom-
ing more soluble and the titanium salt less soluble. This suggests
alternating these solvents in the crystallization of columbium and
potassium double fluorides as one of the best means for removing
titanium.

Recrystallization from hydrofluoric acid in the form of K2CbF.
will remove tin and probably also tungsten from impure potassium
columbium oxyfluoride. Two recrystallizations from that solvent
are sufficient to give an oxide, the ignition of which in a platinum
crucible gave no stain on the crucible. If partially dried oxide
wrapped in the filter paper be ignited directly in a platinum
crucible the presence of a stain on the crucible after removing the
oxide will be a very delicate test for tin. It is likely that when tin
is removed by crystallization tungsten is also, if they are present in
about equal amounts and in such cases where the total amount is
very small. The procedure would remove the necessity for the
tedious sodium carbonate and sulphur fusions used in this work.

BEHAVIOR WITH PRECIPITANTS.

Pennington {/our. Amer. Clicm. Soc, 18, 38) noted that
disodium hydrogen phosphate gave no precipitate in a solution of-
potassium columbium oxyfluoride, while it completely precipitated
titanium from a solution of its double fluoride. This was studied
briefly in order to determine if it might not serve as a quantitative
separation of columbium from titanium. It was found that when
the reagent was added to a solution containing a large excess of
columbium and only a little titanium no precipitate was produced
even on prolonged boiling. If the amount of titanium was increased
slightly both the titanium and the columbium were completely pre-
cipitated by the disodium hydrogen phosphate. This reagent,
therefore, does not separate the two elements.

Geisow found that an alkaline formoxime solution precipitated
zirconium and titanium, but did not precipitate columbium.

Formoxime, or its polymerization product, was prepared by
bringing together solutions of the calculated quantities of formal-

212 HALL AND SMITH— COLUMBIUM. [May 19,

dehyde, sodium carbonate, and hydroxylamine hydrochloride.
The resulting solution gave no precipitate when added to a solu-
tion of titanium as double fluoride, zirconium as double fluoride, or
to a solution of columbium double fluoride. Further, after the
addition of the formoxime solution, ammonium hydroxide failed
to give a precipitate with any of the solutions noted above. It
did, however, give a precipitate with tantalum double fluoride, but
this was only partial.

The statement of Geisow that titanium and zirconium can be
separated from columbium by means of an alkaline formoxime solu-
tion was not verified. The precipitation with tantalum is only
partial, and not complete as stated by him.

It was noted {/our. Amer. Chem. Soc, 26, 1248) that potas-
sium iodate gave a complete precipitation in a solution of potassium
titanium fluoride, and no precipitate with a solution of columbium
double fluoride. Potassium iodate, free from periodate, was pre-
pared, and it was found to give no precipitate with either colum-
bium or titanium, except in acid solution, when both were
precipitated. A solution of a periodate was not tried.

University of Pennsylvania.

MAGELLANIC PREMIUM

Founded in 17S6 by John Hyacinth de Magellan, of London
ICJ05

THE AMERICAL PHILOSOPHICAL SOCIETY

Held at Philadelphia, for Promoting Useful Knowledge

ANNOUNCES THAT IN

DECEMBER, 1905

IT WILL AWARD ITS

MAGELLANIC CxOLD MEDAL

TO THE AUTHOR OF THE BEST DISCOVERY, OR MOST USEFUL INVENTION,
RELATING TO NAVIGATION, ASTRONOMY. OR NATURAL PHILOSOPHY
(MERE NATURAL HISTORY ONLY EXCEPTED) UNDER THE FOLLOWING
CONDITIONS :

i. The candidate shall, on or before November 1, 1905, deliver free of
postage or other charges, his discovery, invention or improvement, ad-
dressed to the President of the American Philosophical Society, No. 104
South Fifth Street, Philadelphia, U. S. A., and shall distinguish his per-
formance by some motto, device, or other signature. With his discovery,
invention, or improvement, he shall also send a sealed letter containing
the same motto, device, or other signature, and subscribed with'the real
name and place of residence of the author.

2. Persons of any nation, sect or denomination whatever, shall 1 e
admitted as candidates for this premium.

3. No discovery, invention or improvement shall be entitled to this
premium, which hath been already published, or for which the author
hath been publicly rewarded elsewhere.

4. The candidate shall communicate his discovery, invention or im-
provement, either in the English, French, German, or Latin language.

5. A full account of the crowned subject shall be published by the
Society, as soon as may be after the adjudication, either in a separate
publication, or in the next succeeding volume of their Transactions, or in
both.

6. The premium shall consist of an oval plate of solid standard gold of
the value of ten guineas, suitably inscribed, with the seal of the Society
annexed to the medal by a ribbon.

All correspondence in relation hereto should he addressed
To the Secretaries of the

AMERICAN PHILOSOPHICAL SOCIETY
No. 104 South Fifth Street

PHILADELPHIA, U. S. A.

TRANSACTIONS

OF THE

American Philosophical Society

HELD AT PHILADELPHIA

For Promoting Useful Knowledge

Vol. XXI, New Series, Part I. Just Published

The Morphology of the Skull of the Pelycosaurian Genus
Dimetrodon. By E. C. Case

Subscription— Five Dollars per Volume
Separate parts are not sold

Address The Librarian of the

AMERICAN PHILOSOPHICAL SOCIETY
No. 104 South Fifth Street

PHILADELPHIA, U. S. A.

\V^>" PROCEEDINGS

OF THE

AMERICAN PHILOSOPHICAL. SOCIETY

HELD AT PHILADELPHIA
FOR PROMOTING USEFUL KNOWLEDGE

Vol. XLIV. August-December. 1905. No. 181.

CONTENTS.

A Study of the Anatomy of Phalaenoptilus, Ridgway. By

Margaret E. Marshall 213

Stated Meeting, April 28 241

Stated Meeting, May j 241

Stated Meeting, May ip 241

Stated Meeting, October 6 242

Stated Meeting, October 20 243

Stated Meeting, November j 243

Stated Meeting, November 17 243

Stated Meeting, December 1 243

Stated Afeeting, December 15 244

Index 245

List of Members 249

w PHILADELPHIA
THE AMERICAN PHILOSOPHICAL SOCIETY
104 South Fifth Street
1906

GENERAL MEETING— 1906

The next General Meeting of the Society will he held on April 17-20.
1906; beginning on the evening of Tuesday, April 17.

Wednesday, April iS. will be devoted to the presentation and discussion
of scientific papers, and Thursday, 19 and Friday, 20, to the ceremonies
connected with the celebration of the 200th Anniversary of the Birth of
Benjamin Franklin.

Members desiring to present papers on subjects of science at the Gen-
eral Meeting are requested to communicate with the Secretaries at once.

CORRIGENDUM

In No. 179 the volume number should be corrected to read Vol. XLIV.

Pages 1-4, including a title page containing the correct volume number, to be
substituted for like pages in No. 179, are inserted at the end of this volume.

Members who have not as yet sent their photographs to the Society will
favor by so doing ; cabinet size preferred.

*... _ — _, — t &.

confer a favor by so doing ; cabinet size preferred.

It is requested that all correspondence be addressed
To the Secretaries of the

AMERICAN PHILOSOPHICAL SOCIETY
104 South Fifth Street

Philadelphia, U. S. A.

PROCEEDINGS

OF THE

AMERICAN PHILOSOPHICAL SOCIETY

HELD AT PHILADELPHIA

FOR PROMOTING USEFUL KNOWLEDGE

Vol. XLIV. August-December, 1905. No. 181.

A STUDY OF THE ANATOMY OF PHAL.ENOPTILUS,
RIDGWAY.1

BY MARGARET E. MARSHALL.

(Plates IV, V and VI.)

( Received June ig, igoj. )

Introduction.

The present paper is a contribution to the knowledge of Phaltz-
noptilus nuttalli nil 'idles (Brewster), the Poorwill, and presents an
account of the alimentary, respiratory and urogenital organs, the
central nervous system and all the muscles of the anterior extremity
and those of the thigh.

As generally defined now the Caprimulgi include the three
families of Steatornithidse, Podargidse and Caprimulgidas. In re-
gard to their distribution Fiirbringer (1888) says that the Capri -
mulgidae represent the largest family (some 100 species) and with
almost cosmopolitan distribution (exclusive of New Zealand, the
pacific subregion and the southern part of South America) ; the
Steatornithid?e, represented by a single species, occur in caves in
the tropical Andean region and the West Indies ; the Podargidse
consisting of about 20 species inhabit the oriental region, particu-
larly New Holland and Papuasia. Of the Caprimulgidge the fol-
lowing genera occur in North America : Antrostomus, Phalcenop-

1 Contributions from the Zoological Laboratory of the University of Texas,
No. 68.

PROC. AMER. PHILOS. SOC. XLIV. l8l G. PRINTED OCTOBER 28, I905.

214 MARSHALL— A STUDY OF THE [June i9,

tilus, Nyctidromus and Chordeiles. Of these genera Phalcenoptilus
extends from Guatemala northward in the western United States
nearly to British Columbia, and is represented by three geograph-
ical races of one species. This genus was first established by Ridg-
way (1880), and is defined as follows by Coues (1903) : "Nos-
trils tubular, cylindrical, opening forward and outward. Rictal
bristles immense, but simple. Tarsus naked except just on the
joint above (as in Nyctidromus), as long as middle toe without
claw. Tail square, much shorter than the rounded wings, which
fold nearly to its end." No anatomical description of this genus
has heretofore appeared so far as the writer knows.

The aim of this study is the interesting question of the homol-
ogies of the Caprimulgi. Fiirbringer (/. c. ) discusses at length
the varying views on their relationship to Ardeidae, Glareolidae,
Strigidae, Cuculidas, Galbulidae, Trogonidae, Coraciidae, Leptoso-
midae, Todidae, Momotidae, Coliidae, Cypselidae, Trochilidae and
Passeres (Eurylcemus, Hirundo), and sums up his position in these
words : " On the ground of the given comparisons, I am inclined
to regard the Caprimulgidae, Steatornithidse and Podargidae as
independent but closely related families, and as united under Capri-
mulgi ; they stand in remarkable genealogical relations in the first
line to the Striges and Coraciae and in the second to the Trogonidae
and Cypselidae, while the relation to the other families coming in
question is less near and direct." Gadow (1891), considers the
Caprimulgi as related ancestrally to the Striges and laterally first
with the Coraciae and second with the Cypseli. It soon became
apparent to the writer that this problem of affinities could not be
settled by the investigation of a single genus. Accordingly this
paper is intended to be the first of a series dealing with these birds,
and is essentially descriptive, general theoretical considerations
being postponed until personal studies have been made upon other
forms. Because there was not time to describe the whole anatomy
it seemed advisable to omit the osteology, since most of the previous
work has been done upon the skeleton.

Special anatomical monographs upon such species of birds are so
few, and yet so much needed, that it is hoped this one may be of
some service to comparative anatomists.

The material used consisted of two entire adult females secured by
Dr. Thos. H. Montgomery, Jr., in the month of June, 1904, in

I9o5.] ANATOMY OF PHAL.-ENOPTILUS, R1DGWAY. 215

Brewster County, Texas, one preserved in alcohol and the other in
formalin.

This work has been done entirely under the direction of Dr.
Montgomery, and I am very much indebted to him for many help-
ful suggestions, and for his unfailing sympathy and encouragement
during the preparation of this memoir.

I. Alimentary Tract.

This bird is remarkable for its enormous month. Arranged in a
regular series along the upper border of the gape there are on each
side of the mouth eight long vibrissas, modified feathers.

The tongue (T, PI. IV., Fig. 10), is slender and pointed.
Posteriorly it is bifid and fimbriated. The hyoid bone (Figs, i
and 10) consists of the following parts: os entglossum (Ent. g.),
basihyal (Has. h.), urohyal (Ur. //.), basibranchial (Bas. Ik),
ceratobranchial (Cer. b.), and epibranchial (Ep. b.). The ent-
glossum is entirely cartilaginous and is bifurcated in the posterior
half, the forks articulating in each side with the basihyal. The
osseous basihyal is a solid piece broadening posteriorly, reaching
its greatest width where the basibranchials come off. It then nar-
rows immediately into the urohyal which has the same structure
except for the cartilaginous tip. The urohyal is about twice the
length of the basihyal. The "horns" of the tongue bone are
also of cartilage. The basal segments, the basi-branchials, are
about one half the entire length of the horn ; the articulating
joints, the ceratobranchials, are a little more than one fourth the
length of the horn ; the last members, the epibranchials, are about
one fourth the length of the horn.

The wide pharynx (Pha., PI. IV, Fig. 10) is succeeded by the
oesophagus (CEs., Fig. 10) which is immediately slightly dilated.
There is no crop. Behind its anterior dilation the oesophagus
gradually narrows until at its posterior end the diameter is little
more than half the diameter of its anterior portion.

The cesophagus passes over into the proventriculus (Prov., PI.
IV, Figs. 9, 10, 15), which opens into the anterior end of the
gizzard (Giz., Figs. 9, 10, 15) somewhat to the right of the mid-
line. The gizzard is overlain anteriorly by the liver lobes, and
extends posteriorly to the region of the cloaca.

The intestine (Int., PI. IV, Figs. 9, 15), arises from the right

216 MARSHALL— A STUDY OF THE

[June 19,

side of the gizzard at the base of the proventriculus, and consists
of four distinct divisions :

1. The duodenum {Duo., Figs. 6, 9, 15) makes up the entire
first loop. It extends from the pylorus almost to the posterior end
of the stomach. It then bends anteriorly, and at the edge of the
right liver lobe passes over into the small intestine. The duodenum
is about 43 mm. in length.

2. The small intestine (Figs. 12, 15) lies between the duodenum
and the insertion of the caeca. It measures about 85 mm.

3. The terminal intestine (Figs. 12, 15) extends from the in-
sertion of the caeca to the anus and is very short. Anteriorly
the diameter is very small but posteriorly it is dilated at the cloaca.
Its length is about 18 mm. Thus the entire length of the main in-
testine from pylorus to anus is 146 mm. The intestine consists of
three closed loops of which the duodenal is the first in course.
The ascending branch of the third loop and the following portion
of the small intestine are covered by the first and second loops.
The descending branches of the second and third loops are to the
left of their respective ascending branches. The intestinal ar-
rangement agrees with the iso-orthoccel type of intestine as defined
by Gadow (/. c. ).

4. Two cozca (Co?., Figs. 12, 15) are present. They are quiet
long and the terminal half of each almost equals the small intestine
in size. At about one third the length of the caeca from their inser-
tion there is on each a constriction, and at this point the diameter
is less than in any part of the alimentary tract. From tip to inser-
tion the caeca measure about 35 mm.

Filling the duodenal loop is a pale, slender organ, the pancreas
(Pan., Figs. 6, 9, 15). It consists of two branches, the main
branch occupying the position mentioned. Extending beyond the
edge of the ascending branch of the duodenum is the smaller division.
The pancreas has two ducts (Pan. d. ), both coming off on the dor-
sal surface. The larger duct, which comes from the main portion,
arises proximal to the branching off of the smaller part, and run-
ning anteriorly close to the descending part of the loop of the duo-
denum enters its ascending branch just about where it begins to
curve along the right liver lobe. The small duct comes from the
smaller division on its inner edge at a point about one third the
length of that division from its anterior end.

igoS.] ANATOMY OF PHALjENOPTILUS, RIDGWAY. 217

Only the merest rudiment of a spleen (Spl., PI. I, Fig. 9) is
present. It is a small, whitish, almost round body lying under the
right lobe of the liver close beside the gall-bladder.

The liver (Ziv., PL IV, Figs. 9, 10, 15) consists of two smooth
lobes, the right being somewhat the larger. The lobes are con-
nected anteriorly. They extend from the heart back over the
stomach for about one half its length. From the right lobe just
above the duodenum comes off the greenish colored gall-bladder
(G. bl., Fig. 9). The figure exhibiting the pancreatic ducts shows
also two others, one situated between the ducts of the pancreas
and another anterior to the smaller pancreatic opening. These
I have taken to be the liver ducts (Ziv. d.). They could not be
traced further on account of the mutilated condition of the bird.

The salivary glands were not found.

II. Respiratory Organs.

The glottis is an oval aperture situated behind the root of the
tongue leading into the trachea. Immediately posterior to the
glottis is a bilobed fimbriated fold of the mucous membrane.

Ventrally, the larynx (Zar., PL IV, Fig. 10) presents two
rather fiat, somewhat triangular, cartilages, the thyroids, which ter-
minate anteriorly at the posterior border of the basihyal. The
thyroids are narrowed in front but not pointed. The two carti-
lages are divided anteriorly by the urohyal which extends almost to
their base.

The length of the trachea (Tra., PL IV, Fig. 10) from the
larynx to the branching of the bronchii is about 5 cm. The rings
of the trachea, about seventy-seven in number, are complete with
two exceptions ; the anterior dorsal has its dorsal edges fused with
the thyoid, and the rings of the posterior dorsal surface are incom-
plete. There are three modifications of these rings on the ventral
surface. Succeeding the base of the larynx there are four simple
rings. The next three are slightly constricted in the middle.
From this point down to where the trachea begins to broaden out
before passing between the forks of the furcula, the rings are inter-
laced, trowel fashion. Between the last of these rings and the
branching of the bronchii we find a repetition of the condition first
described, only the rings are broader and stronger.

The trachea has only two sets of muscles (PL IV, Figs. 7, 8, 10).

218 MARSHALL— A STUDY OF THE [June iy,

One pair comes off on each side from the last tracheal ring, and
continues anteriorly almost to the larynx, at which point it spreads
out fan-like, the delicate fibers being attached to the upper end of
the windpipe ; this muscle is the trachealis lateralis (Tr. Lai.),
named according to the description given by Shufeldt (1890),1
though it does not agree in all points with it; there is a partial
agreement with Gadow's (/. c. ) m. tracheo-bronchialis. The sec-
ond pair of muscles is much stronger but shorter than the last
described, and the origin is more ventral. They arise from the
trachea on each side between the sixth and the tenth rings, count-
ing forward from the last tracheal ring. The muscles become grad-
ually smaller as they approach the insertion which is about the mid-
point of the proximal part of the first rib articulating with the
sternum. This is the m. stemo-trachealis {St. tr.), though it does
not agree in all points with Gadow's (/. c.) description of the
muscle of the same name.

The syrinx (PI. IV, Figs. 7, 8) is tracheo-bronchial. On the
ventral surface (Fig. 7) the last tracheal ring is directed downward
forming with the one above an almost triangular space, of which
the preceding tracheal ring is the base. Corresponding to the last
tracheal ring on the left side there are two on the right, separated
by a small space. These rings and the first bronchial rings are
fused at their inner extremities to a small membrane at the base
of which the bronchii separate. This membrane is stronger than
that between the rings, and is of a yellowish color. The second
bronchial ring bifurcates at its inner extremity, the lower branch
fusing with the following ring, thus causing it to be much enlarged
at its inner termination. Each bronchus is bounded on its inner
surface by a cartilaginous rod, and this rod closes the almost circular
space embraced partially by the above mentioned bifurcation. The
second, third and fourth rings are larger than any of the others, less
flexible, and of a yellowish color.

The membrana tympaniformis externa (Tvm. ex.), is double in
this bird. It is bounded by the second and fourth rings, and
crossed in the middle by the third. This third ring is larger than
the other two. The fourth at its inner extremity loses the yellow-
ish color and for this reason seems shorter than it really is. In all,
the number of rings in the right bronchus is fourteen and in the
left twelve. This may be an individual variation.

1 Myology of the Raven, Philadelphia, 1S90.

*9°5-]

ANATOMY OF PHAL.4ENOPTILUS, RIDGWAY. 219

On the dorsal surface (Fig. 8) we find a condition quite differ-
ent from that shown on the ventral. All of the bronchial rings
are incomplete dorsally. Counting forward from the last tracheal
ring, we find between the third and sixth rings a cartilaginous
bridge situated in the mid-line of the trachea. It is like the carti-
lage of the rings, and is, probably, a fusion of the dorsal ends of
the fifth, fourth and part of the third rings with an extension to
the sixth. This bridge broadens posteriorly, and at the third ring
from the last divides, the branches terminating at the last tracheal
ring. Down the center of the pyramidal-shaped area enclosed by
this fork passes a yellowish rod which is quite resistant to the needle
and is probably bony. It extends beyond the ends of the fork.
This pyramidal area is bounded posteriorly by a cartilaginous ridge
which bends back in the mid-line and gradually fades out on the
side as the bronchial half-rings are reached, forming at this point
the upper boundary of the membrana tympaniformis interna ( Tym.
in.). Below this is another ridge of like structure which forms
the lower boundary of the inner tympaniformis.

Between the dorsal ends of the third, fourth, and fifth half rings
there is a round whitish body covered irregularly with yellowish
brown spots. About half way between the ends of the next four
half rings there is, in the right tympaniformis, a white club-shaped
body (PI. I, Fig. 8, x). The membrane is much thinner around
its edges than elsewhere. Strands of a dark pigment substance are
seen around the edges and over the inner surface of this object
when the bronchus is opened. A similar structure has evidently
been lost from the left tympaniformis, judging from the appear-
ance of the membrane. All of these bodies are probably bits of
cartilage.

III. Female Urogenital Organs (PI. IV, Fig. n).

The left ovary ( Ov. ) is situated anterior to the left kidney. The
oviduct (Z. ovi) , a very much convoluted tube, terminates anteriorly
in an infundibulum {Inf.) facing the left ovary. It lies to the left
of the pelvic cavity and opens posteriorly into the left side of the
cloaca just behind the ureter. The right ovary is absent, but a
very much reduced oviduct (A5, ovi. ) is present. The infundibulum
is readily made out, and slight convolutions of the duct are to be
observed.

220 MARSHALL— A STUDY OF THL [June 19,

The fused kidneys (A'.) extend from the lungs to the pelvic
cavity. The right kidney, slightly larger than the left, consists of
three lobes, the middle one being the smallest. The two lobes of
the left kidney are of about equal size. The ureters { Ur. ) pass
posteriorly to the cloaca which they enter on its dorsal surface
median to the oviducts.

IV. Central Nervous System (PI. I, Figs. 2, 3 ; PI. II, Fig. i6a).

The brain of this bird is notably small as compared with the size
of the head. Its length much exceeds its breadth, resembling in
this respect the brain of a lizard. The large optic lobes ( Op. I.)
are only partially covered by the cerebral hemispheres. The cere-
bellum (67;.), which is comparatively large, covers the medulla
oblongata {Med.) and on each side of it a flocculus {Flo. ) is
found. The greatest length of the cerebrum {Cere.) is about 8
mm., its width 9.5 mm. The longest measurement of the cere-
bellum is approximately 6 mm., its breadth above the flocculi 5
mm. The uncovered portion of the optic lobes measures from
dorsal to ventral surface about 5.5 mm., anterior to posterior border
3 mm. No drawings or measurements of the ventral surface of the
brain could be made on account of its torn condition.

The spinal cord {Sp.), is marked by two important swellings;
one in the cervical region known as the brachial plexus {Br. u.)
(PI. II, Fig. 1 6a), and one in the posterior region as the sacral
plexus. Anterior to the brachial plexus the cord is larger than it
is between this plexus and the succeeding one. The swelling which
indicates the branchial plexus begins at the tenth nerve and termi-
nates with the thirteenth. Three nerves take part in the formation
of this plexus, the eleventh, twelfth and thirteenth. The second
or middle nerve is the largest of the three, the third the smallest.
Soon after leaving the cord the second nerve bifurcates, one branch
going to each of the other two and all intimately related. The
posterior part of the spinal cord was too badly broken for the nerves
of that region to be made out.

V. Sense Organs (PI. IV, Figs. 2, 4, 5).

The nostrils {N~os. ) are tubular and cylindrical, opening for-
ward and outward. Vibrissas, very much shorter and more delicate
than those around the gape, are observed about the nostrils. These

i9°5-]

ANATOMY OF PHAL/ENOPTILUS, RIDGWAY. 221

are arranged in a somewhat circular fashion just posterior to the
external opening. On each side of the median ridge of the palate
is a long, narrow slit bounded by fimbriated folds of mucous mem-
brane, the internal nares.

The pecten of the eye (Figs. 4, 5) consists of four folds. It
measures in height about 2 mm., and its basal breadth is about
1.5 mm. Like the choroid coat it is heavily pigmented.

VI. Myology.

Only muscles of the extremities have been considered, and in
naming them the terminology of Gadow (/. c. ) has been followed
as strictly as possible. There are, however, many deviations from
his definitions.

1. Anterior Extremity.

Here are described all the muscles of the wing proper, also all
coming from the shoulder girdle, ribs and vertebrae and inserting
upon the wing, also all the muscles inserting on the scapula and
coracoid. The metacarpals named by Gadow (/. c. ) I, II and III
are herein termed II, III and IV, for recent embryological investi-
gation show the first and fifth to be the ones lost.

A. Pectoral Muscles.

1 . M. pectoralis. The pars propatagialis and pars abdominalis
are absent.

Pars thoracica {Pect., PI. V, Fig. 24; PI. VI, Fig. 25).
This is the large superficial muscle of the breast, and covers the
other breast muscles. It has an extensive origin, coming from the
clavicle and the membrane between that bone and the sternum ;
from the surface of the keel, the upper half; the posterior border
of the sternum ; and the posterior lateral portion of the breast bone.

It has two points of insertion, both of which are on the humerus.
The short strong tendon which terminates on the ventral projection
of the humerus, just anterior to the biceps, is the posterior inser-
tion. The fibers of the anterior portion converge and pass ob-
liquely to the dorsal crest of the humerus and are there attached
fleshily.

2. M. supracoracoideus {Sup. cor., PI. VI, Figs. 25, 30). This
is a double-feathered muscle arising from that portion of the coraco-
clavicular membrane not occupied by the muscle just described,

222 MARSHALL— A STUDY OF THE [June i9,

from about the lower half of the keel and from that portion of the
body of the sternum not appropriated by the above muscle. The
fibers converge to a line which is dorsal to the mid-line, passing
over into a strong flat tendon that bends around to the inner sur-
face of the coracoid. The tendon goes through the foramen trios-
seum and is attached to the humerus on its dorsal projection.

3. Coraco-brachialis posterior (Cor. br. p., PI. VI, Figs. 25, 26,
30). When the m. pectoralis is turned back this small muscle is
seen extending out from under the supracoracoideus. It arises
from the dorsal proximal half of the border of the coracoid. The
fibers converge to form a short strong tendon which is attached to
the antero-ventral margin of the humerus just anterior to the
pneumatic foramen.

B. Other trunk muscles inserting on wing, scapula and coracoid.

1. M. delioideus major (PI. V, Figs. 16, 22). This muscle
consists of an anterior and posterior portion. The delicate an-
terior part {Del. a.) arises fleshily from the inner surface of the
clavicle and the neighboring portions of the scapula. It emerges
from the foramen triosseum, crosses the tendon of the m. supracor-
acoideus, runs entirely around the projection to which the tendon
of the muscle is attached, and makes a fleshy insertion at the base of
the anterior border of the humeral crest.

The large posterior portion (Del. p.) springs from the dorsal
border of the clavicle and adjoining dorsal surface of the scapula.
It passes obliquely downward and is inserted fleshily on the humeral
crest and along the shaft of the humerus for about one half its
length. This differs somewhat from Gadow's description.

2. M. delioideus minor. This muscle could not be made out,
therefore is probably absent.

3. M. latissimus dorsi (PI. V, Figs. 17, 22 ; PI. VI, Fig. 26).
This is the most superficial muscle of the back, and is revealed by
the removal of the skin. It consists of two portions, a very thin
anterior layer and a much larger and stronger posterior muscle
bundle.

The anterior portion (Lat. d. a.) arises from the spine of the
last cervical vertebra and from the spines of the two following
dorsal vertebras. The fibers pass in a transverse direction over the
scapula, converging somewhat after crossing it, and diverge as they

igo5.] ANATOMY OF PHAL^NOPTILUS, RIDGWAY. 223

approach the humerus. The muscle inserts upon the humerus be-
tween the pars scapuli-cubitalis and the pars humero-cubitalis near
the lower point of the humeral crest.

The posterior part (Zat. d. p.) comes from the last dorsal verte-
bra and the anterior rim of the ilium. The fibers converge rapidly
as they pass anteriorly, and find their insertion on the humerus
beneath the anterior border of the above.

4. M. rhomboideus superficialis (Rh. s., PI. V, Figs. 17, 22).
This flat muscle comes from the last two cervical vertebrae and the
following dorsal vertebrae. It is inserted fleshily on the dorsal
part of the furcula and on the entire dorsal border of the scapula.

5. M. rhomboideus profundus {Rh. p., PI. IV, Fig. 17). This
muscle is covered by the last mentioned muscle and by the pos-
terior portion of the latissimus dorsi. It springs fleshily from the
last cervical and first dorsal vertebrae. The outward directed fibers
find a fleshy insertion on the posterior half of the dorso-median
border of the scapula, the insertion being continued down to the
posterior tip of this bone.

6. M. scapuli-humeralis anterior. This muscle is absent.

7. M. scapuli-humeralis posterior (Sc. hum. p., PI. V, Figs. 17, 22;
PL III, Fig. 26). This large, somewhat rhomboidal-shaped muscle
comes from the outer surface of the posterior two thirds of the
scapula. It is covered by both portions of the latissimus dorsi.
Its fibers are directed forward and downward, converging rapidly
to form a small, round bundle which is attached within the fora-
men pneumaticum.

8. M. subscapularis (S. sc, PI. VI, Fig. 26). This muscle
becomes visible after the removal of them, scapuli-humeralis poste-
rior and m. scapuli-cubitalis. Anteriorly it comes from the dorsal
surface of the scapula just posterior to the origin of the scapuli-
cubitalis, and posteriorly from the lower border of the same bone
where it is overlain by the scapuli-humeralis posterior. In its
middle portion it is divided into two parts by the anterior m. ser-
ratus superficialis, the inner division, subscapularis internus, coming
from the ventral surface of the scapula. The outer portion is the
subscapularis externus.

9. M. serratus superficialis (PI. VI, Fig. 26) is made up of two
parts. The smaller anterior division (Ser. s. a. ) comes from the last
cervical rib and its process uncinatus. The fibers are directed upward

224 MARSHALL— A STUDY OF THE

:;<-

and forward. The muscle terminates tendinously on the ventral
border of the scapula. It divides the subscapularis.

The large posterior division (Ser. sp.) springs with four scallops
or teeth form the second and third dorsal ribs, just below the proc-
esses of these ribs. The fibers of these closely-related bundles are
directed upward and forward. The most posterior bundle is in-
serted on the posterior tip of the scapula. The others do not
reach the scapula but terminate on the ribs and the membrane
connecting them in this region.

10. M. serratus profundus (Ser. p., PI. VI, Fig. 26), occupies
a more dorsal position than the last mentioned muscle, and is
exposed by the removal of the m. scapuli-humeralis posterior. The
edges of the two thin portions composing it overlap. The dorsal
bundle comes from the free cervical rib at the outer end of the ver-
tebral projection. The lower bundle arises from the last cervical
rib and from the membrane connecting this and the preceding rib.
Both bundles are inserted on the posterior median border of the
scapula.

11. M. sterno-coracoideus (St. co., PI. VI, Fig. 26) is covered
at its origin by the abdominal muscles. It arises tendinously from
the first, second and third sternal ribs. This small muscle passes
obliquely to the lateral projection of the sternum below the
coracoid.

12. M. suhcoracoideus (Sub. co., PI. VI, Fig. 26). This deli-
cate fusiform muscle is revealed by the separation of the scapula and
coracoid. It springs tendinously from the inner anterior border
of the coracoid about one third the length of that bone from its
distal end. It is inserted on the humerus proximal to the coraco-
brachialis posterior. The fibers of this muscle are closely associ-
ated with those of the subscapularis. It differs considerably from
Gadow's description.

C. Muscles restricted to the wing.

1. M. propatagialis, pars propatagialis musculi deltoidei (Pro.).
This includes both the long and short tendons, the other parts
being absent. It arises rleshily from the dorsal end of the clavicle
and from the neighboring portions of both coracoid and scapula.
It is a flat muscle, about 14 mm. long and 5 mm. broad. At its
distal end it tapers off into two tendons, the upper and more deli-

i9o5.]

ANATOMY OF PHAL/ENOPTILUS, RIDGWAY. 225

cate being the m. propatagialis longus the lower and stronger ra.
propatagialis brevis.

a. M. propatagialis longus {P. pat. 1., PI. V, Figs. 22, 24).
The tendon of this muscle runs along the anterior margin of the
patagium, with which membrane it is intimately connected. Thence
it continues as a very delicate tendon to the distal end of the radius.
It becomes flattened as it passes over the os radiale, and continues so
to its insertion. The flattened tendon passes to the ventral side of
the os magnum along its base, and is inserted on the posterior
proximal projection of the pollex digit. From this point a pyram-
idal-shaped tendon with its apex on the pollex-digit extends down
to the third metacarpal.

b. M. propatagialis brevis (P. pat. b., PI. V, Figs. 22, 23) is
very complex in this bird. The tendon is larger than the longus
and flattened. It continues distally to the m. extensor metacarpi
ulnaris (radialis?) where it bifurcates, about 5 mm. from the distal
end of the humerus. The longer branch runs back with the m.
extensor metacarpi ulnaris (radialis?) to become inserted on the
humerus just distal to this muscle, and at the base of the tubercle
of the external condyle of the humerus. The shorter one continues
distally about 2 mm., then passes back obliquely to the m. extensor
digitorum communis and here it bifurcates, the proximal short
branch running back with the above muscle to insert itself on
the tubercle above the external condyle of the humerus and above
the origin of the m. ectepicondylo-radialis. The distal extending
branch becomes flattened at its insertion, which is at the base of the
styloid process of the radius on its ulnar side, near the m. extensor
pollicis longus and covered by it. From the second bifurcation
comes off a broad band which passes directly across to the ulna and
is inserted on that bone about 7 mm. or 8 mm. from its proximal
end.

2. The metapatagium was torn away, so I can say nothing about
the ;;/. metapatagialis.

3. M. biceps brachii, pars propatagialis {Pi., PI. VI, Figs. 20,
27 ; PI. II, Fig. 24). This large muscle lies on the anterior sur-
face of the forearm, and arises as two heads. The long head comes
from the anterior end of the coracoid as a strong, flat tendon. The
short head passes immediately into a stout muscle. The two pos-
teriorly unite to form a fusiform muscle which inserts at the elbow

226 MARSHAL!.— A STUDY OF THE

[June 19,

joint, the more delicate portion of the split tendon being attached
to the radius on its inner surface, the other portion to the ulna at
the base of the m. flexor digitorum profundus and dorsal to the m.
brachialis inferior.

4. M. brachialis inferior (Br. inf., PI. V, Fig. 24). This trap-
ezoid-shaped muscle arises fleshily from the distal end of the
humerus, and from its inner surface interior to the origin of the m.
extensor metacarpi ulnaris (radialis?). It crosses to the ulna and
is inserted on that bone beyond the elbow joint and between the
separated portions of the m. flexor digitorum profundus.

5. M. triceps cubiti. This muscle consists of two parts, one long
head and two short ones.

a. Pars scapuli-cubitalis (Pars. sc. cub., PI. V, Figs. 16, 17,
22 ; PI. Ill, Fig. 29). This one arises from the neck of the scap-
ula, posterior to the scapular projection which forms part of the
glenoid fossa. It passes obliquely across the humerus above the
insertion of the latissimus dorsi, continues down the dorsal pos-
terior side of the humerus and near its distal end comes off in a
strong flat tendon which is inserted on the rim of the dorsal prox-
imal process of the ulna.

b. Pars humero-cubitalis (Pars. hu. cub., PI. V, Fig. 24; PI.
IV, Fig. 28) arises by two heads, the inner comes from within the
rim of the humeral head, while the stronger has its origin on the
outer aspect of the head of the humerus, and from about its proximal
quarter. This part ends in a tendon and a broad aponeurosis
inserted on the proximal edge of the olecranon process of the ulna,
and the intervening space between this process and the insertion of
the scapuli cubitalis.

6. Mm. entepicondylo-antibrachiales.
a. Mm. entepicondylo-radiales.

(1) Pronator sublimis (Pron. s., PI. V, Fig. 24). This is the
most superficial muscle of the inner arm. It springs tendinously
from above the internal condyle of the humerus, and interior to the
origin of the brachialis inferior. It passes obliquely across the
interosseus space to become inserted on the ventral side of the radius
for about one third its proximal length.

(2) Pronator profundus. This muscle is smaller than the sub-
limis, and is covered for nearly its entire length by the superficial
muscles. It arises from the lower edge of the internal condyle of the

i9°5-]

ANATOMY OF PHALiENOPTILUS, R1DGWAY.

humerus, and is almost concealed at its origin by the strong tendons
of the flexor digitorum sublimis. It is split in two by the extensor
indicis longus. The fibers of the upper half pass obliquely over
to the radius and are inserted on that bone under the pronator
sublimis, extending about as far distal) y as that muscle. The
lower half bends under the m. extensor indicis longus and is inserted
on the radius in a position corresponding to the upper half. This
muscle is not shown in the drawings.

b. M. entepicondylo-ulnaris is absent in this bird.

7. M. ectepicondylo-ulnaris (Ect. u., PI. V, Fig. 19) arises by
a strong tendon from the posterior projection of the external con-
dyle of the humerus below the m. extensor digitorum communis, and
is covered by the tendon of the m. extensor carpi-ulnaris (radialis ?).
It passes over to the anterior surface of the ulna, and is there
inserted fleshily for fully two thirds the length of that bone.

8. M. ectepicondylo-radialis (Ect. r., PI. V, Fig. 19). This
muscle arises from the posterior projection of the external condyle
of the humerus, below the origin of the m. extensor digitorum com-
munis and below the insertion of the second forward directed
branch of the m. propatagialis brevis. It passes directly over to the
proximal end of the radius and is inserted fleshily along its dorsal
surface for about one third its length.

9. M. flexor carpi ulnaris (E. carp, ul., PI. V, Fig. 24). This
is the largest muscle of the forearm. It arises by a strong, flat ten-
don from the posterior border of the external condyle of the hume-
rus. It is held in place by a ligament which passes from the con-
dyle over and under the tendon to the base of the olecranon process
of the ulna, thus forming a loop. The muscle runs along the ven-
tral surface of the ulna and at about the middle of that bone
separates into two tendons. Both continue distally to become
inserted on the outer border of the os ulnare, the more delicate on
the lower edge.

10. M. ulni metacarpalis veiitralis ([//. met. v., PL V, Fig. 24).
This muscle arises fleshily from about the middle three fifths of the
ventral and posterior surface of the ulna, and is broadest at the
distal end immediately before passing over into the tendon which
crosses in front of the os radiale, to the surface of the third meta-
carpal and is inserted on its dorsal proximal projection. It is cov-
ered by the tendons of other muscles which find their insertion in

22S MARSHALL— A STUDY OF THE

[June 19,

this region. The distal portion of this tendon with that of the m.
flexor digitorum is held in place by a delicate ligament extending
from the distant ventral border of the radius to the ventral projec-
tion on the third metacarpal above the os carpi ulnare.

11. M. ulni metacarpalis dorsalis (UI. met. d., PI. V, Fig. 24),
arises by a short strong tendon from the dorsal distal end of the
ulna at the base of the external condyle. The tendon bends around
the condyle to its posterior border where it swells rapidly into a
thick muscle. The greater part of the muscle is inserted fleshily
on the posterior border of the fourth metacarpal. A small part of
the muscle terminates distally in a broad, flat tendon which fuses
with those that go to the quills.

12. M. extensor metacarpi ulnaris {radialis ?) (£. met. ul. r.,
PI. V, Figs. 19, 20, 22, 24). The origin of this muscle is the most
proximal of all that come from the distal end of the humerus. It
arises by two heads, one tendinous, the other fleshy, from the ante-
rior surface of the humerus superior to the external condyle and
above the upper insertion of the m. propatagialis brevis, the tendin-
ous head being somewhat dorsal. At about 7 mm. from its origin
the tendon passes over into a fusiform muscle. At the same point
is given off a tendinous, sheath which fuses with the tendon of
the propatagialis brevis, above the first bifurcation of that ten-
don. This muscle is smaller and lies dorsal to the one of fleshy
origin. About the mid-point of the radius the two muscles unite
to form a strong, flat tendon which passes over the end of the radius,
across the os radiale, and is inserted on the apex of the os magnum.

13. M. extensor metacarpi ulnaris (E. met. ?(/., PI. V, Fig. 22).
This muscle springs from the external condyle of the humerus close
beside the m. extensor digitorum communis. At its origin it is held
in place by a delicate ligament. It finds attachment on the pos-
terior surface of the third metacarpal about one third the distance
from its proximal end.

14. M. flexor digitorum sublimis (F. dig. s., PI. V, Fig. 24).
This is the central superficial muscle of the inner forearm. It
arises by a strong, flat tendon from the internal condyle of the
humerus. The muscle bundle runs parallel to the ulna, and on the
inner side of the flexor carpi ulnaris for about two thirds the length
of the ulna, and there separates into two tendons. The posterior
tendon passes over the os ulnare, bends under the tendon of the m.

i9°5-l

ANATOMY OF PHAL^NOPTILUS, RIDGWAY. 229

extensor digitorum profundus to the antero-ventral surface of the
third metacarpal, runs along the anterior rim of the first phalanx of
third digit, and is inserted on the proximal end of the second pha-
lanx about one third its length from the proximal end. The ante-
rior tendon continues to the wrist where it merges into a tendin-
ous band which extends from the ventral edge of the styloid proc-
ess of the radius to the anterior border of the os ulnare. From
this latter point come off two other tendons, the upper and more
delicate being inserted at about the mid-point on the ventral bor-
der of the third metacarpal. The thin, flat, posterior tendon runs
along the ventral surface of the fourth metacarpal and is attached
near its distal end. The insertion is quite different from the de-
scription of Gadow (/. c. ).

15. M. flexor digitorum profundus {F. dig. p., PL V, Fig. 24).
This muscle arises fleshily from the proximal half of the ventral sur-
face of the ulna. Proximally it is divided into two almost equal
portions by the brachialis inferior which inserts on the ulna between
them. The surface of origin gradually diminishes and ceases alto-
gether when the broad expansion of the ulni metacarpalis ventralis
is reached. At the wrist the tendon runs under the tendinous
band of the m. flexor digitorum sublimis, passes above the ventral
projection on the proximal end of the third metacarpal, and is here
held in place by a ligament extending from this projection to the
distal ventral edge of the radius. It is inserted on the antero-
ventral rim of the proximal end of the second phalanx of the third
digit.

16. M. extensor digitorum communis {Ex. dig. c., PL V, Fig.
22). This fusiform muscle arises by a short tendon from the ex-
ternal condyle of the humerus between the tendons of origin of the
m. extensor metacarpi ulnaris and m. ectepicondylo radialis. The
muscle becomes tendinous at about two thirds the length of the
radius. Soon after passing the ulna the tendon bifurcates, sending
a delicate slip to the pollex digit, inserting about one third the
length of that bone from its proximal end. The long fork is twice
crossed by the tendon of the m. extensor indicis longus and is finally
inserted on the proximal rim of the first phalanx of the third digit.

17. M. extensor pollicis longus (E.pl. /., PL V, Figs. 19, 22).
Covered by the m. extensor indicis longus, the muscle comes from the
facing surfaces of ulna and radius, from the proximal third of the

230 MARSHALL— A STUDY OF THE

[June 19,

ulna and about the middle third of the radius. At its proximal
extremity it is crossed by a ligamentous band passing from ulna to
radius. It is also held close to the radius by fascia. The tendon
accompanies the m. extensor metacarpi ulnaris (radialis?) to the
apex of the os magnum and is there attached below that muscle.

18. M. extensor indicis longus (E. ind. /. , PI. V, Figs. 22, 24).
This muscle arises by a very short tendon from the internal condyle
of the humerus. It passes directly to the ventral surface of the
radius, and is attached fleshily to the ulna facing surface of that
bone for fully five sixths of its length. The tendon bends under
the radius and becomes dorsal. It crosses the tendon of the m.
extensor digiterum communis, and finds attachment on the base
of the second phalanx of the third digit. It fails to agree with
Gadow's diagnosis.

19. M. interosseus dorsalis {Int. d., PI. V, Fig. 21). Both
interossei spring from the facing surfaces of the third and fourth
metacarpals. In this description the name dorsalis is given to that
muscle which clings to the third metacarpal. At the distal end of
the interosseous space the muscle becomes tendinous and bends
posteriorly, passing along the dorsal surface of the phalanx of the
fourth metacarpal, then to the ventral distal end of the second
phalanx of third digit to become inserted about four fifths the
length of that bone from the proximal end.

20. M. interosseus palmaris {Int. p., PI. V, Fig. 21). This
muscle comes from the anterior surface of the fourth metacarpal,
and terminates tendinously about one half the length of that bone.
The tendon turns dorsally, and is attached to the distal end of the
first phalanx of the third digit on its dorsal surface.

21. M. abductor indicts {Ab. in., PI. V, Fig. 20). This muscle
springs fleshily from the ventral surface of the proximal two thirds
of the third metacarpal, its proximal end being at the base of the
ventral projection of that metacarpal. The round, strong tendon
is inserted on the proximal anterior rim of the second phalanx of
digit three.

22. M. flexor pollicis (Ft. pi, PI. V, Fig. 20). This short
muscle comes from the proximal ventral surface of the third meta-
carpal, lying between the abductor pollicis and the ventral pro-
jection of this metacarpal. It terminates on the posterior proximal
projection of the pollex digit.

igos.] ANATOMY OF PHAL/ENOPTILUS, RIDGWAY. 231

23. M. abductor pollicis (Ab. />/., PL V, Figs. 20, 24). This
rather round muscle arises tendinously from the lower surface of
the tendon of the m. extensor metacarpi ulnaris (radialis?) some-
what proximal to its point of insertion. The muscle then twists
around the base of the pollex digit to its ventral surface, and ter-
minates tendinously about its mid-point.

24. M. extensor pollicis brevis is not present.

25. M. adductor pollicis {Ad. pi, PI. V, Figs. 20, 24). This
fairly well developed muscle lies between the posterior surface of
the pollex digit and the anterior surface of the third metacarpal.
It arises by a strong, fleshy base from the proximal eighth of the
third metacarpal, thence it goes obliquely to the pollex digit and
is attached by a delicate tendon about one third the length of the
digit from its distal end.

26. M. flexor digiti III (F. dig. Ill, PI. V, Figs. 20, 24). This
slender muscle has its origin on the posterior proximal third of the
fourth metacarpal. At its fleshy base is a broad ligament extending
from the anterior rim of the os ulnare to this point. Near the
distal end of this metacarpal the muscle becomes tendinous and
finds attachment about the mid-point of the first phalanx of fourth
digit.

Below are given some muscles found on this bird and not men-
tioned by Gadow (/. c. ).

A. {A., PI. V, Fig. 21.) This is a very delicate muscle extend-
ing along the dorsal surface of the third metacarpal, and at its origin
is covered by tendons of other muscles, fascia and surrounding mem-
branes. It arises by a delicate tendon from the distal dorsal edge
of the radius. The round fusiform carneous portion is covered by
the tendons of the extensor digitorum communis and extensor in-
dicis longus. Its distal hair-like tendon fuses with the m. extensor
indicis longus at a point opposite the middle of the third metacarpal.

B. (B., PI. V, Fig. 24.) This slender muscle extends from the
distal end of the first phalanx of the third digit to the distal end
on the dorsal side of last phalanx of that digit.

C. From the dorsal distal end of the ulna a tendon passes to the
quills. It is not shown in the figures.

D. (B>. , PL V, Fig. 21.) This is a flat muscle which has its car-
neous origin on the proximal dorsal surface of the third metacarpal.
It lies between the proximal projection of that bone and the

232 MARSHALL— A STUDY OF THE [June 19,

pollex digit and os magnum. It is inserted tendinously on the
proximal ridge of the pollex digit.

E. This is a short, stout muscle arising from the ventral and
dorsal end of the coracoid. It passes directly over to the head of
the humerus where it is inserted, one point of the insertion ex-
tending down to the anterior border of the humeral crest. The
long tendon of the biceps passes over this muscle, which does not
appear on the plates.

2. Posterior extremity.

Here are described only those muscles that insert upon and arise
from the femur. The hind limb is so weak in this species and its
other muscles so delicate, that it did not seem worth the time to
work out its whole musculature. They are described in the order
of their occurrence, beginning with the superficial.

i. M. ilio-tibialis intemus or Sartorius {II. tib. int., PI. VI,
Figs. 31, 34, 35). This is the most anterior muscle of the thigh,
of those extending from pelvis to femur. It comes fleshily from the
dorso-lateral border of the ilium and covers the posterior origin of
the posterior portion of the latissimus dorsi, and the anterior edge
of the ilio-trochanterici. It runs free from the muscles of the
pelvis behind it to the femur, gradually diminishing in size and ter-
minating in a flat tendon on the inner surface of the knee joint
where it is covered by a lower leg muscle.

2. M. ilio-trochanterici {It. troch., PI. VI, Figs. 31, 35).
This large, somewhat pyramidal-shaped muscle arises fleshily from
the region of the acetabulum and that portion of the preacetabular
ilium not occupied by the sartorius, the fibers extending even to
its ventral border. These converge and insert by a thin tendon on
the trochanter where it is covered by the m. ilio-tibialis. It has
not the divisions given by Gadow (/. c. ), but is a compact muscle.

3. M. ilio-tibialis (PI. VI, Figs. 31, 34, 35). This thin, broad
muscle is the most superficial one of the thigh. It springs semi-
tendinously from the acetabular and post-acetabular ilium. It
consists of an anterior and posterior portion which are readily dis-
tinguished. The anterior portion (//. tib. ant.) extends about two
thirds the length of the femur, then merges with the underlying
muscle. The posterior portion (//. tib. post.) diminishes in width
distally and inserts aponeurotically upon the muscles covering the
outer surface of the knee joint.

I9oS] ANATOMY OF PHAL/ENOPTILUS, RIDGWAY. 233

4. M. caud-ilio-flexorius (Caud. il. fix., PI. VI, Figs. 31, 32,
33> 34> 35)- Behind the last mentioned muscle this superficial
one is found. It is a small band-shaped muscle, coming from the
posterior border of the ischium. It is partially covered on its
anterior margin by the m. ilio-tibialis and m. ilio-fibularis. The ter-
mination is very peculiar. Coming from the under, distal surface
of the femur is a short, broad muscle, which fuses with the large
muscle mass, the line of fusion being almost at right angles to the
fibers of that portion. From its tibial side comes off a short mus-
cle bundle with fibers directed downward and the tendon of which
fuses with that of one of the leg muscles. This shows great devia-
tion from Gadow's (/. c.) description.

5. M. ischio-flexorius (Isc. fix., PI. VI, Figs. 31, 33, 34,
35). This narrow muscle band comes from the distal border of
the ischium at its union with the pubis. It is covered anteriorly
by the last mentioned layer. Its thin, flat tendon finds insertion
on the anterior borders of the tibial neck.

6. M. ilio-fibularis {II. fib., PI. VI, Figs. 31, 34, 35). This
layer becomes visible after the removal of the m. tibialis anterior
and posterior. It springs from the acetabular ilium. It ends in a
small, round tendon, which, passing through a tendonous loop at
the knee, continues down the leg to become inserted between fibula
and tibia at the point where the former becomes free from the
latter.

7. M. femori-tibialis (Fm. tib., PI. VI, Figs. 33, 34, 35).
This is the largest of the thigh muscles. It is partially covered on
the ventral anterior border by the m. ilio-tibialis internus, dorsally
by the m. ilio-tibialis anterior. Its origin begins at the trochanter
and it is attached fieshily to the femur on both dorsal and ventral
surfaces. It finds a tendinous insertion at the knee joint, being at-
tached to the proximal border of the tibia. The separation into
parts as given by Gadow (/. c. ) can not be made out.

8. M. caud. -ilio-femoralis (Caud. il. fm., PI. VI, Fig. 35).
This is revealed by the removal of the m. ilio-fibularis and m. caud-
ilio-flexorius. Its width where it passes under the m. caud-ilio-
flexorius is equal to that of the above muscle. It comes as a small
rounded tendon from the ventral lateral border of the pygostyle.
Just before reaching the ischium the tendon passes over into the fleshy
muscle. This bends around in a semicircular fashion to the proximal

234 MARSHALL— A STUDY OF THE

[June 19,

third of the femur, and here finds a fleshy insertion on the linea as-
pera, occupying its posterior surface. Pars iliaca is absent and pars
caudalis does not agree in origin with Gadow's (/. e.) description.

9. M. ischio-femoralis (Ise. fin., PL VI, Fig. 35). This
muscle is proximal to the above. It springs from joining surfaces
of ischium and ilium and from neighboring surface of ischium down
to the origin of the m. pub.-ischo-femoralis. This short, thick, flat
muscle there crosses the femur and is inserted by a small, thin
tendon at the base of the trochanter.

10. M. piib-ischio-femoralis {PI), ise.fm., PI. VI, Figs. 33, 35).
This is one of the largest muscles of the thigh. It arises from the
proximal half of the pubis and ischium along their line of union.
It passes somewhat obliquely over to the distal half of the femur
and is there inserted fleshily by its anterior border. Its fibers are
intimately associated with those of the m. caud-ilio-flexorius. It
consists of only one portion, a thick, flat layer.

11. M. obturator (Obt., PL VI, Fig. 35). This is the deepest
lying of the muscles of the outer surface. It springs fleshily from
the edges of the foramen obturatum. Thence it passes to the poste-
rior border of femur, and there is attached semitendinously. The
muscle varies considerably from that of Gadow's (/. e.) of the
same name. It agrees in some points with his mm. accessorii m.
obturatoris.

12. M. ilio-femoralis internus {II. fin. int., PL VI, Fig. 33).
This somewhat triangular muscle comes from the ventral surface,
near its lateral border, of the preacetabular ilium extending almost
to the acetabulum. It passes to the ventral surface of the femur
just distal to the head, and is there attached. The muscle is fleshy
both at origin and insertion.

13. X. (PL VI., Fig. 33). This is a long slender muscle begin-
ning distal to the insertion of the m. ilio-femoralis internus, and is
attached fleshily to the ventral surface of the femur for its remain-
ing length. It terminates distally in a thin, flat tendon which is
inserted on the dorso-ventral border of the proximal end of the
tibia. Gadow (/. c. ) did not describe such a muscle.

The following muscles were not found : m. ilio-femoralis exter-
tuis, m. amine us, mm. aeeesorii m. obturatoris.

19o5.] ANATOMY OF PHAL.ENOPTILUS, RIDGWAY.

235

VII. Comparisons.
Certain characters of the better known genera of the Caprimul
gidas are compared in the following table :

•8

rt

■6

a

■a.

CO

■a

3

3

3

g

a

^

H u

I

0

a

u

U

5

s

m
o

3

3

CO

^1

— t

t

o

Caprimulgus.

2

X

X

X

X

Over
biceps.

Axy

One
notch.

Traches-
bronchial.

X

X

-

Nyctidromus.

2

X

X

X

X

"

Axy

"

"

X

X

—

Chordeiles.

2

X

X

X

—

"

Axy

"

<'

X

X

Antrostomus.

2

X

X

X

X

"

Axy

"

"

X

X

Phalsenoptilus.

2

X

X

—

X

"

(A)y

"

"

X

X

—

The sign "x" denotes occurrence, and " — " absence of a
character. The formulae for the thigh muscles are those given by
Garrod (1S74), slightly modified by Gadow (1891), and denote
the presence of the following muscles :

Pars caudalis m. caud.-il. femoris , = A

Pars iliaca m. caud-il. femoris =B

M. caud-il. flex, inserting only on the tibia = X

M. caud-il. flex, with the " accessorius " inserting on the

femur — Y

All the points of comparison of the first four genera in the above
table were taken from Beddard (1898). It will be noticed that
the amount of difference in these forms is slight. The only char-
acters which differ are the gall bladder, absent in Chordeiles, biceps
slip, absent in Phaloznoptilus, and the difference in the last genus
of the muscle formula for the thigh.

So far as the tabulated characters are concerned, Phalcenoptilus
appears less closely related to Chordeiles than to the other genera.
In two of its muscle characters it differs from all the other genera.

VIII. Aberrant Characters.

In closing it will be well to call attention to the striking varia-
tions from the muscles of the birds studied and described by
Gadow (/. r.).

The following wing muscles were not found: Biceps slip, m.
extensor policis brevis, m. entepicondylo-ulnaris and m. deltoi-
deus minor. Some not mentioned by him were present in the

236 MARSHALL— A STUDY OF THE [June 19,

bird, and in the descriptions and drawings are denoted by the let-
ers A, B, C, D, and E. The complex arrangement of the m.
propatagialis brevis should also be mentioned.

Of the thigh muscles these were missing: m. ilio-femoralis ex-
ternus, m. ambiens, and mm. accessorii m. obturatoris. A muscle
herein denoted by the letter X was not given by Gadow. Pars
caudalis m. caud-ilio-femoralis differs from Gadow's description
and this difference is indicated in the table by placing the letter
representing it in parenthesis. M. caud-ilio-flexorius showed con-
siderable variation.

BIBLIOGRAPHY OF THE ANATOMY OF THE CAPRIMULGI.
Beddard, F. E.

1886. On the Syrinx and Other Points in the Anatomy of the Caprimulgidse.

Proc. Zool. Soc. London.
i8g8. The Structure and Classification of Birds. London.
Blanford, W. T.

1877. Letter on Capriraulgus unwini and some Batrachostomi. Ibis.
Clark, H. L.

1894. The Pterylography of Certain American Goatsuckers and Owls. Proc.

U. S. Nat. Mus.
1 901. The Pterylosis of Podargus : with notes on the Pterylography of Capri-
mulgi. Auk.
Coues, £.
1888. Notes on the Nomenclature of the Muscles of Volation in Birds' Wings.

Auk.
1903. Key to North American Birds. Fifth edition. Boston.
Cuvier, G.
1795. Memoir sur le larynx inferieur des Oiseaux. Millin, Magasin encyclo-

paed. I. Paris.
Fiirbringer, M.

1888. Untersuchungen zur Morphologie und Systematik der Yogel. Amster-
dam.

1891. Anatomie der Vogel. Mem. II. Orn. Congress.
Gadow, H.

1889. On the Taxonomic Value of the Intestinal Convolutions in Birds. Proc.
Zool. Soc. London.

1891. Vogel. Bronn's Klassen und Ordnungen des Thier-Reichs. Leipzig.

Garrod, A. H.

1872. On the Mechanism of the Gizzard in Birds. Proc. Zool. Soc. London.

1873a. On Certain Muscles in the Thigh of Birds, and on their Value in Classifi-
cation. Ibid.

1873b. On the Value in Classification of a Peculiarity in the Anterior Margin of
the Nasal Bones of Certain Birds. Ibid.

I9o5.] ANATOMY OF PHAL^NOPTILUS, RIDGWAY. 237

1873c. On the Carotid Arteries of Birds. Ibid.

1874. On Certain Muscles in Birds and their Value in Classification. Part II.
Ibid.

Goodchild, J G.

1886. Observations on the Disposition of the Cubital Coverts in Birds. Ibid.

Hartert, E.

1892. Notes on the Caprimulgida;. Ibis.

1896. Notes on some Species of the Families Cypselidae, Caprimulgidte, and

Podargidre, with Remarks on Subspecific Forms and their Nomenclature.

Ibis.

Huxley, T. H.

1867. On the Classification of Birds ; and on the Taxonomic Value of the Mod-
ifications of Certain of the Cranial Bones Observable in that Class. Proc.
Zool. Soc. London.

Kessler, K. T.

1841. Osteologie der Vogelfiisse. Bull. Soc. Imp. Natur. Moscou.

Mitchell, P. C.

On the Intestinal Tract of Birds ; with Remarks on the Valuation and
Nomenclature of Zoological Characters. Tr. Linn. Soc. London. Zool.
VIII.

Morse, E.

1872. On the Tarsus and Carpus of Birds. Am. Lye. N. Y.

Miiller, J.

1841. Uber die Anatomie des Steatornis caripensis. Monatsber. d. k. Akad.
der wiss. Berlin.

1842. Anatomische Bemerkungen uber den Quacharo (Steatorins caripensis).
Miiller' s Arch. f. Anat. and Phys.

Nitzsch, C. L.

1867. Pterylography, translated (by W. S. Dallas) from the German. Roy. Soc.
London.

Nitzsch-Burmeister.

1840. System der Pterylographie. Halle.

Nitzsch-Giebel.

1858. Dei Zunge der Vogel und ihr Geriist. Zeitsch. f.d. ges. naturw. XL Berlin.

Parker, W. K.

1868. A Monograph on the Structure and Development of the Shoulder Girdle
and Sternum in the Vertebrata. Roy. Soc. London.

1876a. Memoir on ^Egithognathous Birds. Proc. Zool. Soc. London.

1876b. On the Structure and Development of the Bird's Skull. Trans. Linn.
Soc. Zool. I. London.

1878. On the Skull of the /Egithognathous Birds. Trans. Linn. Soc. Lon-
don. X.

J38

MARSHALL— A STUDY OF THE

[June 19

Sclater, P. L.

1866a. Additional Notes on the Caprimulgid£e. Proc. Zool. Soc. London.

1866b. Notes on the American Caprimulgidae. Ibid.

Shufeldt, R. M.

1885. Contribution to the Comparative Osteology of the Trochilidse, Caprimul-
gidse and Cypselidje. Ibid.

1886. Additional Notes upon the Anatomy of the Trochili, Caprimulgi, and
Cypselidre. Ibid.

1889. Studies of the Macrochires, Morphological and Otherwise, with the View
of Indicating Their Relationships, and Defining Their Several Positions in
the System. Journ. Linn. Soc. Zool. London.
Zoological Laboratory,

University of Texas, 1905.

Description of the Plates.
The following abbreviations have been employed :

Ab. in., Musculus abductor indicis.

Ab. pi., M. abductor pollicis.

Ad.pl., M. adductor pollicis.

A., Anus.

Bas. h., Basihyal.

Bas. b., Basibranchial.

Bi., Biceps brachii, pars propatagi-

alis.
Bi. T., Tendon of biceps brachii.
Bi. inf., Musculus brachialis inferior.
Bi\ «., Brachial nerve.
Bro., Bronchus.
C, Coracoid.
Ccb., Geca.
Caud. il.jlx., Musculus caud.-ilio-flex-

orius.
Caud. il. fm., M. caud.-ilio femoralis.
Cb., Cerebellum.
Cer., Cerebrum.
Cer. b., Cerato-brachial.
CI., Clavicle.
Clo., Cloaca.
Cor. br. p., Musculus coraco-brachia-

lis posterior.
Del. a., M. deltoideus major anterior.
Del. p., M. deltoideus major posterior.
Duo., Duodenum.
Ect. ?-., Musculus ectepicondylo-radi-

alis.
Ect. u., M. ectepicondylo-ulnaris.

E. dig. c., M. extensor digitorum com-
munis.

E. dig. c. t., Tendon of m. extensor
digitorum communis.

E. ind. I., Musculus extensor indicis
longus.

E. ind. 1. L, Tendon of M. extensor
indicis longus.

E. viet. ul., Musculus extensor meta-
carpi ulnaris.

E. met. ul. ?:, M. extensor metacarpi
ulnaris (radialis? ).

E. met. ul. r. t., Tendon of m. ex-
tensor metacarpi ulnaris (radialis? ).

Ent. g., Os entglossum.

E. pi. I., Musculus extensor pollicis
longus.

Ep. b., Epi-branchial.

F. carp, ul., Musculus flexor carpi
ulnaris.

F. dig. p., M. flexor digitorum pro-
fundus.

F. dig. s., M. flexor digitorum sublimis.
/•'. dig. III., M. digiti III.

Fl. pi., M. flexor pollicis.

Flo., Flocculus.

Fm. tib., M. femori-tibialis.

G. bl., Gall bladder.
Giz., Gizzard.

lit., Heart.

ANATOMY OF PHAL/ENOPTILUS. RIDGWAY.

>M9

Hit., Humerus.

II. Jib., Musculus ilio-fibularis.

II. fm. int., M. ilio-femoralis internus.

77. tib. ant., M. ilio-tibialis anterior.

//. tib. post., M. ilio-tibialis posterior.

//. tib. int., M. ilio-tibialis internus.

//. troch., M. ilio-trochanterici.

Inf., Infundibulum.

Int., Intestine.

Int. d., Musculus interosseus dorsalis.

Int. p., M. interosseus palmaris.

Isc. fix., M. ischio-flexorius.

Isc fin., M. ischio-femoralis.

K., Kidney.

Lar., Larynx.

Lat. d. a., Musculus latissimus dorsi

anterior.
Lat.d.p., M. latissimus dorsi posterior.
Liv., Liver.
Liv. d., Liver duct.
L. ovi., Left ovary.
Lu., Lungs.

Med., Medulla oblongata.
Nos., Nostril.
Obt., Musculus obturator.
CEs. , Oesophagus.
Olf. n., Olfactory nerve.
Op. I. , Optic lobes.
Ov. , Ovary.
Pars. hu. cub., Musculus triceps pars

humero-cubitalis.
Pars. sc. cub., M. triceps pars scapuli-

cubitalis.
Pan., Pancreas.
Pan. d., Pancreatic ducts.
Pb., Pubis.
P. pat. b., Musculus propatagialis

brevis.
P. pat. I., M. propatagialis longus.
Pb. isc. fm., M. pub. -ischio-femoralis.
Pect., M. pectoralis, pars thoracica.
Pha., Pharynx.

Pron. s., Musculus pronator sublimis.

Pro., M, propatagialis, pars propata-
gialis musculi deltoidei.

Prov., proven triculus.

R., Os radiale.

Pa., Radius.

Rh. s., Musculus rhomboideus super-
ficialis.

Rh. p., M. rhomboideus profundus.

R. ovi., Right oviduct.

Sc, Scapula.

Sc. hum. p., Musculus scapuli-humer-
alis posterior.

Ser. p., M. serratus profundus.

Ser. s. a., M. serratus superficialis an-
terior.

Ser. s. p., M. serratus superficialis pos-
terior.

Spl., Spleen.

Sp., Spinal cord.

S. sc, Musculus subscapularis.

St. , Sternum.

St. co., Musculus sterno-coracoideus.

St. tr., M. sterno-trachealis.

Sub. co., Musculus subcoracoideus.

Sup. cor., M. supracoracoideus.

T., Tongue.

Tr. lat. , Musculus Tracheo-lateralis.

Tra., Trachea.

Tym. in., Membrana tympaniformis
interna.

Tym. ex., Membrana tympaniformis
externa.

U., Os ulnare.

Ul., Ulna.

Ul. met. d., Musculus ulni metacarpalis
dorsalis.

UL met. v., M. ulni metacarpalis ven-
tralis.

Urh., Urohyal.

Or., Ureter.

240 MARSHALL— A STUDY OF THE

[June 19,

EXPLANATION OF PLATES.

All the figures are from enlarged freehand sketches, and are mostly drawn
to the same scale; they have been reduced almost one half in the reproduction.

PLATE IV.
Fig. I. Tongue bone.

Fig. 2. Dorsal view of brain with outline of head and nostrils.
Fig. 3. Brain viewed from the right side.
Fig. 4. Lateral view of pecten of the eye.
Fig. 5. Pecten seen from its free apex.
Fig. 6. Dorsal view of duodenal loop and pancreas.
Fig. 7. Ventral view of syrinx.
Fig. 8. Dorsal view of syrinx.
Fig. 9. Viscera seen from the right side.
Fig. 10. Ventral view of head, trachea and viscera.
Fig. 11. Female urogenital organs, ventral view.
Fig. 12. Ventral view of posterior portion of alimentary tract.
Fig. 13. Dorsal view of oil gland.
Fig. 14. Lateral view of the same.

Fig. 15. Lateral view of viscera showing intestinal loops. Dotted lines repre-
sent the portion of the intestine covered by superficial folds.

PLATE V.

Fig. 16. Muscles of upper wing.

Fig. 16a. Dorsal view of spinal cord and brachial nerve plexus.
Fig. 17. Superficial muscles of back and upper wing.
Fig. 18. Dorsal superficial muscles of hand.
Fig. 19. Deeper muscles of forearm.
Fig. 20. Muscles of ventral surface of hand.
Fig. 21. Muscles of dorsal surface of hand.
Fig. 22. Superficial muscles of back and outer arm.

Fig. 23. M. propatagialis brevis removed to show more clearly its complex
arrangement.

Fig. 24. Superficial muscles of breast and inner arm.

Fig. 25. Deeper chest muscles.

Fig. 26. Muscles of shoulder and chest.

Figs. 27, 28 and 29. Upper arm muscles.

Fig. 30. Origin of m. coracobrachial posterior.

Fig. 31. Superficial muscles of thigh.

Fig. 32. M. caud.-ilio flexorius removed to show complex arrangement.

Fig. 33. Muscles of thigh viewed from median surface.

Fig. 34. Deeper thigh muscles.

Fig- 35- Deepest muscles of thigh.

Stated Meeting, April 28, 1905.
President Smith in the Chair.

Letters accepting membership were read from Prof. Joseph
S. Ames, President David Starr Jordan, Prof. G. L. Kittridge,
Dr. Robert G. LeConte, Mr. George T. Moore, President
Francis P. Venable and Mr. J. Edward Whitfield.

Dr. J. W. Harshberger read a paper on " Evolution and
Distribution of North American Plants."

Stated Meeting, May 5, 1905.
President Smith in the Chair.

Dr. Robert G. LeConte and Mr. J. Edward Whitfield,
newly elected members, were presented to the Chair and took
their seats in the Society.

Letters accepting membership were read from Mr. R. A. F.
Penrose, Jr., Prof. W. G. Farlow and Prof. Eliakim Hastings
Moore.

Mr. Sydney George Fisher read a paper on " The Military
Strategy of the American Revolution."

Mr. Joseph Willcox exhibited a fossil specimen of bone
from the tail of the Glyptodon found in Florida.

Stated Meeting, May 19, 190J.
President Smith in the Chair.

Mr. R. A. F. Penrose, Jr., a newly elected member, was
presented to the chair and took his seat in the Society.

Letters accepting membership were read from Prof. T. C.
241

242 MINUTES. [Oct. 6,

Chamberlin, Prof. Yves Delage, Prof. W. M. Flinders-Petrie,
Sir W. T. Thiselton-Dyer, and Prof. Otto Nordenskjold.

Dr. Henry Skinner read a paper on " Insects in Relation to
Disease."

Dr. Edgar F. Smith read a paper entitled " Observations
on Columbium."

Stated Meeting, October 6, ipoj.

President Smith in the Chair.

Mr. Henry Carey Baird presented his resignation of mem-
bership which was accepted.

The decease was announced of the following members :
Hon. John Hay, at Newbury, N. H., on July I, 1905,

aet. 66.
Prof. William C. Day, at Swarthmore, Pa., on August 4,

1905, aet. 48.
Prof. Franz Reuleaux, at Berlin, on August 20, 1905,

aet. 76.
Prof. Dr. Jules Oppert, at Paris, on August 21, 1905,

aet. 80.
Gen. Isaac J. Wistar, at Claymont, Del., on September

18, 1905, aet. 78.
Mr. Ellis Yarnall, at Philadelphia, on September 19,
1905, aet. 8y.
The following papers were read :

" The Problems of Human Anatomy," by Dr. George A.
Piersol.

" New Species of Drosera from the Gulf States," by Dr.
John Macfarlane.

" A Study of the Anatomy of Phalaenoptilus, Ridgway,"
by Margaret E. Marshall, communicated by Prof. Thos. H.
Montgomery. (See page 213.)

igosl MINUTES. 243

Stated Meeting, October 20, 1903.
President Smith in the Chair.

The following papers were read :

" Eclipse Problems," by Prof. C. L. Doolittle.

"Some of the Vertebrates of the Florida Keys," by Henry
W. Fowler.

Mr. Samuel Dickson was elected a Councillor to fill the
unexpired term of Gen. Isaac J. Wistar, deceased, and Prof.
Henry F. Osborn was elected a Councillor to fill the unex-
pired term of Mr. Henry Carey Baird, resigned.

Stated Meeting, November j, 1903.
President Smith in the Chair.

A letter was presented from Mr. Bailey Willis, accepting
membership.

Dr. Franz Boas read a paper on " Party Allegiance from the
Anthropological Point of View."

Stated Meeting, November ij, 1903.
President Smith in the Chair.
The decease was announced of Dr. George R. Morehouse,
at Philadelphia, on November 12, 1905, aet. j6.

Dr. Charles Conrad Abbott read a paper on " The Antiquit,
of Man in the Delaware Valley."

Stated Meeting, December 1, 1903.

President Smith in the Chair.

Prof. John M. Macfarlane read a paper on the " Occurrence
Distribution and Hybridization of the American Pitcher Plants,
or Sarracenias."

244 MINUTES. [Dec. is,

Stated Meeting, December ij, ipoj.
President Smith in the Chair.

A letter was read from the Committee on Organization of
the 6th Congres Internationale d' Anthropologic Criminelle, an-
nouncing that the Congress would convene at Turin on April
28, 1906, and inviting the Society to be represented thereat.

The President delivered his Annual Address which included
"The Story of the Isolation of the Metal Calcium."

INDEX.

Abbott, Alexander C, Epidemic cere-
brospinal meningitis, 35

Adler, Cyrus, The present status of the
international catalogue of scientific
literature, 36

Election of officers, 3
Electro-analysis, The use of the rotating

anode and mercury cathode in, 37,

137 ,
Evolution, Development of, 4

Babylonian creation story, 36

Brains of Scymnus, Mitsukurina and
Chlamydoselachus, etc., 39

Bridge truss, Relation between the eco-
nomic depth of a, and the depth that
gives greatest stiffness, 39, 164

Brown, Amos P., The Rocky Moun-
tains, 5

C.

Cambarus, The mutual affinities of the
species of the genus, 37, 91

Catalogue of scientific literature, inter-
national, 36

Cerebro-spinal meningitis, 35

Chittenden, Russell, H., Reason and
intelligence vs. custom and habit in
the nutrition of the body, 37

Color, Normal perception of, 36, 40

Columbium, Some observations on, 177

Columbium and tantalum, Observa-
tions on, 37, 151

Conklin, E. G., Development of evolu-
tion, 4

Mosaic development in ascidian

eggs, 36

Copper foil, The oligodynamic action
of, on some intestinal organisms, 36,
51

D.

Dallas, W. L., Pressure and rainfall
conditions of the trades-monsoon
area, 38, 159

Doolittle, C. L. , Evidence relating to
latitude variation of short-periods, 38

Doolittle, Eric, The secular perturba-
tions of the earth, 37

Douglas, James, Shower of toads, 35

E.

Eggs, Mosaic development in ascidian,
36

Farr, M. S., The mammalian fauna of

the Fort Union beds, 37
Fauna, The mammalian, of the Fort

Union Beds, 37
Fauna, The marsupial, of the Santa

Cruz Beds, 37, 73
Filipino, His customs and characters,

Foods, The effects upon metabolism of

preservatives added to, 37
Four bodies, The problem of, 38

Gas molecules, A possible case of scat-
tering of the ultra-violet light by, 39

H.

Haupt, Lewis M., The emancipation

of the waterways, 36, 42
Haupt, Paul, Biblical pessimism, 4
Heilprin, Angelo, A review of LaCroix's

work on the Montagne Pelee, 37
Hilprecht, Hermann V., Recent re-
searches in the Temple Library at

Nippur, 5
Hobbs, John E., The beginnings o

lumbering as an industry in the New

World, 36
Hall, Roy D., and Edgar F. Smith,

Some observations on columbium,

177

van Ingen, Gilbert, The rounded sands
of Palaeozoic formations, 37

Isobanc charts, construction of, for
upper levels, etc., 38

J-

Jastrow, Morris, Composite character
of the Babylonian creation story, 36

24G

INDEX.

K.

Keasbey, Lindley M., The Weal re-
lation, 36

Kollock, Lily G., and Edgar F. Smith,
The use of the rotating anode and
mercury cathode in electro-analysis,

37. 137
Kraemer, Henry, The oligodynamic
action of copper foil on some intes-
tinal organisms, 36, 51

Lambert, P. A., the straight line con-
cept, 38, 82

Latitude variations, Evidence relating
to, of short periods, 38

Lovett, Edgar Odell, On the problem
of four bodies, 38

McClellan, William, The use of the
falling plate oscillograph as a phase
meter, 39, 166

Macfarlane, John M., New species of
genus Nepenthes, 36

The structure of the lignified cell

wall, 36

Marshall, Margaret E., A study of the
anatomy of Phalaenoptilus, Ridgway,
213

Masque, The English, 36

Mathews, R. H., Sociology of the ab-
origines of western Australia, 5, 32

Matthews, W. D. , Notes on the genus
Sinopa, 37, 69

Meeting, General, 35

Stated, 3, 4, 5, 35

Merriman, Mansfield, Relation between
the economic depth of a bridge truss
and the depth which gives greatest
stiffness, 39, 164

Metabolism, Effects upon, of preserva-
tives, 37, 66

Metzger, J. A., The Filipino, his cus-
toms and character, 5, 6

Members deceased :

Bell, Sir Lowthian, 4
Campbell, John Lyle, 5
Carter, Hon. James C. , 5
Dannefeld, C. Juhlin, 4
Frazier, Benjamin W., 3
Hayes, Richard Somers, 5
Lippincott, James Dundas, 5
Packard, Alpheus Spring, 4
Prescott, Albert Benjamin, 5
Randall, F. A., 35

de Saussure, Henri Louis Fred-
eric, 35
Sellers, William, 4
Weil, Edward IL, 3

Members elected :

Ames, Joseph S., 38
Chamberlin, Thomas Chrowder, 38
Delage, Yves, 39
Farlow, William Gilson, 38
Flinders-Petrie, William Matthew,

39
Frazier Charles H., 38
Jordan, David Starr, 38
Kittredge, George Lyman, 38
LeConte, Robert G., 38
Moore, Eliakim Hastings, 38
Moore, George T., 38
Nordenskjold, Otto, 38
Penrose, Richard A. F. , Jr., 38
Sievers, Edward, 39
Thiselton-Dyer, Sir William, 39
Venable, Francis P., 39
Whitfield, J. Edward, 39
Willis, Bailey, 39

Membership, resignation of :
De Garmo, Charles C, 3

N.

Naturwissenschaftliche V ere in fur
Schleswig-Holstein, 50th Anniver-
sary of, 35
Nepenthes, New species of genus, 36
Nippur, Temple Library at, 5
Nutrition of the body, reason and in-
telligence vs. custom and habit in, 37

0.

Oliver, Charles A., A plea for govern-
mental supervision of posts necessi-
tating normal perception of color,

36, 40

Ortmann, A. E. , The mutual affinities
of the species of the genus Cambarus,

37, 91

Oscillograph, The use of the falling
plate, as a phase meter, 39, 166

P.

Palaeozoic formations, The rounded
sands of, 37

Pelee, A review of LaCroix's work on
Montagne, 37

Pendulum, Theory of the double sus-
pension, 39

247

Perturbations of the earth, the secular,

37.
Pessimism, Biblical, 4
Phalaenoptilus, Anatomy of, 213
Pilsbry, Henry A., The faunal relations

of the Ryu-kyu (Loo Choo) Islands,

37

R.

Radio-activity, Universal, 4

in solar phenomena, 38
Rocky Mountains, 5
Ryu-kyu (Loo Choo) Islands, 37

s.

Sampson, Alden, Thought transference
among animals, by touch and scent,
36

Sandstrom, J. \V., Construction of iso-
baric charts for upper levels, etc., 38

Schelling, Felix E., The English
masque, 36

Sinclair, W. J., The marsupial fauna
of the Santa Cruz Beds, 37, 73

Sinopa, Notes on the Genus, 37, 69

Smith, Edgar F., Observations on col-
umbiurn and tantalum, 37, 151

Snyder, M. B., Universal radio-activ-
ity, 4

Radio-activity in solar phenomena,

38
Sociology of the aborigines of western
Australia, 5, 32

Sodium vapor, dispersion, absorption,
fluorescence and magnetic rotation

of. 39 .
Straight line concept, 38, 82

Thought transference among animals

by touch and scent, 36
Toads, A shower of, 35
Trades-monsoon area, pressure and

rainfall conditions of the, 38, 159

W.

Wall, The structure of the lignified
cell, 36

Waterways, Emancipation of, 36, 42

Weal-relation, 36

Wilder, Burt G. , The brains of Scym-
nus, Mitsukurina and Chlamydose-
lachus, etc., 39

Wiley, Harvey W., The effects upon
metabolism of preservatives added to
foods, 37, 66

Wood, Robert Williams, The disper-
sion, absorption, fluorescence and
magnetic rotation of sodium vapor,
39

A possible case of scattering of

the ultra-violet light by gas mole-
cules, 39

Woodward, Robert S., Theory of the
double suspension pendulum, 39

THE LIST

American Philosophical Society

HELD AT PHILADELPHIA

FOR PROMOTING USEFUL KNOWLEDGE

(Founded 1743)

January, 1906.

OFFICERS

PATRON

The Governor of Pennsylvania

PRESIDENT

Edgar F. Smith

VICE-PRESIDENTS
George F. Barker William B. Scott Simon Newcomb

SECRETARIES

I. Minis Hays
Edwin G. Conklin

Arthur^W. Goodspeed
Morris Jastrow, Jr.

Charles L. Doolittle

Elected in 1904
Richard Wood
Samuel G. Dixon
J. G. Rosengarten
Henry F. Osborn

CURATORS

William P. Wilson

TREASURER

Henry La Barre Jayne

COUNCILLORS

Elected in 1905
George F. Edmunds
James T. Mitchell
Joseph Wharton
William W. Keen

Albert H. Smyth

Elected in 1906
Patterson Du Bois
Samuel Dickson
Ernest W. Brown
William Keith Brooks

MEMBERS.

Elected. Name. Present Address.

1871 Abbe, Prof. Cleveland U. S. Weather Bureau, Wash-
ington, D. C.

1S62 Abbot, Gen. Henry L., U.S.A 23 Berkeley St., Cambridge,

Mass.

1897 Abbott, Alexander C, M.D University of Pennsylvania,

Philadelphia.
1889 Abbott, Charles Conrad, M.D. ... Trenton, N. J.

1876 Ackerman, Prof. Richard Stockholm, Sweden.

1886 Adam, Lucien 41 Bard Sevigne, Rennes,

F ranee.
1901 Adams, Charles Francis, LL.D. . .23 Court St., Boston. 4

1900 Adler, Cyrus, Ph.D Smithsonian Institution, i

Washington, D. C.

1875 Agassiz, Prof. Alexander 36 Quincy St., Cambridge,

Mass.

1869 Agassiz, Mrs. Elizabeth Quincy St., Cambridge, Mass.

1878 Allen, Prof. Joel Asaph Am. Museum of Natural His-
tory, New York City.

1881 Ames, Rev. Charles G 12 Chestnut St., Boston, Mass.

1905 Ames, Joseph Sweetman, Ph.D... Johns Hopkins University,

Baltimore.
1886 Anderson, Maj. Geo. L., U.S.A. .. .Ordnance Board, Governor's

Island, New York City.
1889 Angell, Pres't Jas. Burrill, LL.D . Ann Arbor, Mich.
1893 Appleton, Prof. William Hyde. . .Swarthmore, Pa.

1884 Ashhurst, Richard L 319 S. 11th St., Philadelphia.

1884 Avebury, The Right Hon. Lord High Elms, Down, Kent, Eng.

1884 Bache, R. Meade 4400 Sansom St., Phila.

1877 Bache, Thomas Hewson, M.D 233 S. 13th St., Philadelphia.

1898 Baer, George F 1718 Spruce St., Philadelphia.

1896 Bailey, Prof. L. H Cornell University, Ithaca,

N. Y.

1884 Baird, Prof. Henry M 219 Palisade Ave., Yonkers,

N. Y.

1899 Balch, Edwin Swift 1412 Spruce St., Philadelphia.

1901 Balch, Thomas Willing 1412 Spruce St., Philadelphia.

1897 Baldwin, Prof. James Mark, D.Sc.408 Cathedral St., Baltimore.
1891 Ball, Sir Robert Stawell, LL.D. .Observatory, Cambridge, Eng.

1882 deBar, Hon. Edouard, Seve Ramsgate, England.

252 MEMBERS.

E'ected. Name. Present Address.

1873 Bakkeb, Prof. George F., LL.D.. ..3909 Locust St., Philadelphia.

1884 Barker, Wharton 119 S. 4th St., Philadelphia.

1903 Barnard. Edward E., Sc.D Yerkes Observatory, Williams

Bay, Wisconsin.

1903 Barus, Prof. Carl, Ph.D 30 Elm Grove Ave., Providence,

Rhode Island.
1899 Baugh, Daniel 1601 Locust St., Philadelphia.

1902 Becquerel, Prof. Antoine-Henri . (ime Dumont d'Urville, Paris,

France.

1882 Bell, Prof. Alexander Graham. . . 1331 Connecticut Ave., Wash-
ington, D. C.

1895 Bement, Clarence S 3907 Spruce St., Philadelphia.

1897 deBenneville, James Seguin University Club, Philadelphia.

1895 Berthelot, Marcelin Pierre Eu-
gene, D.es-Sc Palais de l'Institut de France,

Rue Mazarin, No., 3, Vie.,
Paris, France.

1895 Bertin, Georges llbis Rue Ballu, Paris,

France.

1880 Biddle, Cadwalader 1420 Walnut St., Phila.

1877 Biddle, Hon. Craig -2033 Pine Street, Phila.

1887 Billings, John S., M.D 40 Lafayette Place, New York.

1895 Bispham, George Tucker 1805 DeLancey Place, Phila.

1889 Blair, Andrew A 406 Locust St., Philadelphia.

1870 Blake, Prof. Wm. Piiipps I, University Place, Tucson,

Arizona.

1904 Bloomfield, Prof. Maurice, LL.D. 861 Park Ave., Baltimore.

1903 Boas, Franz, Ph.D. , 123 W. 82d Street, New York,

N. Y.

1895 Bonaparte, Prince Roland 10 Ave. d'Jena 22, Paris,

France.

1904 Bowditcii, Henry Pickering, M.D. Sunnyside, Jamaica Plains,

Boston.

1840 Bote, Prof. Martin H Coopersburg, Lehigh Co., Pa.

1877 Brackett, Prof. Cyrus Fogg 4 Prospect Ave., Princeton,

N. J.

1886 Branner, Prof. John C Stanford University, Cal.

1902 Brasiiear, John A., Sc.D 1954 Perryville Ave.. Alle-
gheny, Pa.

1886 Brezina, Dr. Aristides XIII6 St. Veitgasse, 15,

Vienna, Austria.

1886 Brinton, John H., M.D 1423 Spruce St., Philadelphia.

1899 Brock, Robert C. H 1612 Walnut St., Phila.

1899 Broegger, Prof. W. C Christiania, Norway.

1886 Brooks, Prof. William Keith Johns Hopkins University,

Baltimore, Maryland.

MEMBERS. 253

Elected. Name. Present Address.

1901 Brown, Prof. Amos P 20 E. Penn St., Germantown,

Philadelphia.

1879 Brown, Arthur Erwin 1208 Locust St., Philadelphia.

1S9S Brown, Prof. Ernest William ... Haverford College, Haverford,

Pa.

1895 Brubaker, Albert P., M.D 105 N. 34th St., Philadelphia.

1865 Brush, Prof. George J Yale Univ., New Haven, Conn.

1898 Bryant, Henry Grier, F.R.G.S 805 Land Title Building, Phil-

adelphia.
1895 Bryce, Right Hon. James 54 Portland Place, London,

W., England.

1895 Budge, E. A. Walllis, Litt.D British Museum, London, Eng.

1881 Butler, Hon. William West Chester, Pa.

1899 Cadwalader, John 1519 Locust St., Philadelphia.

1903 Campbell, Wm. Wallace, LL.D..Lick Observatory, Mt. Hamil-

ton, California.

1885 Cannizzaro, Tomaso Santa Maria fuori cinta, Casa

Roffa, Messina, Sicily.

1873 Capellini, Prof. Giovanni Portovenere pres Spezia, Italy.

1875 Carll, Prof. John Franklin Pleasantville, Venango Co., Pa.

1902 Carnegie, Andrew, LL.D 2 E. 91st St., New York, N. Y.

1880 Carson, Hampton L., LL.D 1033 Spruce St., Philadelphia.

1872 Cassatt, Alexander Johnson .... Haverford, Delaware Co., Pa.

1887 Castner, Samuel, Jr 3729 Chestnut St., Phila.

1888 Cattell, Prof. J. McKeen Garrison-on-Hudson, N. Y.

1905 Chamberlin, Thomas Chrowder,

LL.D Univ. of Chicago, Chicago, 111.

1880 Chance, Henry Martyn, M.D 819 Drexel Building, Phila.

1S75 Chandler, Prof. C. F Columbia Univ., N. Y. City.

1875 Chapman, Henry C, M.D 2047 Walnut St., Phila.

1886 deCharencey, Comte Hyacinth . . 25 Rue Barbet de Jouy, Paris,

France.

1904 Cheyney, Prof. Edward Potts 259 S. 44th St., Philadelphia.

1904 CHiTTENDEN,Prof. Russel H., Ph.D. 83 Trumbull St., New Haven,

Conn.

1889 Clark, Clarence H 42d and Locust Sts., Phila.

1902 Clark, Prof. William Bullock. .Johns Hopkins University,

Baltimore, Maryland.
1904 Clark, Frank Wigglesworth,

Sc.D U. S. Geological Survey,

Washington, D. C.

1883 Claypole, Prof. E. W Pasadena, Cal.

1895 Cleemann, Richard A., M.D 2135 Spruce St., Philadelphia.

1897 Cleveland, Hon. Grover Westland, Princeton, N. J.

1899 Coles, Edward 2010 DeLancey Place, Phila.

1902 Collitz, Prof. Hermann, Ph.D. . . . Brvn Mawr, Pa.

254 MEMBERS.

Elected. Name. Present Address.

1897 Conklin, Prof. Edwin Grant University of Perm., Phila.

1898 Converse, John H 500 X. Broad St., Phila.

1895 Cook, Joel 849 N. Broad St., Phila.

1886 Cora, Prof. Guido 2 Via Goito, Rome Italy.

1892 Cramp, Charles H Aldine Hotel, Philadelphia.

1877 Crane, Prof. Thomas Frederick .. Cornell Univ., Ithaca, N. Y.

1886 Crookes, .Sir William 7 Kensington Park Gardens,

London, W., England.

1898 Crowell, Prof. Edward P 21 Amity St., Amherst, Mass.

1897 Culin, Stewart Brooklyn Institute of Arts and

Sciences, Brooklyn, N. Y.

1904 Da Costa, John Chalmers, M.D. .2045 Walnut St., Phila.

1897 Dall, Prof. William H U. S. National Museum, Wash-

ington, D. C.

1899 Dana, Charles E 2013 DeLancey Place, Phila.

1896 Dana, Prof. Edward S Yale Univ., New Haven, Conn.

1902 Darboux, Jean-Gaston 36 Rue Gay-Lussac, Paris,

France.

1898 Darwin, Sir George Howard, K.C.B.Newnham Grange, Cambridge,

England.

1876 Davenport, Sir Samuel Beaumont, Adelaide, S. Aus-
tralia.

1866 Davidson, Prof. George 2221 Washington St.,

San Francisco, Cal.

1899 Davis, Prof. William Morris .... Cambridge, Mass.

1880 Dawkins, Prof. William Boyd. . . Woodhurst, Fallowfield, Man-

chester, England.

1899 Day, Frank Miles Allen's Lane, Mount Airy,

Philadelphia.

1905 Delage, Prof. Yves Universite de Paris, Station

Zoologique de Roscoff, Paris,
France.

1904 Delitzsch, Prof. Friedrich, Ph.D. .University of Berlin, Berlin,

Germany.

1892 Dercum, Francis X., M.D 1719 Walnut St., Phila.

1899 Dewar, Prof. James, LL.D The Royal Institution, Lon-
don, England.

1884 Dickson, Samuel 901 Clinton St., Philadelphia.

1892 Dixon, Samuel G., M.D Black Rock Farm, Ardmore,

Pa.

1903 Dohrn, Dr. Anton Marine Zoological Station,

Naples, Italy.

1886 Dolley, Charles S., M.D 3707 Woodland Ave., Phila.

1886 Donner, Prof. Otto Helsingfors, Finland.

1881 Doolittle, Prof. C. L Upper Darby, Delaware Co.,

Pa.

MEMBERS. 255

Elected. Navie. Present Address.

1903 Doolittle Eric University of Pennsylvania,

Philadelphia.

1899 Dougherty, Thomas Harvey School House Lane, German-
town, Philadelphia.

1877 Douglass, James, LL.D Spuytenduyvil, New York,

N. Y.

18S0 Draper, Daniel, Ph.D Meteorological Observatory,

Central Park, New York,
N. Y.

1880 Du Bois, Patterson 401 S. 40th St., Philadelphia.

1879 Dudley, Charles Benj., Ph.D.... Box 156, Altoona, Blair Co.,

Pa.

1886 Duncan, Louis, Ph.D., U.S.N 56 Pine St., New York, N. Y.

1867 Dunning, George F 500 Madison Ave., New York,

N. Y.
1873 DuPont, Edouard Royal Museum, Bruxelles, Bel-

gique.

1894 DuPont, Col. Henry A Winterthur, Del.

1871 Dutton, Maj. Clarence E., U.S.A . Englewood, N. J.

1880 Eckfeldt, Jacob B U. S. Mint, Philadelphia.

1877 Eddy, Prof. H. Turner University of Minnesota, Min-
neapolis, Minn.

1896 Edison, Thomas Alva, Ph.D Orange, N. J.

1895 Edmunds, Hon. George F Aiken, S. C.

1871 Eliot, Pres't Charles W 17 Quincy St., Cambridge,

Mass.
1895 Elliott, Prof. A. Marshall Johns Hopkins University,

Baltimore, Md.

1897 Ely, Theodore N., C.E 115 Broad St. Station, Phila.

1897 Emerson, Prof. Benj. Kendall ... Amherst, Mass.

1898 Emmet, W. L. R 48 Washington Ave.,

Schenectady, N. Y.
1883 Emmons, Prof. S. F 1721 H St., Washington, D. C.

1881 Evans, Sir John, K.C.B Nash Mills, Hemel Hemp-

stead, England.
1895 Ewell, Marshall D., M.D., LL.D. 59 Clark St., Chicago, 111.
1905 Farlow, Prof. William Gilson ... Cambridge, Mass.
1895 Fennel, C. A. M., Litt.D 139 Chesterton Road,

Cambridge, England.

1890 Field, Robert Patterson 218 S. 42d St., Philadelphia.

1897 Fine, Prof. Henry B Princeton, N. J.

1897 Fisher, Sydney George, LL.D 328 Chestnut St., Phila.

1901 Flexner, Simon, M.D Rockefeller Institute, 50th and

Lexington Ave., New York.
1880 Flint, Austin, M.D 60 E. 34th St.,New York, N. Y.

256 MEMBERS.

Elected. Name. Present Address.

1891 Forbes, Prof. Geoege, F.R.S 34 Great George St., 8. W„

London.
1902 Foster, Sib Michael, K.C.B.,

F.R.S., D.C.L Nine Wells, Great Shelford,

Cambridge, Eng.

1880 Fraley, Joseph C 1833 Pine St., Philadelphia.

1904 Francke, Prof. Kuno, Ph.D Harvard University,

Cambridge, Mass.

1872 Frazee, Persifor, Dr. es-Sc. Nat.. 928 Spruce St., Philadelphia.

1889 Friebis, George, M "C 1906 Chestnut St., Phila.

1890 Fullertox, Rev ..eorge S Columbia University,

New York, N. Y

1873 Fulton, John 136 Park PI., Johnstown, Pa.

1880 Furness, Horace Howard, LL.D . . Wallingford, Del. Co., Pa.

1897 Furness, Horace Howard, Je 2034 DeLancey Place, Phila.

1897 Furness, William H., 3d, M.D ... The Warwick, 1906 Sansom

St., Philadelphia.
1901 Garnett, Richard, C.B., LLD . . . . 27 Tanza Road, Hampstead,

London, England.
1S86 Gates, Merrill E.. LL.D 1315 New Hampshire Ave.,

Washington, D. C.
1884 Gatschet, Albebt S., Ph.D 2020 Fifteenth St., N. W.,

Washington, D. C.
1880 Geikie, Sir Archibald 28 Jermyn St., London, S. W.,

England.
1876 Geikie, Prof. James 83 Colinton Rd.

Edinburgh, Scotland.

1886 Genth, Prof. F. A., Jr 222 Walnut St., Phila.

1854 Gibbs, Prof. Oliver Wolcott 158 Gibbs Ave., Newport, R. I.

1901 Giglioli, Prof. Henry H 19 Via Romana,Florence,Italy.

1902 Gilbert, Gbove Kabl, LL.D U. S. Geological Survey,

Washington, D. C.

1903 Gildersleeve, Prof. Basil L., LL.D. 1002 Belvidere Terrace,

Baltimore, Md.
1867 Gill, Theodore N., M.D., Ph.D. . . .Smithsonian Institution,

Washington, D. C.
1876 Gilman, Daniel C, LL.D 614 Park Ave., Baltimore, Md.

1895 Glazebbook, Richard T., F.R.S ... Bushey House, Teddington.

Middlesex, Eng.
1893 Goodale, Prof. George Lincoln... 10 Craigie St., Cambridge,

Mass.

1896 Goodspeed, Prof. Arthur W Univ. of Pennsylvania, Phila.

1892 Goodwin, Harold 133 S. 12th St., Philadelphia.

1895 Goodwin, Prof. W. W Cambridge, Mass.

1900 Gray, George, Hon Wilmington, Del.

1904 Greely, Gen. Adolphus W., U.S.A. 1914 G St., Washington, D. C.

MEMBERS. 257

Elected. Name. Present Address.

1893 Gbeen, Samuel A., M.D Historical Soc, Boston, Mass.

1879 Gbeene, William H., M.D N. E. Cor. Arch and 16th Sts.,

Philadelphia.
1899 Geeenman, Milton J., M.D Wistar Institute, 36th St. and

Darby Road, Philadelphia.
1888 di Gregobio, Marquis Antonio... Al Molo, Palermo, Sicily.
1891 Gregory, Prof. Caspar Rene Naunhoferstrasse 5 Marien-

hohe, Leipzig-Stotteritz,

Germany.

1886 de Gubernatis, Prof. Angelo Florence, Italy.

1903 Gummere, Prof. Francis Barton,

Ph.D Haverford College, Haverford,

Pa.

1902 Hadley, Pres't Arthur T Yale University, New Haven,

Conn.
1885 Haeckel, Prof. Dr. Eenst University, Jena, Germany.

1903 Hague, Abnold, D.Sc 1724 I St., Washington, D. C.

1870 Hale, Rev. Edwabd Evebett 39 Highland St., Roxbury,

Mass.
1902 Hale, Prof. Geobge E Yerkes Observatory, Williams

Bay, Wis.

1878 Hall, Prof. Asaph South Norfolk, Conn.

1875 Hall, Chables Edwabd Instituto Geologico de Mexico,

Santa Maria, Mexico, Mex.

1898 Hall, Chables M 136 Buffalo Ave.,

Niagara Falls, N. Y.

1885 Hall, Prof. Lyman B Haverford Coll., Haverford,

Pa.

1891 Hamy, Dr. Ebnst T 40 Rue Liibeck, Ave. du Troca-

dero, Paris, France.

1887 Habbis, Joseph S 144 School Lane, Germantown,

Philadelphia.
1895 Habbison, Provost Chables C. . . .400 Chestnut St., Phila.

1877 Hart, Prof. James Morgan 1 Reservoir Ave., Ithaca, N. Y.

1878 Haupt, Prof. Lewis M 107 N. 35th St., Philadelphia.

1902 Haupt, Prof. Paul 2511 Madison Ave., Baltimore.

1886 Hays, I. Minis, M.D 266 S. 21st St., Philadelphia.

1883 Heilprin, Prof. Angelo 1801 Arch St., Philadelphia.

1893 Hewett, Prof. Waterman T 31 East Ave., Ithaca, N. Y.

1895 Heyse, Paul, Ph.D Munich, Bavaria.

1903 Hill, George William, LL.D West Nyack, N. Y.

1897 Hiller, H. M., M.D Kohoka, Mo.

1886 Hilprecht, Prof. Hermann V Free Museum of Art,

Univ. of Penn., Phila.
1874 Himes, Prof. Charles Francis. . . .Dickinson Coll., Carlisle, Pa.

1899 Hirst, Barton Cooke, M.D 1821 Spruce St., Philadelphia.

258 MEMBERS.

Elected. Name. Present Address.

1870 Hitchcock, Prof. Ciias. Henry. . .Dartmouth College.

Hanover, N. H.

1897 Holden, Prof. Edward S U. S. Military Academy,

West Point, N. Y.

1886 Holland, James W., M.D 2006 Chestnut St., Phila.

1899 Holmes, Prof. William H Bureau of Ethnology, Smith-
sonian Institution, Wash-
ington, D. C.

1869 Hooker, Sir Joseph D., LL.D The Camp, Sunningdale, Eng.

1893 Hoppin, Prof. J. M New Haven, Conn.

1886 Horner, Inman 1811 Walnut St., Phila.

1872 Hough, Prof. George W Northwestern University,

Evanston, 111.

1872 Houston, Prof. Edwin J 1809 Spring Garden St., Phila.

1897 Howe, Prof. Henry M 27 W. 73d St., New York City.

1903 Howell, Prof. William Henry. . . .232 W.Lanvale St., Baltimore.
1895 Huggins, Sir William, K.C.B 90 Upper Tulse Hill, S. W.,

London, England.
1877 Humphrey, H. C ?

1895 Hunter, Richard S 1413 Locust St., Phila.

1898 Hutchinson, Emlen Aldine Hotel, Philadelphia.

1875 Ingham, Wm. Armstrong 320 Walnut St., Phila.

1893 d'Invilliers, Edward Vincent. . . .506 Walnut St., Phila.

1884 James, Pres't Edmund J Urbana, 111.

1897 Jastrow, Prof. Morris, Jr 248 S. 23d St., Philadelphia.

1898 Jayne, Henry LaBarre 1826 Chestnut St., Phila.

1885 Jayne, Horace, M.D 318 S. 19th St., Philadelphia.

18S2 Jefferis, William W 474 Central Park West,

New York City.

1905 Jordan, Pres't David Starr Stanford Univ., Cal.

1884 Jordan, Francis, Jr Ill N. Front St., Phila.

1883 Kane, Elisha Kent Kushequa, Pa.

1897 Karpinsky, Prof. Alex. Petro-

vitch Geological Survey,

St. Petersburg, Russia.
1889 Keane, Right Rev. John J Dubuque, Iowa.

1899 Keasbey, Prof. Lindley M Univ. of Texas, Austin, Texas.

1897 Keen, Gregory B 3237 Chestnut St., Phila.

1884 Keen, William W., M.D., LL.D.

(Edin.) 1729 Chestnut St., Phila.

1898 Keiser, Prof. Edward H Washington University,

St. Louis, Mo.

1900 Keller, Prof. Harry F Central High School, Phila.

1873 Kelvin, Right Hon. Lord The Library, The University,

Glasgow, Scotland.

1896 Kennelly, A. E., D.Sc Harvard University,

Cambridge, Mas<.

MEMBERS. 25(J

Elected. Name. Present Address.

1905 Kittredge, George Lyman, LL.D . 8 Hilliard St., Cambridge,

Mass.

1898 Knight, Prof. William A Holmleigh, Malvern, Eng.

1874 Konig, Prof. George A School of Mines,

Houghton, Mich.

1899 Kraemer, Prof. Henry 14.5 N. 10th St., Philadelphia.

1889 Krauss, Friedrich S., Ph.D VII2 Neustiftgasse 12,

Vienna, Austria.

1872 Lambert, Prof. Guillaume 42 Boulevard Bischoffsheim,

Brussels, Belgium.

1904 Lambert, Prof. Preston A Lehigh University,

Bethlehem, Pa.
1899 Lamberton, Prof. William A University of Penna., Phila.

1898 de Lancey, Edward F 20 E. 28th St., New York.

1897 Lanciani, Prof. Rodolfo 2 Via Goito, Rome, Italy.

1878 Landreth, Burnet Bristol, Pa.

1875 Langley, Samuel P., LL.D Smithsonian Institution,

Washington, D. C.

1903 Lankester, Edwin Ray, LL.D.,

F.R.S British Museum, Cromwell

Rd., London, S. W., Eng.

1873 La Roche, C. Percy, M.D 1518 Pine St., Philadelphia.

1867 Lea, Henry Charles, LL.D 2000 Walnut St., Phila.

1899 Learned, Prof. Marion D University of Penna., Phila.

1905 LeConte, Robert G., M.D 1625 Spruce St., Phila.

1883 Lehman, Ambrose E 506 Walnut St., Phila.

1889 Le Moine, Sir James M Spencer Grange, Quebec, Can.

1881 Leroy-Beaulieu, Prof. Paul 27 Ave. du Bois de Boulogne,

Paris, France.
1886 Levasseur, Prof. Emile 20 Rue Mons. le Prince, Paris,

France.

1896 Lewis, G. Albert 1834 DeLancey Place, Phila.

1897 Libbey, Prof. William 20 Bayard Ave.,Princeton,N.J.

1897 Lister, The Right Hon. Lord.... 12 Park Crescent, Portland

Place, London, England.

1874 Lockyer, Sir Joseph Norman,

K.C.B Royal College of Science, S.

Kensington, London, S. W.,

England.
1901 Lodge, Sir Oliver Joseph, LL.D.. The University,

Birmingham, England.
1899 Loeb, Dr. Jacques University of California.

Berkeley, Cal.
1878 Longstreth, Morris, M.D 1416 Spruce St., Phila.

1904 Lovett, Prof. Edgar Odell, Ph.D. .Princeton, N. J.

1892 Low, Hon. Seth 30 E. 46th St., New York.

1897 Lowell, Percival 53 State St., Boston, Mass.

260 MEMBERS.

Elected. Xame Present Address.

1869 Lyman, Benjamin Smith. 708 Locust St., Phila.

1S97 Mabery, Prof. Charles F 57 Adelbert St., Cleveland, 0.

1886 MacAlister, Pres't James 4031 Walnut St., Phila.

1897 McCay, Prof. Leroy W 257 Nassau St., Princeton,N.J.

1897 McClure, Prof. Charles F. W Princeton, N. J.

1896 McCook, Rev. Henry C, D.D Devon, Pa.

1879 McCreath, Andrew S 121 Market St., Harrisburg,

Pa.

1892 Macfarlane, Prof. John M Univ. of Pennsylvania, Phila.

1899 Mackenzie, Prof. Arthur S., Ph.D.Dalhousie University,

Halifax, Nova Scotia.

1896 Magie, Prof. Wm. Francis Princeton, N. J.

1897 Mahan, Capt. Alfred T., U.S.N. . . 1U0 W. 86th St., New York.

1885 Mallet, John Wm., M.D University of Virginia,

Charlottesville, Va.

1878 Mansfield, Ira Franklin Beaver, Beaver Co., Pa.

1878 March, Prof. Francis Andrew. . . .Lafayette College, Easton, Pa.

1901 Marconi, Guglielmo 18 Finch Lane, E. C, London.

1901 Marcovnikoff, Prof. Vladimir. . . .Imp. Moskovsy Universitet,

Moscow, Russia.
1878 Marks, Prof. William D Art Club Philadelphia.

1886 Marshall, John, M.D 1718 Pine St., Philadelphia.

1890 Mascart, Prof. E 176 Rue de l'Universite,

Paris, France.

1867 Mason, Andrew 30 Wall St., New York City.

1899 Mason, Prof. Otis T U. S. National Museum,

Washington, D. C.

1896 Mason, Prof. Wm. Pitt, M.D Rensselaer Polytechnic Insti-

tute, Troy, N. Y.

1891 Maspero, Prof. Gaston Camille. . College de France, Paris,

France.

1899 Matthews, Albert , 483 Beacon St., Boston.

1899 Meigs, Arthur V., M.D 1322 Walnut St., Phila.

1901 Meigs, William M 1815 Pine St., Phila.

1886 von Meltzel, Prof. Dr. Hugo Koloszvar, Hungary.

1897 Melville, Rear Admiral Geo. W..Navy Dept., Washington, D. C.
1899 Mendenhall, Prof. Thomas C .... Worcester, Mass.

1898 Mengarini, Prof. Guglielmo Rome, Italy.

1895 Mercer, Henry C Doylestown, Pa.

1902 Merriam, Dr. C. Hart 1919 16th St., Washington,

D. C.

1880 Merrick, John Vaughan Roxborough, Philadelphia.

1881 Merriman, Prof. Mansfield Lehigh Univ., Bethlehem, Pa.

1899 Meyer, Prof. Adolph B K. Zoologisches u. Antliropolo-

gisch-Ethnographisches Mu-
seum, Dresden, Germany.

MEMBERS. 261

Elected. Name. Present Address.

1902 Michelson, Prof. Albert A., Sc.D.

(Cantab.) Univ. of Chicago, Chicago, 111.

1899 Millek, Prof. Leslie \V N. W. cor. Broad and Pine

Sts., Philadelphia.

1896 Minot, Chas. Sedgwick, M.D Harvard University,

Cambridge, Mass.

1890 Mitchell, Hon. James T 1722 Walnut St., Phila.

1862 Mitchell, S. Weir, M.D 1.324 Walnut St., Phila.

1895 Montegaza, Paolo Florence, Italy.

1898 Montgomery, Prof. Thos. H., Jr. . .Univ. of Texas, Austin, Texas.
1S97 Moore. Clarence B 1321 Locust St., Phila.

1905. Moore, Prof. Eliakim H Univ. of Chicago, Chicago, 111.

1905 Moore, George T., Ph.D Hammond Court, 30th and Q

Sts., N. W., Washington, D. C.

1885 Moore, Prof. James W., M.D Lafayette College, Easton, Pa.

1903 Morley, Edward W., LL.D Adelbert Coll., Cleveland, O.

1897 Morley, Prof. Frank '. . . . Johns Hopkins University,

Baltimore.

1899 Morris, Harrison S Oak Lane P. O., Philadelphia.

1S99 Morris, Israel W 225 S. 8th St., Philadelphia.

1883 Morris, J. Cheston, M.D 1514 Spruce St., Philadelphia.

1901 Morris, John T S26 Drexel Building, Phila.

1895 Morse, Prof. Edward S Essex Institute, Salem, Mass.

1903 Morse, Harmon N., Ph.D 1117 N. Eutaw St., Baltimore.

1886 Much, Prof. Matil^us, Ph.D XIIP Penzingerstrasse, 84,

Vienna, Austria.

1901 Munro, Prof. Dana C University of Wisconsin,

Madison, Wis.

1891 Munroe, Prof. Charles E Columbian University,

Washington, D. C.

1886 Mtjrdock. Com. J. B., U.S.N Navy Dept.r Washington, D. C.

1881 Murray, James A. H., LL.D Sunnyside, Banbury Road,

Oxford, England.

1897 Nansen, Prof. Fridtjof Godthaab, Lysaker, Norway.

1S78 Newcomb, Prof. Simon • 1620 P St.. N. W.,

Washington, D. C.

1904 Nichols, Prof. Edward L Cornell Univ., Ithaca, N. Y.

1872 Nichols, Rev. Starr Hoyt 128 Main St., Danbury, Conn.

1866 Nikitin, Prof. Sergi ■. . . .Geological Survey,

St. Petersburg, Russia.

1905 Nordenskjold, Prof. Otto Univ. of Upsal, Upsal, Sweden.

1872 Norris, Isaac, M.D Fair Hill, Bryn Mawr, Pa.

1895 Nuttall, Mrs. Zelia . . Casa de Alvarado, Coyoncan,

D. F. Mexico.

18S6 Oliver, Charles A., M.D 1507 Locust St., Phila.

1897 Olney, Hon. Richard 23 Court Street, Boston.

262 MEMBERS.

Elected. Same. Present Address.

1897 OBTMANNj Prof. Arnold E Carnegie Museum, Annex, 419

Craft Ave., Pittsburg, Pa.
1887 Osbobn, Prof. Henry F American Museum of Natural

History, New York, N. Y.
1885 Osler, William, M.D 7 Norham Gardens, Oxford,

Eng.
1867 Packard, John H., M.D University Club, Philadelphia.

1898 Pancoast, Henry S 78 Vernon St., Hartford, Conn.

1885 Patterson, C. Stuart 1000 Walnut St., Phila.

1900 Patterson, Hon. Edward Supreme Court, Appellate Div.,

1st Dept., New York City.

1898 Patterson, Lamar Gray Guano, Amherst Co., Va.

1851 Patterson, Robert 329 Chestnut St., Phila.

1897 Patton, Pres't Francis L., D.D. . .Princeton, N. J.

1899 Paul, J. Rodman 903 Pine St., Philadelphia.

1875 Pearse, John B 317 Walnut Ave.,

Roxbury, Mass.

1897 Peckiiam, Prof. S. F 150 Halsey St., Brooklyn.

1878 Peirce, C. Newlin, D.D.S 3316 Powelton Ave., Phila.

1873 Pemberton, Henry 1947 Locust St., Phila.

1886 Penafiel, Dr. Antonio City of Mexico, Mexico.

1901 Penniman, Prof. Josiah H 4326 Sansom St., Phila.

1886 Pennypacker, Hon. Samuel W ... Executive Mansion,

Harrisburg, Pa.

1S63 Penrose, R. A. F., M.D 1331 Spruce St., Phila.

1905 Penrose, R. A. F., Jr., Ph.D 460 Bullitt Bld'g, Phila.

1886 Pepper, Edward, M.D Care of Drexel, Harjes & Co.,

31 Boulevard, Haussman,

Paris, France.
1897 Pepper, George Wharton, LL.D..701 Drexel Building, Phila.
1905 Petrie, W. M. Flinders, D.C.L.,

LL.D., F.R.S University College, Gower St.,

W. C, London, Eng.

1895 Pettit, Henry Pelican Lodge,

• Palm Beach, Florida.

1899 Phillips, Prof. Francis C P. O. Box 126, Allegheny, Pa.

1896 Pickering, Prof. Edw. C Harvard University,

Cambridge, Mass.

1897 Piersol, George A., M.D Chester Ave. and 49th St.,

Philadelphia.

1895 Pilsbry, Prof. Henry A Academy of Natural Sciences,

Philadelphia.

1898 Platt, Charles 237 S. 18th St., Phila.

1899 Poincare, Prof. Henri 63 Rue Claude Bernard,

Paris, France.

1885 Pomialowsky, Prof. John St. Petersburg, Russia.

1886 Postgate, Prof. John P Cambridge, England.

MEMBERS. 2G3

Elected. Name. Present Address.

1899 Peeece, Sir Wm. Henky, F.R.S. ... 13 Queen Anne's Gate,

London, S. W., England.

1875 Prime, Frederick 1008 Spruce St., Phila.

1899 Pritchett, Pres't Henry S., LL.D. Massachusetts Institute of

Technology, Boston.

1874 Pumpelly, Prof. Raphael Newport, R. I.

1896 Pupin, Prof. Michael 1 7 Highland PI., Yonkers, N. Y.

1895 Putnam, Prof. Frederick W Peabody Museum, Cambridge,

Mass.
1886 Rada, Juan de Dios-y Delgado. . .Calle de la Corredera baja de

S. Pablo No. 12, Madrid,

Spain.
1899 Ramsay, Sir William, K.C.B University College, Gower St.,

W. C. London, Eng.

1901 Ravenel, Mazyck P., M.D 908 Pine St., Phila.

1898 Rawle, Francis 328 Chestnut St., Phila.

1899 Rawle, William Brooke 230 S. 22d St., Phila.

1886 Rayleigh, The Right Hon. Lord. . .Terling PL, Witham, Essex,

England.

1875 Raymond, Capt. Rossiter W 99 John St., New York, N. Y.

1898 Redwood, Sir Boverton, F.R.S. . ..4, Bishopsgate St. Within,

E. C, London, England.

1899 Remington, Prof. Joseph, P 1832 Pine St., Phila.

1879 Remsen, President Ira Johns Hopkins University,

Baltimore, Md.

1879 Renevier, Prof. E Univ., Lausanne, Switzerland.

1886 Reville, Prof. Dr. Albert 21 Rue Guenegaud,

Paris, France.

1902 Richards, Prof. Theo. William. ..15 Follen St., Cambridge,

Mass.
1882 Robins, Rev. James W., D.D 2115 Pine St., Phila.

1890 Rogers, Prof. Robert W Drew Theological Seminary,

Madison, N. J.
1862 Rohrig, Prof. F. L. Otto 402 S. Oakland Ave.,

Pasadena, Cal.
1904 Roosevelt, Hon. Theodore The White House,

Washington, D. C.

1903 Roscoe, Sir Henry E., F.R.S.,

D.C.L Woodcote Lodge, West Hors-

ley, Leatherhead, England.

1891 Rosengarten, Joseph G 1704 Walnut St., Phila.

1882 de Rosny, Prof. Leon 28 Rue Mazarin, Paris,

France.
1877 Rothrock, Prof. Joseph T West Chester, Pa.

1904 Rutherford, Prof. Ernest, F.R.S. .McGill Univ. Montreal, Can.
1894 Sachse, Julius F., Litt.D 4428 Pine St., Phila.

2(54 MEMBERS.

Elected. Xante. Present Address.

1874 Sadtler, Prof. Samuel P X. E. cor. 10th and Chestnut

Sts., Philadelphia.

1888 Sajous, Charles E., M.D 2043 Walnut St., Phila.

1897 Sampson, Alden Haverford, Pa.

1866 Sandberger, Prof. Fredolix University of Wiirzburg,

Wiirzburg, Bavaria.

1897 Sanders, Richard H 1225 Locust St., Phila.

1882 Sargent, Prof. Charles Sprague. .Jamaica Plain, Mass.
1902 Sciiellixg, Prof. Felix E., Ph.D.. 4211 Sansora St., Phila.
1901 Schiaparelli, Prof. Giovanni Royal Observatory,

Milan, Italy.

1878 Schurz, Hon. Carl 54 William St., New York.

1873 Sclater, Philip Lutley, Ph.D.,

F.R.S 3 Hanover Square, London,

W., England.

1898 Scott, Charles F Westinghouse Electric Co.,

Pittsburg, Pa.

1886 Scott, Prof. William B Princeton, N. J.

1878 Scudder, Samuel Hubbard Cambridge, Mass.

1897 See, Thomas J. J., LL.D Observatory, Mare Island, Cal.

1872 Sellers, Coleman, Sc.D 3301 Baring St., Phila.

1899 Sellers, Coleman, Jr 410 N. 33d St., Phila,

1885' Sergi, Prof. Giuseppe Universita Romana,

Rome, Italy.

1886 Sharp, Benjamin, M.D Academy of Natural Sciences,

Philadelphia.
1882 SHARPLES,Prof. Stephen Paschall.26 Broad St., Boston, Mass.

1875 Sherwood, Andrew Mansfield, Tioga Co., Pa.

1899 Sigsbee, Admiral Charles D.,

U.S.N Navy Dept., Washington, D. C

1900 Sinkler, Wharton, M.D 1606 Walnut St., Phila.

1897 Smith, A. Donaldson, M.D 1820 Chestnut St., Phila.

1887 Smith, Prof. Edgar F Univ. of Pennsylvania, Phila.

1875 Smith, Stephen, M.D 57 W. 42d St., New York, N. Y.

1897 Smock, Prof. John C Trenton, N. J.

1887 Symth, Prof. Albert H 5214 Main St., Germantown,

Philadelphia.
1894 Snellen, Herman, Jr., Ph.D Utrecht, Netherlands.

1873 Snowden, A. Loudon, LL.D 1812 Spruce St., Phila.

1884 Snyder, Prof. Monroe B 2402 N. Broad St., Phila.

1873 Spofford, A. R., LL.D Library of Congress,

Washington, D. C.

1903 Stengel, Alfred, M.D 1811 Spruce St., Phila.

1897 Stepiiexs, Prof. H. Morse University of California.

Berkeley, Cal.

MEMBERS. 265

Elected. Name. Present Address.

1884 Stevens, Prof. Walter LeConte. .Lexington, Va.

1877 Stevenson, Prof. John James .... Univ. Heights, New York City.

1895 Stevenson, Sara Y., Sc.D 237 S. 21st St., Philadelphia.

1898 Stillwell, Lewis B 100 Broadway, New York City.

1902 Stoney, Prof. G. Johnstone, F.R.S. 30 Ledbury Road, Bayswater,

London, W., England.

1904 Stratton, Prof. Samuel W Bureau of Standards,

Washington, D. C.

1886 Suess, Prof. Eduard K. K. Geologische Reichsan-

stalt, Vienna, Austria.

1895 Sulzberger, Hon. Mayer 1303 Girard Ave., Phila.

1886 Szombathy, Prof. Josef Burgring 7, Vienna, Austria.

1897 Tatham, William 1811 Walnut St., Phila.

1886 Temple, Col. Richard Carnac . . . . Port Blair, Andaman Islands,

Bengal, India.

1896 Tesla, Nikola, LL.D Warclenclyffe, Long Island.

N. Y.

1905 Thiselton-Dyer, Sir William T. . Royal Botanic Gardens,

Kew, England.
1884 Thomas, Prof. Allen C Haverford, Pa.

1884 Thompson, Heber S Sheafer Build'g, Pottsville, Pa.

1902 Thompson, Prof. Silvanus P.,

F.R.S Technical College, Finsbury,

Leonard St., City Road,
London, E. C, England.
1876 Thomson, Prof. Elihu Swampscott, Mass.

1903 Thomson, Joseph John, D.Sc,

F.R.S Trinity Coll., Cambridge, Eng.

1880 Thomson, William, M.D 1426 Walnut St., Phila.

1885 im Thurn, Everard F Govt. House, Sura, Fiji,

Western Pacific.

1864 Thury, Prof. A University of Geneva,

Geneva, Switzerland.

1902 Tilghman, Benjamin Chew 1126 S. 11th St., Phila.

1886 Topinard, Prof. Dr. Paul 105 Rue de Rennes,

Paris, France.

1895 Tower. Hon. Charlemagne, Jr.,

LL.D U. S. Embassy, Berlin, Ger.

1903 Trelease, William Sc.D Missouri Botanical Garden,

St. Louis, Mo.

1899 Trevelyan, Rt. Hon. Sir George

Otto 8 Grosvenor Crescent, S. W.

London, England.

1896 Trowbridge, Prof. John Harvard University,

Cambridge, Mass.
1899 True, Dr. Frederick William . . . . U. S. National Museum,

Washington, D. C.

266 MEMBERS.

Elected. Xame. Present Address.

1882 Tschermak, Gustav Universitat, Vienna, Austria.

1897 Tschernyschew, Prof. Theodore. .Geological Survey,

St. Petersburg, Russia.

1864 V. Tunner, Prof. Peter R Leoben, Austria.

1890 Turrettini, Prof. Theodore Geneva, Switzerland.

1889 Tuttle, David K., Ph.D U. S. Mint, Philadelphia.

1889 Tyler, Hon. Lyon G., Pres't Williamsburg, Va.

1887 Tyson, James, M.D 1506 Spruce St., Phila.

1890 Unwin, Prof. William C 7 Palace Gate Mansions,

London, England.

1904 Van't Hoff, Prof. Jakob H Univ. of Berlin, Berlin, Ger.

1899 Vauclain, Samuel M 1533 Green St., Phila.

1897 Vaux, George, Jr 404 Girard Building, Phila.

1905 Venable, Pres't Francis P Chapel Hill, N. C.

1870 Vose, Prof. George L Brunswick, Maine.

1890 Vossion, Louis Consulate of France,

Cape Town, South Africa.

1903 de Vries, Prof. Hugo University of Amsterdam,

Amsterdam, Netherlands.
1885 Wagner, Samuel Greenbank Farm,

West Chester, Pa.

1874 Wahl, William H., Ph.D 15 S. 7th St., Philadelphia.

1897 Walcott, Charles D., LL.D U. S. Geological Survey,

Washington, D. C.

1904 Waldeyer, Prof. Wilhelm 56 Luisenstrasse, Berlin,

N. W., 6, Germany.
1873 Wallace, Alfred Russel, LL.D. . . Parkstone, Dorset, England.
1889 Ward, Lester F., LL.D 1464 Rhode Island Ave.,

Washington, D. C.
1881 Ware, Lewis S 54 Rue de la Bienfaisfince,

Paris, France.

1897 Warfield, Pres't Ethelbert D. . . .Easton, Pa.

1896 Welch, William H., M.D 935 St. Paul St., Baltimore,

Md.

1869 Wharton, Joseph P. 0. Box 1332, Phila.

1869 White, Hon. Andrew D White Library, Cornell Univ.,

Ithaca, N. Y.
1878 Wthite, Prof. Israel C 119 Wiley St., Morgantown,

W. Va.

1905 Whitfield, J. Edward 406 Locust St., Phila.

1898 Whitfield, Prof. R. P American Museum of Natural

History, New York, N. Y.

1899 Whitman, Prof. Charles Otis,

Ph.D., LL.D Univ. of Chicago, Chicago, 111.

1878 Wilder, Prof. Burt G 60 Cascadilla PI., Ithaca, N. Y.

1904 Wiley, Harvey W., M.D., LL.D ... U. S. Dept. of Agriculture,

Washington, D. C.

MEMBERS. 267

Elected. Name. Present Address.

1895 Willcox, Joseph " The Clinton," 10th and Clin-

ton Sts., Philadelphia.
1897 Williams, Prof. Edward H., Jr. . .53 Phillips St., Andover, Mass.

1888 Williams, Talcott, LL.D 910 Pine St., Phila.

1905 Willis, Bailey, E.M., C.E U. S. Geological Survey,

Washington, D. C.
1890 Willis, Prof. Hexry 4036 Baring St., Phila.

1885 Wilson, James Cornelius, M.D. . . 1511 Walnut St., Phila.
1887 Wilson, Prof. William Powell,

M.D Commercial Museum, Phila.

1897 Wilson, Pres't Woodrow Prospect, Princeton, N. J.

1897 Wister, Owen 328 Chestnut St., Phila.

1897 Witmer, Prof. Ligiitner, Ph.D 2426 Spruce St., Phila.

1879 Wood, Richard 1620 Locust St., Phila.

1899 Wood, Stuart 1620 Locust St., Phila.

1874 Woodward, Henry, LL.D., F.R.S. . British Museum, Cromwell

Road, London, S. W., Eng.

1902 Woodward, Pres't Robert S., Ph.D. Carnegie Institution,

Washington, D. C.

1896 Wright, Prof. Arthur W., Ph.D.. 73 York Square, New Haven,

Conn.

1900 Wright, William Aldis, LL.D.,

D.C.L Trinity Coll., Cambridge, Eng.

1895 Wundt, Prof. William Leipzig, Germany.

1899 Wurts, Alexander Jay Carnegie Technical School,

Pittsburg, Pa.
1881 Wurts, Charles Stewart, M.D... 1701 Walnut St., Phila.

1886 Wyckoff, Lieut. A. B., U.S.N Navy Dep't, Washington. D. C.

1874 Young, Prof. Charles Augustus.. 16 Prospect Ave.,

Princeton, N. J.

MAGELLANIC PREMIUM

Founded in 17S6 by John Hyacinth de Magellan, of London
I9O6

THE AMERICAL PHILOSOPHICAL SOCIETY

Held at Philadelphia, for Promoting Useful Knowledge

ANNOUNCES THAT IN

DECEMBER, 1906

IT WILL AWAKD ITS

MAGELLANIC GOLD MEDAL

TO THE AUTHOR OF THE BEST DISCOVERY, OR MOST USEFUL INVENTION,
RELATING TO NAVIGATION ASTRONOMY, OR NATURAL PHILOSOPHY
(MERE NATURAL HISTORY ONLY EXCEPTED) UNDER THE FOLLOWING
CONDITIONS :

i. The candidate shall, on or before November 1, 1906, deliver free of
postage or other charges, his discovery, invention or improvement, ad-
dressed to the President of the American Philosophical Society, No. 104
South Fifth Street, Philadelphia, U. S. A., and shall distinguish his per-
formance by some motto, device, or other signature. With his discovery,
invention, or improvement, he shall also send a sealed letter containing
the same motto, device, or other signature, and subscribed with the real
name and place of residence of- the author.

2. Persons of any nation, sect or denomination whatever, shall be
admitted as candidates for this premium.

3. No discovery, invention or improvement shall be entitled to this
premium, which hath been already published, or for which the author
hath been publicly rewarded elsewhere.

4. The candidate shall communicate his discovery, invention or im-
provement, either in the English, French, German, or Latin language.

5. A full account of the crowned subject shall be published by the
Society, as soon as may be after the adjudication, either in a separate
publication, or in the next succeeding volume of their Transactions, or in
both.

6. The premium shall consist of an oval plate of solid standard gold of
the value of ten guineas, suitably inscribed, with the seal of the Society
annexed tothe medal by a ribbon.

All correspondence in relation hereto should be addressed
To the Secretaries of the

AMERICAN PHILOSOPHICAL SOCIETY
No. 104 Soith Fifth Street

PHILADELPHIA, U. S. A.

TRANSACTIONS

OF THE

American Philosophical Society

HELD AT PHILADELPHIA

For Promoting Useful Knowledge

Vol. XXI, New Series, Part II. Just Published

On the Construction of Isobaric Charts for High Levels in tht

Earth's Atmosphere and their Dynamic Significance.

By J. YV. Sandstrom, of Stockholm.

Subscription— Five Dollars per Volume
Separate parts are not sold

Address The Librarian of the

AMERICAN PHILOSOPHICAL SOCIETY
No. 104 South Fifth Street

PHILADELPHIA, U S. A.

Proceedings Am. Philos. Soc. Vol. XLIV. No. 180

Plate III

7

Map to illustrate the centers of origin and the chief directions of migration of the different subdi
of the genus Cambarus.

Proceedings Am. Philos. Soc. Vol. XLIV. No. 181

An- '--~Clo.

Anatomy of Phalrenoptilus, Ridgway.

Proceedings Am. Philos. Soc. Vol. XLIV. No. 181

Plate V

M y,kum.

Anatomy of Phalsenoptilus, Ridgway.

Proceedings Am. Philos. Soc. Vol. XLIV. No. 181

Plate VI

Anatomy of Phalaenoptilus, Ridgvvay.

e

^\

V V

M^