COSMOS

A SKETCH

A PHYSICAL DESCRIPTION OF THE UNIVERSE.

ALEXANDER VON HUMBOLDT. / 7^ 1 - / S Slf

TRANSLATED FROM THE GERMAN,

BY E. C. 0TT:6 and W. S. DALLAS, F.L.S.

Naturae vero rerum vis atque majestas in omniljus momentis fide caret, si quia moffo
partes ejus ac non totam complectatur animo. — Plin., HUt. Nat.^ lib. vii., c. 1.

VOL. V.

NE W YORK:

HARPER & BROTHERS, PUBLISHERS,
329 & 331 PEARL STREET,

FRANKLIN SQUARE.

18(56.

q \52

Bmnctoh Ubcar>

GENERAL SUMMARY OF CONTENTS

OF
VOLUME V. OF COSMOSw

Introduction to the special results of observation in the domain of
telluric phenomena ]?age 5-14

EiRST Section ...., 14-156

Size, form, and density of the earth 14-37

Internal heat of the earth 37-50

Magnetic activity of the earth 50-156

Historical portion 50-88

Intensity 87-100

Inclination 100-115

Declination 115-146

Polar light 146-156

Second Section 157-

Reaction of the interior of the earth upon its surface. 157, etc-
Earthquakes; dynamic action, waves of concussion... 160-176

Thermal springs 177-198

Gas springs, salses, mud volcanoes, naptha springs .... 198-214

Volcaiioes with and without slructiiral frames {conical and bell-shaped
mountains) 214-451

Range of volcanoes from north (\^\° N. lat.) to south, as far as 46°
south latitude : Mexican volcanoes, p. 266 and 375 (Jorullo, p. 292,
304, note at p. 293) ; Cofre de Perote, p. 307, Cotopaxi, notes p. 317-
321. Subterranean eruptions of vapor, p. 322-324. Central America,
p. 255-263. New Granada and Quito, p. 266-270, and notes (Anti-
sana, p. 311-316; Sangay, p. 416; Tungurahua, p. 415; Cotopaxi,
p. 318-320; Chimborazo, p. 431, note *); Peru and Bolivia, p. 270,
note; Chili, p. 272, note 1| (Antilles, p. 394, note *).

Enumeration of all the active volcanoes in the Cordilleras, p. 270.
Relation of the tracts without volcanoes to those abounding in them,
p. 280, note * at 268 ; volcanoes in the Northwest of America, to the
north of the parallel of the Rio Gila, p. 377-392 ; review of all the
volcanoes not belonging to the New Continent, p. 270-377; Europe,
p. 328, 329 ; islands of the Atlantic Ocean, p. 330 ; Africa, p. 332 ;
Asia— Continent, p. 334-344; Thian-shan, p. 336, 337, 405, and
notes p. 327 to 330 (peninsula of Kamtschatka, p. 340-344) ; Eastern
Asiatic Islands, p. 344 (island of Saghahn, Tarakai or Karafuto, note^

SYNOPSIS.

p. 288 and 289; volcanoes of Japan, p. 350; islands of Southern
Asia, p. 354-358) ; Java, p. 281-290. The Indian Ocean, p. 358-
363 ; the South Sea, p. 363-376.

Probable number of volcanoes on the globe, and their distribution on the
continents and islands Pace 393-403

Distance of volcanic activity from the sea, p. 279, 404, 405. Ee-
gions of depression, p. 403-407 ; Maars, Mine funnels, p. 221, 222.
Different modes in which solid masses may reach the surface from
the interior of the earth, through a net-work of fissures in the cor-
rugated soil, without the upheaval or construction of conical or dome-
shaped piles (basalt, phouolite, and some layers of pearl-stone and
pumice, seem to owe their appearance above the surface, not to sum-
mit-craters, but to the effects of fissures). Even the effusions from
volcanic summits do not in some lava streams consist of a continuous
fluidity, but of loose scoriae, and even of a series of ejected blocks and
rubbish ; there are ejections of stones which have not all been glow-
ing, p. 291, 311, 312-315, 322-326, note * (p. 289), note * (page 315).
^ Mineralogical composition of the volcanic rock : generalization of
the term trachyte, p. 423 ; classification o'f the trachytes, according to
their essential ingredients, into six groups or divisions in conformity
with the definitions of Gustav Rose ; and geographical distribution of
these gi'oups, p. 423-436; the designations andesite and andesine,
p. 422-437, note, 440. Along Avith the characteristic ingredients of
the trachyte formations there are also unessential ingredients, the
abundance or constant absence of which in volcanoes frequently very
near each other deserves great attention, p. 441 ; Mica, ibid. ; glassy
feldspar, p. 442; hornblende and augite, p. 443 ; leucite, p. 44*4 ; oli-
vin, p. 444 ; obsidian, and the difference of opinion on the formation
of pumice, p. 447 ; subterranean pumice-beds, remote from volcanoes,
at Zumbalica, in the Cordilleras of Quito, at Huichapa in the Mexican
Highland, and at Tschigem in the Caucasus, p. 320-324. Diversity
of the conditions^ under which the chemical processes of volcanicily
proceed in the formation of the simple minerals and their association
into trachytes, p. 440, 441, 451.

INTRODUCTION.

SPECIAL RESULTS OF OBSERVATION IN THE DOMAIN
OF TELLURIC PHENOMENA.

In a work embracing so wide a field as the Cosmos, which
aims at combining perspicuous comprehensibility with gen-
eral clearness, the composition and co-ordination of the whole
are, perhaps, of greater importance than copiousness of detail.
This mode of treating the subject becomes the more desira-
ble because, in the Book of Nature, the generalization of
views, both in reference to the objectivity of external phe-
nomena and the reflection of the aspects of nature upon the
imagination and feelings of man, must be carefully separated
from the enumeration of individual results. The first two
volumes of the Cosmos were devoted to this kind of general-
ization, in which the contemplation of the Universe was con-
sidered as one great natural whole, while at the same time
care was taken to show how, in the most widely remote zones,
mankind had, in the course of ages, gradually striven to dis-
cover the mutual actions of natural forces. Although a great
accumulation of phenomena may tend to demonstrate their
causal connection, a General Picture of Nature can only pro-
duce fresh and vivid impressions when, bounded by narrow
limits, its perspicuity is not sacrificed to an excessive aggre-
gation of crowded facts.

As in a collection of graphical illustrations of the surface
and of the inner structure of the earth's crust, general maps
precede those of a special character, it has seemed to me that
in a physical description of the Universe it would be most
appropriate, and most in accordance with the plan of the
present work, if, to the consideration of the entire Universe
from general and higher points of view, I were to append in
the latter volumes those special results of observation upon
which the present condition of our knowledge is more partic-
ularly based. These volumes of my work must, therefore,
in accordance with a remark already made {Cosmos, vol. iii.,
p. 5-9), be considered merely as an expansion and more
careful exposition of the General Picture of Nature ( Cosmos,

6 COSMOS. -

vol. i., p. 56-359), and, as the uranological or sidereal sphere
of the Cosmos was exclusively treated of in the two last
volumes, the present volume will be devoted to the consid-
eration of the telluric sphere. In this manner the ancient,
simple, and natural separation of celestial and terrestrial ob-
jects has been preserved, which we find by the earliest evi-
dences of human knowledge to have prevailed among all na-
tions.

As in the realms of space, a transition to our own planet-
ary system from the region of the fixed stars, illumined by
innumerable suns, whether they be isolated or circling round
one another, or whether they be mere masses of remote neb-
ulae, is indeed to descend from the great and the universal to
the relatively small and special — so does the field of our con-
templation become infinitely more contracted when we pass
from the collective solar system, which is so rich in varied
fornxs, to our own terrestrial spheroid, circling round the
sun. The distance of even the nearest fixed star, a Centauri,
is 263 times greater than the diameter of our solar system,
reckoned to the aphelion distance of the comet of 1680 ; and
yet this aphelion is 853 times further from the sun than our
earth {Cosmos, vol. iv., p. 190). These numbers, reckoning
the parallax of a Centauri at 0^^-9187, determine approxi-
mately both the distance of a near region of the starry heav-
ens from the supposed extreme solar system and the distance
of those limits from the earth's place.

Uranology, which embraces the consideration of all that
fills the remote realms of space, still maintains the character
it anciently bore, of impressing the imagination most deeply
and powerfully by the incomprehensibility of the relations
of space and numbers which it embraces ; by the known or-
der and regularity of the motions of the heavenly bodies ;
and by the admiration which is naturally yielded to the
results of observation and intellectual investigation. This
consciousness of regularity and periodicity was so early im-
pressed upon the human mind, that it was often reflected in
those forms of speech which refer to the ordained course of
the celestial bodies. The known laws which rule the celes-
tial sphere excite, perhaps, the greatest admiration by their
simplicity, based, as they solely are, upon the mass and distri-
bution of accumulated ponderable matter and upon its forces
of attraction. The impression of the sublime, when it arises
from that which is immeasurable and physically great, pass-
es almost unconsciously to ourselves beyond the mysterious

INTRODUCTION. 7

boundary which connects the metaphysical with the physical,
and leads us into another and higher sphere of ideas. The
image of the immeasurable, the boundless, and the eternal, is
associated with a power which excites within us a more earn-
est and solemn tone of feeling, and which, like the impres-
sion of all that is spiritually great and morally exalted, is not
devoid of emotion.

The effect which the aspect of extraordinary celestial phe-
nomena so generally and simultaneously exerts upon entire
masses of people, bears witness to the influence of such an
association of feelings. The impression produced in excita-
ble minds by the mere aspect of the starry vault of heaven
is increased by profounder knowledge, and by the use of those
means which man has invented to augment his powers of vi-
sion, and at the same time enlarge the horizon of his observ-
ation. A certain impression of peace and calmness blends
with the impression of the incomprehensible in the universe,
and is awakened by the mental conception of normal regu-
larity and order. It takes from the unfathomable depths of
space and time those features of terror which an excited im-
agination is apt to ascribe to them. In all latitudes man,
in the simple natural susceptibihty of his mind, prizes '* the
calm stillness of a starlit summer night."

Although magnitude of space. and mass appertains more
especially to the sidereal portion of cosmical delineation, and
the eye is the only organ of cosmical contemplation, our tel-
luric sphere has, on the other hand, the preponderating ad-
vantage of presenting us with a greater and a scientifically
distinguishable diversity in the numerous elementary bodies
of which it is composed. All our senses bring us in contact
with terrestrial nature ; and while astronomy, which, as the
knowledge of moving luminous celestial bodies is most acces-
sible to mathematical treatment, has been the means of in-
creasing in the most marvelous manner the splendor of the
higher forms of analysis, and has equally enlarged the lim-
its of the extensive domain of optics, our earthly sphere, on
the other hand, by its heterogeneity of elements, and by the
complicated play of the expressions of force inherent" in
matter, has formed a basis for chemistry, and for all those
branches of physical science which 'treat of phenomena
that have not as yet been found to be connected with vibra-
tions generating heat and light. Each sphere has, there-
fore, in accordance with the nature of the problems which
it presents to our investigation, exerted a different influence

8 COSMOS.

on the intellectual activity and scientific knowledge of man-
kind.

All celestial bodies, excepting our own planet and the
aerolites which are attracted by it, are, to our conception,
composed only of homogeneous gravitating matter, without
any specific or so-called elementary difierence of substances.
Such a simple assumption is, however, not by any means
based upon the inner nature and constitution of these remote
celestial orbs, but arises merely from the simplicity of the
hypotheses which are capable of explaining and leading us to
predict the movements of the heavenly bodies. This idea
arises, as I have already had occasion frequently to remark
(Cosmos, vol. i., p. 62-67, and p. 135-137 ; vol. iii., p. 6-20,
and 22-24), from the exclusion of all recognition of "hetero-
geneity of matter, and presents us with the solution of the
great problem of celestial mechanics, in which all that is va-
riable in the uranological sphere is subjected to the sole con-
trol of dynamical laws.

Periodical alternations of light upon the surface of the
planet Mars do indeed point, in accordance with its different
seasons of the year, to various meteorological processes, and
to the polar precipitates excited by cold in the atmosphere
of that planet (Cosmos, vol. iv., p. 160). Guided by analo-
gies and i-easoning, we may indeed here assume the presence
of ice or snow (oxygen and hydrogen), as in the erup|,ive
masses or the annular plains of the moon we assume the ex-
istence of different kinds of rock on our satellite, but direct
observation can teach us nothing in reference to these points.
Even Newton ventured only on conjectures regarding the
elementary constitution of the planets which belong to our
own solar system, as we learn from an important conversa-
tion which he had at Kensington with Conduit (Cosmos, vol.
i., p. 132). The uniform image of homogeneous gravitating
matter conglomerated into celestial bodies has occupied the
fancy of mankind in various ways, and mythology has even
linked the charm of music to the voiceless regions within the
realms of space (Cosmos, vol. iv., p. 108-110).

Amid the boundless wealth of chemically varying sub-
stances, M'ith their numberless manifestations of force — amid
the plastic and creatirt^e energy of the whole of the organic
world, and of many inorganic substances — amid the meta-
morphosis of matter which exhibits an ever-active appear-
ance of creation and annihilation, the human mind, ever
striving to grasp at order, often yearns for simple laws of

INTRODUCTION. V

motion in the investigation of the terrestrial sphere. Even
Aristotle, in his Physics, states that " the fundamental prin-
ciples of all nature are change and motion ; he who does not
recognize this truth recognizes not Nature herself" {Phys.
Auscult, iii., 1, p. 200, Bekker), and, referring to the differ-
ence of matter ("a diversity in essence"), he designates mo-
tion, in respect to its qualitative nature, as a metamorphosis,
dXXoto)Gig, very different from mere mixture, fJil^ig, and a
penetration which does not exclude the idea of subsequent
separation {De Gener. et Corrujit., i., 1, p. ,327).

The unequal ascent of fluids in capillary tubes — the endos-
mosis which is so active in all organic cells, and is probably
a consequence of capillarity — the condensation of different
kinds of gases in porous bodies (of oxygen in spongy plati-
num, with a pressure which is equal to a force of more than
700 atmospheres, and of carbonic acid in boxwood charcoal,
when more than one third is condensed in a liquid state on
the walls of the cells) — the chemical action of contact-sub-
stances, which by their presence occasion or destroy (by ca-
talysis) combinations without themselves taking any part in
them — all these phenomena teach us that bodies at infinitely
small distances exert an attraction upon one another, which
depends upon their specific natures. We can not conceive
such attractions to exist independently of motions, which
must be excited by them although inappreciable to our eyes.

We are still entirely ignorant of the relations which recip-
rocal molecular attraction as a cause of unceasing motion
on the surface, and very probably also in the interior of the
earth's body, exerts upon the attraction of gravitation, by
which the planets as well as their central body are main-
tained in constant motion. Even the partial solution of this
purely physical problem would yield the highest and most
splendid results that can be attained in these paths of in-
quiry, by the aid of experimental and intellectual research.
I purposely abstain in these sentences from associating (as is
commonly done) the name ^f Newton with that law of at-
traction which rules the celestial bodies in space at bound-
less distances, and which is inversely as the square of the
distance. Such an association implies almost an injustice
toward the memory of this great man, who had recognized
both these manifestations of force, although he did not sepa-
rate them with sufficient distinctness; for we find — as if in
the felicitous presentiment of future discoveries — that he at-
tempted, in the Queries to his Optics, to refer capillarity, and

A 2

10 COSMOS.

the little that was then known of chemical affinity, to univers-
al gravitation (Laplace, Expos, du Syst. dit 3Ionde, p. 384.
Cosmos, vol. iii., p. 23).

As in the physical world, more especially on the borders
of the sea, delusive images often appear which seem for a
time to promise to the expectant discoverer the possession of
some new and unknown land; so, on the ideal horizon of
the remotest regions of the world of thought, the earnest in-
vestigator is often cheered by many sanguine hopes, which
vanish almost as^quickly as they have been formed. Some
of the splendid discoveries of modern times have undoubtedly
been of a nature to heighten this expectation. Among these
we may instance contact-electricity — magnetism of rotation,
which may even be excited by fluids, either in their aqueous
form or consolidated into ice — the felicitous attempt of con-
sidering all chemical affinity as the consequence of the elec-
trical relations of atoms with a predominating polar force —
the theory of isomorphous substances in its application to
the formation of crystals — many phenomena of the electrical
condition of living muscular fibre — and, lastly, the knowledge
which we have obtained of the influence exerted by the sun's
position, that is to say, the thermic force of the solar rays,
upon the greater or lesser magnetic capacity and conducting
power of one of the constituents of our atmosphere, namely,
oxygen. When light is unexpectedly thrown upon any pre-
viously obscure group of phenomena in the physical world,
we may the more readily believe that we are on the thresh-
old of new discoveries, when we find that these relations ap-
pear to be either obscure, or even in opposition to already
established facts.

I have more particularly adduced examples in which the
dynamic actions of attracting forces seem to show the course
by which we may hope to approximate toward the solution
of the problem of the original, unchangeable, and hence
named the elementary heterogeneity of substances (for in-
stance, oxygen, hydrogen, sulpjjur, potassium, phosphorus,
tin, etc.), and of the amount of their tendency to combine ;
in other words, their chemical affinity. Differences of form
and mixture are, I would again repeat, the only elements of
our knowledge of matter; they are the abstractions under
which we endeavor to comprehend the all-moving universe,
both as to its size and composition. The detonation of the
fulminates under a slight mechanical pressure, and the still
more formidable explosion of terchloride of nitrogen, which

INTRODUCTION. 11

is accompanied by fire, contrast with the detonating combi-
nation of chlorine and hydrogen, which explodes when the
sun's rays fall directly upon it (more especially the violet
rays). Metamorphosis, union, and separation afford evi-
derlce of the eternal circulation of the elements in inorganic
nature no less than in the living cells of plants and animals.
"The quantity of existing matter remains, however, the
same ; the elements alone change their relative positions to
one another."

We thus find a verification of the ancient axiom of Anax-
agoras, that created things neither increase nor decrease in
the Universe, and that that which the Greeks termed the
destruction of matter was a mere separation of parts. Our
earthly sphere, within which is comprised all that portion
of the organic physical world which is accessible to our ob-
servation, is apparently a laboratory of death and decay ; but
that great natural process of slow combustion, which we call
decay, does not terminate in annihilation. The liberated
bodies combine to form other structures, and through the
agency of the active forces which are incorporated in them
a new life germinates from the bosom of the earth.

COSMOS.

RESULTS OF OBSERVATION IN THE TELLURIC PORTION
OF THE PHYSICAL DESCRIPTION OF THE UNIVERSE.

In the attempt to grasp the inexhaustible materials aiFord-
^d by the study of the physical world ; or, in other words,
to group phenomena in such a manner as to facilitate our in-
sight into their causal connection, general clearness and lu-
cidity can only be secured where special details — more par-
ticularly in the long and successfully cultivated fields of ob-
servation— are not separated from tlie higher points of view
of cosmical unity. The telluric sphere, as opposed to the
uranological, is separable into two portions, namely, the in-
organic and the organic departments. The former comprises
the size, form, and density of our terrestrial planet ; its in-
ternal heat ; its electro-magnetic activity ; the mineral con-
stitution of the earth's crust ; the reaction of the interior of
the planet on its outer surface which acts dynamically by
producing earthquakes, and chemically by rock-forming, and
rock-metamorphosing processes ; the partial covering of the
solid surface by the liquid element — the ocean ; the contour
and articulation of the upheaved earth into continents and
islands ; and, lastly, the general external gaseous investment
(the atmosphere). The second or organic domain comprises
not the individual forms of life which we have considered in
the Delineation of Nature, but the relations in space which
they bear to the solid and fluid parts of the earth's surface,
the geography of plants and animals, and the descent of the
races and varieties of man from one common, primary stock.

This division into two domains belongs, to a certain extent,
to the ancients, who separated from the vital phenomena of
plants and animals such material processes as change of form
and the transition of matter from one body to another. In
the almost total deficiency of all means for increasing the
powers of vision, the difference between the two organisms*
was based upon mere intuition, and in part upon the dogma
* See CosmoSf voL iii., p. 42.

14 ^ COSMOS.

of self-nutrition (Ar'istot., De Anima, ii., 1, t. i., p. 412, a 14,
Bekker), and of a spontaneous incentive to motion. Tliis
kind of mental comprehension which I have named intuition,
together with that felicitous acumen in the power of combin-
ing his ideas, which was so characteristic of the Stagyrite,
led him to the assumption of an apparent transition from
the inanimate to the living, from the mere element to the
plant, and induced him even to adopt the view that in the
ever-ascending processes of plastic formation there Vere grad-
ual and intermediate stages connecting plants with the low-
er animals (Aristot., De part Animal., iv., 5, p. 681, a 12,
and Hist. Animal, viii., 1, p. 588, a 4, Bekker). The history
of organims (taking the word history in its original sense,
and therefore in relation to the faunas and floras of earlier
periods of time) is so intimately connected with geology,
with the order of succession of the superimposed terrestrial
strata, and with the chronometrical annals of the upheaval
of continents and mountains, that it has appeared most ap-
propriate to me, on account of the connection of great and
widely diffused phenomena, to avoid establishing the natural
division of organic and inorganic terrestrial life as the main
element of classification in a work treating of the Cosmos.
We are not here striving to give a mere morphological rep-
resentation of the organic world, but rather to arrive at bold
and comprehensive views of nature, and the forces which
she brings into play.

1.

SIZE, CONFIGURATION, AND DENSITY OP THE EARTH— THE HEAT
IN THE INTERIOR OF THE EARTH, AND ITS DISTRIBUTION.— MAG-
NETIC ACTIVITY, MANIFESTED IN CHANGES OF INCLINATION,
DECLINATION, AND INTENSITY OF THE FORCE UNDER THE IN-
FLUENCE OF THE SUN'S POSITION IN REFERENCE TO THE HEAT
AND RAREFACTION OF THE AIR. — MAGNETIC STORMS. —POLAR
LIGHT.

That which in all languages is comprehended under
etymologically differing symbolical forms by the expression
Mature, and which man, who originally refers every thing
to his own local habitation, has further designated as Ter-
restrial Nature, is the result of the silent co-operation of a
system of active forces, whose existence we can only recog-
nizo by means of that which they move, blend together, and

THE EARTH. 15

again dissever; and which they in part develop into organic
tissues (living organisms), which have the power of repro-
ducing like structures. The appreciation of nature is ex-
cited in the susceptible mind of man through the profound
impression awakened by the manifestation of these forces.
Our attention is at first attracted by the relations of size in
space exhibited by our planet, which seems only like a hand-
ful of conglomerated matter in the immeasurable universe.
A system of co-operating forces, which either tend to com-
bine or separate (through polar influences), shows the de-
pendence of every part of nature upon other parts, both in
the elementary processes (as in the formation of inorganic
substances) and in the production and maintenance of life.
The size and form of the earth, its mass, that is to say, the
quantity of its material parts, which, when compared with
the volume, determines its density, and by means of the lat-
ter, under certain conditions, both the constitution of the in-
terior of the earth and the amount of its attraction, are rela-
tions which stand in a more manifest, and a more mathe-
matically-demonstrable dependence upon one another than
we observe in the case of the above-named vital processes,
in the distribution of heat, in the telluric conditions of elec-
tro-magnetism, or in the chemical metamorphoses of matter.
Conditions, which we are not yet able to determine quanti-
tatively on account of a complication of phenomena, may
nevertheless be present, and may be demonstrated through
inductive reasoning.

Although the two kinds of attraction, namely, that which
acts at perceptible distances, as the force of gravity (the
gravitation of the celestial bodies toward one another), and
that which is manifested at immeasurably small distances,
as molecular or contact-attraction, can not, in the present
condition of science, be reduced to one and the same law,
yet it is not on that account the less credible that capillary
attraction and endosmosis, which is so important in refer-
ence to the ascent of fluids, and in respect to animal and
vegetable physiology, may be quite as much aflected by the
force of gravitation, and its local distribution, as electro-
magnetic processes and the chemical metamorphosis of mat-
ter. To refer to extreme conditions, we may assume that if
our planet had only the mass of the moon, and therefore al-
most six times less intensity of gravity, the meteorological
processes, the climate, the hypsometrical relations of up-
heaved mountain chains, and the physiognomy of the vege-

16 COSMOS.

tation would be quite different from what they now are.
The absolute size of our planet, which we are here consider-
ing, maintains its importance in the collective economy of
nature merely by the relations which it bears to mass and
rotation ; for even in the universe, if the dimensions of the
planets, the quantitative admixture of the bodies which com-
pose them, their velocities and distances from one another,
were all to increase or diminish in one and the same propor-
tion, all the phenomena depending upon relations of gravita-
tion would remain unchanged in this ideal macrocosmos, or

a. Size, Figure, Ellipticity, and Density of the Earth.

(Expansion of the Picture of Nature, Cosmos, vol. i., p. 163-171.)

The earth has been measured and weighed in order to de-
termine its form, density, and mass. The accuracy which
has been incessantly aimed at in these terrestrial determina-
tions has contributed, simultaneously with the Solution of
the problems of astronomy, to improve instruments of meas-
urement and methods of analysis. A very important part
of the process involved in the measurement of a degree is
strictly astronomical, since the altitudes of stars determine
the curvature of the arc, whose length is found by the solu-
tion of a series of triangles. The higher departments of
mathematics have succeeded, from given numerical data, in
solving the difficult problems of the figure of the earth, and
the surface of equilibrium of a fluid homogeneous, or dense
shell-like heterogeneous mass, which rotates uniformly round
a solid axis. Since the time of Newton and Huygens, the
most distinguished geometricians of the eighteenth century

* "The law of reciprocal attraction which acts inversely as the
square of the distance is that of emanations, proceeding from a cen-
tre. It appears to, be the law of all those forces whose action is per-
ceptible at sensible distances, as in the case of electrical and magnet-
ic forces. One of the remarkable properties of this law is that, if the
dimensions of all the bodies in the universe, together with their mu-
tual distances and their velocities, were proportionally increased or
diminished, they would still describe curves precisely similar to those
which they now describe ; so that the universe, after being thus suc-
cessively reduced to the smallest conceivable limits, would still always
present the same appearance to the observer. These appearances are
consequently independent of the dimensions of the universe, as, in vir-
tue of the law of the ratio which exists between force and velocity,
they are independent of absolute movement in space." — Laplace, Ex-
posi^on du Syst. du Monde (Seme ed.), p. 385.

THE FIGURE OP THE EARTH. 17

have devoted themselves to the solution of these problems.
It is well that we should bear in mind that all the great re-
sults which have been attained by intellectual labor and by-
mathematical combinations of ideas, derive their importance
not only from that which they have discovered, and which
has been appropriated by science, but more especially from
the influence which they have exerted on the development
and improvement of analytical methods.

" The geometrical figure of the earth, in contradistinction
to the physical,^ determines the surface which the superficies
of waiter would assume in passing through a net-work of
canals connected with the ocean, and covering and intersect-
ing the earth in every direction. The geometrical surface
intersects the directions of the forces vertically, and these
forces are composed of all the attractions emanating from
the individual particles of the earth, combined with the cen-
trifugal force, which corresponds with its velocity of rota-
tion.! This surface must be generally considered as approx-
imating very closely to an oblate spheroid, for irregularities
in the distribution of the masses in the interior of the earth
will also, where the local density is altered, give rise to ir-
regularity in the geometrical surface, which is the product
of the co-operation of unequally distributed elements. The
physical surface is the direct product of the surface of the
solid and fluid matter on the outer crust of the earth." Al-
though, while it is not improbable, judging from geological
data, that the incidental alterations which are readily brought
about in the fused portions of the interior of the earth, when
they are moved by a change of position of the masses, may
even modify the geometrical surface by producing curvature
of the meridians and parallels in small spaces, and at very
widely separated periods of time ; the physical surface of the
oceanic parts of our globe is periodically subjected to a
change of place in the masses, occasioned by the ebbing and
flowing (or, in other words, the local depression and eleva-
tion) of the fluid element. The inconsiderable amount of

* Gauss, Besthnmung des Breitenunterschiedes zwischen den Stem-
warten von Gottingen und Altona, 1828, s. 73. (These two observato-
ries, by a singular chance, are situated within a few yards of the same
meridian.) ' '

t Bessel, Ueher den Einfluss der Unregelmassigkciten der Figur der
Erde avf geodatische Arheiten und ihre Vergkichuvg mit astronomischen
Bestimniungen, in Schumacher'' s Astron. Nachr., bd. xiv., No. 329, s.
270; and Bessel and Baeyer, Gradmessung in OstpreusseUj 1838, s.
427-442.

18 COSMOS.

the effects of gravity in continental regions may indeed ren-
der a gradual change inappreciable to actual observation ;
and, according to Bessel's calculation, in order to increase
the latitude of a place by a change of only V^, it must be
assumed that there is a transposition in the interior of the
earth of a mass whose weight (its density being assumed to
be that of the mean density of the earth) is that of 7296 ge-
ographical cubic miles. =^ However large the volume of this
transposed mass may appear to us when we compare it with
the volume of Mont Blanc, or Chimborazo, or Kintschind-
jinga, our surprise at the magnitude of the phenomenon soon
diminishes when we remember that our terrestrial spheroid
comprises upward of 1696 hundreds of millions of such cubic
miles.

Three different methods have been attempted, although
with unequal success, for solving the problem of the figure
of the earth, whose connection with the geological question
of the earlier liquid condition of the rotating planetary
bodies was known at the brilliant epoch of Newton, Huy-
gens, and Hooke.f These methods were the geodetico-as-
Ironomical measurement of a degree, pendulum experiments,
and calculations of the inequalities in the latitude and lon-
gitude of the moon. In the application of the first method
two separate processes are required, namely, measurements
of a degree of latitude on the arc of a meridian, and meas-
urements of a degree of longitude on diflferent parallels.

Although seven years have now passed since I brought
forward the results of Bessel's important labors ia reference
to the dimensions of our globe, in my General Delineation
of Nature, his work has not yet been supplanted by any one
of a more comprehensive character, or based upon more re-
cent measurements of a degree. An important addition and
great improvements in this department of inquiry may, how-

* Bessel, Ueher den Einfluss der Verdnderungen des Erdhorpers avf
die Polhohen^ in Lindenau und Bohnenberger, Zeitschiift fur Astrono-
mie, bd. v., 1818, s. 29. "The weight of the earth, expressed in
German pounds=9933xl0-^', and that of the transposed mass=^947
X lO-i*."

t The theoretical labors of that time were followed by those of
Maclaurin, Clairaut, and D'Alembert, by Legendre, and by Laplace.
To this latter period we may add the" theorem advanced by Jacobi, in
1834, that ellipsoids of three unequal axes may, under certain condi-
tions, represent the figures of equilibrium no less than the two pre-
viously-indicated ellipsoids of rotation. — See the treatise of this writer,
whose early death has proved a severe loss to science, in PoggendorfTs
A7imlen der Physik und Chemie, bd. xxxiii., 1834, s. 229-233.

THE SIZE OF THE EARTH. 19

ever, be expected on the completion of the Russian geodetic
measurements, which are now nearly finished, and which, as
they extend almost from the North Cape to the Black Sea,
will aiFord a good basis of comparison for testing the accu--
racy of the results of the Indian survey.

According to the determinations published by Bessel in
the year 1841, the mean value of the dimensions of our
planet was, according to a careful investigation * of ten

* The first accurate comparison of a large number of geodetic meas-
urements (including those made in the elevated plateau of Quito, two
East Indian measurements, together with the French, English, and
recent Lapland observations) was successfully effected by Walbeck, at
Abo, in 1819. He found the mean value for the earth's ellipticity to
^® ^iru^TST' ^"^ ^^^^^ of a meridian degree 57009'758 toises, or 324,'628
feet. Unfortunately his work, entitled De Forma et Magnitudine Tel-
luris, has not been published in a complete form. Excited by the en-
couragement of Gauss, Eduard Schmidt was led to repeat and correct
his results in his admirable Hand-book of Mathematical Geography,
in which he took into account both the higher powers given for the
ellipticity, and the latitudes observed at the intermediate points, as
well as the Hanoverian measurements, and those which had been ex-
tended as far as Formentera by Biot and Arago. The results of this
comparison have appeared in thi-ee forms, after undergoing a gradual
correction, namely, in Gauss's Bestivimung der Breitenunterschiede von
Gottingen und Altona, 1828, s. 82 ; in Eduard Schmidt's Lehrbuch der
Mathem. und Fhys. Geographic, 1829, Th. 1, s. 183, 194-199 ; and, last-
ly, in the preface to the latter work (s. 5). The last result is, for a
meridian degree, 57008-655 toises, or 324,201 feet ; for the ellipticity,
■jjTfT-^Tj-. Bessel's first work of 1830 had been immediately preceded
by Airy's treatise on the Figure of the Earth, in the Encyclopasdia
Metropolitana, ed. of 1849, p. 220^239. (Here the semi-polar axis
was given at 20,853,810 feet=3949-585 miles ; the semi-equatorial
axis at 20,923,713 feet=3962-824 miles; the meridian quadrant at
32,811,980 feet, and the ellipticity at .^tjI.^. The great astronomer
of Konigsberg was uninterruptedly engaged, from 1836 to 1842, in cal-
culations regarding the figure of the earth ; and, as his earlier works
were amended by subsequent corrections, the admixture of results of
investigations at different periods of time has, in many works, proved
a source of great confusion. In numbers, which, from their very na-
ture, are dependent on one another, this admixture is rendered still
more confusing from the erroneous reduction of measurements ; as,
for instance, toises, metres, English feet, and miles of 60 and 69 to
the equatorial degree ; and this is the more to be regretted, since
many works, which have cost a very large amount of time andjabor,
are thus rendered of much less value than they otherwise would be.
In the summer of 1837 Bessel published two treatises, one of which
was devoted to the consideration of the influence of the irregulai-ity
of the earth's figure upon geodetic measurements, and their compai-
ison with astronomical determinations, while the other gave the axes
of the oblate spheroid, which seemed to correspond most closely to
existing measurements of meridian arcs (Schum,, Astr. Nachr., bd.
xiv., No. 329, s. 269, No. 333, s. 345). The results of his calculation

20

COSMOS.

measurements of degrees, as follows : The semi-axis major
of a rotating spheroid, a form that approximates most close-
ly to the irregular figure of our earth, was 3272077-14
toises, or 20,924,774 feet; the semi-axis minor, 3261139-33
toises, or 20,854,821 feet; the length of the earth's quad-
rant, 5131179-81 toises, or 32,811,799 feet; the length of
a mean meridian degree, 57013-109 toises, or 364,596 feet;
the length of a parallel degree at 0° latitude, and conse-
quently that of an equatorial degree, 57108-52 toises, or
365,186 feet; the length of a parallel degree at 45°, 40449-371
toises, or 258,657 feet ; the ellipticity of the earth, .^^-^^^ ;
and the length of a geographical mile, of which sixty go to
an .equatorial degree, 951-8 toises, or 6086-5 feet.

The table on page 21 shows the increase of the length of
the meridian degree from the equator to the pole, as it has
been found from observations, and therefore modified by the
local disturbances of attraction :

were, 3271953'8o4 toises for the semi-axis major ; 3261072-900 toises
for the semi-axis minor ; and for the length of a mean meridian de-
gree— that is to say, for the ninetieth part of the earth's quadrant
(vertically to the equator) — 57011-453 toises. An error of 68 toises,
or 440-8 feet, which was detected by Puissant, in the mode of calcula-
tion that had been adopted, in 1808, by a Commission of the Nation-
al Institute for determining the distance of the parallels of Montjouy,
near Barcelona, and Mola, in Formentera, led Bessel, in the year
1841, to submit his previous calculations regarding the dimensions of
the earth to a new revision. (Schum., Ast?-. Nachr., bd. xix., No. 438,
s. 97-116). This correction yielded for the.length of the earth's quad-
rant 5131179-81 toises, instead of 5130740 toises, which had been ob-
tained in accordance with the first determination of the metre ; and
for the mean length of a meridian degree, 57013-109 toises, which
is about 0-611 of a toise more than a meridian degree at 45° lat.
The numbers given in the text are the result of Bessel's latest calcu-
lations. The length of the meridian quadrant, 5131180 toises, with a
mean error of 255-63 toises, is therefore = 10000856 metres, which
would therefore give 40003423 metres, or 21563-92 geographical miles,
for the entire circumference of the earth. The difference between the
original assumption of the Commission des Poids et 3fesures, according
to which the metre was the forty-millionth part of the earth's circum-
ference, amounts, for the entire circumference, to 3423 metres, or
1756-27 toises, which is almost two geographical miles, or, more ac-
curately speaking, 1-84. According to the earliest determinations,
the length of the metre was determined at 0-5130740 of a toise, while
according to Bessel's last determination it ought to be 0-5131180 of a
toise. The diff'erence for the length of the metre is, therefore, 0-038
of a French line. The metre has, therefore, been established by Bes-
sel as equal to 443-334 French lines, instead of 443-296, which is its
present legal value. (Compare also, on this so-called natural stand-
ard, Faye, Lemons de Cosmographie, 1852, p. 93.)

THE SIZE OF THE EARTH.

21

Countries.

Geographical
Latitude of the

Middle of the
measured Arc.

Length of the
measured Arc.

The Length of a
Degree for the
Latitude of the
Middle Arc as
obtained from

Observations, and
given in Feet.

Observers;

Sweden

Russia

Prussia

Denmark

Hanover

England

France

North America
East Indies . . ,

Quito (s. t.) ..

Cape of Good
Hope (s. L.).

(66° iiO' 10"
ioG 19 37
56 3 55-5
54 58 26 0
54 8 13 -T
52 32 16-6
|'52 35 45-0
\52 2 19-4

44 51 2-5

39 12 0

fl6 8 21-5

\12 32 20-8

1 31 0-4

(33 18 30
135 43 20

1° 37' l;>"-6

0 57 30-4
8 2 289

1 30 29-0

1 31 53-3

2 0 57-4

3 57 13-1

2 50 23-5

12 22 12-T

1 28 45-0

15 57 40-7

1 34 50-4

3 T 3-5

1 13 17-5
3 34 34 7

3a54T3-4
365S82-1
365368-0
3t;5396 0
365087-0
3U5400-0
365071-2 \
364951-1 /

3G4671-5 1

S637S5-1
3.3044-0
36295!) -6

S63625-2 1

364819-2
304160-0

Svanberg.
.Maupertiiis.
Struve, Tenner.
Bessel, Baeyer.
Schumacher.
Gauss.

Roy, Mudge, Kater.

Delambre, Mechain,

Biot, Arago.
Mason, Dixon.
Lanibton, Everest.
Lambton.
La (Jondamine,

Bouguer.
r,acaille.
Maclear.

The determination of the figure of the earth by the meas-
urement of degrees of longitude on different parallels requires
very great accuracy in fixing the longitudes of diflferent places.
Cassini de Thury and Lacaille employed, in 1740, powder
signals to determine a perpendicular line at the meridian of
Paris. In more recent times, the great trigonometrical sur-
vey of England has determined, by the help of far better in-
struments and with greater accuracy, the lengths of the arcs
of parallels and the differences of the meridians between
Beachy Head and Dunnose, as well as between Dover and
Falmouth. These determinations were, however, only made
for differences of longitude of 1° 26^ and 6° 22^* By far
the most considerable of these surveys is the one that was
carried on between the meridians of Marennes, on the west-
ern coast of France, and Fiume. It extends over the west-
ern chain of the Alps, and the plains of Milan and Padua,
in a direct distance of 15° 32^ 27^^, and was executed under
the direction of Brousseaud and Largeteau, Plana and Car-
lini, almost entirely under the so-called mean parallel of 45°.
The numerous pendulum experiments which have been con-
ducted in the neighborhood of mountain chains have con-
firmed in the most remarkable manner the previously-recog-
nized influences of those local attractions which were inferred
from the comparison of astronomical latitudes with the re^
suits of geodetic measurements, f

* Airy, Figure of the Earth, in the Encycl Metrop., 1849, p. 214s
216.

t Biot, Astr. Physique, t. ii., p. 482, and t. iii., p. 482. A very ac^

22

COSMOS.

In addition to the two secondary methods for the direct
measurement of a degree on meridian and parallel arcs, we
have still to refer to a purely astronomical determination of
the figure of the earth. This is based upon the action which
the earth exerts upon the motion of the moon, or, in other
words, upon the inequalities in lunar longitudes and latitudes.
Laplace, who was the first to discover the cause of these in-
equalities, has also taught us their application by ingenious-
ly showing how they afford the great advantage which indi-
vidual measurements of a degree and pendulum experiments
are incapable of yielding, namely, that of showing in one
single result the mean figure of the earth.* We would here,
again, refer to the happy expression of the discoverer of this
method, " that an astronomer, without leaving his observa-
tory, may discover the individual form of the earth in which
he dwells, from the motion of one of the heavenly bodies."
After his last revision of the inequalities in the longitude
and latitude of our satellite, and by the aid of several thou-
sand observations of Biirg, Bouvard, and Burckhardt,t La-
place found, by means of his lunar method, a compression

curate geodetical measurement, which is the more important from its
serving as a comparison of the levels of the Mediterranean and At-
lantic, has been made on the parallel of the chain of the Pyrenees by
Coraboeuf, Delcros, and Peytier.

* Cosmos, vol. i., p. 168. "It is very remarkable that an astrono-
mer, without leaving his observatory, may, merely by comparing his
observations with analytical results, not only be enabled to determine
with exactness the size and degree of ellipticity of the earth, but also
its distance from the sun and moon — results that otherwise could only
be arrived at by long and arduous expeditions to the most remote parts
of both hemispheres. The moon may, therefore, by the observation
of its movements, render appreciable to the higher departments of as-
tronomy the ellipticity of the earth, as it taught the early astronomers
the rotundity of our earth by means of its eclipses." (Laplace, Expos,
du Syst. du Monde, p. 230.) We have already in Cosmos, vol. iv., p.
145-146, made mention of an almost analogous optical method sug-
gested by Arago, and based upon the observation that the intensity
of the ash-colored light — that is to say, the terrestrial light in the moon
— might afford us some information in reference to the transparency
of our entire atmosphere. Compare also Airy, in the Encycl. Meti-op.,
p. 189, 236, on the deternaination of the earth's ellipticity by means
of the motions of the moon, as well as at p. 231-235, on the infer-
enc.es which he draws regarding the figure of the earth from preces-
sion and nutation. According to Biot's investigations, the latter de-
termination would only give, for the earth's ellipticity, limiting and
widely differing values (^^ and -j^). Astron. Physique, 3^me ed.,
t. ii., 1844, p. 463.

t Laplace, Mecanique Celeste, ed. de 1846, t. v., p. 16, 53.

THE FIGURE OF THE EARTH. 23

amounting to -g-^, which is very nearly equal to that yield-
ed by the measurements of a degree of latitude (^^).

The vibrations of the pendulum yield a third means of de-
termining the figure of the earth (or, in other words, the re-
lation of the major to the minor axis, on the supposition of
our planet being of a spheroidal form), by the elucidation of
the law according to which gravity increases from the equa-
tor toward the pole. The Arabian astronomers, and more
especially Ebn-Junis, at the close of the tenth century, and
during the brilliant epoch of the Abbassidian Califs,* first
employed these vibrations for the determination of time, and,
after a neglect of six hundred years, the same method was
again adopted by Galileo, and Father Riccioli, at Bologna. t
The pendulum, in conjunction with a system of wheels used
to regulate the clocks (which were first employed in the im-
perfect experiments of Sanctorius at Padau, in 1612, and
then in the more perfect observations of Huygens in 1656),
gave the first material proof of the different intensity of gravi-
ty at different latitudes in Eicher's comparison of the beats of
the same astronomical clock at Paris and Cayenne, in 1672.
Picard was, indeed, engaged in the equipment of this import-
ant voyage, but he does not on that account assume to him-
self the merit of its first suggestion. Richer left Paris in
October, 1671 ; and Picard, in the description of his meas-
urement of a degree of latitude, which appeared in the same
year,{ merely refers to "a conjecture which was advanced

* Cosmos, vol. i., p. 166. Edward Bernard, an Englishman, was
the first who recognized the application of the isochronism of pendu-
lum-oscillations in the writings of the Arabian astronomers. (See
his letter, dated Oxford, April, 1683, and addressed to Dr. Robert
Huntington, in Dublin. Philos. Transac, vol. xii., p. 567.)

+ Fr^ret de f Etude de la Philosophie Ancienne in the Mem. de VAcad.
des Inscr., t. xviii. (1753), p. 100.

X Picard, Mesure de h Terre, 1671, Art. 4. It is scarcely probable
that the conjecture which was advanced in the Paris Academy even
before the year 1671, to the effect that the intensity of gravity varies
with the latitude (Lalande, Astronomie, t. iii., p. 20 § 2668), 'should
have been made by the illustrious Huygens, who had certainly pre-
sented his Discours sur la Cause de la Gravite to the Academy in the
course of the year 1669. There is no mention made in this treatise
of the shortening of the seconds-pendulum, which was being observed
by Richer at Cayenne, although a reference to it occurs in the supple-
ments to this work (one of which must have been completed after the
publication of Newton's Principia, and consequently later than 1687).
Huygens writes as follows: "Maxima pars hujus libelli scripta est,
cum Lutetiaj degerem (to 1681) ad eum usque locum, ubi de altera-
tione, quae pendulis accidit e motu Terrae." See also the explanation

24 COSMOS.

by one of the members, at a meeting of the Academy, accord-
ing to which the weight of a body must be less at the equa-
tor than at the pole, in consequence of the rotation of the
earth." He adds, doubtfully, that although it would appear,
from certain experiments made in London, Lyons, and Bo-
logna, as if the seconds-pendulum must be shortened the
nearer we approach to the equator ; yet, on the other hand,
he was not sufficiently convinced of the accuracy of the meas-
urements adduced, because at the Hague, notwithstanding
its more northern latitude, the pendulum lengths were found
to be precisely the same as at Paris. The periods at which
Newton first became acquainted with the important pendu-
lum results that had been obtained by Richer as early as
1672, although they were not printed until 1679, and at
which he first heard of the discovery that had been made by
Cassini, before the year 1666, of the compression of Jupiter's
disk, have unfortunately not been recorded with the same
exactness as the fact of his very tardy acquaintance with

which I have given in Cosmo.% vol. ii,, p. 351. The observations made
by Richer at Cayenne were not published until 1679, as I have ailready
observed in the text, and therefore not until fully six years after his
return, and, what is more remarkable, the annals of the Academie des
Inscriptions contain no notice during this long period of Richer's im-
portant double observations of the pendulum clock and of the simple
seconds-pendulum. We do not know the time when Newton first be-
came acquainted with Richer's results, although his own earliest the-
oretical speculations regarding the figure of the earth date farther back
than the year 1665. It would appear that Newton did not become
acquainted until 1682 with Picard's geodetic measurement, which had
been published in 1671, and even then "he accidentally heard of it at
a meeting of the Royal Society, which he was attending." His knowl-
edge of this fact, as Sir David Brewster has shown {Memoirs of Sir I.
Newton, vol. i., p. 291), exerted a very important influence on his de-
termination of the earth's diameter, and of the relation which the fall
of a body upon our planet bears to the force which retains the moon
in its orbit. Newton's views may have been similarly influenced by
the knowledge of the spheroidal form of Jupiter, which had been as-
certained by Cassini prior to 1666, but was first described in 1691, in
the Mhnoires de l' Academie des Sciences, t. ii., p. 108. Could Newton
ha/e learned any thing of a much earlier pulalication, of which some
of the sheets were seen by Lalande in the possession of Maraldi?
(Compare Lalande, Astr., t. iii., p. 335, § 3345, with Brewster, Mem-
oirs of Sir I. Newipn, vol. i., p. 322, and Cosmos, vol. i., p. 165.) Amid
the simultaneous labors of Newton, Huygens, Picard, and Cassini, it
is often very difiicult to arrive, with any certainty, at a just apprecia-
tion of the diffusion of scientific knowledge, owing to the tardiness
with which men at that day made known the result of their observa-
tions, the publication of which was, moreover, frequently delayed by
accidental circumstances.

THE FIGURE OF THE EARTH. 25

Picard's measurement of a degree. In an age so remarkable
for the successful emulation that distinguished the cultivators
of science, and when theoretical views led to the prosecution
of observations which, by their results, reacted in their turn
upon theory, it is of great interest to the history of the math-
ematical establishment of physical astronomy that individual
epochs should be determined with accuracy.

Although direct measurements of meridian and parallel
degrees (the former especially in the case of the French meas-
urement of a degree* between the latitudes 44° 42^ and 47°
30'', and the latter by the comparison of points lying to the
east and west of the Italian and Maritime Alps)t exhibit
great deviations from the mean ellipsoidal figure of the earth,
the variations in the amount of ellipticity given by pendulum
lengths (taken at different geographical points and in differ-
ent groups) are very much more striking. The determina-
tion of the figure of the earth obtained from the increase or
decrease of gravity (intensity of local attraction), assumes
that gravity af the surface of our rotating spheroid must have
remained the same as it was at the time of our earth's con-
solidation from a fluid state, and that no later alterations can
have taken place in its density.f Notwithstanding the great
improvements which have been made in the instruments and
methods of measurement by Borda, Kater, and Bessel, there
are at present in both lierhispheres, from Spitzbergen in 79°
50^ north latitude, to the Falkland Islands, in 51° 35'' south
latitude, where Freycinet, Duperrey, and Sir James Ross
su€cessi^ly made their observations, only from 65 to 70 ir-
regularl^cattered points § at which the length of the simple

* Delambre, Base du Syst, M^trique, t. ill., p. 548.

t Cosmos, vol. i., p. 167. Plana, Operations Geodesiques et Astrono-
miques pour la Mesure d'un Arc du Parallele Moyen, t. ii., p. 847 ;
Carlini in the Effemeridi Astronomiche di Milano per tanno 1842, p. 57.

X Compare Biot, Astronomie Physique, t. ii., 1844, p. 464, with Cos-
m»s, vol. i., p. 168, and vol. iv., p. 105, where I have considered the
difficulties presented by a comparison of the periods of rotation of
planets, and their observed compression. Schubert (^Astron., Th. iii.,
§ 316) has also drawn attention to this difficulty; and Bessel, in his
treatise On Mass and Weight, says expressly that the supposition of
the invariability of gravity at any one point of observation has been
rendered somewhat uncertain by the recent experiments made on the
slow upheaval of large portions of the earth's surface.

§ Airy, in his admirable treatise on the Figure of the Earth {Encycl.
MetropoL, 1849, p. 229), reckoned fifty different stations where trust-
worthy results had been obtained up to the year 1830, and fourteen
others (those of Bouguer, Legentil, Lacaille, Maupertuis, and Lft

Vol. v.— B

26 COSMOS.

pendulum has been determined with as much accuracy as the
position of the place in respect to its latitude, longitude, and
elevation above the level of the sea.

The pendulum experiments made by the French astrono-
mers on the measured part of a meridian arc, and the observ-
ations of Captain Kater in the trigonometrical survey of
Great Britain, concurred in showing that the results do not
individually admit of being referred to a variation of gravity
proportional to the square of the sine of the latitude. On
this account the English government determined, at the sug-
gestion of the Vice-president of the Eoyal Society, Davies
Gilbert, to fit out a scientific expedition, which was intrust-
ed to my friend Edward Sabine, who had accompanied Cap-
tain Parry on his first polar voyage in the capacity of as-
tronomer. In the course of this voyage, which was con-
tinued through the years 1822 and 1823, he coasted along
the western shores of Africa, from Sierra Leone to the Isl-
and of St. Thomas, near the equator, then by Ascension to
South America, from Bahia to the mouth of tne Orinoco, on
his way to the West Indies and the New England States,
after which he penetrated into the Arctic regions as far as
Spitzbergen, and a hitherto unexplored and ice-bound por-
tion of East Greenland (74° 3?''). This brilliant and ably-
conducted expedition had the advantage of being mainly di-
rected to one sole object of investigation, and of embracing
points which are separated from one another by 93° of lati-
tude.

The field of observation in the French expediti||| for the
measurements of degrees was more remote from tne^quinoc-
tial and arctic zones ; but it had the great advantage of pre-
senting a linear series of points of observation, and of aflPord-
ing direct means of comparison with the partial curvature
of the arcs obtained by geodetico-astronomical observations.
Biot, in 1824, carried the line of pendulum measurements
from Formentera (38° 39^ 56^^), where he had already made
observations conjointly with Arago and Chaix, as far as
Unst, the most northerly of the Shetland Islands (60° 45^
25^^), and with Mathieu he extended it to the parallels of
Bordeaux, Figeac, and Padua, as far as Fiume.* These

Croyere), which, however, do not bear comparison with the former in
point of accuracy.

* Biot and Arago, Recueil dOhserv. Gdodisiques et Astronomiques,
1821, p. 526-540 ; and Biot, TraiU d'Astr. Physique, t. ii., 1844, p.
465-473.

THE FIGURE OF THE EARTH. 27

pendulum results, when compared with those of Sabine, cer-
tainly give Tj^ for the compression of the whole northern
quadrant ; but when separated into two halves, they yield a
still more varying result, giving -^-^ from the equator to
45°, and ^y- from 45° to the pole.* It has been shown in
many instances, and in both hemispheres, that there is an
appreciable influence exerted by surrounding denser rocks
(basalt, green-stone, diorite, and melaphyre, in opposition to
specifically lighter secondary and tertiary formations), in the
same manner as volcanic islandsf influence gravity and aug-
ment its intensity. Many of the anomalies which presented
themselves in these observations do not, however, admit of
being explained by any visible geological characters of the
soil.

For the southern hemisphere we possess a small number
of admirable, but very widely-diffused observations, made by
Freycinet, Duperrey, Fallows, Liitke, Brisbane, and Riimker.

* Op. a7.,p. 488. Sabine {Exper.for determining the Variation in
the Length of the Pendulum vibrating Seconds, 1825, p. 352) finds -^^.-^
from all the thirteen stations of his pendulum expedition, notwith-
standing their great distances from one another in the northern hem-
isphere ; and from these, increased by all the pendulum stations of the
British survey and of the French geodetic measurement from Formen-
tera to Dunkirk, comprising, therefore, in all a comparison of twenty-
five points of observation, he again found ^-jV.t,-. It is still more strik-
ing, as was already observed by Admiral Liitke, that far to the west
of the Atlantic region, in the meridians of Petropawlowski and New
Archangel, the pendulum lengths yield a much greater ellipticity,
namely, -^i-y. As the previously applied theory of the influence of the
air surrounding the pendulum led to an error in the calculation, and
had rendered a correction necessary as early as 1786 (when a some-
what obscure one was given by the Chevalier de Buat;, on account of
the difference in the loss of weight of solid bodies, when they are either
at rest in a fluid, or impelled in a vibratory motion, Bessel, with his
usual analytical clearness, laid down the following axiom in his Unter-
suchungen uber die Lange des einfachen Secundenpendels, s. 32, G3, 126-
129 : " When a body is moving in a fluid (the atmosphere), the latter
belongs with it to the moved system, and the moving force must be
distributed not only over the particles of the solid moved body, but
also over all the moved particles of the fluid." On the experiments
of Sabine and Baily, which originated in Bessel's practically import-
ant pendulum correction (reduction to a vacuum), see John Herschel
in the Memoir of Francis Baily, 1845, p. 17-21.

t Cosmos, vol. i., p. 167. Compare, for the phenomena occurring
in islands, Sabine, Pend. Exper., 1825, p. 237 ; and Lutke, Obs. du Pen-
dule invariable, exdcut^es de 1826-1829, p. 241. This work contains a
remarkable table, p. 239, on the nature of the rocks occurring at 16
pendulum stations, from Melville Island (79° 50' N. lat.) to Valparai-
so (32° 2' S. lat.).

28 COSMOS.

These observations have confirmed a fact which had been
strikingly demonstrated in the northern hemisphere, namely,
that the intensity of gravity is not the same for all places
having the same latitude, and that the increase of gravity
from the equator toward the poles appears to be subjected
to different laws under different meridians. Although the
pendulum measurements made by Lacaille at the Cape of
Good Hope, and those conducted in the Spanish circumnav-
igating expedition by Malaspini, may have led to the belief
that the southern hemisphere is, in general, much more com-
pressed than the northern, comparisons made between the
Falkland Islands and New Holland on the one hand, and
New York, Dunkirk, and Barcelona on the other, have,
however, by their more exact results, shown that the con-
trary is the case, as I have already elsewhere indicated.*

From the above data it follows that the pendulum (al-
though it is by no means an unimportant instrument in
geognostic observations, being as it were a sort of plummet
cast into the deep and unseen strata of the earth) does not
determine the form of our planet with the same exactitude

* Cosmos, vol. i., p. 169. Eduard Schmidt (Mathem. und Phys. Geo-
graphie, Th. i., s. 394) has separated from a large number of the pen-
Tiulum observations which were made on board the corvettes Descubi-
erta and Atrevida, under the command of Malaspina, those thirteen
stations which belong to the southern hemisphere, from which he ob-
tained a mean compression of -^-g^.-g^. Mathieu obtained tj-;^.^ from
a comparison of Lacaille's observations at the Cape of Good Hope and
the Isle of France with Paris, but the instruments of measurement
used at that day did not afford the same certainty as we now obtain
by the appliances of Borda and Kater, and the more modern methods
of observation. The present would seem a fitting place to notice the
beautiful experiments of Foucault, which afford so high a proof of the
ingenuity of the inventor, and by which we obtain ocular evidence of
the rotation of the earth on its axis by means of the pendulum, whose
plane of vibration slowly rotates from east to west. ( Comphs rendus
de VAcad. des Sc; Seance du 3 Fevrier, 1851, t. xxxii., p. 135.) Ex-
periments for noticing the deviation toward the east in observations
of falling bodies, dropped from church towers or into mines, as sug-
gested by Benzenberg and Reich, require a very great height, while
Foucault's apparatus makes the effects of the earth's rotation percep-
tible with a pendulum only six feet long. We must not confound the .
phenomena which may be explained by rotation (as, for instance,
■Richer's clock experiments at Cayenne, diurnal aberration, the devia-
tion of projectiles, trade-winds, etc.) with those that may at any time
be produced by Foucault's apparatus, and of which the members of
the Academia del Cimento appear to have had some idea, although they
did not farther develop it (Antinori, in the Comptes rendus^ t. xxxii.,
p. 635).

THE FIGURE OF THE EARTH. 29

as the measurement of a degree or the movements of our
satellite. The concentric, elliptical, and individually hotno-
geneous strata, which increase in density according to certain
functions of distance from the surface toward the centre of
the earth, may give rise to local fluctuations in the intensity
of gravity at individual^oints of the earth's surface, which
differ according to the character, position, and density of the
several points. If the conditions which produce these devi-
ations are much more recent than the consolidation of the
outer crust, the figure of the surface can not be assumed to
be locally modified by the internal motion of the fused masses.
The difference of the results of pendulum measurements is,
however, much too great to be ascribed at the present day
to errors of observation. Even where a coincidence in the
results, or an obvious regularity, has been discovered by
the various grouping and combination of the points of ob-
servation, the pendulum always gives a greater §]lipticity
(varying between the limits -^^5 ^^^ niu) ^^^^^ could have
been deduced from the measurements of a degree.

If we take the ellipticity which, in accordance with Bes-
sel's last determination, is now generally adopted, namely,
2T¥-T5^> we shall find that the bulging* at the equator

* In Grecian antiquity two regions of the earth were designated as
being characterized, in accordance with the prevalent opinions of the
time, by remarkable protuberances of the surface, namely, the high
north of Asia and the land lying under the equator. *'The high and
naked Scythian plains," says Hippocrates (De AereetAgvis, § xix., p.
72, Littre), "without being crowned by mountains, stretch far upward
to the meridian of the Bear." A similar opinion had previously been
ascribed to Empedocles (Phit., De Plac. Philos., ii., 3). Aristotle {Me-
teor., i., 1 a 15, p. 66, Ideler) says that the older meteorologists, ac-
cording to whose opinions the sun "did not go under the earth, but
passed round it," considered that the protuberances of the earth to-
ward the north were the cause of the disappearance of the sun, or of
the production of night. And in the compilation of the Problems
(xxvi., 15, p. 941, Bekker), the cold of the north wind was ascribed to
the elevation of the soil in this region of the earth, and in all these
passages there is no reference to mountains, but merely to a bulging
of the earth into elevated plateaux. I havie already elsewhere shown
{Asie Centrale, t. i., p. 58) that Strabo, who alone makes use of the
very characteristic word dpoiridia, says that the difference of climate
which arises from geographical position must every where be distin-
guished fftm that which we ascribe to elevation above the sea, in
Armenia (xi., p. 522, Casaub.), in Lycaonia, which is inhabited by
wild asses (xii., p. 568), and in Upper India, in the auriferous country
of the Derdi (xv., p, 706). "Even in southern parts of the world,"
says the geographer of Amasia, " every high district, if it be also a
plain, is cold" (ii., p. 73). Eratosthenes andPolybius ascribe the very

30 COSMOS.

amounts to about 645,457 feet ; about lli, or, more accu-
rately, 1 1*492 geographical miles. As a comparison has

moderate temperature which prevails under the equator not only to the
more rapid transit of the sun (Geminus, Elem. Astron., c. 13; Cleom,,
Ci/cl. Theor., 1, 6), but more especially to the bulging of the earth (see
ray Exameri Cnt. de la Geogr., t. iii., p. 19i-152). Both maintain, ac-
cording to the testimony of Strabo (ii., p. 97), " that the district lying
immediately below the equator is the highest, on which account much
rain falls there, in consequence of the very large accumulation of
northern clouds at the period when those winds prevail, which change
with the season of the year." Of these two opinions regarding the
elevation of the land in Northern Asia (the Scythian Europe of Herodo-
tus) and in the equatorial zone, the former of the two, with the perti-
nacity characteristic of error, has kept its ground for nearly two thou-
sand years, and has given occasion to the geological myth of an un-
interrupted plateau in the Tartar district lying to the north of the
Himalayas, while the other opinion could only be justified in reference
to a portion of Asia, lying beyond the tropical zone, and consequently
applies only to the colossal, " elevated or mountain plateau, Meru,"
which is celebrated in the most ancient and noblest memorials of In-
dian poetry. (See Wilson's Did. Sanscrit and English, 1832, p. 674,
where the word Meru is explained to signify an elevated plateau.) I
have thought it necessary to enter thus circumstantially into this ques-
tion, in order that I might refute the hypothesis of the intellectual
Freret, who, without indicating any passages from Greek writers, and
merely alluding to one which seemed to treat of tropical rain, inter-
prets the opinion advanced regarding bulgings of the soil as having
reference to compression or elongation at the poles. In the M^m. de
VAcad. des Inscriptions, t. xviii., 1753, p. 112, Freret expresses him-
self as follows : " To explain the rains which prevailed in those equi-
noctial regions, which the conquests of Alexander first made known,
it was supposed that there were currents which drove the clouds from
the poles toward the equator, where, in default of mountains to stop
their progress, they were arrested by the general elevation of the soil,
whose surface at the equator is farther removed from the centre than
under the poles. Some physicists have ascribed to the globe the figure
of a spheroid, which bulges at the equator and is flattened toward the
poles; while on the contrary, in the opinion of those of the ancients
who believed that the earth was elongated toward the poles, the polar
regions are farther removed than the equatorial zone from the centre
of the earth." I can find no evidence in the works of the ancients to
justify these assertions. In the third section of the first book of Strabo
(p. 48, Casaub.), it is expressly stated that, " after Eratosthenes has
observed that the whole earth is spherical, although not like a sphere
that has been made by a turning- lathe (an expression that is borrowed
from Herodotus, iv., 36), and exhibits many deviations from this form,
he adduces numerous modifications of shape which have been produced
by the action of water and fire, by earthquakes, subterranean currents
of wind (elastic vapors ?), and other causes of the same kind, which,
however, are not given in the order of their occurrence, for the rotun-
dity of the entire earth results from the co-ordination of the whole, such
modifications in no degree affecting the general form of our earth, the
lesser vanishing in the greater." Subsequently we read, also in Gros-

THE FIGURE OF THE EARTH, 31

very frequently been made from the earliest times of astro-
nomical inquiry between this swelling or convex elevation
of the earth's surface and carefully measured mountain
masses, I will select as objects of comparison the highest of
the known peaks of the Himalayas, namely, that of Kin-
tschindjinga, which was fixed by Colonel Waugh at 28,174
feet, and that portion of the elevated plateau of Thibet which
is nearest to the sacred lakes of Eakas-Tal and Manassa-
rova, and which, according to Lieutenant Henry Strachey,
is situated at the mean height of 15,347 feet. The bulging
of our planet at the equatorial zone is, therefore, not quite

kurd's admirable translation, " that the earth, together with the sea, is
spherical, the two constituting one and the same surface. The projec-
tion of the land, which is inconsiderable and may remain unnoticed, is
lost in such magnitudes, so that in these cases we are unable to determ-
ine its spherical form with the same accuracy as in the case of a sphere
made by a turning-lathe, or as well as the sculptor, who judges from
his conceptions of form, for here we are obliged to determine by phys-
ical and less delicate perception." (Strabo, ii., p. 112.) " The world
is at once a work of nature and of providence — a work of nature, inas-
much as all things tend toward one point, the centre of the whole, round
which they group themselves, the less dense element (water) containing
the denser (earth)." (Strabo, xvii., p. 809.) Wherever we find the fig-
ure of the earth described by the Greeks, it is compared (Cleom., CijcL
Theor., i., 8, p. 51) with a flat or centrally depressed disk, a cylinder
(A^aximander), a cube or pyramid ; and, lastly, we find it generally held
to be a sphere, notwithstanding the long contest of the Epicureans, who
denied the tendency of attraction toward the centime. The idea of com-
pression does not seem to have presented itself to their imagination.
The elongated earth of Democritus was only the disk of Thales length-
ened in one direction. The drum-like form, to cxviia, rvinzavosLdeg,
which seems more especially to have emanated from Leucippus (Plut.,
De Plac.Philos., iii.,10; Galen. Hist. Phil, cap. 21 ; Aristotle, De CoeJo,
ii., 13, p. 293 Bekker), appears to have been founded upon the idea of
a hemisphere with a flat basis, which probably represented the equator,
while the curvature was regarded as the oiKovjuevr]. A passage in Pliny,
regarding Pearls (xi., 54), elucidates this form, while Aristotle merely
compares the segments of the sphere with the drum {MeteoroL, ii., 5,.
a 10, Ideler, t. i., p. 563), as we also find from the commentary of
Olympiodorus (Ideler, t. i., p. 301). I have here purposely avoided re-
ferring to two passages, which are well known to me, in Agathemerus
(DeGeographia, lib. i., cap. 1, p. 2, Hudson), and inEusebius {Evangel.
Proeparat., t. iv., p. 125, ed. Gaisford, 1843), because they prove with
what inaccuracy later writers have often ascribed to the ancients views
which were totally foreign to them. According to these versions,
"Eudoxus gave for the length and breadth of the earth's disk values
which stood in relation to one another as 1 to 2 ; the same is said in
reference to Dicsearchus, the pupil of Aristotle, who, however, advanced
his own special proofs of the spherical form of the earth (Marcian, Ca~
pella, lib. vi., p. 192). Hipparchus regarded the earth as TpaTve^oeidj^Ct
and Thales held it to be a sphere!"

32 COSMOS.

three times as great as the elevation of the highest ol our
mountains above the sea's level, but it is almost five times
as great as that of the eastern plateau of Thibet.

We ought here to observe that the results of the earth's
compression, which have been obtained by mere measure-
ments of a degree, or by combinations of the former with
pendulum measurements, show far less* considerable differ-
ences in the amount of the equinoctial bulging than we
should have been disposed at first sight to conclude from the
fractional numbers. The difference of the polar compres-
sions {j]-^ and -ji^) amounts to only about 7000 feet in the
difference of the major and minor axes, basing the calcula-
tion on both extreme numerical limits ; and this is not twice
the elevation of the small mountains of the Brocken and of
Vesuvius ; the difference being only about one tenth of the
bulging which would be yielded by a polar compression

As soon as it had been ascertained by more accurate meas-
urements of a degree, made at very different latitudes, that

* It has often seemed to me as if the amount of the compression of
the earth was regarded as somewhat doubtful merely from our wish
to attain an unnecessary degree of acciuacy. If we take the values
of the compression at -j^iT' 'S'hj^ ¥517' TSTT' ^® ^"'^ *^^^* *^® difference
of both radii is equal to 10,554, 10,905, 11,281, 11,684 toises^ or
67,488, 69,554, 73,137, 74,714 feet. The fluctuation of 30 units in
the denominator produces only a fluctuation of 1130 toises, or 7126
feet, in the polar radius, an amount which, when compared with the
visible inequalities of the earth's surface, appears so very inconsid-
erable, that I am often surprised to find that the experiments coin-
cide within such closely approximating limits. Individual observa-
tions scattered over wide surfaces will indeed teach us little more than
what we already know, but it would be of considerable importance to
connect together all the measurements that have been made over the
entire surface of Europe, including in this calculation all astronomic-
ally determined points. (Bessel, in a letter addressed to myself, De-
cember, 1828.) Even if this plan were carried out, we should then
only know the form of that portion of the earth, which may be re-
garded as a peninsular projection, extending westward, about sixty-
six and a half degrees from the great Asiatic Continent. The steppes
of Northern Asia, even the middle Kirghis steppe, a considerable por-
tion of which I have myself seen, are often interspersed with hills,
and in respect to uninterrupted levels, can not be compared with the
Pampas of Buenos Ayres, or the Llanos of Venezuela. The latter,
which are far removed from all mountain chains, and consist immedi-
ately below the surface of secontlary and tertiary strata, having a very
uniform and low degree of density, might, by differences in the results
of pendulum vibrations, yield very decisive conclusions in reference
to the local constitution of the deep internal strata of the earth. —
Compare my Views of Nature, p. 2-8, 29-32.

THE FIGURE OF THE EARTH. 33

the earth could not be uniformly dense in its 'interior (be-
cause the results showed that the compression was very
much less than had been assumed by Newton (yto)? and
much greater than was supposed by Huygens (57^), who
considered that all forces of attraction were combined in the
centre of the earth),- the connection between the amount of
compression and the law of density in the interior of our
earth necessarily became a very important object of analyt-
ical calculation. Theoretical speculations regarding gravity
very early led to the consideration of the attraction of large
mountain masses, which rise freely and precipitously into the
atmosphere from the dried surface of our planet. Newton,
in his Treatise of the System of the World in a Popular Way,
1728, endeavored to determine what amount of deviation
from the perpendicular direction the pendulum would experi-
ence from a mountain 2665 feet in height and 5330 feet in
diameter. This consideration very probably gave occasion
to the unsatisfactory experiments which were made by Bou-
guer on Chimborazo,* by Maskelyne and Hutton on She-
hallien, near Blair- Athol, in Perthshire ; to the comparison
of pendulum lengths on a plain lying at an elevation of 6000

* Bouguer, who had been induced by La Condamine to institute
experiments on the deviation of the plummet near the mountain of
Chimborazo, does not allude, in his Figure de la Terre, p. 364-394, to
Newton's proposition. Unfortunately the most skillful of the two trav-
elers did not observe on the east and western sides of the colossal
mountain, having limited his experiments (December, 1738) to two
stations lying on the same side of Chimborazo, first in a southerly di-
rection 61° 30' West, about 4572 toises, or 29,326 feet, from the centre
of the mountain, and then to the South 16° West (distance 1753 toises,
or 1 1,210 feet). The first of these stations lay in a district with which
I am well acquainted, and probably at the same elevation as the small
alpine lake of Yana-cocha, and the other in the pumice-stone plain of
the Arenal (La Condamine, Voyage d I'Equateur, p, 68-70). The
deviation yielded by the altitudes of the stars was, contrary to all ex-
pectation, only 7'' "5, which was ascribed by the observers themselves
to the difficulty of making observations so immediately in the vicinity
of the limit of perpetual snow, to the want of accuracy in their instru-
ments, and, above all, to the great cavities which were conjectured to
exist within this colossal trachytic mountain. I have already ex-
pressed many doubts, based upon geological grounds, as to this as-
sumption of very large cavities, and of the very inconsiderable mass
of the trachytic dome of Chimborazo. South-southeast of this mount-*
ain, near the Indian village of Calpi, lies the volcanic cone of Yana-
urcu, which I carefully investigated in concert with Bonpland, and
which is certainly of more recent origin than the elevation of the
great dome-shaped trachytic mountain, in which neither I nor Bous-
singault could discover auy thing analogous to a crater. See the
Ascent of Chimborazo in my Kleine Schriften, bd. i., s. 138.

F> 2

34 COSMOS.

feet and at the level of the sea (as, for instance, Carlini's
observations at the Hospice of Mont Cenis, and Biot and
Mathieu's at Bordeaux) ; and, lastly, to the delicate and
thoroughly decisive experiments undertaken in 1837 by
Reich and Bailey with the ingeniously constructed torsion-
balance which was invented by John Mitchell, and subse-
quently given to Cavendish by Wollaston.* The three
modes of determining the density of our planet (by vicinity
to a mountain mass, elevation of a mountainous plateau,
and the balance) have already been so circumstantially de-
tailed in a former part of the Cosmos (vol. i., p. 157), that it
only remains for us to notice the experiments given in
Reich's new treatise, and prosecuted by that indefatigable
observer during the interval between the years 1847 and
1850.t The whole may, in accordance with the present
state of our knowledge, be arranged in the following man-

Shehallien, according to the mean of the maximum 4:"867 and

the minimum 4*559, as found by Playfair 4-713

Mont Cenis, observations of Carlini, with the correction of

Giulio 4-950

* Baily, Exper. with the Torsion Rod for determining the mean Density
of the Earth, 1843, p. 6; John Herschel, Memoirof Francis Baily, 1845,
p. 24.

t Keich, Neue Versuche mit der Drehwage, in the Ahhandl. der ma-
thern. physischen Classe der Kbn. Sdchsischen Gesellschaft der Wissen-
schaften zu Leipzig, 1852, bd. i., s. 405, 418. The most recent experi-
ments of my respected friend Professor Reich approximate somewhat
more closely to the results given in Baily's admirable work. I have
obtained the mean 5-5772 from the whole series of experiments : (a)
with the tin ball and the longer thicker copper wire, the result was
5-5712, with a probable error of 0-0113 ; (b) with the tin ball, and with
the shorter thinner copper wire, as well as with the tin ball and the
bi-filar iron Avire, 5-5832, with a probable error of 0-0149. Taking
this error into account, the mean in (a) and (b) is 5-5756. The re-
sult obtained by Baily, and which was certainly deduced from a larger
number of experiments (5-660), might indeed give us a somewhat
higher density, as it obviously rose in proportion to the greater light-
ness of the balls that were xised in the experiments, which were either
of glass or ivory. (Reich, in Poggend., Annalen, bd. Ixxxv., s. 190.
Compare also Whitehead Hearn, in the PhUos. Transact, for 1847, p.
217-229.) The motion of the torsion-balance was observed by Baily
by means of the reflection of a scale obtained from a mirror, which
was attached to the middle of the balance, a method that had been
first suggested by Reich, and was employed by Gauss in his magnetic
observations. The use of such a mirror, which is of great importance,
from the exactness with which the scale may be read off, was proposed
by PoggendoriF as early as the year 1826. {Annalen der Physik., bd.
rii., s. 121.)

THE DENSITY OF THE EARTH. 35

The torsion-balance, Cavendish (according to Baily's calcula-
tion) 5-448

Reich, 1838 5-440

Baily, 1832 5-660

Reich, 1847-1850 5-577

A far more important result in reference to the density of
the earth than that obtained by Baily (1842) and Eeich
(1847-1850) has been brought out by Airy's experiments
with the pendulum, conducted with such exemplary care in
the Mines of Harton, in the year 1854. According to these
experiments the density is 6-566, with a probable error of
0-182 (Airy, in the Fhilos. Transact, for 1856, p. 342). A
slight modification of this numerical value, made by Pro-
fessor Stokes on account of the effect of the rotation and el-
lipticity of the earth, gives the density for Harton, which
lies at 54° 48' north latitude, at 6*565, and for the equator
at 6-489.

The mean of the two last results gives 5-62 for the density
of the earth (taking that of water as 1), and consequently
much more than the densest finely granular basalt, which,
according to the numerous experiments of Leonhard, varies
from 2-95 to 3-67, and more than that of magnetic iron (4-9
to 5*2), and not much less than that of the native arsenic of
Marienberg or Joachimsthal. We have already elsewhere
observed (Cosmos, vol. i., p. 167) that from the great distribu-
tion of secondary and tertiary fomiations, and of those up-
heaved strata which constitute the visible continental part
of our earth's surface (the Plutonic and volcanic upheavals
being scattered in the form of islands over a small area of
space), the solid portion of the upper part of the earth's crust
possesses a density scarcely reaching from 2*4 to 2-6. If we
assume with Rigaud that the relation of the solid to the fluid
oceanic surface of our^obe is as 10 : 27, and if further we
consider that the latteWias been found by experiments with
the sounding-lead to extend to a depth of 27,700 feet, the
whole density of the upper strata, which underlie the dry
and oceanic surfaces, scarcely equals 1*5. The distinguished
geometrician Plana has correctly observed that the author of
the Mecanique Celeste was in error when he ascribed to the
upper stratum of the earth a density equal to that of granite,
which, moreover, he estimated somewhat highly at 3, which
would give him 10-047 for the density of the centre of the
earth.* This density would, according to Plana, be 16*27,

* Laplace, Micanlque Celeste^ ed. de 1846, t. v., p. 57. The mean

36 COSMOS.

if we assume that of the upper strata =1'83, which differs
but slightly from the total density of 1-5 or 1*6 of the earth's
crust. The vertical pendulum, no less than the horizontal
torsion-balance, may certainly be designated as a geognostic
instrument ; but the geology of the inaccessible parts of the
interior of our globe is, like the astrognosy of the unillumin-
ated celestial bodies, to be received with considerable cau-
tion. In a portion of my work, which treats of volcanic
phenomena, I can not wholly pass in silence those problems
which have been suggested by other inquirers in reference to
the currents pervading the general fluid in the interior of
our planet, or the probable or improbable periodically ebb-
ing and flowing movement in individual and imperfectly filled
basins, or the existence of portions of space, having a very

specific weight of granite can not be set down at more than 2-7, since
the bi-axial white potash-mica, and green uni-axial magnesia-mica
range from 2-85 to 3-1, while the other constituents of this rock,
namely, quartz and feldspar, are 2-56 and 2-65. Even oligoclase is
only 2-68. If hornblende rises as high as 3*17, syenite, in which feld-
spar always predominates, never rises above 2*8. As argillaceous
schist varies from 2*69 to 2*78, while pure dolomite, lying below lime-
stone, equals only 2*88, chalk 2-72, and gypsum and rock-salt only 2*3,
I consider that the density of those continental parts of the crust of
our earth, which are appreciable to us, should be placed at 2*6 rather
than at 2 4. Laplace, on the supposition that the earth's density in-
creases in arithmetical progression from the surface toward the cen-
tre, and on the assumption (which is assuredly erroneous) that the
density of the upper stratum is equal to 3, has found 4*7647 for the
mean density of the whole earth, which deviates very considerably
from the results obtained by Reich (5*577) and by Baily (5*660) ; this
deviation being much greater than could be accounted for by the prob-
able error of observation. In a recent discussion on the hypothesis
of Laplace, which will soon form a very interesting paper in Schu-
macher's Astr. Nachrichten, Plana has arrived at the result that, by a
different method of treating this hypothesis, Reich's mean density of
the earth, and the density of the dry and oceanic superficial strata,
which I estimated at 1*6, as well as the^llipticity, within the limits
that seem probable for the latter value, may be very closely approxi-
mated to. "If the compressibility of the substances of which the
earth is formed," writes the Turin geometrician, "has given rise to
regular strata nearly elliptical in form, and having a density which
increases from the surface toward the centre, we may be allowed to
suppose that these strata, in the act of becoming consolidated, have
experienced modifications which, although they are actually very
small, are nevertheless large enough to preclude the possibility of our
deducing, with all the precision that we could desire, the condition of
tlie solid earth from its prior state of fluidity. This reflection has
made me attach the greater weight to the first hypothesis advanced by
the author of the Mecanique Celeste^ and I have consequently determ-
ined upon submitting it to a new investigation."

THE HEAT OF THE EARTH. 37

low Specific gravity and underlying the upheaved mountain
chains.* In a work devoted to cosmical phenomena no
question should be overlooked on which actual observations
have been instituted, or which may seem to be elucidated by
close analogies.

b. The Existence and Distribution of Heat in the interior of
our Globe.

(Expansion of the Delineation of Nature, Cosmos, vol. i.,
pt 1G8-176.)

Considerations regarding the internal heat of our earth,
the importance of which has been greatly augmented by the
connection which is now generally recognized to exist be-
tween it and phenomena of upheavals and of volcanic action,
are based partly upon direct, and therefore incontrovertible
measurements of temperature in springs, borings, and sub-
terranean mines, and partly upon analytical combinations
regarding the gradual cooling of our planet, and the influence
which the decrease of heat may have exercised in primeval
ages upon the velocity of rotation and upon the direction
of the currents of internal heat.f The figure of the com-
pressed terrestrial spheroid is further dependent upon the
law, according to which density increases in concentric su-
perimposed non-homogeneous strata. The first or experi-
mental, and therefore the more certain portion of the inves-
tigation to which we shall limit ourselves in the present
place, throws light only upon the accessible crust of the
earth, which is of viry inconsiderable thickness, while the
second or mathematical part, in accordance with the nature
of its applications, yields rather negative than positive results.
This method of inquiry, which possesses all the charm of
ingenious a*d intellectual combinations of thought, J leads
to problems, which can not be wholly overlooked when we
touch upon conjectures regarding the origin of volcanic
forces, and the reaction of the fused interior upon the solid
external crust of our earth. Plato's geognostic myth of the
Pyriphlcgethon,§ as the origin of all thermic springs, as well

* See Petit su7' la latitude de t Observatoire de Toulouse, la densite
moyenne de la chaine des Pyrenees, et la probabilite qu'il existe un vide
sous sette chaine, in the Comptes rendus de tAcad. des Sc, t. xxix., 1849,
p. 730. t Cosmos, vol. i., p. 176.

% Hopkins, Physical Geology, in the Report of the British Association
for 1838, p. 92; Philos. Transact., 1839, pt. li., p. 381, and 1840, pt.
1., p. 193; Hennessey {Terrestrial Physics), in the Philos. Transact.,
1851, pt. ii., p. 504-525. § Cosmos, vol. i., p. 237.

38 COSMOS.

as of volcanic igneous currents, emanated from the early and
generally felt requirement of discovering some common cause
for a great and complicated series of phenomena.

Amid the multiplicity of relations presented by the earth's
surface, in respect to insolation (solar action) and its capacity
of radiating heat, and amid the great differences in the ca-
pacity for conducting heat, which varies in accordance with
the composition and density of heterogeneous rocks, it is
worthy of notice, that wherever ^le observations have been
conducted with care, and under favorable circumstances, the
increase of the temperature with the depth has been found
to present for the most part very closely coinciding results,
even at very different localities. For very great depths we
obtain the most certain results from Artesian wells, especial-
ly when they are filled with fluids that have been rendered
turbid by the admixture of clay, and are therefore less favor-
able to the passage of internal currents, and when they do
not receive many lateral affluents flowing into them at differ-
ent elevations through transverse fissures. On account of
their depth, we will begin with two of the most remarkable
Artesian wells, namely, that of Grenelle, near Paris, and
that of the New Salt-works at Oeynhausen, near Minden.
We will proceed in the following paragraph to give some
of the most accurate results which they have yielded.

According to the ingenious measurements of Walferdin,*
to whom we are indebted for a complete series of very deli-
cate apparatus for determinations of temperature at great
depths in the sea and in springs, the ^rface of the basin of
the well at Grenelle lies at an elevation of 36*24 metres, or
119 feet, above the level of the sea. The upper outlet of
the ascending spring is 33'33 metres, or 109-3 feet, higher.
This total elevation of the ascending water (69*57 metres, or
228*2 feet) is, when compared with the level of the sea, about
196*8 feet lower than the outbreak of the green sandstone
strata in the hills near Lusigny, southeast of Paris, to whose
infiltrations the rise of the waters in the Artesian wells at
Grenelle have been ascribed. The borings extend to a depth
of 547 metres, or 1794-6 feet, below the base of the Grenelle
basin, or about 510-76 metres, or 1675 feet, below the level

* The observations of Walferdin were made in the autumn of 1847,
and deviate very slightly from the results obtained with the same ap-
paratus by Arago, in 1840, at a depth of 1657 feet, when the borer
had left the chalk and was beginning to penetrate through the gault.
See Cosmos, vol. i., p. 174, and Comptes rendus, t. xi., 1840, p. 707.

INTERNAL HEAT OF THE EARTH. 39

of the sea ; the waters, consequently, rise to a total height of
580-33 metres, or 1904 feet. The temperature of the spring
is 81°-95 F. ; consequently the increase of heat marks 1° F.
for about every 59 feet.

The boring at the New Salt-works at Rehme is situated
231 feet above the level of the sea (above the water-mark at
Amsterdam). It has penetrated to an absolute depth of
2281 feet below the surface of the earth, measuring from the
point where the operations were begun. The salt spring,
which, when it bursts forth, is impregnated with a large
quantity of carbonic acid, lies, therefore, 2052 feet below the
level of the sea — a relative depth which is perhaps the great-
est that has ever been reached by man in the interior of the
earth. The temperature of the salt spring at the New Salt-
works of Oeynhausen is 91° 04 F. ; and, as the mean annual
temperature of the air at these works is about 49°'3 F., we
may assume that there is an increase of temperature of 1° F.
for every 54-68 feet. The boring at these Salt-works* is,
therefore, 491 feet absolutely deeper than the boring at Gre-
nelle ; it sinks 377 feet deeper below the surface of the sea,
and the temperature of its waters is 9°*18 F. higher. The
increase of the heat at Paris is about 1° F. for 59 feet, and
therefore scarcely -jJ^th greater. I have already elsewhere
drawn attention to the fact that a similar result was obtained
by Auguste de la Rive and Marcet, at Bre'gny, near Geneva,
in investigating a boring which was only 725 feet in depth,
although it was situated at an elevation of more than 1600
feet above the Mediterranean Sea.f

If to these three springs, which possess an absolute depth
varying between 725 feet and 2285 feet, we add another,
that of Monkwearmouth, near Newcastle (the water rising
through a coal-mine which, according to Phillips, is worked

* According to the manuscript results given by the superintendent
of the mines of Oeynhausen. See Cosmos, vol. i., p. 157, 174; and
.Bischof, Lehrbuch der Chem. und Phys. Geologie, bd. i., abth. 1, s. 154-
163. . In regard to absolute depth the borings at Mondorf, in the
Grand Duchy at Luxemburg (2202 feet), approach most nearly to those
at the New Salt-works at Oeynhausen.

t Cosmos, vol. i., p. 174; and Mcmoires de la Socicte d'Hist. Naturelle '
de Geneve, t. vi., 1833, p. 243. The comparison of a number of Arte-
sian wells in the neighboi'hood of Lille with those of Saint Ouen and
Geneva would, indeed, lead us to assume, if we were quite certain as
to the accuracy of the numerical data, that the different conductive
powers of terrestrial and rocky strata exert a more considerable in-
fluence than has generally been supposed (Poisson, Thcorie Mathema-
tique de la Chaleur, p. 421).

40 COSMOS.

at a depth of 1496 feet below the level of the sea), we shall
find this remarkable result, that at four places widely sepa-
rated from one another an increase of heat of 1° F. varies
only between 54 and 58-6 feet;* such a coincidence in the
results can not, however, be always expected to occur when
we consider the nature of the means which are employed for
determining the internal heat of the earth at definite depths.
Although we may assume that the water which is infiltrated
in elevated positions through hydrostatic pressure, as in con-
nected tubes, may influence the rising of springs at points of
great depth, and that the subterranean waters acquire the
temperature of the terrestrial strata with which they are
brought in contact, the*water that is obtained through bor-
ings may, in certain cases, when communicating with vertic-
ally descending fissures, obtain some augmentation of heat
from an inaccessible depth. An influence of this kind, which
is very diflferent from that of the varying conductive power
of different rocks, may occur at individual points widely dis-
tant from the original boring. It is probable that the waters
in the interior of our earth move in some cases within limit-
ed spaces, flowing either in streams through fissures (on which
account it is not unusual to find that a few only of a large
number of contiguous borings prove successful), or else follow
a horizontal direction, and thus form extensive basins — a re-
lation which greatly favors the labor of boring, and in some
rare cases betrays, by the presence of eels, muscles, or vege-
table remains, a connection with the earth's surface. Al-
though, from the causes which we have already indicated,
the ascending springs are sometimes warmer than the slight
depth of the boring would lead us to anticipate, the afflux
of colder water which flows laterally through transverse fis-
sures leads to an opposite result.

It has already been observed that points situated on the
same vertical line, at an inconsiderable depth within the in-

* In a table of fourteen borings, which were more than one hundred
yards in depth, and wliich were situated in various parts of Fvance,
Bravais, in his very instructive encyclopedic memoir in the Pati'ia^
1847, p. 145, indicates nine in which an increase of temperature of
1° F. is found to occur for every 50-70 feet of depth, which would
give a deviation of about 10 feet in either direction from the mean
value given in the text. See also Magnus, in Poggen., Ann,, bd. xxii.,
1831, s. 146. It would appear, on the whole, that the increase of
temperature is most rapid in Artesian wells of very considerable depth,
although the very deep wells of Monte Massi, in Tuscany, and Neuffen,
on the northwest part of the Swabian Alps, present a remarkable ex-
ception to this rule.

INVARIABLE TEMPERATURE. 41

terior of our earth, experience at very different times the
maximum and minimum of atmospheric temperature, which
is modified by the sun's place and by the seasons of the year.
According to the very accurate observations of Quetelet,
daily variations of temperature are not perceptible at depths
of Sfths feet below the surface ;* and at Brussels the high-
est temperature was not indicated until the 10th of Decem-
ber, in a thermometer which had been sunk to a depth of
more than 25 feet, while the lowest temperature was ob-
jferved on the 15th of June. In like manner, in the admira-
ble experiments made by Professor Forbes, in the neighbor-
hood of Edinburgh, on the conductive power of different
rocks, the maximum of heat was not observed until the 8th
of January in the basaltic trap of Calton Hill, at a depth of
24 feet below the surface.f It would appear, from the ob-
servations which were carried on for many years by Arago
in the garden of the Paris Observatory, that very small dif-
ferences of temperature were perceptible 30 feet below the
surface. Bravais calculated one degree for about every 50
feet on the high northern latitude of Bossekop, in Finmark
(69° 58^ N. lat.). The difference between the highest and
lowest annual temperature diminishes in proportion with
the depth, and according to Fourrier this difference dimin-
ishes in a geometrical proportion as the depth increases in
an arithmetical ratio.

The stratum of invariable temperature depends, in respect to
its depth, conjointly upon the latitude of the place, the con-
ductive power of the surrounding strata, and the amount of
difference of temperature between the hottest and the coldest
seasons of the year. In the latitude of Paris (48° 50^ the
depth and temperature of the Caves de V Ohservatoire (86 feet
and 53°'30 F.) are usually regarded as affording the amount
of depth and temperature of the invariable stratum. Since
Cassini and Legentil, in 1783, placed a very correct mercurial
thermometer in these subterranean caves, which are portions
of old stone quarries, the mercury in the tube has risen about
0°-4.{ Whether the cause of this rising is to be ascribed to

* Quetelet, in the Bulletin de VAcad. de Brnxelles, 1836, p. 75.

f Forbes, Exper. on the Temperature of the Earth at different Depths,
in the Trans, of the Royal Soc. of Edinburgh,. yo\. xvi., 1849, pt. ii.,
p. 189.

X All numbers referring to the temperature of the Caves de VOb-
servatoire have been taken from the work of Poisson, Th'corie Mathc-
matique de la Chaleur, p. 415 and 462. The Annuaire Mctcorologiq^ie
de la France, edited by Martins and Haeghens, 1849, p. 88, contains

42 COSMOS.

V

an accidental alteration in the thermometrical scale which,
however, was adjusted by Arago in 1817 with his usual care,
or whether it indicates an actual increase of heat, is still
undecided. The mean temperature of the air at Paris is
51°*478 F. Bravais is of opinion that the thermometer in
the Caves de V Observatoire stands below the limit of invari-
able temperature, although Cassini believes that he has founc^
a difference of x^injtl^s of a degree (Fahr.) between the winter
and summer temperature, the higher temperature being found
to prevail in the winter.* If we now take the mean o^
many observations of the temperature of the soil between
the parallels of Zurich (47^ 22") and Upsala (59° 51^, we
obtain an increase of 1° F. for every 40 feet. Differences
of latitude can not produce a difference of more than 12 or
15 feet, which is not marked by any regular alteration from
south to north, because the influence which the latitude un-
doubtedly exerts is masked within these narrow limits by
the influence of the conductive power of the soil, and by
errors of observation.

As the terrestrial stratum in which we first cease to ob-
serve any alteration of temperature through the whole year
lies, according to the theory of the distribution of heat, so
much the nearer the surface, as the maxima and minima of
the mean annual temperature approximate tcf one another, a
consideration of this subject has led my friend Boussingault
to the inojenious and convenient method of determinino; the
mean temperature of a place within the tropical regions (es-
pecially between 10 degrees north and south of the equator)
by observing a thermometer which has been buried 8 or 12
inches below the surfiice of the soil in some well-protected
spot. At different hours and different months of the year,
as in the experiments of Captain Hall near the coast of the
Choco in Tumaco, those at Salaza in Quito, and those of
Boussingault in la Vega de Zupia, Marmato, and Anserma
Nuevo in the Cauca valley, the temperature scarcely varied
one tenth of a degree ; and almost within the same limits it
was identical with the mean temperature of the air at those
places in which it had been determined by horary observa-
tions. It was, moreover, very remarkable that this identity

corrections by Gay-Lussac for Lavoisier's subterranean thermometer.
The mean of three readings, from June till August, was 53°'95 F. for
this thermometer, at a time when Gay-Lussac found the temperature
to be 53°-32, which was therefore a difference of 0°-63.

* Cassini, in the Mem. de VAcad. des Sciences^ 1786, p. 511.

INVARIABLE *STRATUM. 43

remained perfectly uniform, whether the thermometric sound-
ings (of less than one foot in depth) were made on the torrid
shores of Guayaquil and Payta, on the Pacific, or in an
Indian village on the side of the volcano of Purace, which I
found from my barometrical measurements to be situated at
an elevation of 1356 toises, or 8671 feet above the sea. The
mean temperatures differed by fully 25° F. at these different
stations.*

I believe that special attention is due to two observations
which I made on the mountains of Peru and Mexico, in
mines which lie at a greater elevation than the summit of
the Peak of TenerifFe, and are therefore the highest in which
a thermometer has ever been placed. At a height of be-
tween 12,000 and 13,000 feet above the level of the sea I
found the subterranean air 25° F. warmer than the external
atmosphere. Thus, for instance, the little Peruvian town of
Micuipampaj lies, according to my astronomical and hypso-

* Boussingault, Sur la j^^ofondem- a laquelle on trouve dans la zone
torride la couche de temperatm-e invariable, in the Annates de Chimie et
de Physique, t. liii,, 1833, p. 225-247. Objections have been advanced
by John Caldecott, the astronomer to the Rajah of Travancore, and by
Captain Newbold, in India, against the method recommended in this
memoir, although it has been employed in South America in many
very accurate experiments. Caldecott found at Trevandrum (Edin.
Transact., vol. xvi., part iii., p. 379-393) that the temperature of the
soil, at a depth of three feet and more below the surface (and there-
fore deeper than Boussingault's calculation), was 85° and 86° F., while
the mean temperature of the air was 80°*02. Newbold's experiments
{Philos. Transact for the Year 1845, pt. i., p. 133), which were made
at Bellary, lat, 15° 5', showed an increase of temperature of 4° F. be-
tween sunrise and 2 P.M. for one foot of depth ; but at Cassargode, lat.
12° 29', there was only an increase of l°-30 F., under a cloudy sky.
Is it quite certain that the thermometer in this case Avas sufficiently
covered to protect it from the influence of the sun's rays ? Compare
also Forbes, Exper. on the Temp, of the Earth at different Depths, in the
Edin. Transact., vol. xvi., part ii., p. 189. Colonel A. Costa, the ad-
mirable historian of New Granada, has made a prolonged series of ob-
servations, which fully confirm Boussingault's statement, and which
were completed, about a year ago, at Guadua, on the southwestern
side of the elevated plateau of Bogota, where the mean annual tem-
perature is 43° -94 F. at the depth of one foot, and at a carefully pro-
tected spot. Boussingault thus refers to these experiments: "The
observations of Colonel A. Costa, whose extreme precision in every
thing which is connected with meteorology is well known to you, prove
that, when fulli/ sheltered from all disturbing influences, the temperature
within the tropics remains constant at a very small depth below the
surface."

t In reference to Gualgayoc (or Minas de Chota) and Micuipampa,
see Humboldt, Recueil d^ Observ. Astron., vol. i., p. 324.

44 COSMOS.

metrical observations, in the latitude 6° 43^ S., and at an
elevation of 1857 toises, or 11,990 feet, at the base of Cerro
de Gualgayoc, celebrated for the richness of its silver mines.
The summit of this almost isolated fortress-like and pictur-
esquely situated mountain rises 240 toises, or 1504 feet, high-
er than the streets of Micuipampa ; the external air at a dis-
tance from the mouth of the pit of the Mina del Purgatorio
was 42°*26 F. ; but in the interior of the mine, which lies
more than 2057 toises, or 13,154 feet above the sea, I saw
that the thermometer every where indicated a temperature
of 67°*64 F., there being thus a difference of 25°-38 F. The
limestone rock was here perfectly dry, and very few men
were working in the mine. In the Min£W de Guadalupe,
which lies at the same elevation, I found that the temper-
ature of the internal air was 57°*9 F., showing, therefore, a
difference of 15°*64 F. when compared with the external
air. The water which flowed out from the very damp mine
stood at 52°*34 F. The mean annual temperature of Micui-
pampa is probably not more than 45° -8 F. In Mexico, in
the rich silver mines of Guanaxuato,* I found, in the Mina
de Valenciana, the external temperature in the neighborhood
of the Tiro Nuevo (which is 7590 feet above the sea) 70°-16 F.,
and the air in the deepest mines — for instance, in the Planes
de San Bernardo — 1630 'feet below the opening of the shaft
of Tiro Nuevo, fully 80°*6 F., which is about the mean tem-
perature of the littoral region of the Gulf of Mexico. At a
point 147 feet higher than the mouth of the Planes de San
Bernardo, a spring of water issues from the transverse rock,
in which the temperature is 84°*74 F. I determined the
latitude of the mountain town of Guanaxuato to be 21° O'N.,
with a mean annual temperature varying between 60°*44 and
61°-26 F. The present is not a fitting place in which to
advance conjectures, which it might be diflicult to establish
in relation to the causes of probably an entirely local rise of
the subterranean temperature at mountain elevations, varying
from 6000 to more than 12,000 feet.

A remarkable contrast is exhibited in the steppes of
Northern Asia, by the conditions of the frozen soil, whose
very existence was doubted, notwithstanding the early testi-
mony of Gmelin and Pallas. It is only in recent times that
correct views in relation to the distribution and thickness of
the stratum of subterranean ice have been established by

* Essai Polit. sur le Roy. de la Nouv, Espagne (2eme ed., t. iii.,
p. 201).

THE FROZEN SOIL. 45

means of the admirable investigations of Erman, Baer, and
MiddendorfF. In accordance with the descriptions given of
Greenland by Cranz, of Spitzbergen by Martens and Phipps,
and of the coasts of the sea of Kara by Sujew, the whole of
the most northern part of Siberia was described by too hasty
a generalization as entirely devoid of vegetation, always froz-
en on the surface, and covered with perpetual snow, even in
the plains. The extreme limit of vegetation in Northern
Asia is not, as was long assumed, in the parallel of 67°, al-
though sea-winds and the neighborhood of the Bay of Obi
make this estimate true for Obdorsk ; for in the valley of
the great River Lena high trees grow as far north as the
latitude of 71°. Even in the desolate islands of New Si-
beria, large herds of rein-deer and countless lemmings find an
adequate nourishment.* Middendoi'iF's two Siberian expe-
ditions, which are distinguished by a spirit of keen observa-
tion, adventurous daring, and the greatest perseverance in a
laborious undertaking, were extended, from the year 1843 to
1846., as far north as the Taymir land in 75° 45' lat., and
southeast as far as the Upper Amoor and the Sea of Ochotsk.
The former of these perilous undertakings led the learned in-
vestigator into a hitherto unvisited region, whose exploration
was the more important in consequence of its being situated
at equal distances from the eastern and western coasts of the
old Continent. In addition to the distribution of organisms
in high northern latitudes, as depending mainly upon climat-
ic relations, it was directed by the St. Petersburg Academy
of Sciences that the accurate determination of the tempera-
ture of the ground and of the thickness of the subterranean
frozen soil should be made the principal objects of the expe-
dition. Observations were made in borings and mines, at a
depth of froHi 20 to 60 feet, at more than twelve points (near
Turuchansk, on the Jenisei, and oa the Lena), at relative dis-
tances of from 1600 to 2000 geographi^lLmiles.

The most important seat of these georaermic observations
was, however, Schergin's shaft at Jakutsk, 62° 2' N. lat.f

* E. von Baer, in Middendorff 's Reise in Sib., bd. i., s. 7.

t The merchant Fedor Schergin, cashier to the Russian-American
Trading Company, began, in the year 1828, to dig a well in the com-t-
yard of a house belonging to the company. As he had only found
frozen earth and no water at the depth of 90 feet, which he reached in
1830, he determined to give up the attempt, until Admiral Wrangel,
who passed through Jakutsk on his way to Sitcha, in Russian America,
and who saw how interesting it would be, in a scientific point of view,
to penetrate through this subterranean stratum of ice, induced Scher-

20°

•26 ''

21°

•43 «

23°

•27 "

24°

•49 "

26°

•60 "

46 COSMOS.

Here a subterranean stratum of ice was pierced to a depth of
more than 382 feet. The thermometer was sunk at eleven
points along the lateral walls of the shaft, between the surface
and the greatest depth, which was reached in 1837. The
observer was obliged to be let down standing in a bucket,
with one arm fastened to a rope, while he read off the ther-
mometric scale. The series of observations, whose mean
error does not amount to more than 0°-45 F., embrace the
interval between April, 1844, and June, 1846. The decrease
of cold was not proportional to the depth at individual points,
but nevertheless the following results were obtained for the
total increase of the mean temperatures for the different
superimposed frozen strata :

50 feet 17°-13Fahr.

100 "

150 "

200 "

250 "

382 "

After a very careful consideration of all these observa-
tions, Middendorff* determined the general increase of tem-
perature to be 1° F. for every space varying from 44°*5 to
52 feet.* This result shows a more rapid increase of heat

gin to continue the boring ; and up to 1837, although an opening had
been made to a depth of 382 feet below the surface, it had not pene-
trated beyond the ice.

* Middendorfl', Reise in Sib., bd. i., s. 125-133. "If we exclude,"
says Middendorff, " those depths which' did not quite reach 100 feet,
on the ground that they were influenced by annual deviations of tem-
perature, as was determined by experiments previously made in Si-
beria, we shall still find certain anomalies in the partial increase of
heat. Thus, for instance, between the depths of 150-200 feet the
temperatui-e rises at a ratio of 1° F. for only 29^3 feet;, while between
250-300 feet the correspondipg increase is 96^4 feet. We may, there-
fore, venture to assi^^iat the results of observations that have hith-
erto been obtained flPlBhergin's shaft are by no means sufficient to
determine with certainty the amount of the increase of temperature,
and that, notwithstanding the great variations which may depend upon
the different conductive powers of the terrestrial strata, and the dis-
turbing influence of the air or water which enters from above, an in-
crease of 1° F. occurs for every 44-52 feet. The result of 52 feet is
the mean of six partial increases of temperature, measured at intervals
of 50 feet between the depths of 100 and 382 feet. On comparing
the mean annual temperature of Jakutsk, 13°-71 F., with that which
was found from observation to be the mean temperature of the ice
(26°-6) at the greatest depth of the mine (382 feet), I find 29^6 feet
for every increase of 1 ° F. A comparison of the- temperature at the
deepest part with that at a depth of 100 feet would give 44 4 feet for

THE TEMPERATURE OF THE EARTH. 47

in Schergin's shaft than has been obtained from different
borings in Central Europe, whose results approximate closely
to one another (see p. 39). The difference fluctuates be-
tween ^th and ^th. The mean annual temperature of Ja-
kutsk was determined at 13°*7 F. The oscillation between
the summer and winter temperature is so great, according to
Newerow's observations, which were continued for fifteen
years (from 1829 to 1844), that sometimes for fourteen days
consecutively, in July and August, the atmospheric tempera-
ture rises as high as 77°, or even 84° -6 F. ; while during 120
consecutive winter days, from November to February, the
cold falls to between — 42°-3 F. and —69° F. In estimat-
ing the increase of temperature which was found on boring
through the frozen soil, we must take into account the depth
below the surface at which the ice exhibits the temperature
of 32° F., and which is consequently the nearest to the lower
limit of the frozen soil ; according to Middendorff 's results,
which entirely agree with those that had been obtained much
earlier by Erman, this point was found in Schergin's shaft to
be 652, or 684 feet below the surface. It would appear,
however, from the increase of temperature which was ob-

this increase. From the acute investigations of Middendoiff and
Peters, in reference to the velocity of transmission of changes of at-
mospheric temperature, including the maxima of cold and heat (Mid-
dend., s. 133-157, 168-175), it follows that in the different borings,
which do not exceed the inconsiderable depth of from 8 to 20 feet,
" the temperature rises from March to October, and falls from Novem-
ber to April, because the spring and autumn are the seasons of the
year in which the changes of atmospheric temperature are most con-
siderable" (s. 142-145). Even carefully covered mines in Northern
Siberia become gradually cooled, in consequence of the walls of the
shafts having been for years in contact with the air ; this cause, how-
ever, has only made the temperature fall about 1° F. in Schergin's
shaft, in the course of eighteen years. A remarkable and hitherto un-
expITlined phenomenon, which has also presented itself in the Scher-
gin shaft, is the warmth occasionally observed in the winter, although
only at the lowest strata, without any appreciable influence from with-
out (s. 156-178). It seems still more striking to me, that in the bor-
ings at Wedensk, on the Pasina, when the atmospheric temperature
is —31° F., it should be 26°'-4 at the inconsiderable depth of 5 or 10
feet ! The isogeothermal lines, whose direction was first pointed out
by KupflPer in his admirable investigations (Cosmos, vol. i., p. 219), will
long continue to present problems that we are unable to solve. The
solution of these problems is more especially difficult in those cases
in which the complete perforation of the frozen soil is a work of con-
siderable time ; we can, however, no longer regard the frozen soil at
Jakutsk as a merely local phenomenon, which, in accordance with
Slobin's view, is produced by the terrestrial strata deposited from wa-
ter (Middend., s. 167).

48

COSMOS.

served in the mines of Mangan, Shilow, and Dawydow, which
are situated at about three or four miles from Irkutsk, in the
chain of hills on the left bank of the Lena, and which are
scarcely more than 60 feet in depth, that the normal stratum
of perpetual frost seems to be situated at 320 feet below the
surface.* Is this inequality only apparent in consequence
of the uncertainty which attaches to a numerical determina-
tion, based on so inconsiderable a depth, and does the in-
crease of temperature obey diflferent laws at different times?
Is it certain that if we were to make a horizontal section of
several hundred fathoms from the deepest part of Schergin's
shaft into the adjoining country, we should find in every di-
rection and at every distance from the mine frozen soil, in
which the thermometer would indicate a temperature of 4°*5
below the freezing point ?

Schrenk has examined the frozen soil in 67° 30^ N. lat.,in
the country of the Samojedes. In the neighborhood of
Pustojenskoy Gorodok, fire is employed to facilitate the
sinking of wells, and in the middle of summer ice was found
at only 5 feet below the surface. This stratum could be
traced for nearly 70 feet, when the works were suddenly
stopped. The inhabitants were able to sledge over the
neighboring lake of Usteje throughout the whole of the sum-
mer of 1813. t During my Siberian expedition with Ehren-
berg and Gustav Rose, we caused a boring to be made in a
piece of turfy ground near Bogoslowsk (59° 44'' N. lat.),
among the Ural Mountains, on the road to the Turjin mines. {
We found pieces of ice at the depth of 5 feet, which were
imbedded, breccia-like, in the frozen ground, below which
bagan a stratum of thick ice, which we had not penetrated
at the depth of 10 feet.

The geographical extension of the frozen ground, that is to
say, the limits within which ice and frozen earth are fbund
at a certain depth, even in the lAonth of August, and conse-

* MiddendorfF, bd. i., s. 160, 164, 179. In these numerical data and
conjectures regarding the thickness of the frozen soil, it is assumed
that the temperature increases in arithmetical progression with the
depth. Whether a retardation of this increase occurs in greater depths
is theoretically uncertain, and hence there is no use in entering upon
deceptive calculations regarding the temperature of the centre of the
earth in the fused heterogeneous rocky masses which give rise to cur-
rents.

t Schrenk's Eeise durch die Tmdern der Samojeden, 1848, th. i.,
s. 597.

X Gustav Rose, Reise nack dem Ural, bd. i., s. 428,

THE FROZEN SOIL. 49

quently throughout the whole year, in the most northern
parts of the Scandinavian peninsula, as far east as the coasts
of Asia, depends, according to MiddendorfF's acute observa-
tions (like all geothermal relations), more upon local influ-
ences than upon the temperature of the atmosphere. The
influence of the latter is, on the whole, no doubt, stronger
than any other ; but the isogeothermal lines are not, as Kupf-
fer has remarked, parallel in their convex and concave curves
to climatic isothermal lines, which are determined by the
means of the atmospheric temperature. The infiltration of
liquid vapors deposited by the air, the rising of thermal
springs from a depth, and the varying conductive powers of
the soil, appear to be especially active.* "On the most
northern point of the European continent, in Finmark, be-
tween the high latitudes of 70° and 71°, there is as yet
no continuous tract of frozen soil. To the eastward, im-
pinging upon the valley of the Obi, 5° south of the North
Cape, we find frozen ground at Obdorsk and Beresow. To
the east and southeast of this point the cold of the soil in-
creases, excepting at Tobolsk, on the Irtisch, where the tem-
perature of the soil is colder than at Witimsk, in the valley
of the Lena, which lies 1° farther north. Turuchansk (G5° 54^
N. lat.) on the Jenisei, is situated upon an unfrozen soil, al-
though it is close to the limits of the ice. The soil at Am-
ginsk, southeast of Jakutsk, presents as low a temperature
as that of Obdorsk, which lies 5° farther north; the same
being the ca^e with Oleminsk, on the Jenisei. From the
Obi to the latter river the curve formed by the limits of the
frozen soil seems to rise a couple of degrees farther north,
after which it intersects, as it turns southward, the Lena
valley, almost 8° south of the Jenisei. Farther eastward,
this line again rises in a northerly direction."! KupfFer,
who has visited the mines of Nertshinsk, draws attention to
the fact that, independently of the continuous northern mass

* Compare my friend G. von Helmersen's experiments on the rela-
tive conductive powers of different kinds of rocks {Mem, de V Academie
de St. Petershourg : Melanges Physiques et Chimiques, 1851, p. 32).

t Middendorff, bd. i., s. 166. Compare also s. 179. " The curve
representing the commencement of the freezing of the soil in North-
ern Asia exhibits two convexities, inclining southward, one on the
Obi, which is very inconsiderable, and the other on the Lena, which
is much more strongly marked. The limit of the frozen soil passes
from Berresow, on the Obi, toward Turuchansk, on the Jenisei; it
then runs between "Witimsk and Oleminsk, on the right bank of the-
Lena, and, ascending northward, turns to the east."

Vol. v.— C

50 ' COSMOS.

of frozen soil, the phenomenon occurs in an island-like form
in the more southern districts, but in general it is entirely
independent of the limits of vegetation, or of the growth of
timber.

It is a very considerable advance in our knowledge, when
we are able gradually to arrive at general and sound cosmical
views of the relations of temperature of our earth m the
northern portions of the old continent, and to recognize the
fact that under different meridians the limits of the frozen
soil, as well as those of the mean annual temperature and
of the growth of trees, are situated at very different lati-
tudes ; whence it is obvious that continuous currents of heat
must be generated in the interior of our planet. Franklin
found in the northwest part of America that the ground was
frozen even in the middle of August at a depth of 16 inches ;
while Richardson observed, upon a more eastern point of the
coast, in 71° 12'' lat., that the ice-stratum was thawed in
July as low as three feet beneath the herb-covered surface.
Would that scientific travelers would afford us more general
information regarding the geothermal relations in this part
of the earth and in the southern hemisphere ! An insight
into the connection of phenomena is the most certain means
of leading us to the causes of apparently involved anomalies,
and to the comprehension of that which we are apt too
hastily to regard as at variance with normal laws.

c. Magnetic Activity of the Earth in its three Manifestations of
Force — Intensity, Inclination, and Variation. — Points {called
the Magnetic Poles) in which the Inclination is 90°. — Curves
on ivhich no Inclination is observed (Magnetic Equator). —
The Four different 3Iaxima of Intensity. — Curve of weakest
Intensity. — Extraordinat^ Disturbances of the Declination
{Magnetic Storms). — Polar Light.

(Extension of the Picture of Nature, Cosmos, vol. i., p. 176-202; vol.
ii., p. 333-336 ; and vol. iv., p. 82-86.)

The magnetic constitution of our planet can only be de-
duced from the many and various manifestations of terres-
trial force in as far as it presents measurable relations in
space and time, These manifestations have the peculiar
property of exhibiting perpetual variability of phenomena to
a much higher degree even than the temperature, gaseous
• admixture, and electrical tension of the lower strata of the
atmosphere. Such a constant change in the nearly-allied

THE MAGNETIC NEEDLE. 51

magnetic and electrical conditions of matter, moreover, es-
sentially distinguishes the phenomena of electro-magnetism
from those which are influenced by the primitive fundament-
al force of matter — its molecular attraction and the attrac-
tion of masses at definite distances. To establish laws in
that which is ever varying is, however, the highest object of
every investigation of a physical force. Although it has.
been shown by the labors of Coulomb and Arago that the
electro-magnetic process may be excited in the most vari-
ous substances, it has nevertheless been proved by Faraday's
brilliant discovery of diamagnetism (by the differences of the
direction of the axes, whether they incline north and south,
or east and west) that the heterogeneity of matter exerts an
influence distinct from the attraction of masses. Oxygen
gas, when inclosed in a thin gfass tube, will show itself un-
der the action of a magnet to be paramagnetic, inclining
north and south like iron ; and while nitrogen, hydrogen,
and carbonic acid gases remain unaffected, phosphorus,
leather, and wood show themselves to be liiamagnetic, and
arrange themselves equatorially from east to west.

The ancient Greeks and Romans were acquainted with
the adhesion of iron to the magnet, attraction and repulsion,
and the transmission of the attracting action through brass
vessels as well as through rings, which were strung together
in a chain-like form, as long as one of the rings was kept in
contact with the magnet ;* and they were likewise acquaint-
ed with the non-attraction of wood and of all metals, except-
ing iron. The force of polarity, which the magnet is able
to impart to a movable body susceptible of its influence,
was entirely unknown to the Western nations (Phoenicians,
Tuscans, Greeks, and Komans). The first notice which we
meet with among the nations of Western Europe of the
knowledge of this force of polarity, which has exerted so im-
portant an influence on the improvement and extension of
navigation, and which, from its utilitarian value, has led so
continuously to the inquiry after one universally diffused,
although previously unobserved force of nature, does not
date farther back than the 11th and 12th centuries. In the
history and enumeration of the principal epochs of a physic-

* The principal passage referring to the magnetic chain of rings
occurs in Plato's Ion., p. 533, D.E, ed. Steph. Mention has been made
of this transmission of the attracting action not only by Pliny (xxxiv.,
14) and Lucretius (vi., 910), but also by Augustine {De civitate Dei,
XX., 4) and Philo (Z>e Mtmdi opificio, p. 32 D, ed. 1G91).

52 COSMOS.

al contemplation of the universe, it has been found necessa-
ry to divide into several sections, and to notice, the sources
from which we derive our knowledge of that which we have
here summarily arranged under one common point of view.*
We find that the application among the Chinese of the
directive power of the magnet, or the use of the north and
south direction of magnetic needles floating on the surface
of water, dates to an epoch which is^ probably more ancient
than the Doric migration and the return of the Heraclidae
into the Peloponnesus. It seems, moreover, very striking
that the use of the south direction of the needle should have
been first applied in Eastern Asia not to navigation but to
land traveling. In the anterior part of the magnetic wagon
a freely floating needle moved the arm and hand of a small
figure, which pointed toward the south. An apparatus of
this kind (caMed fse-nan, indicator of the south) was present-
ed during the dynasty of the Tscheu, 1100 years before our
era, to the embassadors of Tonquin and Cochin-China, to
guide them ove^ the vast plains which they would have to
cross in their homeward journey. The magnetic wagon was
used as late as the loth century of our era.f Several of
these wagons were carefully preserved in the imperial pal-
ace, and were employed in the building of Buddhist monas-
teries in fixing the points toward which the main sides of
the edifice should be directed. The frequent application of
magnetic apparatus gradually led the more intelligent of the
people to physical considerations regarding the nature of
magnetic phenomena. The Chinese eulogist of the magnet-
ic needle, Kuopho (a writer of the age of Constantine the
Great), compares, as I have already elsewhere remarked, the
attractive force of the magnet with that of rubbed amber.
This force, according to him, is " like a breath of wind

* Cosmos, vol. i., p. 188 ; vol. ii., p. 253.

t Humboldt. Asie Centrak, t. i., p. xl.-xlii. ; and Examen Crit. de
tHist. de la Geographie, t. iii,, p. 35. Eduard Biot, who has extend-
ed and confirmed by his own careful and bibliographical studies, and
with the assistance of my learned friend Stanislas Julien, the inves-
tigations made by Klaproth in reference to the epoch at which the
magnetic needle was first used in China, adduces an old tradition,
according to which the magnetic wagon was already in use in the reign
of the Emperor Hoang-ti. No allusion to this tradition can, however,
be found in any writers prior to the early Christian ages. This cele-
brated monarch is presumed to have lived 2600 years before our era
(that is to say, 1000 years before the expulsion of the Hyksos from
Egypt). Ed. Biot, sur la direction de Vaiguille aimantee en Cliine in
the Comptes rendus de VAcad. des Sciences, t. xix., 1844, p. 822.

THE MAGNETIC NEEDLE. 53

which mysteriously breathes through these two bodies, and
has the property of thoroughly permeating them with the
rapidity of an arrow." The symbolical expression of " breath
of wind" reminds us of the equally symbohcal designation of
soul, which in Grecian antiquity was applied by Thales, the
founder of the Ionian School, to both these attracting sub-
stances— soul signifying here the inner principle of the mov-
ing agent.*

As the excessive mobility of the floating Chinese needles
rendered it difficult to observe and note down the indications
which they afforded, another arrangement was adopted in
their place as early as the 12th century of our era, in which
the needle that was freely suspended in the air was attached
to a fine cotton or silken thread exactly in the same manner
as Coulomb's suspension, which was first used^by William
Gilbert in Western Europe. By means of this more perfect
apparatus,! the Chinese as early as the beginning of the 12th
century determined the amount of the western valuation,
which in that portion of Asia seems only to xmdergo very in-
considerable and slow changes. From its use on land, the
compass was finally adapted to maritime purposes, and* under
the dynasty of Tsin, in the 4th century of our era, Chinese
vessels under the guidance of the compass visited Indian ports
and the eastern coast of Aft-ica.

Fully 200 years earlier, under the reign of Marcus Aure-
lius Antoninus, who is called An-tun by the writers of the

. * Cosmos, vol. i., p. 188. Aristotle {De Anima, i., 2) speaks only of
the animation of the magnet as of an opinion that originated with
Thales. Diogenes Laertius interprets this statement as applying also
distinctly to amber, for he says, "Aristotle and Hippias maintain as
to the doctrine enounced by Thales." . . . The sophist Hippias of
Elis, who flattered himself that he possessed universal knowledge, oc-
cupied himself with physical science and with the most ancient tradi-
tions of the physiological school. "The attracting breath," which, ac-
cording to the Chinese physicist, Kuopho, "permeates both the mag-
net and amber," reminds us, according to Buschmann's investigations
into the Mexican language, of the aztec name of the magnet tlaihio-
anani tetl, signifying "the stone which attracts by its breath" (from
i/iioil, breath, and ana, to draw or attract).

t The remarks which Klaproth has extracted from the Penthsaoyan
regarding this singular apparatus are given more fully in the Munrj-
khi-])i-than, Cowptes rendus, t. xix., p. 365. We may here ask why,
in this latter treatise, as well as in a Chinese book on plants, it is
stated that the cypress turns toward the west, and, more generally,
that the magnetic needle points toward the south ? Does this imply
a more luxuriant development of the- branches on the side nearest
the sun, or in consequence of the direction of the prevalent winds ?

54 COSMOS.

dynasty of Han, Roman legates came by sea by way of Ton-
quin to China. The application of the magnetic needle to
European navigation was, however, not owing to so transient
a source of intercourse ; for it was not until its use had be-
come general throughout the whole of the Indian Ocean,
along the shores of Persia and Arabia, that it was introduced
into the West in the 12th century, either directly through
the influence of the Arabs or through the agency of the Cru-
saders, who since 1096 had been brought in contact with
Egypt and the true Oriental regions. In historical investi-
gations of this nature, we can only determine with certainty
those epochs which must be considered as the latest limits
beyond which it would be impossible for us to urge our in-
quiries. In the politico-satirical poem of Guyot of Provins,
the mariner's compass is spoken of (1199) as an instrument
that had been long known to the Christian world ; and this
is also the case in the description of Palestine, which we owe
to thelBishop of Ptolemais, Jaques de Vitry, and which was
completed between the years 1204 and 1215. Guided by
the magnetic needle, the Catalans sailed along the northern
island* of Scotland as well as along the western shores of
tropical Africa, the Basques ventured forth in search of the
whale, and the Northmen made their way to the Azores (the
Bracir islands of Picigano). The Spanish Leyes de las Par-
tidas {del sahio Hey Don Alonso el nono), belonging to the first
half of the 13th century, extolled the magnetic needle as " the
true mediatrix {medianera) between the magnetic stone {la
piedra) and the north star." Gilbert also, in his celebrated
work De Magnete Physiologia Nova, speaks of the mariner's
compass as a Chinese invention, although he inconsiderately
adds that Marco Polo, " qui apud Chinas artem pyxidis di-
dicit,"- first brought it to Italy. As, however, Marco Polo
began his travels in 1271, and returned in 1295, it is evident,
from the testimony of Guyot of Provins and Jaques de Vi-
tiy, that the compass was, at all events, used in European
seas from 60 to 70 years before Marco Polo set forth on his
journeyings. The designations zohron and aphron, which
Vincent of Beauvais applied, in his Mirror of Nature, to the
southern and northern ends of the magnetic needle (1254),
seem to indicate that it was through Arabian pilots that Eu-
ropeans became possessed of the Chinese compass. These
designations point to the same learned and industrious nation
of the Asiatic peninsula whose language too often vainly ap-
peals to us in our celestial maps and globes.

VARIATION CHARTS.

56

From the remarks which I have already made, there can
scarcely be a doubt that the general application of the mag-
netic needle by Europeans to oceanic navigation as early as
the 12th century, and perhaps even earlier in individual cases,
originally proceeded from the basin of the Mediterranean.
The most essential share in its use seems to have belonged
to the Moorish pilots, the Genoese, Venetians, Majorcans,
and Catalans. The latter people, under the guidance of
their celebrated countryman, the navigator, Don Jaime Fer-
rer, penetrated, in 1346, to the mouth of the Rio de Ouro
(23° 40^ N. lat.), on the western coast of Africa; and, ac-
cording to the testimony of Raymundus Lullus (in his nauti-
cal work, JF^enix de las Maravillas del Orbe, 1286), the Barce-
lonians employed atlases, astrolabes, and compasses, long be-
fore Jaime Ferrer.

The knowledge of the amount of magnetic variation is of
a very early date, and was simultaneously imparted by the
Chinese to Indian, Malay, and Arabian seamen, through
whose agency it must necessarily have spread along the
shores of the Mediterranean. This element of navigation,
which is so indispensable to the correction of a ship's reck-
oning, was then determined less by the rising and setting of
the sun than by the polar star, and in both cases the determ-
ination was very uncertain ; notwithstanding which, we find
it marked down upon charts, as, for instance, upon the very
scarce atlas of Andrea Bianco, which was drawn out in the
year 1436. Columbus, who had no more claim than Sebas-
tian Cabot to be regarded as the first discoverer of the vari-
ation of the. magnetic needle, had the great merit of determ-
ining astronomically the position of a line of no variation
2i° east of the island of Corvo, in the Azores, on the 13th
of September, 1492. He found, as he penetrated into the
western part of the Atlantic Ocean, that the variation pass-
ed gradually from northeast to northwest. This observation
led him to the idea, which has so much occupied navigators
in later times, of finding the longitude by the position of the
curves of variation, which he still imagined to be parallel to
the meridian. We learn from his ship's log that when he
was uncertain of his. position during his second voyage
(1496), he actually endeavored to steer his way by observ-
ing the declination. The insight into the possibility of such
a method was undoubtedly that uncommunicable secret of
longitude which Sebastian Cabot boasted on his death-bed
of having acquired through special divine manifestation.

56 COSMOS.

The idea of a curve of no declination in the Atlantic was
associated in the easily excited fancy of Columbus with oth-
er somewhat vague views of alterations of climate, of an
anomalous configuration of the earth, and of extraordinary
motions of the heavenly bodies, in which he found a motive
for converting a physical into a political boundary line. Thus
the ra2/a, on which the agujas de marear point directly to the
polar star, became the line of demarkation between the king-
doms of Portugal and Castillo ; and from the importance of
determining with astronomical exactness the geographical
length of such a boundary in both hemispheres, and over ev-
ery part of the earth's surface, an arrogant Papal decree, al-
though it failed in effecting this aim, nevertheless exerted a
beneficial effect on the extension of astronomico-nautical
science and on the improvement of magnetic instruments.
(Humboldt, Examen Crit. de la Geog., t. iii., p. 54.) Felipe
Guillen, of Seville, in 1525, and probably still earlier the
cosmographer Alonso de Santa Cruz, teacher of mathematics
to the young Emperor Charles V., constructed new variation
compasses by which solar altitudes could be taken. The lat-
ter in 1530, and therefore fully 150 years before Halley, drew
up the first general variation chart; nlthough it was certain-
ly based upon very imperfect materials. We may form
some idea of the interest that had been excited in reference
to terrestrial magnetism in the IGth century, after the death
of Columbus, and during the contest regarding the line of
demarkation, when we find that Juan Jayme made a voyage
in 1585, with Francisco Gali, from the Philipines to Aca-
pulco, for the sole purpose of testing by a long trial in the
South Sea a Declinatorium of his own invention.

Amid this generally diffused taste for practical observa-
tion we trace the same tendency to theoretical speculations
which always accompanies or even more frequently precedes
the former. Many old traditions current among Indian and
Arabian sailors speak of rocky islands which bring death and
destruction to the hapless mariner, by attracting, through
their magnetic force, all the iron which connects together
the planks of the ship, or even by immovably fixing the en-
tire vessel. The effect of such delusions as these was to
give rise to a conception of the concurrence, at the poles, of
lines of magnetic variation, represented materially under the
image of a high magnetic rock lying near one of the poles.
On the remarkable chart of the New Continent, which was
added to the Latin edition of 1508 of tlie Geography of

FIRST USE OF THE LOG. 57

Ptolemy, we find that north of Greenland (Gruentlant),
which is represented as belonging to the eastern portion of
Asia, the north magnetic pole is depicted as an insular
mountain. Its position was gradually marked as being far-
ther south in the Breve Compendio de la Sphera, by Martin
Cortez, 1545, as well as in the Geograpliia di Tolomeo^ of
Liveo Sanuto, 1588. The attainment of this point, called
el calamitico, was associated with great expectations, since it
was supposed in accordance with a delusion, which was not
dissipated till long afterward, that some miraculoso stupendo
effetto would be experienced by those who reached it.

Until toward the end of the 16th century men occupied
themselves only with those phenomena of variation which
exerted a direct influence on the ship's reckoning and the de-
termination of its place at sea. Instead of the one line of no
variation, which had been found by Columbus in 1492, the
learned Jesuit, Acosta, who had been instructed by Portu-
guese pilots (1589), expressed the belief, in his admirable
Historia Natural de las IndiaSf that he was able to indicate
four such lines. As the ship's reckoning, together with the
accurate determination of the direction (or of the angle
measured by the corrected compass), also requires the dis-
tance the ship had made, the introduction of the log, al-
though this mode of measuring is even at the present day
very imperfect, nevertheless marked an important epoch in
the history of navigation. I believe that I have proved, al-
though contrary to previously adopted opinions, that the first
certain evidence of the use of the log* (la cadena de la popa,
la corredera) occurs in the journal which was kept by An-
tonio Pigafetta during the voyage of Magellan, and which
refers to the month of January, 1521. Columbus, Juan de
la Cosa, Sebastian Cabot, and Vasco de Gama, were not ac-
quainted with the log and its mode of application, and they

* Cosmos, vol. ii., p. 256-258. In the time of King Edward III. of
England, when, as Sir Harris -Nicolas (^History of the Royal Navy,
184:7, vol. ii., p. 180) has shown, ships were guided by the compass,
which was then called the sail-stone dial, sailing -needle, or adamant,
we find it expressly stated in the accounts of the expenses for equip-
ping the king's ship. The George, in the year 1345, that sixteen hour-
glasses had been bought in Flanders. This statement, however, is
by no means a proof of the use of the log. The ampoUetas (or hour-
glasses) of the Spaniards were, as we most plainly find from the
statements of Enciso in Cespides, in use long before the introduc-
tion of the log, " echando punto por fantasia in la corredera de Io8
perezosos."

C2

58 COSMOS.

estimated the ship's speed merely by the eye, while they found
the distance they had made by the running down of the sand
in the glasses known as ampolletas. For a considerable pe-
riod the horizontal declincUion from the north pole was the
only element of magnetic force that was made use of, but at
length (in 157G) the second element, inclination, began to be
first measured. Kobert Norman was the first who determ^
ined the inclination of the magnetic needle in London, which
he noted with no slight degree of accuracy by means of an
inclinatorium, which he had himself invented. It was not
until 200 years afterward that attempts were made to meas-
ure the third element, the intensity of the magnetic terrestrial
force.

About the close of the 16th century, William Gilbert, a
man who excited the admiration of Galileo, although his
merits were wholly unappreciated by Bacon, first laid down
comprehensive views of the magnetic force of the earth.*
He clearly distinguished magnetism from electricity by their
several effects, although he looked upon both as emanations
of one and the same fundamental force, pervading all matter.
Like other men of genius, he had obtained many happy re-
sults from feeble analogies, and the clear views which he had
taken of terrestrial magnetism (de magno magnete tellure)
led him to ascribe the magnetization of the vertical iron rods
on the steeples of old church towers to the effect of this force.
He, too, was the first in Europe who showed that iron might
be rendered magnetic by being touched with the magnet, al-
though the Chinese had been aware of the fact nearly 500
years before him.'j' Even then, Gilbert gave steel the pref-
erence over soft iron, because the former has the power of
more permanently retaining the force imparted to it, and of
thus becoming for a longer time a conductor of magnetism.

In the course of the I7th century, the navigation of the

* Cosmos, vol. i., p. 177. Calamitico was the name given to these
instruments in consequence of the first needles for the compass hav-
ing been made in the shape of a frog.

t See Gilbert, Physiologia Nova de Magnete, lib. iii., cap. viii., p, 124.
Even Pliny (Cosmos, vol. i., p. 177) remarks generally, without, how-
ever, referring to the act of touching, that magnetism may be impart-
ed for a long period of time to iron. Gilbert expresses himself as
follows in reference to the vulgar opinion of a magnetic mountain :
" Vulgaris opinio de montibus magneticis aut rupe aliqua magnetica,
de polo phantastico a polo mundi distante" (1. c. p. 42-98). The va-
riation and advance of the magnetic lines were entirely unknown to
him. "Varietas uniuscujusque loci constans est" (1. c. 42, 98, 152,
153).

THE MAGNETIC POLES. 59

Netherlanders, British, Spaniards, and French, which had
been so widely extended by more perfect methods of determ-
ining the direction and length of the ship's course, increased
the knowledge of those lines of no variation which, as I have
already remarked. Father Acosta had endeavored to reduce
into a system.* Cornelius Van Schouten indicated, in 1616,
points lying in the midst of the Pacific and southeast of the
Marquesas Islands in which the variation was null. Even noAy
there lies in this region a singular, closed system of isogonic
lines, in which every group of the internal concentric curves
indicates a smaller amount of variation.t The emulation
which was exhibited in trying to find methods for determin-
ing longitudes, not only by means of the variation, but also
by the inclination (which, when it was observed under a
cloudy, starless sky, aere caliginoso^X was said by Wright to
be "worth much gold"), led to the multiplication of instru-
ments for magnetic observations, while it tended, at the same
time, to increase the activity of the observers. The Jesuit
Cabeus of Ferrara, Ridley, Lieutaud (1668), and Henry Bond
(1676), distinguished themselves in this manner. Indeed,
the contest between the latter and Beckborrow, together
with Acosta's view that there were four lines of no variation
which divided the entire surface of the earth, may very prob-
ably have had some influence on the theory advanced in 168.3
by Halley, of four magnetic poles or points of convergence.

Halley is identified with an important epoch in the history
of terrestrial magnetism. He assumed that there was in
each hemisphere a magnetic pole of greater and lesser intens-
ity, consequently four points with 90° incHnation of the
needle, precisely as we now find among the four points of
greatest intensity an analogous inequality in the maximum
of intensity for each hemisphere, that is to say, in the rapid-
ity of the oscillations of the needle in the direction of the
magnetic meridian. The pole of greatest intensity was situ-

* Historia Natural de las Indias, lib. i., cap. 17. *

t Cosmos, vol. i., p. 181.

X In the very careful obsen'ations of inclination which I made on the
Pacific, I demonstrated the conditions under which an acquaintance
with the amount of the inclination may be of important practical util-
ity in the determination of the latitude during the prevalence, on the
coasts of Peru, of the Garua, when both the sun and stars are obscured
(Cosmos, vol. i., p. 180). The Jesuit Cabeus, author of the Philoso-
phia Magnetica (in qua nova quaedam pyxis explicatur, <\n&i poll eleva-
tionem ubique demonstrat), drew attention to this fact during the first
half of the 1 7th ccnturv.

60 COSMOS.

ated, according to Halley, in 70° S. lat., 120° east of Green-
wich, and therefore almost in the meridian of King George's
Sound in New Holland (Nuyts Land).* Halley's three voy-
ages, which were made in the years 1698, 1699, and 1702,
were midertaken with the view of elaborating a theory which
must have owed its origin solely to the earlier voyage which
he had made seven years before to St. Helena, and to the
imperfect observations of variation made by Baffin, Hudson,
and Cornelius van Schouten. These were the first expedi-
tions which were equipped by any government for the estab-
lishment of a great scientific object — that of observing one of
the elements of terrestrial force on which the safety of navi-
gation is especially dependent. As Halley penetrated to 52°
south of the equator, he was able to construct the first cir-
cumstantial variation chart, which affords to the theoretical
labors of the 19 th century a point of comparison, although
certainly not a very remote one, of the advancing' movement
of the curves of variation.

Halley's attempt to combine graphically together by lines
different points of equal variation was a very happy one,t
since it has given us a comprehensive and clear insight into
the connection of the results already accumulated. My iso-
thermal lines (that is to say, lines of equal heat or mean an-
nual summer and winter temperature), which were early re-
ceived with much favor by physicists, have been formed on
a similar plan to Halley's isogonic curves. These lines, es-
pecially since they have been extended and greatly improved
by Dove, are intended to afford a clear view of the distribu-
tion of heat on the earth's surface, and of the principal de-
pendence of this distribution on the form of the solid and fluid
parts of the earth, and the reciprocal position of continental
and oceanic masses. Halley's purely scientific expeditions
stand so much the more apart from others, since they were
not, like many later expeditions, fitted out at the expense of
the government with the object of making geographical dis-
coveries, Ih addition to the results which they have yielded

* Edmund Halley, in the Philos. Transact, for 1683, vol. xii., No.
118, p. 216.

t Lines of tins kind, which he called tractus chalyhoeUtlcos, were
marked down upon a chart by Father Christopher Burrus in Lisbon,
and offered by him to the King of Spain for a large sum of money ;
these lines being drawn for the purpose of showing and determining
longitudes at sea. See Kircher's Magnes, ed. 2, p. 443. The first va-
riation chart, which was made in 1530, has already been referred to
in the text (p. 56).

THE MAGNETIC POLES. 61

in respect to terrestrial magnetism, they were also the means
of affording us an important catalogue of southern stars as
the fruits of Haliey's earlier sojourn in the island of St. He-
lena in the years 1677 and 1678. This catalogue was, more-
over, the first that was drawn up after telescopes had been
combined, according to Morin's and Gascoigne's methods,
with instruments of measurement.*

As the 17th century had been distinguished by an advance
in a more thorough knowledge of the position of the lines of
variation, and by the first theoretical attempt to determine
their points of convergence, viz., the magnetic poles, the 18th
century was characterized by the discovery of horary period-
ical alterations of variation. Graham has the incontestable
merit of being the first to observe (London, 1722) these hour-
ly variations with accuracy and persistency. Celsius and Hi-
orter in Upsala,t who maintained a correspondence with him,
contributed to the extension of our knowledge of this phe-
nomenon. Brugmans, and after him Coulomb, who was en-
dowed with higher mathematical powers, entered profoundly
into the nature of terrestrial magnetism (1784-1788). Their
ingenious physical experiments embraced the magnetic attrac-
tion of all matter, the local distribution of the force in a mag-
netic rod of a given form, and the law of its action at a dis-
tance. In order to obtain accurate results, the vibrations of
a horizontal needle suspended by a thread, as well as deflec-
tions by a torsion balance, were in turn employed.

The knowledge of the difference of intensity of terrestrial
magnetism at different points of the earth's surface by the
measurement of the vibrations of a vertical needle in the
magnetic meridian is due solely to the ingenuity of the Cheva-
lier Borda — not from any series of specially successful ex-
periments, but by a process of reasoning, and by the decided
influence which he exerted on those who were equipping
themselves for remote expeditions. Borda's long-cherished
conjectures were first confirmed by means of observations

* Twenty years after Halley had drawn up his catalogue of south-
ern stars at St. Helena (which, unfortunately, included none under
the sixth magnitude) Hevelius boasted, in his Firmamentwn Sohescia-
num, that he did not employ any telescope, but observed the heavens
through fissures. Halley, who, during his visit to Dantzic in 1679,
was present at these observations, praises their exactness somewhat
too highly. Cosmos, vol. iii., p. 42.

t Traces of the diurnal and horary variations of the magnetic force
had been observed in London as early as 1634. by Hellibrand, and in
Siam by Father Tachard, in 1682.

62 COSMOS.

made from the year 1785 to 1787, by Lamanon, the com-
panion of La Perouse. These results remained unknown,
unheeded, and unpublished, although they had been commu-
nicated as early as the summer of the last-named year to
Condorcet, the Secretary of the Academic des Sciences. The
first, and therefore certainly an imperfect knowledge of the
important law of the variability of intensity in accordance
with the magnetic latitude, belongs undoubtedly* to the un-
fortunate but scientifically equipped expedition of La Perouse ;
but the law itself, as I rejoice to think, was first incorporated
in science by the publication of my observations, made from
1798 to 1804, in the south of France, in Spain, the Canary
Islands, the interior of tropical America both north and
south of the equator, and in the Atlantic and Pacific oceans.
The successful expeditions of Le Gentil, Feuillee, and La-
caille; the first attempt made by Wilke, in 1768, to con-
struct an inclination chart; the memorable circumnaviga-
tions of Bougainville, Cook, and Vancouver, have all tended,
although by the help of instruments possessing veiy unequal
degrees of exactness, to establish the previously neglected but
very important element of inclination at various intervals of
time, and at many different points — the observations being
made more at sea, and in the immediate vicinity of the ocean,
than in the interior of continents. Toward the close of the
18th century, the stationary observations of declination which
were made by Cassini, Gilpin, and Beaufoy (from 1784 to
1790), with more perfect instruments, showed definitely that
there is a periodical influence at difierent hours of the day,
no less than at different seasons of the year — a discovery
which imparted a new stimulus to magnetic investigations.

In the 19th century, half of which has now expired, this
increased activity has assumed a special character differing
from any that has preceded it. We refer to the almost si-
multaneous advance that has been made in all branches of
the theory of terrestrial magnetism, comprising the numeric-
al determination of the intensity, inclination, and variation
of the force ; in physical discoveries in respect to the excita-
tion and the amount of the distribution of magnetism ; and

* Cosmos, vol. i., p. 185-187. The admirable construction of the
inclination compass made by Lenoir, according to Borda's plan, the
possibility of having long and free oscillations of the needle, the much
diminished friction of the pivots, and the correct adjustment of instru-
ments provided with scales, have been the means of enabling us accu-
ratelv to measure the amount of the terrestrial force in different zones.

PROGRESS IN MAGNETISM. 63

in the first and brilliant suggestions of a theory of terrestrial
magnetism, which has been based by its founder, Friedrich
Gauss, upon strictly mathematical combinations. The means
which have led to these results are improvements in the in-
struments and methods employed ; scientific maritime expe-
ditions, which in number and magnitude have exceeded those
of any other century, and which have been carefully equipped
at the expense of their respective governments, and favored
by the happy choice both of the commanders and of the ob-
servers who have accompanied them ; and various expeditions
by land, which, Jiaving penetrated far into the interior of
continents, have been able to elucidate the phenomena of
terrestrial magnetism, and to establish a large number of
fixed stations situated in both hemispheres in corresponding
north and south latitudes, and often in almost opposite lon-
gitudes. These observatories, which are both magnetic and
meteorological, form, as it were, a net-work over the earth's
surface. By means of the ingenious combination of the ob-
servations which have been published at the national expense
in liussia and ^igland, important and unexpected results
have been obtained. The establishment of a law regulating
the manifestation of force which is a proximate, although
not the ultimate, end of all investigations, has been satis-
factorily effected in many individual phases of the phenome-
non. All that has been discovered by means of physical ex-
periments concerning the relations which terrestrial magnet-
ism bears to excited electricity, to radiating heat and to light,
and all that we may assume in reference to the only lately
generalized phenomena of diamagnetism, and to that specific
property of atmospheric oxygen — polarity — opens, at all
events, the cheering prospect that w.e are drawing nearer to
the actual nature of the magnetic force.

In order to justify the praise which we have generally ex-
pressed in reference to the magnetic labors of the first half
of our century, 1 will here, in accordance with the nature
and form of the present work, briefly enumerate the principal
sources of our information, arranging them in some cases
chronologically, and in others in groups.*"

1803-1806. Krusenstern's voyage round the world (1812);

* The dates with which the following table begins (as, for instance,
from 1803-180G) indicate the epoch of the observation, while ihe fig-
ures which are marked in parenthesis, and appended to the titles of
the works, indicate the date of their pubUcation, which was frequentlj
much later.

64 COSMOS.

the magnetic and astronomical portion was by Horner (bd.
iii., s. 317).

1804. Investigation of the law of the increase in the in-
tensity of terrestrial magnetic force from the magnetic equa-
tor northward and southward, based upon observations made
from 1799 to 1804. (Humboldt, Voyage aux Regions Equi-
noxiales du Nouveau Continent, t. iii., p. 615-623 ; Lametherie,
Journal de Physique, t. Ixix., 1804, p. 433 ; the first sketch
of a chart showing the intensities of the force, Cosmos, vol. i.,
p. 185.) Later observations have shown that the minimum
of the intensity does not correspond to the magnetic equator,
and that the increase of the intensity in both hemispheres
does not extend to the magnetic pole.

1805-1806. Gay-Lussac and Humboldt, Observations of
Intensity in the south of France, Italy, Switzerland, and
Germany. Memoires de la Societe d'Arcueil, t. i., p. 1-22.
Compare the observations of Quetelet, 1830 and 1839, with
a *' Carte de I'intensite magnetique horizontale entre Paris
et Naples," in the Mem. de'VAcad. de Bruxelles, t. xiv. ; the
observations of Forbes in Germany, Flanders, and Italy, in
1832 and 1837 {Transact, of the Royal Soc of Edinburgh, \o\.
XV., p. 27); the extremely accurate observations of Eiidberg
in France, Germany, and Sweden, 1832 ; the observations
of Dr. Bache (Director of the Coast Survey of the United
States), 1837 and 1840, at twenty-one stations, both in refer-
ence to inclination and intensity.

1806-1807. A long series of observations at Berlin on
the horary variations of declination and the recurrence of
magnetic storms (perturbations), by Humboldt and Oltmanns,
mainly at the periods of the solstices and equinoxes for five
and six, or even sometimes nine days, and as many nights
consecutively, by means of Prony's magnetic telescope, which
allowed arcs of seven or eight seconds to be distinguished.

1812. Morichini, of Rome, maintained that non-magnetic
steel-needles become magnetic by contact with the violet rays
of light. Regarding the long contention excited by this as-
sertion, and the ingenious experiments of Mrs. Somerville,
together with the wholly negative results of Riess and Moser,
see Sir David Brewster, Treatise on Magnetism, 1837, p. 48.

1815-1818.") ^, , . . ,. r.^,,

^ono_i oop f The two circumnavigation voyages ot Otto

von Kotzebue : the first in the Ruric ; the second, five years
later, in the Predprijatie.

1817-1848. The series of great scientific n/aritime expe-

ARCTIC EXPEDITIONS. 66

ditions equipped by the French government, and which yield-
ed such rich results to our knowledge of terrestrial magnet-
ism— beginning with Freycinet's voyage in the corvette
Uranie, 1817-1820; and followed by Duperrey in the frig-
ate La Coquille, 1822-1825 ; Bougainville in the frigate
Thetis, 1824-1826 ; Dumont d'Urville in the Astrolabe, 1826-
1829, and to the south pole in the Zelee, 1837-1840 ; Jules
De Blosseville to India, 1828 (Herbert, Asiat Researches, vol.
xviii., p. 4 ; Humboldt, Asie Cent, t. iii., p. 468), and to Ice-
land, 1833 (Lottin, Voy. de la liecherche, 1836, p. 376-409);
Du Petit Thouars with Tessan in the Venus, 1837-1839 ;
De Vaillant in the Bonite, 1836-1837 ; the voyage of the
" Commission Scientifique du Nord" (Lottin, Bravais, Mar-
tins, Siljestrom) to Scandinavia, Lapland, the Faroe Islands,
and Spitzbergen in the corvette La Recherche, 1835-1840;
Be'rard to the Gulf of Mexico and North America, 1838 —
to the Cape of Good Hope and St. Helena, 1842 and 1846
(Sabine, in the Phil. Transact, for 1849, pt. ii., p. 175) ; and
Francis de Castlenau, Voyage dans les parties Centrales de
VAmerique du Sud, 1847-1850.

1818-1851. The series of important and adventurous ex-
peditions in the Arctic Polar Seas through the instrument-
ality of the British government, first suggested by the praise-
worthy zeal of John Barrow ; Edward Sabine's magnetic
and astronomical observations in Sir John Ross's voyage to
Davis' Straits, Baffin's Bay, and Lancaster Sound in 1818,
as well as in Parry's voyage in the Hecla and Griper, through
Barrow Straits to Melville Island, 1819-1820; Franklin,
Richardson, and Back, 1819-1822, and again frpm 1825-
1827 ; Back alone from 1833-1835, when almost the only
food that the expedition could obtain for weeks together was
a lichen {Gyrophora pustulatd), the "Tripe de Roche" of the
Canadian hunters, which has been chemically analyzed by
John Stenhouse in the Phil. Transact, for 1849, pt. ii., p. 393 ;
Parry's second exjJfedition with Lyon in the Fu7y and Hecla,
1821-1823; Parry's third voyage with James Ross, 1824-
1825 ; Parry's fourth voyage, when he attempted, with Lieu-
tenants Foster and Crozier, to penetrate northward from
Spitzbergen on the ice in 1827, when they reached the lati-
tude 82° 45^ ; John Ross, together with his accomplished
nephew James Ross, in a second voyage undertaken at the
expense of Felix Booth, and which was rendered the more
perilous on account of protracted detention in the ice, name-
ly, from 1829 to 1833 ; Dease and Simpson of the Hudson's

66 COSMOS.

Bay Company, 1838-1839 ; and more recently, in search of
Sir John Franklin, the expeditions of Captains Ommanney,
Austin, Fenny, Sir John lioss, and Phillips, 1850 and 1851.
The expedition of Captain Penny reached the northern lat-
itude of 77° 6^ Victoria Channel, into which Wellington
Channel opens.

1819-1821. Bellinghausen's Voyage into the Antarctic
Ocean.

1819. The appearance of the great work of Hansteen On
the Magnetism of the Earth, which, however, was completed
as early as 1813. This work has exercised an undoubted
influence on the encouragement and better direction of geo-
magnetic studies, and it was followed by the author's gener-
al charts of the curves of equal inclination and intensity for
a considerable part of the earth's surface.

1819. The observations of Admirals Koussin and Givry
on the Brazilian coasts, between the mouths of the rivers
Maranon and La Plata.

1819-1820. Oersted made the great discovery of the fact
that a conductor that is being traversed by a closed electric
current exerts a definite action upon the direction of the
magnetic needle according to their relative positions, and as
long as the current continues uninterrupted. The earliest
extension of this discovery (together with that of the ex-
hibition of metals from the alkalies and that of the two
kinds of polarization of light — probably the most brilliant
discovery of the century)* was due to Arago's observation,
that a wire through which an electrical current is passing,
even when made of copper or platinum, attracts and holds
fast iron filings like a magnet, and that needles introduced
into the interior of a galvanic helix become alternately
charged by the opposite magnetic poles in accordance with
the reversed direction of the coils {Ann. de Chim. et de Phys.,
t. XV., p. 93). The discovery of these phenomena, which
were exhibited under the most varied moflifications, was fol-
lowed by Ampere's ingenious theoretical combinations re-
garding the alternating electro-magnetic actions of the mole-
cules of ponderable bodies. These combinations were con-
firmed by a series of new and highly ingenious instruments,
and led to a knowledge of the laws of many hitherto appar-
ently contradictory phenomena of magnetism.

1820-1824. Ferdinand von Wrangel's and Anjou's expe-

* Malus's (1808) and Arago's (1811) ordinary and chromatic polari-
zation of Light. See Cosmos, vol. ii., p. 332.

' MAGNETIC OBSERVATIONS. 67

•

dition to the north coasts of Siberia and to the Frozen Ocean.
(Important phenomena of polar light; see th. ii., s. 259.)

1820. Scoresby's Account of the Arctic Regions ; experi-
ments of magnetic intensity, vol. ii., p. 537-554.

1821. Seebeck's discovery of thermo-magnetism and ther-
mo-electricity. The contact of two unequally warmed metals
(especially bismuth and copper), or differences of temperature
in the individual parts of a homogeneous metallic ring, were
recognized as sources of the production of magneto-electric
currents.

1821-1823. Weddell's Voyage into the Antarctic Ocean
as far as lat. 74° 15^.

1822-1823. Sabine's two important expeditions for the
accurate determination of the magnetic intensity and the
length of the pendulum in different latitudes (from the east
coasts of Africa to the equator, Brazil, Havana, Greenland
as far as lat. 74° 23^, Norway and Spitzbergen in lat. 79°
.50^). The results of these very comprehensive operations
were first published in 1824, under the title of Account of
Experiments to determine the Figure of 'the Earthy p. 460-509.

1824. Erikson's Magnetic Observations along the shores
of the Baltic.

1825. Arago discovers Magnetism of dotation. The first
suggestion that led tO this unexpected discovery was afford-
ed by his observation on the side of the hill in Greenwich
Park of the decrease in the duration of the oscillations of an
inclination-needle by the action of neighboring non-magnetic
substances. In Arago's rotation experiments the oscillations
of the needle were affected by water, ice, glass, charcoal, and
mercury.*

1825-1827. Magnetic Observations by Boussingault in
different parts of South America (Marmato, Quito).

1826-1827. Observations of Intensity by Keilhau at 20
stations (in Finmark, Spitzbergen, and Bear Island), by
Keilhau and Boeck, in Southern Germany and Italy (Schum.,
Ast7\ Nachj\, No. 146).

1826-1829. Admiral Lutke's Voyage Round the World;
the magnetic part was most carefully prepared in 1834 by
Lenz (see Partie Nautique du Voyage, 1836).

1826-1830. Captain Philip Parker King's Observations
in the southern portions of the eastern and western coasts
of South America (Brazil, Montevideo, the Straits of Ma-
gellan, Chili, and Valparaiso).

* Cosmoa, vol. i., p. 1^9.

68

COSMOS.

1827-1839. Quetelet, Etat du Magnetise Terrestre {Brvx-
elks) pendant douze dnnees. Very accurate observations.

1827. Sabine, On the determination of the relative in-
tensity of the magnetic terrestrial force in Paris and London.
An analogous comparison between Paris and Christiana was
made by Hansteen in 1825-1828 {Meeting of the British As-
sociation at Liverpool, 1837, p. 19-23). The many results
of intensity which had been obtained by French, English,
and Scandinavian travelers now first admitted of being
brought into numerical connection with oscillating needles,
which had been compared together at the three above-
named cities. These numbers, which could, therefore, now
be established as relative values, were found to be for Paris,
1-348, as determined by myself; for London, 1-372, by Sa-
bine; and for Christiana, 1-423, by Hansteen. They all
refer to the intensity of the magnetic force at one point of
the haagnetic equator (the curve of no inclination), which in-
tersects the Peruvian Cordilleras between Micuipampa and
Caxamarca, in south latitude 7° 2^, and western longitude
78° 48^, where the intensity was assumed by myself as —
1-000. This assumed standard (Humboldt, Recueil d' Ohserv.
Astr., vol. ii., p. 382-385 ; and ' Voyage. aiix Regions Equin,,
t. iii., p. 622) formed the basis, for forty years, of the reduc-
tions given in all tables of intensity (Gay-Lussac, in the
Mem. de la Societe d'Arcueil, t. i., 1807, p. 21 ; Hansteen,
On the Magnetism of the Earth, 1819, p. 71; Sabine, in the
Report of the British Association at Liverpool, p. 43-58). It
has, however, in recent times been justly objected to on ac-
count of its want of general applicability, because the line
of no inclination* does not connect together the points of

* "Before the practice was adopted of determining absolute values,
the most generally used scale (and which still continues to be very fre-
quently referred to) was founded on the time of vibration observed by
M. de Humboldt, about the commencement of the present century, at
a station in the An^es of South America, where the direction of the
dipping-needle was horizontal, a condition which was for some time
erroneously supposed to be an indication of the minimum of magnetic
force at the earth's surface. From a compai-ison of the times of vibra-
tion of M. de Humboldt's needle in South America and in Paris, the
ratio of the magnetic force at Paris to what Avas supposed to be its
minimum was inferred (1-348), and from the results so obtained, com-
bined with a similar comparison made by myself between Paris and
London in 1827, with several magnets, the ratio of the force in Lon-
don to that of M. de Humboldt's original station in South America
has been inferred to be 1-372 to 1*000. This is the origin of the num-
ber 1-372, which has been generally employed by British observers.

MAGNETIC OBSERVATIONS. . 69

feeblest intensity (Sabine, in the Phil. Transact, for 1846, pt.
iii., p. 254 ; and in the Manual of Scient. Inquiry for the use
of the British Navy, 1849, p. 17).

1828-1829. The Voyage of Hansteen and Due: Magnetic
observations in European Russia, and in Eastern Siberia as
far as Irkutsk.

1828-1830. Adolf Erman's voyage of circumnavigation,
v^^ith his journey through Northern Asia, and his passage
across both oceans, in the Russian frigate Krotkoi. The
identity of the instruments employed, the uniformity of the
methods, and the exactness of the astronomical determina-
tions of position, will impart a permanent scientific reputa-
tion to this expedition, which was equipped at the expense
of a private individual, and conducted by a thoroughly well-
informed and skillful observer. See the General Declination
Chart, based upon Erman's observations in the Report of the
Committee relative to the Arctic Expedition, 1840, pi. 3.

1828-1829. Humboldt's continuation of the observations
begun in 1800 and 1807, at the time of the solstices and
equinoxes regarding horary declination and the epochs of
extraordinary perturbations, carried on in a magnetic pavil-
ion specially erected for the purpose at l5erlin, and provided
with one of Gambey's compasses. Corresponding measure-
ments were made at St. Petersburg, Nikolajew, and in the
mines of Freiberg, by Professor Reich, 227 feet below the
surface of the soil. Dove and Riess continued these observ-
ations in reference to the variation and intensity of the hori-
zontal magnetic force till November, 1830 (Poggend.,^w7ia-
len, bd. xv., s. 318-336 ; bd. xix., s. 375-391, with 16 tab. ;
bd. XX., s. 545-555).

1829-1834. The botanist David Douglas, who met liis
death in Owhyhee by falling into a trap in which a wild
bull had previously been caught, made an admirable series of
observations on declination and intensity along the north-
west coast of America, and upon the Sandwich Islands as
far as the margin of the crater of Kiraueah (Sabine, Rep. of
the Meeting of the British Association at Liverp9ol, p. 27-32).

«

By absolute measurements we are not only enabled to compare numer-
ically with one another the results of experiments made in the most
distant parts of the globe, with apparatus not previously compared,
but we also furnish the means of comparing hereafter the intensity
which exists at the present epoch with that which may be found at
future periods." Sabine, in the Manual for the use of the British Navy,
1849, p. 17.

70 COSMOS.

1829. KupfFer, Voyage au Mont Elhrouz dans le Caucase,
p. 68-115.

1829. Humboldt's magnetic observations on terrestrial
magnetism, with the simultaneous astronomical determina-
tions of position in an expedition in Northern Asia, under-
taken by command of the Emperor Nicholas, between the
longitudes 11° 3^ and 80° 12^ east of Paris, near the Lake
Dzaisan, as well as between the latitudes of 45° 43'' (the
island of Birutschicassa, in the Caspian Sea) to 58° 52^, in
the northern parts of the Ural district, near Werchoturie
(Asie Centrak, t. iii., p. 440-478).

1829. The Imperial Academy of Sciences at St. Peters-
burg acceded to Humboldt's suggestion for the establish-
ment of magnetic and meteorological stations in the different
climatic zones of European and Asiatic Russia, as well as
for the erection of a physical central observatory in the capi-
tal of the empire under the efficient scientific direction of
Professor Kupffer. (See Cosmos, vol. i., p. 190. KupfFer,
Rapport Adresse a VAcad. de St. Petershourg relatif a VObser-
vatoire physique central, fonde aupres da Coips des Mines, in
Schum., Astr. Nachr., No. 726; and in his Annates Magne-
tiques, p. xi.) Through the continued patronage which the
Finance Minister, Count Cancrin, has awarded to every
great scientific undertaking, a portion of the simultaneously
corresponding observations* between the White Sea and the

* The first idea of the utility of a systematic -and simultaneously
conducted series of magnetic observations is due to Celsius, and, with-
out referring to the discovery and measurement of the influence of
polar light on magnetic variation, which was, in fact, due to his as-
sistant, Olav Hiorter (March, 1741), we may mention that he was the
means of inducing Graham, in the summer of 1741, to join him in his
investigations for discovering whether certain extraordinary perturba-
tions, which had from time to time exerted a horary influence on the
course of the magnetic needle at Upsala, had also been observed at
the same time by him in London. A simultaneity in the perturba-
tions afforded a proof, he said, that the cause of these disturbances is
extended over considerable portions of the earth's surface, and is not
dependent upon accidental local actions (Celsius, in Svenska Veten-
skaps Academiens Handlingar for 1740, p. 44; Hiorter, op. cit., 1747,
p. 27). As Arago had recognized that the magnetic perturbations,
owing fo polar light, are diffused over districts in which the phenom-
ena of light which accompany magnetic storms have not been seen,
he devised a plan by which he was enabled to carry on simultaneous
horary observations (in 1823) with our common friend Kupffer at
Kasan, which lies almost 47° east of Paris. Similar simultaneous ob-
servations of declination were begun in 1828 by myself, in conjunction
with Arago and Reich, at Berlin, Paris, and Freiberg (see Poggend.,
Annalen^ bd. xix., s. 337).

MAGNETIC OBSERVATIONS. 71

Crimea, and between the Gulf of Finland and the shores of
the Pacific, in Russian America, were begun as early as
1832. A permanent magnetic station was established in
the old monastery at Pekin, which from time to time, since
the reign of Peter the Great, has been inhabited by monks
of the Greek Church. The learned astronomer, Fuss, who
took the principal part in the measurements for the determ-
ination of the difference of level between the Caspian and
the Black Sea, was chosen to arrange the first magnetic es-
tablishments in China. At a subsequent period, Kupffer, in
his Voyage of Circumnavigation, compared together all the
instruments that had been employed in the magnetic and
meteorological stations as far east as Nertschinsk in 119°
36^ longitude, and with the fundamental standards. The
magnetic observations of Fedorow, in Siberia, which are no
doubt highly valuable, are still unpublished.

1830-1845. Colonel Graham, of the topographical engi-
neers of the United States, made observations on the mag-
netic intensity at the southern boundary of Canada {Phil
Transact for 1846, pt. iii., p. 242).

1830. Fuss, Magnetic, Astronomical, and Plypsometrical
Observations on the journey from the Lake of Baikal,
through Ergi-Oude, Durma, and the Gobi, which lies at an
elevation of only 2525 feet, to Pekin, in order to establish
the magnetic and meteorological observatory in that city,
where Kovanko continued for ten years to prosecute his ob-
servations {Rep. of the Seventh Meeting of the Brit. Assoc,
1837, p. 497-499 ; and Humboldt, Asie Centrale, t. i., p. 8 ;
t. ii., p. 141 ; t. iii., p. 468, 477>

1831-1836. Captain Fitzroy, in his voyage round the
world in the Beagle, as well as in the survey of the coasts
of the most southern portions of America, with a Gambey's
inclinatorium and oscillation needles supplied by Hansteen.

1831. Dimlop, Director of the Observatory of Paramatta,
Observations on a voyage to Australia {Phil. Transact, for
1840, pt. i., p. 133-140).

1831. Faraday's induction-currents, whose theory has
been extended by Nobili and Antinori. The great discov-
ery of the development of light by magnets.

1833 and 1839 are the two important epochs of the first
enunciation of the theoretical views of Gauss : (1) Intensitas
vis magneticae terrestris ad mensuram absolutam revocata,
1833 ; (p. 3 : '• elementum tertium, intensitas, usque ad
tempora recentiora penitus neglectum mansit") ; (2) the im-

72 COSMOS.

mortal work on *' the general theory of terrestrial magnet-
ism" (see Results of the Observations of the Magnetic As-
sociation in the year 1838, edited by Gauss and Weber,
1839, p. 1-57).

1833. Observations of Barlow on the attraction of the
ship's iron, and the means of determining its deflecting ac-
tion on the compass ; Investigation of electro-magnetic cur-
rents in Terrellas ; Isogonic atlases. (Compare Barlow's
Essay on Magnetic Attraction, 1833, p. .89, with Poisson, sur
les deviations de la boussole produite par le fer des vaisseaux, in
the Jlem. de Vlnstitut, t. xvi., p. 481-555 ; Airy, in the Phil.
Transact, for 1839, pt. i., p. 167; and for 1843, pt. ii., p.
146 ; Sir James Ross, in the Phil. Transact, for 1849, pt.
ii., p. 177-195).

1833. Moser's methods of ascertaining the position and
force of tha variable magnetic pole (Poggend., Annalen, bd.
xxviii., s. 49-296).

1833. Christie on the Arctic observations of Captain Back,
Phil. Transact, for 1836, pt. ii., p. 377. (Compare also his
earlier and important treatise.in the Phil. Transact, for 1825,
pt. i., p. 23.)

1834. Parrot's expedition to Ararat {Magnetismus, bd. ii.,
s. 53-64).

1836. Major Estcourt, in the expedition of Colonel Ches-
ney on the Euphrates. A portion of the observations on
intensity were lost with the steamer Tigris, which is the
n^ore to be regretted, since we are entirely deficient in accu-
rate observations of this portion of the interior of Western
Asia, and of the regions lying south of the Caspian Sea.

1836. Letter from M. A. de Humboldt to his Royal High-
ness Duke of Sussex, President of the Royal Society of
London, on the proper means of improving our knowledge
of terrestrial magnetism by the establishment of magnetic
stations and corresponding observations (April, 1836"). On
the happy results of this appeal, and its influence on the
great Antarctic expedition of Sir James Ross, see Cosmos,
vol. i., p. 192, and Sir James Ross's Voyage to the Southern
and Antarctic Regions^ 1847, vol. i., pt. xii.

1837. Sabine, On the Variations of the Magnetic Intensity
of the Earth, in the Report of the Seventh Meeting of the Brit-
ish Association at Liverpool, p. 1-85 : the most complete work
of the kind.

1837-1838. Erection of a magnetic observatory at Dub-
lin, by Professor Humphrey Lloyd. On the observations

MAGNETIC OBSERVATIONS. 73

made there from 1840 to 1846 (see Transact, of tlie Royal
Irish Academy, vol. xxii., pt. i., p. 74-96).

1837. Sir David Brewster, A Treatise on Magnetism, p.
185-263.

1837-1842. Sir Edward Belcher's Voyage to Singapore,
the Chinese Seas, and the western coasts of America {Phil.
Transact, for 1843, pt. ii., p. 113, 140-142). These observ-
ations of inclination, when compared with my own, which
were made at an earlier date, show a very unequal advance
of the curves. Thus, for instance, in 1803, I found the in-
clinations at Acapulco, Guayaquil, and Callao de Lima to
be +38° 48", +10° 42", and -9° 54^ while Sir Edward
Belcher found +37° 57", +9° V, and -9° 54^ Can the
frequent earthquakes upon the Peruvian coasts exert a local
influence upon the phenomena which depend upon magnetic
force of the earth ?

1838-1842. "Charles Wilkes's Narrative of the United
States Exploiting Expedition, vol. i., p. xxi.

1838. Lieutenant James Sullivan's Voyage from Fal-
mouth to the Falkland Islands (Fhil. Transact, for 1840, pt.
i., p. 129, 140-143).

1838 and 1839. The establishment of magnetic stations
under the admirable superintendence of General Sabine in
both hemispheres, at the expense of the British government.
The instruments were dispatched in 1839, and the observa-
tions were begun at Toronto and in Van Diemen's Land in
1840, and at the Cape in 1841 (see Sir John Herschel in
the Quarterly Review, vol. Ixvi., 1840, p. 297 ; and Becque-
rel, Traite d' Electricite et de Magnetisme, t. vi., p. 173). By
the careful and thorough elaboration of these, valuable ob-
servations, which embrace all the elements or variations of
the magnetic activity of the earth. General Sabine, as super-
intendent of the Colonial observatories, discovered hitherto
unrecognized laws, and disclosed new views in relation to the
science of magnetism. The results of his investigations
were collected by himself in a long series of separate mem-
oi i (Contributions to Terrestrial Magnetism) in the Philo-
sophical Transactions of the Royal Society of London, and in
separate works, which constitute the basis of this portion of
the Cosmos. We will here indicate only a few of the most
important: (1) Observations on Days of unusual Magnetic Dis-
turbances (Storms) in the Years 1840 and 1841, p. 1-107 ;
and as a continuation of this treatise. Magnetic Storms from
1843-1845, in the Phil. Transact, for 1851, pt. i., p. 123-

VoL. v.— D

74

COSMOS.

139 ; (2) Observations made at the Magnetical Observatory at
Toronto, 1840, 1841, and 1842 (43° 39^ N. lat., and 81° 41''
W. long.), vol. i., p. xiv.-xxviii. ; (3) The very variable Direc-
tion of Magnetic Declination in one half of the Year at Long-
wood House, St. Helena (15° 55^ S. lat., 8° 3^ W. long.),
Philosophical Transactions for 1847, pt. i., p. 54; (4) Observ-
ations made at the Magnetical and Meteorological Observatoiy
at the Cape of Good Hope, 1841-1846 ; (5) Observations made
at the Magnetical and Meteorological Observatory at Hobarton
(42° 52" S. lat., 145° 7" E. long.), in Van DiemerHs -Land
and the Antarctic Expedition, vol. i. and ii. (1841-1848) ; On
the Separation of the Eastern and Western Disturbances, see
vol. ii., p. ix.-xxxvi. ; (6) Magnetic Phenomena within the
Antarctic Polar Circle, in KerguelerHs and Van DiemerCs Land
{Phil Transact, for 1843, pt. ii., p. 145-231); (7) On the
Isoclinal and Isodynamic Lines in the Atlantic Ocean, their Con-
dition in 1837 (Phil. Transact, for 1840, pt. T., p. 129-155);
(8) Basis of a chart of the Atlantic Ocean, which exhibits
the lines of magi\etic variation between 60° N. lat. and 60°
S. lat. for the year 1840 (Phil. Transact, for 1849, pt. ii., p.
173-233); (9) Methods of determining the absolute Values,
secular Change, and annual Variation of the Magnetic Force
(Phil. Transact, for 1850, pt. i.,»p. 201-219); Coincidence
of the epochs of the greatest vicinity of the sun with the
greatest intensity of the force in both hemispheres, and of
the increase of inclination, p. 216; (10) On the Amount of
Magnetic Intensity in the most Northern parts of the New Con-
tinent, and upon the Point of greatest Magnetic Force found by
Captain Lefroy in 52° 19" lat. (Phil. Transact, for 1846, pt.
iii., p. 237-336); (11) The periodic Alterations of the three
Elements of terrestrial Magnetism, Variation, Inclination, and
Intensity at Toronto and Hobarton, and on the Connection of the
decennial Period of Magnetic Alterations with the decennial
Period of the frequency of Solar Spots, discovered by Schwabe
at Dessau (Phil. Transact, for 1852, pt. i., p. 121-124). The
observations of variation for 1846 and 1851 are to be con-
sidered as a continuation of those indicated in No. 1, as be-
longing to the years 1840-1845.

1839. Representation of magnetic isoclinal and isodynam-
ic lines, from observations of Humphrey Lloyd, John Phil-
lips, Robert Were Fox, James Ross, and Edward Sabine.
As early as 1833 it was determined, at the meeting of the
British Association in Cambridge, that the magnetic inclin-
ation and intensity should be determined at several parts of

MAGNETIC OBSERVATIONS. 75

the empire, and in the summer of 1834 this suggestion was
fully carried out by Professor Lloyd and General Sabine,
and the operations of 1835 and 1836 were then extended to
Wales and Scotland {Report of the Meeting of the Brit. Assoc,
held at Newcastle, 1838, p. 49-196), with an isoclinal and
isodynamic chart of the British islands, the intensity at
London being taken as =1.

1838-1843. The^great exploring voyage of Sir James
Ross to the South Pole, which is alike remarkable for the
additions which it afforded to our knowledge by proving the
existence of hitherto doubtful polar regions, as well as for
the new light which it has diffused over the magnetic con-
dition of large portions of the earth's surface. It embraces
all the three elements of terrestrial magnetism numerically
determined for almost two thirds of the area of all the high
latitudes of the southern hemisphere.

1839-1851. Kreil's observations, which were continued
for twelve years, at the Imperial Observatory at Prague, in
reference to the variation of all the elements of terrestrial
magnetism, and of the conjectured soli-lunar influence.

1840. Horary magnetic observations with one of Gam-
bey's declination compasses during a ten years' residence in
Chili, by Claudio Gay (see his Historia fhica y politica de
Chile, 1847).

1840-1851. Lamont, Director of the Observatory at Mu-
nich. The results of his magnetic observations, compared
with those «f Gottingen, which date back as far as 1835.
Investigation of the important law of a decennial period* in

* Arago has left behind him a treasury of magnetical observations
(upward of 52,600 in number) carried on from 1818 to 1835, which
have been carefully edited by M. Fedor Thoman, and published in the
(Euvres Completes de Francois Arago (t. iv., p. 498). In these observ-
ations, for the series of years from 1821 to 1830, General Sabine has
discovered the most complete confirmation of the decennial period of
magnetic declination, and its correspondence with the same period, in
the alternate frequency and rarity of the solar spots {Meteorological Es-
says, London, 1855, p. 350). So early as the year 1850, when Schwabe
published at Dessau his notices of the periodical return of the solar
spots {Cosmos, vol. iv., p. 83), two years before Sabine first showed
the decennial period of magnetic declination to be dependent on tlie
solar spots (in March, 1852, Phil. Tr. for 1852, pt. i., p. 116-121 ; Cos-
mos, vol. v., p. 76, note), the latter had already discovered the import-
ant result that the sun operates on the earth's magnetism by the mag-
netic power proper to its mass. He had discovered (Phil. Tr. for 1850,
pt. i., p. 216; Cosmos, vol. v., p. 136) that the magnetic intensity is
greatest, and that the needle approaches nearest to the vertical direc-
tion, when the earth is nearest to the sun. The knowledge of such a

76 COSMOS.

the alterations of declination (see Lamont in Poggend., A}i?i.
der Phys., 1851, bd. 84, s. 572-582; and Relshuber, 1852,
bd. 85, s. 179-184). The already-indicated conjectural con-
nection between the periodical increase and decrease in the
annual mean for the daily variation of declination in the
magnetic needle, and the periodical frequency of the solar
spots, was first made known by General Sabine in the Phil.
Transact, for 1852; and four or five n^onths later, without
any knowledge of the previous observations, the same re-
sult was enunciated by Rudolf Wolf, the learned Director
of the Observatory at Berne.* Lamont's manual of terres-
trial magnetism, 1848, contains a notice of the newest meth-
ods of observation, as well as of the development of these
methods.

1840-1845. Bache, Director of the Coast Survey of the
United States, Observ. made at the Magn, and Meteorol. Ob-
servatoi-y at Girard College, Philadelphia (published in 1847).

1840-1842. Lieutenant Gilliss, U. S., Magnetical and Me-
teorological Observations made at Washington, published 1847,
p. 2-319; Magnetic Storms, p. 336.

1841-1843. Sir Robert Schomburgk's observations of
declination in the woody district of Guiana, between the
mountain Roraima and the village Pirara, between the par-
allels of 4° 57^ and 3° 39^ {Phil. Transact, for 1849, pt. ii.,
p. 217).

1841-1845. Magnet, and Meteorol. Observations made at
Madras.

magnetical operation of the central body of our planetary system, not
by its heat-producing quahty, but by its own magnetic power, as well
as by changes in the Photosphere (the size and frequency of funnel-
shaped openings), gives a higher cosmical Interest to the study of the
earth's magnetism, and to the numerous magnetic observatories {Cos-
mos, vol. i., p. 190; vol. v., p. 72) now planted over Kussia and North-
ern Asia, since the resolutions of 1829, and over the colonies of Great
Britain since 1840-1850. (Sabine, in the Proceedings of the Roy. Soc,
vol. viii., No. 25, p. 400; and in the Phil. Trans. Jor 1856, p. 362.)

* The treatise of Rudolf Wolf, referred to in the text, contains
special daily observation of the sun's spots (from January 1 to June
30, 1852) and a table of Lamont's periodical variations of declina-
tion, with Schwabe's results on the frequency of solar spots (1835-
1850). These results were laid before the meeting of the Physical
Society of Berne on the 31st of July, 1852, while the more compre-
hensive treatise of Sabine {Phil. Transact., 1852, p. 116-121) had
been presented to the Royal Society of London in the beginning of
March, and read in the beginning of May, 1852. From the most re-
cent investigations of the observations of solar spots, Wolf finds that
between the years 1600 and 1852 the mean period was 11-11 years.

MAGNETIC OBSERVATIONS. It

1843-1844. Magnetic observations in Sir Thomas Bris-
bane's observatory at Makerston, Koxburghshire, 55° 34^
N. lat. (see Transact, of the Royal Society of Edinb., vol. xvii.,
pt. ii., p. 188 ; and vol. xviii., p. 46).

1843-1849. Kreil, On the Ivfaence of the Alps upon the
Manifestations of the Magnetic Force (see Solium., Astr.
Nachr., No. 602).

1844-1845. Expedition of the Pagoda into high antarc-
tic latitudes, as far as 64° and 67°, and from 4° to 117° E.
long., embracing all the three elements of terrestrial mag-
netism, under the command of Lieutenant Moore, who had
already served in the Terror, in the polar expedition ; and
of Lieutenant Clerk, of the Royal Artillery, and formerly
Director of the Magnetic Observatory at the Cape. A
worthy completion of the labors of Sir James Ross at the
South Pole.

1845. Proceedings of the Magn. and Meteorol. Conference
held at Cambridge.

1845. Observations made at the Magn. and Meteorol. Observ-
atoi^ at Bombay, under the superintendence of Arthur Bed-
ford Orlebar. This observatory was erected in 1841, on the
little island of Colaba.

1845-1850. Six volumes of the Pesidts of the Magn. and
Meteorol. Observations made at the Royal Observatory at
Greenwich. The magnetic house was erected in 1838.

1845. SimonofF, Professor at Kazan, Pecherches sur V action
magnetique de la Terre.

1846-1849. Captain Elliot, Madras Engineers, Magnetic
Survey of the Eastern Archipelago. Sixteen stations, at each
of which observations were continued for several months in
Borneo, Celebes, Sumatra, the Nicobars, and Keeling isl-
ands, compared with Madras, between 16° N. lat. and 12°
S. lat., and 78° and 123° E. long. {Pkil. Transact, for 1851,
pt. i., p. 287-331, and also p. i.-clvii.). Charts of equal in-
clination and declination, which also expressed the horizon-
tal and total force, were appended to these observations,
which also give the position of the magnetic equator and of
the line of no variation, and belong to the most distinguish-
ed and comprehensive that had been drawn up in modern
times.

1845-1850. Faraday's brilliant physical discoveries: (1)
In relation to the axial or equatorial (diamagnetic*) direc-

* See Cosmos, vol. iv., p. 84. Diamagnetic repulsion and an equa-
torial, that is to say, an east and -west position in respect to a power-

78 COSMOS.

tion assumed by freely-oscillating bodies under external mag-
netic influences {Phil. Transact, for 1846, § 2420, and Phil.
Transact, for 1851, pt. i., § 2718-2796); (2) Kegarding the
relation of electro-magnetism to a ray of polarized light, and
the rotation of the latter by means of the altered molecular
condition of the bodies through which the ray of polarized
light and the magnetic current have both been transmitted
{Phil Transact, for 1846, pt. i., § 2195 and § 2215-2221);
(3) Regarding the remarkable property which oxygen (the
only gas which is paramagnetic) exerts on the dements of
terrestrial magnetism, namely, that like soft iron, although
in a much weaker degree, it assumes conditions of polarity
through the diffused action of the body of the earth, which
represents a permanently present magnet* {Phil Transact,
for 1851, pt. i., § 2297-2967).

ful magnet, are exhibited by bismuth, antimony, silver, phosphorus,
rock-salt, ivory, wood, apple-shavings, and leather. Oxygen gas, either
pure or when mixed with other gases, or when condensed in the inter-
stices of charcoal, is paramagnetic. See, in reference to crystallized
bodies, the ingenious observations made by Plucker concerning the
position of certain axes (Poggend., AnnaL^ bd. Ixxiii., s. 178; and
Phil Transact, for 1851, § 2836-2842). The repulsion by bismuth
was first recognized by Brugmans in 1788, next by Le Bailiff in 1827,
and, finally, more thoroughly tested by Seebeck in 1828. Faraday
himself (§ 2429-2431), Reich, and Wilhelm Weber, who, from the
year 1836, has shown himself so incessantly active in his endeavors to
promote the progress of terrestrial magnetism, have all endeavored to
exhibit the connection of diamagnetic phenomena with those of induc-
tion (Poggend., Annalen, bd. Ixxiii., s. 241-253). Weber has, more-
over, tried to prove that diamagnetism derives its source from Am-
pere's molecular currents. (Wilh. Weber, Ahhandlungen iiber electro-
dynamische Maassbesthnmungen, 1852, s. 545-570.)

* In order to excite this polarity, the magnetic fluids in every par-
ticle of oxygen must be separated, to a certain extent, by the actio in
distans of the earth in a definite direction, and with a definite force.
Every particle of oxygen thus represents a small magnet, and all these
small magnets react upon one another as well as upon the earth, and,
finally, in connection with the latter, they further act upon a magnet-
ic needle, which may be assumed to be in or beyond the atmosphere.
The envelope of oxygen that encircles our terrestrial sphere may be
compared to an armature of soft iron upon a natural magnet or a
piece of magnetized steel ; the magnet may further be assumed to be
spherical, like the earth, while the armature is assumed to be a hollow
shell, similar to the investment of atmospheric oxygen. The magnet-
ic power which each particle of oxygen may acquire by the constant
force of the earth diminishes with the temperature and the rarefaction
of the oxygen gas. When a constant alteration of temperature and
an expansion follows the sun around the earth from east to west, it
must proportionally alter the results of the magnetic force of the earth,
and of the oxygen investment ; and this, according to Faraday's opiu'

MAGNETIC OBSERVATIONS. 79

1849. Emory, Magnetic observations made at the Isth-
mus of Panama.

1849. Professor William Thomson, of Glasgow, A Mathe-
matical Theory of Magnetism, in the Phil. Transact, for 1851,
pt. i., p. 243-285. (On the problem of the distribution of
magnetic force, compare § 42 and 6Q, with Poisson, in the
Mem. de V Institute 1811, pt. i., p. 1 ; pt. ii., p. 163.)

1850. Airy, On the present state and prospects of the
science of Terrestrial Magnetism — the fragment of what
promises to be a most admirable treatise.

1852. Kreil, Influence of the Moon on Magnetic Declina-
tion at Prague in the years 1839-1849. On the earlier la-
bors of this accurate observer, between 1836 and 1,838, see
Osservazioni sulV intensita e sulla direzione della forza magnet-
ica instituite negli anni 1836-1838 aW I. R. Osservatorio di
Milano, p. 171 ; and also his Magneiical and Meteorological
Observations at Prague, vol. i., p. 59.

1852. Faraday, On Lines of Magnetic Force, and their
definite character.

1852. Sabine's new proof deduced from observations at
Toronto, Hobarton, St. Helena, and the Cape of Good Hope
(from 1841 to 1851), that every where between the hours of
seven and eight in the morning the magnetic declination ex-
hibits an annual period ; in which the northern solstice pre-
sents the greatest eastern elongation, and the southern sol-
stice the greatest western elongatien, without the temperature
of the atmosphere or the earth's crust evincing a maximum
or minimum at these turning periods. Compare the second
volume of the Observations made at Toronto, p. xvii., with
the two treatises of Sabine, already referred to, on the Influ-
ence of the sun's vicinity {Phil Transact, for 1850, pt. i.,
p. 216), and of the solar spots {Phil. Transact, for 1852,
pt. i., p. 121).

The chronological enumeration of the progress of our
knowledge of terrestrial magnetism during half a century,
which I have uninterruptedly watched with the keenest in-
terest, exhibits a successful striving toward the attainment

ion, is the origin of one part of the variations in the elements of ter-
restrial magnetism. Plucker finds that, as the force with which the
magnet acts upon the oxygen is proportional to the density of this gas,
the magnet presents a simple eudiometric means of recognizing the
presence of free oxygen gas in a gaseous mixture even to the 100th or
200th part.

80

COSMOS.

of a two-fold object. The greater number of these labors
have been devoted to the observation of the magnetic activi-
ty of our planet in its numerical relations to time and space,
while the smaller part belongs to experiments, and to the
manifestation of phenomena which promise to lead us to the
knowledge of the character of this activity, and of thcin-
ternal nature of the magnetic force. Both these methods —
the numerical observation of the manifestation of terrestrial
magnetism, both in respect to its direction and intensity —
and physical experiments on the magnetic force generally,
have tended reciprocally to the advancement of our physical
knowledge. Observations alone, independently of every hy-
pothesis regarding the causal connection of phenomena, or
regarding the hitherto immeasurable and unattainable recip-
rocal action of molecules in the interior of substances, have
led to important numerical laws. Experimental physicists
have succeeded, by the display of the most wondrous inge-
nuity, in discovering in solid and gaseous bodies polarizing
properties, whose presence had never before been suspected,
and which stands in special relation to the temperature and
pressure of the atmosphere. However important and un-
doubted these discoveries may be. '^hcj can not, in the pres-
ent condition of our knowledge, be regarded as satisfactory
grounds of explanation for the laws which have already been
recognized in the movements of the magnetic needle. The
most certain means of enabling us thoroughly to comprehend
the variable numerical relations of space, as well as to ex-
tend and complete that mathematical theory of terrestrial
magnetism which was so nobly sketched by Gauss, is to pros-
ecute simultaneous and continuous observations of all the
three elements of the magnetic force at numerous well-se-
lected points of the earth's surface. I have, however, else-
where illustrated, by example, the sanguine hopes which I
entertained of the great advantages that may be derived from
the combirtntion of experimental and mathematical investi-
gation.*

Nothing that occurs upon our planet can be supposed to
be independent ot cosmical influences. The word planet in-
stinctively leads us to the idea of dependence upon a central
body, and of a connection with a group of celestial bodies
of very different masses, which probably have a similar or-
igin. The influence of the sun's position upon the manifest-
ation of the magnetic force of the earth was recognized at a
* See p. 10.

HORARY VARIATION. 81

very early period. The most distinct intimation of this
relation was afforded by the discovery of horary variation,
although it had been obscurely perceived by Kepler, who, a
century before, had conjectured that all the axes of the plan-
ets were magnetically directed toward one portion of the uni-
verse. He says expressly, " that the sun may be a magnetic
body, and that on tliat account the force which impels the
planets may be centred in the sun."^ The attraction of
masses and gravitation appeared at that time under the
semblance of magnetic attraction. Horrebow,t who did not
confound gravitation with magnetism, was the first who
called the process of light a perpetual northern light, pro-
duced in ttie solar atmosphere by means of magnetic forces.
Nearer our own times (and this difference of opinion is very
remarkable) two distinct views were promulgated in refer-
ence to the nature of the influence exerted by the sun.

Some physicists, as Canton, Ampere, Christie, Lloyd, and
Airey, have assumed that the sun, without being itself mag-
netic, acts upon terrestrial magnetism merely by producing
changes of temperature, while others, as Coulomb, believed
the sun to be enveloped by a magnetic atmosphere,{ which
exerts an action on terrestrial magnetism by distribution.
Although Faraday's splendid discovery of the paramagnetic
property of oxygen gas has removed the great difficulty of
having to assume, with Canton, that the temperature of the
solid crust of the earth and of the sea must be rapidly and
considerably elevated from the immediate effect of the sun's
transit through the meridian of the place, the perfect co-or-
dination and an ingenious analysis of all the measurements
and observations of General Sabine have yielded this result,
that the hitherto observed periodic variations of the magnetic
activity of the earth %an not be based upon periodic changes

* Kepler, in Stella Martis, p. 32-34 (and compare with it his treat-
ise, Mysterium Cosmogr., cap. xx., p. 71).

t Cosmos, vol. iv., p. 77, where, however, in consequence of an
error of the press, in the place of Basis Astronomice we should read
Clavis Astronomice. The passage (§ 226) in which the luminous pro-
cess of the sun is characterized as a perpetual northern lin;ht does not
occur in the first edition of the Clavis Astr., by Horrebow (Havn.,
1730), but is only found in the second and enlarged new edition of
the work in Horrebow's Operum Mathematico-Physicorum, t. i., Havn.,
1740, p. 317, as it belongs to this appended portion of the Clavis.
Compare with Horrebow's view the precisely similar views of Sir Will-
iam and Sir John Herschel (Cosmos, vol. iii., p. 34).

J M^moires de Math^m, et de Phys. presentes d VAcad. Roy. des Sc.^
t. ix., 1780, p. 262.

D2

82 COSMOS.

of temperature in those parts of the atmosphere which are
accessible to us. Neither the principal epochs of diurnal and
annual alterations of declination at the different hours of the
day and night, nor the periods of the mean intensity of the
terrestrial force* coincide with the periods of the maxima
and minima of the temperature of the atmosphere, or of the
upper crust of the earth. We may remark that the annual
alterations were first accurately represented by Sabine from
a very large number of observations. The turning points in
the most important magnetic phenomena are the solstices
and the equinoxes. The epoch at which the intensity of the
terrestrial force is the greatest, and that at which the dip-
ping-needle most nearly assumes the vertical position in

* " So far as these four stations (Toronto, Hobarton, St. Helena,
and the Cape), so widely separated from each other and so diversely
situated, justify a^generalization, we may arrive at the conclusion that
at the hour of 7 to 8 A.M. the magnetic declination is every where
subject to a variation of which the period is a year, and which is every
where similar in character and amount, consisting of a movement of
the north end of the magnet from east to west between the northern
and the southern solstice, and a return from west to east between the
southern and the northern solstice, the ampHtude being about 5 min-
utes of arc. The turning periods of the year are not, as many might
be disposed to anticipate, those months in which the temperature at th^
surface of our planet, or of the subsoil, or of the atmosphere (as far as
we possess the means of judging of the temperature of the atmosphere)
attains its maximum and minimum. Stations so diversely situated would,
indeed, present in these respects thermic conditions of great variety ;
whereas uniformity in the epoch of the turning periods is a not less
conspicuous feature in the annual variation than "similarity of char-
acter and numerical value. At all the stations tKe solstices are the
turning periods of the annual variation at the hour of which we are
treating. The only periods of the year in which the diurnal or horary
variation at that hour does actually disappear are at the equinoxes,
when the sun is passing from the one hemisphere to the other, and
when the magnetic direction, in the course of its annual variation
from east to west, or vice versa, coincides with the direction which is
the mean declination of all the months and of all the hours. The
annual variation is obviously connected with, and dependent on, the
earth^s position in its orbit relatively to the sun around which it re-
volves ; as the diurnal variation is connected with, and dependent on,
the rotation of the earth on its axis, by which each meridian success-
ively passes through every angle of inclination to the sun in the round
of 24 hours." Sabine, On the Annual and Diurnal Variations, in the
second volume of Observations made at the Magnetic and Meteorological
Observatory at Toronto, p. xvii.-xx. See also his memoir, On the An-
nual Variation of the Magnetic Declination at different periods of the
Day, in the Philos. Transact, for 1851, pt. ii., p. 635, and the Intro-
duction of his Observations made at the Observatory at Hobarton, vol. i.,
p. xxxiv.-xxxYi.

MAGNETIC INTENSITY. 83

♦

both hemispheres, is identical with the period at which the
earth is nearest to the sun,* and consequently when its ve-
locity of translation is the greatest. At this period, however,
when the earth is nearest to the sun, namely, in December,
January, and February ; as well as in May, June, and July,
when it is farthest from the sun, the relations of temperature
of the zones on either side of the equator are completely re-
versed, the turning points of the decreasing and increasing
intensity, declination and inclination can not, therefore, be
ascribed to the sun in connection with its thermic influence.
The annual means deduced from observations at Munich
and Gottingen have enabled the active director of the Koyal
Bavarian Observatory, Professor Lament, to deduce the re-
markable law of a period of 10^ years in the alterations of
declination.| In the period between 1841 and 1850, the
mean of the monthly alterations of declination attained very
uniformly their minimum in 1843^, and their maximum in
1848^. Without being acquainted with these European re-
sults. General Sabine was led to the discovery of a periodic-
ally active cause of disturbance from a comparison of the
monthly means of the same years, namely from 1843 to 1848,
which were deduced from observations made at places which
lie almost as far distant from one another as possible (Toron-
to in Canada, and Hobarton in Van Diemen's Land). This
cause of disturbance was found by him to be of a purely cos-
mical nature, being also manifested in the decennial periodic
alterations in the sun's atmosphere.f Schwabe, who has ob-
served the spots upon the sun with more constant attention
than any other living astronomer, discovered (as I have al-
ready elsewhere observed),§ in a long series of years (from

* Sabine, On the Means adopted for determining the Absolute Values,
Secular Change^ and Annual Variation of the Terrestrial Magnetic Force,
in the Phil. Transact, for 1850, pt. i., p. 216. In his address to the
Association at Belfast (^Meeting of the Brit. Assoc, in 1852), he hke-
wise observes, " that it is a remarkable fact which has been estab-
lished that the magnetic force is greater, in both the northern and
southern hemispheres, in the months of December, January, and
February, when the sun is nearest to the earth, than in those of May,
June, and July, when he is most distant from it ; whereas, if the ef-
fects were due to temperature, the two hemispheres should be oppo-
sitely, instead of similarly, affected in each of the two periods re-
ferred to."

t Lament, in Poggend., Annalen, bd. Ixxxiv., s. 579.

% Sabine, On peHodical Laws discoverable in the mean Effects of the
Iwger Magnetic Disturbances, in the Phil. Transact, for 1852, pt. i., p.
121. Vide supra, p. 75. § Cosmos, vol. iv., p. 85.

84

COSMOS.

1826 to 1850), a periodically-varying frequency in the oc-
currence of the solar spots, showing that their maxima fell
in the years 1828, 1837, and 1848, and their minima in the
years 1833 and 1843. "I have not had the opportunity,"
he writes, " of investigating a continuous series of older ob-
servations, but I willingly subscribe to the opinion that this
period may itself be variable." A somewhat analogous kind
of variability — periods ivithin periods — is undoubtedly observ-
able in the processes of light of other self-luminous suns. I
need here only refer to those complicated changes of intensi-
ty which have been shown by Goodricke and Argelander to
exist in the light of /3 Lyrae and Mira Ceti.*

If, as Sabine has shown, the magnetism of the sun is
manifested by an increase in the terrestrial force when the
earth is nearest to that luminary, it is the more striking
that, according to Kreil's very thorough investigations of the
magnetic influence of the moon, the latter should hitherto
not have been perceptible, either during the different lunar
phases, or at the different distances assumed by the satellite
in relation to the earth. The vicinity of the moon does not
appear, when compared with the sun,t to compensate in this

* Op. cit., vol. iii., p. 228.

t Though the nearness of the moon in comparison with the sun
does not seem to compensate the smallness of her mass, yet the al-
ready well-ascertained alteration of the magnetic declination in the
course of a lunar day, the lunar^diumal magnetic variation (Sabine, in
the Report to the Brit. Assoc, at Liverpool, 1854, p. 11, and for Ho-
hart Town in the Phil. Tr. for 1857, Art. i., p. 6), stimulates to a per-
severing observation of the magnetic influence of the earth's satellite.
Kreil has the great merit of having pursued this occupation with great
care, from 1839 to 1852 (see his treatise Ueher den Elnjiuss des Mondes
aitfdie horizontale Component der Magnetischen Erdkraft, in the Denk-
schriften der Wiener Akademie der Wiss. Mathem. Naturwiss, C^asse,
vol. v., 1853, p. 45, and Phil. Trans, for 1856, Art. xxii.). His ob-
servations, which were conducted for the space of many years, both
at Milan and Prague, having given support to the opinion that both
the moon and the solar spots occasioned a decennial period of decli-
nation, led General Sabine to undertake a very important work. He
found that the exclusive influence of the sun on a decennial period,
previously examined in relation to Toronto, in Canada, by the em-
ployment of a peculiar and very exact form of calculation, may be
recognized in all the three elements of terrestrial magnetism {Phil.
Trans, for 1856, p. 361), as shown by the abundant testimony of hour-
ly observations caiTied on for a course of «ight years at Hobart Town,
from January, 1841, to December, 1848. Thus both hemispheres fur-
nished the same result as to the operation of the sun, as well as the
certainty " that the lunar-diurnal variation corresponding to different
years shows no conformity to the inequality manifested in those of
the solar-diurnal variation. The earth's inductive action, refieoted

MAGNETIC VARIATION. 85

respect for the smallness of its mass. The main result of
the investigation, in relation to the magnetic influence of the
earth's satellite, which, according to Melloni, exhibits only a
trace of calorification,* is that the magnetic declination in
our planet undergoes a regular alteration in the course of a
lunar day, during which it exhibits a two-fold maximum and
a two-fold minimum. Kreil very correctly observes, " that
if the moon exerts no influence on the temperature on the
surface of our earth (which is appreciable by the ordinary
means of measuring heat), it obviously can not in this way
effect any alteration in the magnetic force of the earth ; but
if, notwithstanding, an alteration of this kind is actually ex-
perienced, we must necessarily conclude that it is produced
by some other means than through the moon's heat." Ev-
ery thing that can not be considered as the product of a sin-
gle force must require, as in the case of the moon, that all
foreign elements of disturbance should be eliminated, in or-
der that its true nature may be recognized.

Although hitherto the most decisive and considerable va-
riations in the manifestations of terrestrial magnetism do not
admit of being satisfactorily explained by the maxima and
minima in the variations of temperature, there can be no
doubt that the great discovery of the polar property of oxy-
gen in the gaseous envelope of our earth will, by a more
profound and comprehensive view of the process of the mag-
netic activity, speedily aiFord us a most valuable assistance
in elucidating the mode of origin of this process. It would
be inconceivable if, amid the harmonious co-operation of all
the forces of nature, this property of oxygen and its modifi-
cation by an increase of temperature should not participate
in the production and manifestation of magnetic phenomena.

If, according to Newton's view, it is very probable that
the substances which belong to a group of celestial bodies (to
one and the same planetary system) are for the most part
identical,! we may, from inductive reasoning, conclude that

from the moon, must be of a very little amount." (Sabine, in the
Phil. Tr. for 1857, Art. i., p. 7, and in the Proceedings of the Royal
Soc, vol. viii.. No. 20, p. 404.) The magnetic portion of this volume
having been printed almost three years ago, it seemed especially nec-
essary, with reference to a subject whidh has so long been a favorite
one with me, that I should supply what was wanting by some addi-
tional remarks.

* Kreil, Einjluss des Mondes avf die Magnetische Declination^ 1852,
B. 27, 29, 46,

t Cosmos, vol, i., p. 133, 134; also vol, iv,, p. 206.

86 COSMOS.

the electro-magnetic activity is not limited to the gravitating
matter on our own planet. To adopt a different hypothesis
would be to limit cosmical views with arbitrary dogmatism.
Coulomb's hypothesis regarding the influence of the mag-
netic sun on the magnetic earth is not at variance with anal-
ogies based upon the observation of facts.

If we now proceed to the purely objective representation
of the magnetic phenomena which are exhibited by our
planet on different parts of its surface, and in its different
positions in relation to the central body, we must accurately
distinguish, in the numerical results of our measurements,
the alterations which are comprised within short or very
long periods. All are dependent on one another, and in this
dependence they reciprocally intensify, or partially neutral-
ize and disturb each other, as the wave-circles in moving
fluids intersect one another. Twelve objects here present
themselves most prominently to our consideration.

Two magnetic poles^ which are unequally distant from the
poles of rotation of the earth, and are situated one in each
hemisphere ; these are points of our terrestrial spheroid at
which the magnetic inclination is equal to 90°, and at which,
therefore, the horizontal force vanishes.

The magnetic equator, the curve on which the inclination
of the needle =zO°.

The Bies of equal declination, and those on which the dec-
lination = 0 (isogonic lines and li?ies of no variation).

The lines of equal inclination (isoclinal lines).

The four points of greatest intensity of the magnetic force,
two of unequal intensity in each hemisphere.

The lines of equal terrestrial force {isodynamic lines).

The undulating line which connects together on each me-
ridian the points of the weakest intensity of the terrestrial
force, and which has sometimes been designated as a dynamic
equator.* This undulating line does not coincide either with
the geographical or the magnetic equator.

The limitation of the zone where the intensity is generally
very weak, and in which the horary alterations of the mag-

* See Mrs. Somerville's short but lucid description of terrestrial
maj?netism, based upon Sabine's works {Physical Geography, vol. ii., p.
102). Sir James Ross, who intersected the curve of lowest intensity
in his great Antarctic expedition, December, 1839, in 19° S. lat. and
29° 13' W. long., and who has the great merit of having first determ-
ined its position in the southern hemisphere, calls it "the equator of
less intensity." See his Voyage to the Southern and Antarctic Regions^
vol. i., p. 22.

MAGNETIC INTENSITY. 87

netic needle participate, in accordance with the different sea-
sons of the year, in producing the alternating phenomena
observed in both hemispheres.*

In this enumeration I have restricted the use of the word
pole to the two points of the earth's surface at which the
horizontal force disappears, because, as I have already re-
marked, these points, which are the true magnetic poles, but
which by no means coincide with the maxima of intensity,
have frequently been confounded in recent times with the
four terrestrial points of greatest intensity.! Gauss has also
shown that it would be inappropriate to attempt to distin-
guish the chord which connects the two points at which the
dip of the needle =90°, by the designation of magnetic axis
of the earth.f The intimate connection which prevails be^
tween the objects here enumerated fortunately renders it pos-
sible to concentrate, under three points of view, the compli-
cated phenomena of terrestrial magnetism in accordance with
the three manifestations of one active force — Intensity, Incli-
nation, and Declination.

Intensity.

The knowledge of the most important element of terres-
trial magnetism, the direct measurement of the intensity of
•the terrestrial force, followed somewhat tardily the knowl-
edge of the relations of the direction of this force in horizon-
tal and vertical planes (declination and inclination). Oscil-
lations, from the duration of which the intensity is deduced,
were first made an object of experiment toward the close of
the 18th century, and yielded matter for an earnest"^ and con-
tinuous investigation during the first half of the 19th centu-
ry. Graham, in 1723, measured the oscillations of his dip-
ping-needle with the view of ascertaining whether they were
constant,§ and in order to find the ratio which the force di-
recting them bore to gravity. The first attempt to determ-
ine the intensity of magnetism at widely different points of

* " Stations of an intermediate character, situated between the
northern and southern magnetic hemispheres, partaking, although in
opposite seasons, of those contrary features which separately prevail
(in the two hemispheres) throughout the year." Sabine, in the Phil.
Transact, for 1847, pt. i., p. 53-57.

t The pole of intensity is not the pole of verticity. Pliil. Transact,
for 1846, pt. iii., p. 255.

X Gauss, Allgem. Theorie des Erdmagnetismus^ § 31.

§ Phil. Transact., \oL xxxiii., /or 1724-1725, p. 332 ("to try if the
dip and vibrations were constant and regular").

88 COSMOS.

the earth's surface, by counting the number of oscillations in
equal times, was made by Mallet in 1769. He found, with
a very imperfect apparatus, that the number of the oscilla-
tions at St. Petersburg (59° 56^ N. lat.) and at Ponoi (67°
4^) were precisely equal ;* and hence arose the erroneous
opinion, which was even transmitted to Cavendish, that the
intensity of the terrestrial force was the same under all lati-
tudes. Borda, as he has himself often told me, was prevent-
ed, on theoretical grounds, from falling into this error, and
the same had previously been the case with Le Monnier ; but
the imperfection of the dipping-needlfe, the friction which ex-
isted between it and the pivot, prevented Borda (in "his expe-
dition to the Canary Islands in 1776) from discovering any
difference in the magnetic force between Paris, Toulon, Santa
Cruz de TeneriiFe, and Goree, in Senegambia, over a space
of 35° of latitude. {Voyage de La Perouse, t. i., p. 162.)
This difference was for the first time detected, with im-
proved instruments, in the disastrous expedition of La Pe-
rouse in the years 1785 and 1787, by Lamanon, who com-
municated it from Macao to the Secretary of the French
Academy. This communication, as I have already stated
(see p. 62), remained unheeded, and, like many others, lay
buried in the archives of the Academy.

The first published observations of intensity, which, more-
over, were instituted at the suggestion of Borda, are those
which I made during my voyage to the tropical regions of
the New Continent between the years 1798 and 1804. The
results obtained at an earlier date (from 1791 to 1794), re-
garding the magnetic force, by ray friend De Rossel, in the
Indian Ocean, were not printed till four years after my re-
turn from Mexico. In the year 1829 I enjoyed the advant-
age of being able to prosecute my observations of the mag-
netic intensity and inclination over a space of fully 188° of
longitude from the Pacific eastward as far as the Chinese
Dzungarei, two thirds of this portion of the earth's surface
being in the interior of continents. The differences in the
latitudes amounted to 72° (namely, • from 60° N. to 12° S.
lat.).

When we carefully follow the direction of the closed iso-
dynamic lines (curves of equal intensity), and pass from the
external and weaker to the interior and gradually stronger

* Novi Comment. Acad. Sclent. Petropol, t. xiv., pro anno 1769, pars
2, p. 33. See also Le Monnier, Lois du , Magnetisme comparces aux
Observations^ 1776, p. 50.

MAGNETIC INTENSITY. 89

curves, we shall find, in considering the distribution of the
magnetic force in each hemisphere, that there are two points,
or foci, of the maxima of intensity, a stronger and a weaker
one, lying at very unequal distances both from the poles of
rotation and the magnetic poles of the earth. Of these four
terrestrial points the stronger, or American, is situated in
the northern hemisphere,* in 52° 19^ N. lat. and in 92° W.
long. ; while the weaker, which is often called th^ Siberian,
is situated in 70° N. lat. and in 120° E. long., or perhaps a
few degrees less to the eastward. In the journey from Far-
schinsk to Jakutsk, Erman found, in 1829, that the curve
of greatest intensity (1*742) was situated at Beresowski Os-
trow, in 117° 5V E. long, and 59° 44^ N. lat. (Erman,
Magnet. Beob., s. 172-540; Sabine, in the Phil. Transact,
for 1850, pt. i., p. 218). Of these determinations that of
the American focu^ is the more certain, especially in respect
to latitude, while in respect "to longitude it is probably
somewhat too far west." The oval which incloses the stron-
ger northern focus lies, consequently, in the meridian of the
western end of Lake Superior, between the southern extrem-
ity of Hudson's Bay and that of the Canadian lake of Win-
nipeg. We owe this determination to the important land
expedition, undertaken in the year 1843, by Captain Lefroy,
of the Royal Artillery, and formerly director of the Magnetic
Observatory at St. Helena. " The mean of the lemniscates
which connect the stronger and the weaker focus appears to
be situated northeast of Behring's Straits, and somewhat
nearer to the Asiatic than to the American focus."

When I crossed the magnetic equator, the line on which
the inclination =0, between Micuipampa and Caxamarca, in
the Peruvian chain of the Andes, in the southern hemisphere,
in 7° 2^ lat. and 78° 48^ W. long., and when I observed that
the intensity increased to the north and south of this remark-
able point, I was led, from an erroneous generalization of
my own observations, and in the absence of all points of
comparison (which were not madeiltill long afterward), to
the opinion that the magnetic force of the earth increases
uninterruptedly from the magnetic equator toward both
magnetic poles, and that it was probable that the maximum
of the terrestrial force was situated at these points, that is

* In those cases in which individual treatises of General Sabine
have not been specially referred to in these notes, the passages have
been taken from manuscript communications, which have been kind-
ly placed at my disposal by this learned physicist.

90 COSMOS.

to say, where the inclination =90°. "When we first strike
upon the trace of a great physical law, we generally find that
the earliest opinions adopted require subsequent revision.
Sabine,* by his own observations, which were made from
1818 to 1822 in very different zones of latitude, and by the
able arrangement and comparison of the numerous oscilla-
tion-experiments with the vertical and horizontal needles,
which of late years have gradually become more general, has
shown that the intensity and inclination are very variously
modified ; that the minimum of the terrestrial force at many
points lies far from the magnetic equator ; and that in the
most northern parts of Canada and in the Arctic regions
around Hudson's Bay, from 52° 20^ lat. to the magnetic pole
in 70° lat. and from about 92° to 93° W. long., the intensi-
ty, instead of increasing, diminishes. In the Canadian focus
of greatest intensity, in the northern hefhisphere, found by
Lefroy, the dip of the needle in 1845 was only 73° 7^, and
in both hemispheres we find the maxima of the terrestrial
force coinciding with a comparatively small dip.f

However admirable and abundant are the observations of
intensity which we owe to the expeditions of Sir James Ross,
Moore, and Clerk, in the Antarctic polar seas, there is still
much doubt in reference to the position of the stronger and
weaker focus in the southern hemisphere. The first of these
navigators has frequently crossed the isodynamic curves of
greatest intensity, and, from a careful consideration of his
observations, Sabine has been led to refer one of the foci to
64° S. lat. and 137° 30^ E. long. Koss himself, in the ac-
count of his great voyage, J conjectures that the focus lies in

* Fifth Report of the British Association^ p. 72 ; Seventh Report, p.
64-68. Contributions to Terrestrial Magnetism, No. vii., in the Phil.
Transact, for 18i6, pt. iii., p. 254.

t Sabine, in the Seventh Report of the Brit. Assoc, p. 77.

X Sir James Ross, Voyage in the Southern and Antarctic Regions, vol.
i., p. 322. This great navigator, in sailing between Kerguelen's Land
and Van Diemen's Land, twice crossed the curve of greatest intensity,
first in 46° 44' S. lat. 128°. |g' E. long., where the intensity increased
to 2-034, and again diminished, further east, near Hobarton, to 1*824
( Voy., vol. i., p. 103-104) ; then again, a year later, from January
1st to April 3d, 1841, during which time it would appear, from the
log of the Erebus, that they had gone from 77° 47' S. lat. 175° 41'
E. long, to 51° 16' S. lat. 136° 50' E. long., where the intensities were
found to be uninterruptedly more than 2-00, and even as much as 2*07
(Phil. Transact, for 1843, pt. ji., p. 211-215). Sabine's result for the
one focus of the southern hemisphere (64° S. lat. 137° 30' E. long.),
which I have already given in the text, was deduced from observations
made by Sir James Ross between the 19th and 27th of March, 1841

MAGNETIC INTENSITY. 91

the neighborhood of the Terre d'Adelie, discovered by D'Ur-
ville, and therefore in about 67° S. lat. and 140° E. long.
He thought that he had approached the other focus in 60°
S. lat. and 125° W. long.; but he was disposed to place it
somewhat further south, not far from the magnetic pole, and
therefore in a more easterly meridian.*

Having thus established the position of the four maxima
of intensity, we have next to consider the relation of the
forces. These data can be obtained from a much earlier
source, to which I have already frequently referred ; that is
to say, by a comparison with the intensity which I found at
a point of the magnetic equator in the Peruvian chain of the
Andes, which it intersects in 7° 2^ lat. and 78° 48^ W. long.,
or, according to the earliest suggestions of Poisson and Gauss,
by absolute measurement.! If we assume the intensity at
the above-indicated point of the magnetic equator =r 1-000 in
the relative scale, we find, from the comparison made be-
tween the intensity of Paris and that of London in the year
1827 (see page 68), that the intensities of these two cities
are 1-348 and 1-372. If we express these numbers in ac-
cordance with the absolute scale they will stand as about
= 10-20 and 10-38, and the intensity, which was assumed
to be 1-000 for Peru, would, according to Sabine, be 7-57
in the absolute scale, and therefore even greater than the
intensity at St. Helena, which, in the same absolute scale,
=6*4. All these numbers must be subjected to a revision
on account of the different years in which the comparisons
were made. They can only be regarded as provisional,
whether they are reckoned in the relative (or arbitrary) scale
or in the absolute scale, which is to be preferred to the for-
mer ; but even in their present imperfect degree of accuracy
they throw considerable light on the distribution of the mag-
netic force — a subject which, till within the last half cen-
tury, was shrouded in the greatest obscurity. They afford

(while crossing the southern isodynamie ellipse of 2-00, about midway
between the extremities of its principal axis), between the southern
latitudes 58° and 64° 26', and the eastern longitudes of 128° 40' and
148° 20' {Contrih. to Terr. Magn., in the Phil. Transact, for 1846, pt.
iii., p. 252).

* Ross, Voyage, vol. ii., p. 224. In accordance with the instructions
drawn up for the expedition, the two southern foci of the maximum
of intensity were conjectured to be in 47° S. lat. 140° E. long, and in
60° S. lat. 235 E. long. (vol. i., p. xxxvi.).

t l^hil. Transact, for 1850, pt. i., p. 201 ; Admiralty Manual^ 1849,
p. 16 ; Erman, Magnet. Beob., s. 437-454.

92 COSMOS.

what is cosmically of very great importance, historical points
of departure for those alterations in the force which will be
manifested in future years, probably through the dependence
of the earth upon the magnetic force of the sun, by which it
is influenced.

In the northern hemisphere the stronger or Canadian
focus, in 5^° 19^ N. lat. and 92° W. long., has been most
satisfactorily determined by Lefroy. This intensity is ex-
pressed in the relative scale by 1*878, the intensity of Lon-
don being 1*372, 'while in the absolute scale it would be ex-
pressed by 14-21.* Even in New York, lat. 40° 42^, Sabine
found the magnetic force not much less (1*803). For the
weaker northern or Siberian focus, 70° lat., 120° E. long.,
it was found by Erman to be 1 74 in the relative scale, and
by Hansteen 1*76 ; that is to say, about 13-3 in the absolute
scale. The Antarctic expedition of Sir James Ross has
shown us that the difference of the two foci in the southern
hemisphere is probably less than in the northern, but that
each of the two southern foci exceeds both the northern in
intensity. The intensity in the stronger southern focus, 64°
lat., 137° 30^ E. long., is at least 2*06 in the relative or ar-
bitrary scale,t while in the absolute scale it is 15*60 ; in the
weaker southern focus, 60° lat, 129° 40^ W. long., we find
also, according to Sir James Ross, that it is 1*96 in the ar-
bitrary scale and 14*90 in the absolute scale. The greater or
lesser distance of the two foci from one another in the same
hemisphere has been recognized as an important element of
their individual intensity, and of the entire distribution of
the magnetic force. " Even although the foci of the south-
ern hemisphere exhibit a strikingly greater intensity (name-
ly, 15*60 and 14*90 in the absolute scale) than the foci of

* On the map of isodynamic lines for North America, which occurs
in Sabine's Contributions to Terrestrial Magnetism, No. vii., we find, by
mistake, the value 14*88 instead of 14-21, although the latter, which
is the true number, is given at page 252 of the text of this memoir.

t I follow the value given in Sabine's Contributions, No. vii.; p. 252,
namely, 15*60. We find from the Magnetic Journal of the Erebus
(Phil Transact, for 1843, pt. ii., p. 169-172) that several individual
obsen'ations, taken on the ice on the 8th of Februai-v, 1841, in 77° 47'
S. lat. and 172° 42' W. long., yielded 2*124. The value of the intens-
ity 15*60 in the absolute scale would lead us to assume provisionally
that the intensity at Hobarton was 13*51 {Magn. and Meteor oL Observ.
made at Hobarton, vol. i., p. 75). This value has, however, lately
been slightly augmented (to 13*56) (vol. ii., xlvi.). In the Admiralty
Manual, p. 17, I find the southern focus of greatest intensity changed
to 15*8.

MAGNETIC INTENSITY. 93

the northern hemisphere (which are respectively 14*21 and
13 •30), the total magnetic force of the one hemisphere can
not be esteemed as greater than that of the other." •

"The result is, however, totally different when we sepa-
rate the terrestrial sphere into an eastern and western part,
in accordance with the meridians of 100° and 280° E. long.,
reckoning from west to east in such a manner that the east-
ern or more continental sphere shall inclose South America,
the Atlantic Ocean, Europe, Africa, and Asia, almost as far
as Baikal ; while the western, which is the more oceanic and
insular, includes almost the whole of North America, the
broad expanse of the Pacific, New Holland, and a portion
of Eastern Asia." These meridians lie the one about 4°
west of Singapore, the other 13° west of Cape Horn, in the
meridian of Guayaquil. All four foci of the maximum of
the magnetic force, and even the two magnetic poles, fall
within the western hemisphere.*

Adolph Erman's important observation of least intensity
in the Atlantic Ocean, east of the Brazilian province of Es-
piritu Santo (20° S. lat., 35° 0.2^ W. long.), has been already
mentioned in our Delineation of Kature.f He found in the
relative scale 0*7062 (in the absoli||| scale 5-35). This re-,
gion of weakest intensity was also twice crossed by Sir James
Ross, in his Antarctic expedition, J between 19° and 21° S.
lat., as well as by Lieutenant Sulivan and Dunlop in their
voyage to the Falkland Islands.§ In his isodynamic chart
of the entire Atlantic Ocean, Sabine has drawn the curve of
least intensity, which Ross calls the equator of less intensity,
from coast to coast. It intersects the West African shore
of Benguela, near the Portuguese colony of Mossamedes (15°
S. lat.) ; its summits are situated in the middle of the ocean,
in 18° W. long., and it rises again on the Brazilian coast as
high as 20° S. lat. Whether there may not be another zone

* See the interesting Map of the World, divided into hemispheres by
a plane coinciding with the meridians o/'lOO anof 280 east of Greenwich,
exhibiting the unequal distribution of the magnetic intensity in the two
hemispheres, plate v., in the Proceedings of the Brit. Assoc, at Liver-
pool, 1837, p. 72-74. Erman found that the intensity of the terres-
trial force was almost constantly below 0-76, and consequently very
small in the southern zone between latitudes 24° 25' and 13° 18', and
between the western longitudes of 34° 50' and 32° 44',

t Cosmos, vol. i., p. 187.

t Voyage in the Southern Seas, vol, i., p. 22, 27 ; vide supra, p. 96,

§ See the Journal of Sulivan and Dunlop, in the Phil. Transact.
for 1840, pt. i., p. 143, They found as the minimum only 0-800.

94

COSMOS.

of tolerably low intensity (0*97) lying north of the equator
(10° to 12° lat.), and about 20° east of the Philipines, is a
question that must be left for future investigations to eluci-
date.

I do not think that the ratio which I formerly gave of the
weakest to the strongest terrestrial force requires much mod-
ification in consequence of later investigations. This ratio
falls between 1:2^ and 1 : 3, being somewhat nearer to the
latter number, and the difference of the data* arises from the
circumstance that in some cases the minima alone, and in
others the minima and maxima together, have been altered
somewhat arbitrarily. Sabinef has the great merit of having
first drawn attention to the importance of the dynamic equa-
tor, or curve of least intensity. " This curve connects the
points of each geographical meridian at which the terrestrial
intensity is the smallest. It describes numerous undulations
in passing round the earth, on both sides of which the force
increases with the higher latitudes of each hemisphere. It
in this manner indicates the limits between the two magnetic
hemispheres more definite^ than the magnetic equator, on
which the direction of the magnetic force is vertical to the
, direction of gravity. Jlrespect to the theory of magnetism,
that which refers directly to the force itself is of even greater
importance than that which merely refers to the direction of
the needle, its horizontal or vertical position. The curves
of the dynamic equator are numerous, in consequence of their
depending upon forces which produce four points (foci) of
the greatest terrestrial force, which are unsymmetrical and
of unequal intensity. We are more especially struck in these
inflections with the great convexity in the Atlantic Ocean
toward the South Pole, between the coasts of Brazil and the
Cape of Good Hope."

* We obtain 1 : 2-4:4: on comparing in the absolute scale St. Helena,
which is 6-4, with the focus of greatest intensity at the south pole,
which is 15-60, and 1 : 2*47 by a comparison of St. Helena with the
higher southern maximum of 15*8, as given in the Admiralty Manual,
p. 17, and 1 : 2-91 by a comparison in the relative scale of Erman's ob-
seri^ation in the Atlantic Ocean (0*706), with the southern focus (2-06) ;
indeed, even 1 : 2-95, when we compare together in the absolute scale
the lowest value given by this distinguished traveler (5-35), with the
highest value fiir the southern focus (15-8). The mean resulting ratio
would be 1 : 2-69. Compare for the intensity of St. Helena (6'4 in the
absolute, or 0-845 in the arbitrary scale) the earliest observations of
Fitzroy (0-836), Phil. Transact, for 1847, pt. i., p. 52, and Proceedings
of the Meeting at Liverpool, p. 56.

t See Contributions to Ten-estrial Magnetism^ No. vii., p. 256.

MAGNETIC INTENSITY. 95

Does the intensity of the magnetic force perceptibly de-
crease at such heights as are accessible to lis, or does it per-
ceptibly increase in the interior of the earth"? The problem
which is suggested by these questions is extremely complica-
ted in the case of observations which are made either in or
upon the earth, since a comparison of the effect of considera-
ble heights on mountain journeys is rendered difficult, because
the upper and lower stations are seldom sufficiently near
one another, owing to the great mass of the mountain ; and
since, further, the nature of the rock and the penetration of
veins of minerals, which are not accessible to our observation,
together with imperfectly understood horary and accidental
alterations in the intensity, modify the results, where the ob-
servations are not perfectly simultaneous. In this manner
we often ascribe to the height or depth alone conditions which
by no means belong to either. The numerous mines of con-
siderable depth which I have visited in Europe, Peru, Mexi-
co, and Siberia have never afforded localities which inspired
me with any confidence.* Then, moreover, care should be
taken, in giving the depths, not to neglect the perpendicular
differences above or below the level of the sea, which consti-
tutes the mean surface of the earth. The borings at the
mines of Joachimsthal, in Bohemia, are upward of 2000
feet in absolute depth, and yet they only reach to a stratum
of rock which lies between 200 and 300 feet above the level
of the sea.t Very different and more favorable conditions
are afforded by balloon ascents. Gay-Lussac rose to an ele-
vation of 23,020 feet above Paris ; consequently, therefore,
the greatest relative depth that has been reached by borings
in Europe scarcely amounts to -jirth of this height. My
own mountain observations, between the years 1799 and
1806, led me to believe that the terrestrial force gradually
decreases with the elevation, although, in consequence of the
causes of disturbance already indicated, several results are
at variance with this conjectural decrease. I have collected
in a note individual data, taken from 125 measurements of
intensity made in the Andes, in the Swiss Alps, Italy, and

* We may ask what kind of eiTor can have led, in the coal-mines
of Flenu, to the result that in the interior of the earth, at the depth
of 87 feet, the horizontal intensity had increased O'OOl ? Journal de
rinstitut, 1845, Avril, p. 146, In an English mine, which is 950 feet
below the level of the sea, Henwood did not find any increase in the
intensity (Brewster, Treatise on Magn., p. 276).

t Cosmos, vol. i., p. 159.

96 COSMOS.

Germany.* These observations extended from the level of
the sea to an elevation of 15,944 feet, and therefore to the
very limits of 'perpetual snow, but the greatest heights did
not afford me the most reliable results. The most satis-
factory were obtained on the steep declivity of the Silla de
Caracas (8638 feet), which inclines toward the neighboring
coasts of La Guayra; the Santuario de Nostra Senora de
Guadalupe, which rises immediately over the town of Bogota,
upon the declivity of a steep wall of limestone rock, with a

* A diminution of the intensity with the height is shown in ray
observations from the comparisons of the Silla de Caracas (8638 feet
above the- sea, intensity 1-188) with the harbor of Guayra (height 0
feet, intensity 1-262) and the town of Caracas (height 2648 feet, in-
tensity 1*209) ; from a comparison of the town of Santa Fe de Bogota
(elevation 8735 feet, intensity 1*147) with the chapel of Neustra Se-
nora da Guadalupe (elevation 10,794 feet, intensity 1*127), which
seems to hang over the town like a swallow's nest, perched upon a
steep ledge of rock ; from a comparison of the volcano of Purace (ele-
vation 14,548 feet, intensity 1*077) with the mountain village of Pu-
race (elevation 8671 feet, intensity 1*087) and with the neighboring
town of Popayan (elevation 5825 feet, intensity 1*117); from a com-
parison of the town of Quito (elevation 9541 feet, intensity 1*067)
with the village of San Antonio de Lulumbamba (elevation 8131 feet,
intensity 1*087), lying in a neighboring rocky fissure directly under
the geographical equator. The oscillation experiments, which I made
at the highest point at which I ever instituted observations of the kind,
namely, at an elevation of 15,944 feet, on the declivity of the long-
since extinct volcano of Antisana, opposite the Chussulongo, were
quite at variance with this result. It was necessary to make this ob-
servation in a large cavern, and the great increase in the intensity
was no doubt the consequence of a magnetic local attraction of the
trachytic rock, as has been shown by the experiments which I made
with Gay-Lussac within, and on the margin of, the crater of Vesuvius.
I found the intensity in the Cave of Antisana increased to 1*188,
while in the neighboring lower plateau it was scarcely 1*068. The
intensity at the Hospice of St. Gotthard (1*313) was greater than that
at Airolo (1*309), but less than that at Altorf (1*322). Airolo, on the
other hand, exceeded the intensity of the Ursern Lake (1*307). In
the same manner Gay-Lussac and myself found that the intensity was
1*344 at the Hospice of Mont Cenis, while at the foot of the same
mountain, at Lans le Bourg, it was 1*323, and at Turin 1*336. The
greatest contradictions were necessarily presented by the burning vol-
cano of Vesuvius, as we have already- remarked.- While in 1 805 the
terrestrial force at Naples was 1*274, and at Portici 1*288, it rose in
the Monastery of St. Salvador to 1*302; while it fell in the crater of
Vesuvius lower than any where else throughout the whole district,
namely, to 1*193. The iron contained in the lava, the vicinity of
magnetic poles, and the heat of the soil, which probably has the effect
of diminishing this force, combined to produce the most opposite local
disturbances. See my Voyage aux Regions Equinox iales, t. iii., p. 619^^
626, and Mem. de la Societe d'Argueil, t. i., 1807, p. 17-19.

MAGNETIC OBSERVATIONS. 97

difference of elevation amounting to upward of 2000 feet ;
and the volcano of Purace, which rises 8740 feet above the
Plaza Mayor of the town of Popayan. KupiFer in the Cau-
casus,* Forbes in many parts of Europe, Laugier and Mau-
vais on the Canigou, Bravais and Martins on the Faulhorn,
and during their very adventurous sojourn in the immediate
vicinity of the summit of Mont Blanc, have certainly ob-
served that V the intensity of the magnetic force diminished
with the height, and this decrease appeared from Bravais's
general consideration of the subject to be more rapid in the
Pyrenees than in the chain of the Alps.f

Quetelet's entirely opposite results, obtained in an excur-
sion from Geneva to the Col de Balme and the Great St-
Bernard, make it doubly desirable, for the final and decisive
settlement of so important a question, that observations
should be made at some distance from the surface of the
earth; and these observations can only be carried on by
means of balloon ascents, such as were, employed in 1804 by
Gay-Lussac, first in association with Biot, on the 24th of
August, and subsequently alone on the 16th of September.
Oscillations measured at elevations of 19,000 feet can, how-
ever, only afford us certain information regarding the trans-
mission of the terrestrial force in the free atmosphere when
care is taken to obtain corrections for temperature in the
needles that are employed both before and after the asceht.
The neglect of such a correction has led to the erroneous
result deducible from Gay-Lussac's experiments, that the
magnetic force remains the same to an elevation^bf more

* Kupffer's observations do not refer to the summit of the Elbruz,
but to the difference of height (4796 feet) between two stations, v'z.,
the bridge of Malya and the mountain decHvity of Kharbis, which un-
fortunately differ considerably in longitude and latitude. Eegarding
the doubts which Necker and Forbes have advanced in relation to
this result, see Transact, of the Royal Soc. o/Edin., vol. xiv., 1840, p.
23-25-

t Compare Laugier and Mauvais, in the Comptes rendns, t. xvi.,
1843, p. 1175 ; and Bravais, Observ.de V hitensite du Magnetisme Ter-
restre en France, en Suisse, et en Savoie, in the Annales de Chcmie et de
Phys., 3eme Serie, t. xviii., f846, p. 214; Kreil, Einfluss dcr Alpen
auf die Intensitdt, in the Denkschrifien der Wiener Akad. der Wiss.
Mathem. Naturwiss. Classe, bd. i., 1850, s. 265, 279, 290. It is very
remarkable that so accurate an observer as Quetelet should have found,
in a tour which he made in the year 1830, that the horizontal intensity
increased with the heigh tffin ascending from Geneva (where it was
1-080) to the Col de Balne (where it was 1-091) and to the Hospice
of St. Bernard (where it was as high as 1-096). See Sir David Brew-
ster, Treatise on Magn., p. 275.

Vol. v.— E

98 COSMOS.

than 22,000 feet,* while conversely the experiment showed
a decrease in the force on account of the shortening of the
oscillating needle in the upper cold reg'ion.f Faraday's
brilliant discovery of the paramagnetic force of oxygen must
not be disregarded in the discussion of this subject. This
great physicist shows that in the upper strata of the atmos-
phere the decrease in the intensity can not be sought merely
in the original source of the force, namely, the solid earth,
but that it may equally arise from the excessively rarefied
condition of the air, since the quantity of oxygen in a cubic
foot of atmospheric air must differ in the upper and lower
strata. It seems to me, however, that we are not justified
in assuming more than this — that the decrease of the para-
magnetic property of the oxygenous parts of the atmosphere,
which diminish with the elevation and with the rarefaction
of the air, must be regarded as a co-operating modifying
cause. Alterations of temperature and density through the
ascending currents of air may further alter the amount of
this influence.f Such disturbances assume a variable and
specially local character, and they operate in the atmosphere
in the same manner as different kinds of rocks upon the sur-
face of the earth. With every advance which we may re-
joice in having made in our knowledge of the gaseous en-
velope of our planet and of its physical properties, we at the
game time learn to know new causes of disturbance in the
^5ternating mutual action of forces, which should teach us
r^w cautiously we ought to draw our conclusions.

The llitensity of the terrestrial force, when measured at
definite points of the surface of our planet, has, like all the
phenomena of terrestrial magnetism, its horary as well as its
secular variations. The horary variations were distinctly
recognized by Parry during his third voyage, and also, con-
jointly with him, by Lieutenant Foster (1825), at Port
13owen. The increase of intensity from morning till evening
in the mean latitudes has been made an object of the most
careful investigation by Christie,§ Arago, Plansteen, Gauss,
and Kupffer. As horizontal oscillations, notwithstanding
the great improvements which have been made in the pres-

* Annales de amnie, t. lii., 1805, p. 86, 87.

f Arago, in the Anntiaire -du Bureau des Longitudes pour 1836^ p.
287; Forbes, in the Edin. Transact., vol. Hi v., 1840, p'. 22.

X Earaday, Exper. Researches in Electricity, 1851, p. 53, 77, § 2881,
2961.

§ Christie, in the rUl Transact, for 1825, p. 49.

MAGNETIC OBSERVATIONS. 99

ent day in the dipping-needle, are preferable to oscillations
of the latter kind, it is not possible to ascertain the horary
variation of the total intensity without a very accurate knowl-
edge of the horary variation of the dip. The establishment
of magnetic stations in the northern and the southern hemis-
phere has afforded the great advantage of yielding the most
abundant, and comparatively the most accurate results. It
will be sufficient here to instance two points of the earth's
surface, which are both situated without the tropics, and al-
most in equal latitudes on either side of the equator — name-
ly, Toronto, in Canada, 43° 39^ N. lat., and Hobarton, in
Van Diemen's Land, in 42° 53^ S. lat., with a difference of
longitude of about 15 hours. The simultaneous horary mag-
netic observations belong at the one station to the" winter
months, while at the other they fall within the period of the
summer months. While measurements are made at the one
place during the day, they are being simultaneously carried
on at the other station, for the most part, during the night.
The variation at Toronto is 1° 33^ West ; at Hobarton it is
9° 57' East ; the inclination and the intensity are similar to
one another; the former is, at Toronto, about 75° 15' to
the north, and at Hobarton about 70° 34' to the south,
while the total intensity is 13*90 in the absolute scale at
Toronto, and 13*56 at Hobarton.* It would appear, from
Sabine's investigation, that these well-chosen stations ex-
hibitf four turning-points for the intensity in Canada, and
only two such points for Van Diemen's Land. At Toronto
the variation in intensity reaches its principal maximum at
6 P.M., and its principal minimum at 2 A.M. ; a weaker
secondary maximum at 8 A.M., and a weaker secondary
minimum at 10 A.M. The intensity at Hobarton, on the
contrary, exhibits a simple progression from a maximum be-
tween 5 and 6 P.M. to a minimum between 8 and 9 A.M. ;
although the inclination there, no less than at Toronto, ex-
hibits four turning-points.J By a comparison of the varia-

* Sabine, On Periodical Laws of the larger Magnetic Disturbances,
in the Phil. Transact, for 1851, pt. i., p. 126; and on the Annual Va-
riation of the Magn. Declin., in the Phil. Transact, for 1851, pt. ii., p.
636.

t Observations made at the Magn. and Meteorol. Observatory at To-
ronto, vol. i. (1840-1842), p. Ixii. ••

X Sabine, in Magn. and Meteor. Observations at Hobarton, vol. i., p.
Ixviii. " There is also a correspondence in the ran£!;e and turning
hours of the diurnal variation of the total force at Hobarton and at
Toronto, although the progression is a double one at Toronto and a

100 COSMOS.

tions of inclination with those of the horizontal force, it has
been established that in Canada, during the winter months,
when the sun is in the southern signs of the zodiac, the total
terrestrial force has a greater intensity than in the summer
months, while in Van Diemen's Land the intensity is great-
er than the mean annual value — that is to say, the total ter-
restrial force — from October to February, which constitutes
the summer of the southern hemisphere, while it is less from
April to August. According to Sabine,* this intensity of
tlie terrestrial magnetic force is not dependent on differences
of temperature, but on the lesser distance of the magnetic
solar body from the earth. At Hobarton the intensity dur-
ing the summer is 13*574: in the absolute scale, while during
the winter it is 13 '543. The secular variation of intensity
has hitherto been deduced from only a small number of ob-
servations. At Toronto it appears to have suffered some de-
crease between 1845 and 1849, and the comparison of my
own observations with those of Rudberg, in the years 1806
and 1832, give a similar result for Berlin.f

Inclination.

The knowledge of the isoclinal curves, or lines of equal in-
clination, as well as the more rapid or slower increase of the
inclination by which they are determined (reckoning from
the magnetic equator, where the inclination =0, to the
northern and southern magnetic pole, where the horizontal
force vanishes), has acquired additional importance in mod-
ern times, since the element of the total magnetic force can
not be deduced from the horizontal intensity, which requires
to be measured with excessive accuracy, unless we are pre-
viously well acquainted with the inclination. The knowl-
edge of the geographical position of both magnetic poles is

single one at Hobarton." The time of the maximum of intensity falls
at Hobarton between 8 and 9 A.M. ; while the secondary or lesser
minimum falls at Toronto about 10 A.M., and consequently the in-
crease and diminution of the intensity fall within the same hours in
accordance with the time of the*place, and not at opposite hours, as
is the case with respect to the inclination and the decUnation. See,
regarding the causes of this phenomenon, p. bcix. (compare also Far-
aday, Atmospheric Maqnetism, § 3027-3034).

* Phil. Transact, for 1850, pt. i., p. 215-217; Magnet. Ohserv. at
Hobarton, vol. ii.T 1852, p. xlvi. See also p. 26 of the present volume.
At the Cape of Good Hope the intensity presents less difference at
opposite periods of the year than the inclination {Magnet. Ohserv.
made at the Cape of Good Hope, vol. i., 1851, p. Iv.).

t See the magnetic part of my work on Asie Centrale, t. iii., p. 442.

MAGNETIC INCLINATION. 101

due to the observations and scientific energy of the adven-
turous navigator, Sir James Ross. His observations of the
northern magnetic pole were made during the second expe-
dition of his uncle, Sir John Ross (1829-1833),* and of the
southern during the Antarctic expedition under his own
command (1839-1843). The northern magnetic pole in
70° 5' lat., 96° 43^ W. long., is 5° of latitude farther from
the ordinary pole of the earth than the southern magnetic
pole, 75° 35^ lat., 154° 10^ E. long., while it is also situated
farther west from Greenwich than the northern magnetic
pole. The latter belongs to the great island of Boothia Fe-
lix, which is situated very near the American continent, and
is a portion of the district which Captain Parry had pre-
viously named North Somerset. It is not far distant from
the western coast of Boothia Felix, near the promontory of
Adelaide, which extends into King William's Sound and
Victoria Strait.f The southern magnetic pole has not been
directly reached in the same manner as the northern pole.
On the 17th of February, 1841, the Erebus penetrated as
far as 76° 12^^ S. lat., and 164° E. long. As the incHnation
was here only 88° 40^, it was assumed that the southern
magnetic pole was about 160 nautical miles distant. J Many
accurate observations of declination, determining the inter-
section of the magnetic meridian, render it very probable that
the south magnetic pole is situated in the interior of the great
antarctic region of South Victoria Land, west of the Prince
Albert mountains, which approach the south pole, and are
connected with the active volcano of Erebus, which is 12,400
feet in height.

The position and change of form of the magnetic equator,
that is to say, the line on which the dip is null, were very
fully considered in the Picture of Nature, Cosmos, vol. i,, p.
183. The earliest determination of the African node (the
intersection of the geographical and magnetic equators) was

* Sir John Barrow, Arctic Voyages of Discovery, 1846, p. 521-529.

t The strongest inclination which has as yet been observed in the
Siberian continent is 82° 16', which was found by Middendorf, on the
River Taimyr, in 74° 17' N. lat., and 95° 40' E. long. (Middend., Si-
her. Reise, th. i., s. 194).

X Sir James Ross, Voyage to the Antarctic Regions, vol. i., p. 246.
"I had so long cherished the ambitious hope," says this navigator,
"to plant the flag of my country on both the magnetic poles of out
globe; but the obstacles which presented themselves being of so in-
surmountable a character was some degree of consolation, as it left us
no grounds for self-reproach" (p. 247).

102 COSMOS.

made by Sabine* at the beginning of his pendulum expedi-
tion in 1822. Subsequently, in 1840, the same learned ob-
server noted down the results obtained by Duperrey, Allen,
Dunlop, and Sulivan, and constructed a chart of the magnet-
ic equator^* from the west coast of Africa at Biafra (4° N.
lat., 9° 30^ E. long.), through the Atlantic Ocean, and Bra-
zil (16° S. lat., between Porto Seguro and Rio Grande), to
the point where, upon the Cordilleras, in the neighborhood
of the Pacific, I saw the northern inclination assume a south-
ern direction. The African node, as the point of intersection
of both equators, was situated, in 1837, in 3°£. long., while
in 1825 it had been in 6° 57^ E. long. The secular motion
of the node, turning from the basaltic island of St. Thomas,
which rises to an elevation of more than 7000 feet, was,
therefore, somewhat less than half a degree westward in the
course of a year ; after which the line of no inclination tu'ned
toward the north on the African coast, while oii the Brazil-
ian coast it is inclined southward. The convexity of the
magnetic equatorial curve is persistently turned toward the
south pole, while in the Atlantic Ocean it passes at a dis-
tance of about 16° from the geographical equator. For the
interior of South America, the terra incognita of Matto Grosso
between the large rivers of Xingu, Madera, and Ucayle, we
have no observations of the dip until we reach the chain of
the Andes, where, 68 geographical miles east of the shores
of the Pacific, between Montan, Micuipampa, and Caxa-
marca, I determined astronomically the position of the mag-
netic equator, which rises toward the northwest (7° 2^ S.
lat., and 78° 46^ W. long.).}:

* Sabine, Pendul. Exper., 1825, p. 476.

t Sabine, in the Phil. Transact, for 1840, pt. i., p. 136, 139, 146.
I follow, for the progression of the African node, the map which is
appended to this treatise.

X I here give, in accordance with my usual practice, the elements
of this not wholly unimportant determination : Micuipampa, a Peru-
vian mountain town at the foot of Cerro de Guelgayoc, celebrated for
its rich silver mines, 6° 44' 25'' S. lat., 78° 33' 3 " W. long., elevation
above the Pacific 11,872 feet, magnetic inclination 0°'42 north (ac-
cording to the centesimal division of the circle) ; Caxamarca, a town
situated on a plateau at an elevation of Q362 feet, 7° 8' 38" S. lat,
5h. 23' 42" long., inclination 0*15 south; Montan, a farm-house (or
hacienda), surrounded by Llama flocks, situated in the midst of mount-
ains, 6° 33' 9" S. lat., 5h. 26' 51" W. long., elevation 8571 feet, in-
clination 0-70 north; Tomependa, on the mouth of the Chinchipe, on
the River Amazon, in the province of Jaen de Bracamoros, 5° 31' 28"
S. lat., 78° 37' 30" W. long., elevation 1324 feet, inclination 3°-55
north; Truxillo, a Peruvian town on the Pacific, 8° 5' 40" S. lat.,

MAGNETIC INCLINATION. lUo

The most complete series of observations which we pos-
sess in reference to the position of the magnetic equator was
made by my old friend Duperrey during the years 1823-
1825. He crossed the equator six times during his voyages
of circumnavigation, and he was enabled to determine this
line by his own observations over a space of 220°.* Accord-
ing to Duperrey's chart of the magnetic equator, the two
nodes are situated in long. 5° 50^ E. in the Atlantic Ocean,
and in long. 177° 20^ E. in the Pacific, between the merid-
ians of the Fejee and Gilbert islands. While the magnet-
ic equator leaves the western coasts of the South American
continent, probably between Punta de la Aguja and Payta,
it is constantly drawing nearer in the west to the geograph-
ical equator, so that it is only at a distance of 2° from it,
in the meridian of the group of the Mendana Islands.!
About 10^ farther west, in the meridian which passes
through the western part of the Paumotu Islands (Low
Archipelago), lying in 153° 50^ E. long.. Captain Wilkes
found that the distance from the geographical equator in
1840 was still fully 2°.J The intersection of the nodes in
the i^acific is not as much as 180° from that of the Atlantic
nodes; that is to say, it does not occur in 174° 10' W. long.,
but in the meridian of the Fejee Islands, situated in about
177° 20^ E. long.'- If, therefore, we pass from the west coast

79° 3' 37" W. long., inclination 2*'15 south. Humboldt, Recueil
d'Observ. Astron. (Nivellement Barometrique et Geodesique), vol. i.,
p. 316, No. 242, 244-254. For the basis of astronomical determina-
tions, obtained by altitudes of the stars and by the chronometer, see
the same work, vol. ii., p. 379-891. The result of my observations
of inclination in 1802, in 7° 2' S. lat., and 78° 48' W. long., accords
pretty closely by a singular coincidence, and notwithstanding the sec-
ular alteration, with the conjecture of Le Monnier, which was based
upon theoretical calculation. He says, "the magnetic equator must
be in 7° 45' north of Lima, or at most in 6° 30' S. lat., in 177G" {Lois
du Magnetisme comparees aux Observations, pt. ii., p. 59).

* Saigey, M^m. sur VEqiiatmr Magnctique d'apres les Ohserv. du
Capitaine Duperrey, in the Annales Maritimes et Coloniales, Dec, 1833,
t. iv., p. 5. Here it is observed that the magnetic equator is not a
curve of equal intensity, but that the intensity varies in different parts
of this equator from 1 to 0-8G7.

t This position of the magnetic equator was confirmed by Erman
for the year 1830. On his return from Kamtschatka to Europe, he
found the inclination almost null at 1° 30' S. lat., 132° 37' W. long. ;
in 1° 52' S. lat., 135° 10' W. long.; in 1° 54' lat., in 133° 45' W.
long. ; in 2° 1' S. lat., 139° 8' W. long. (Erman, Magnet. Beob., 1841,
s. 636).

X Wilkes, United States Exploring Expedition, vol. iv., p. 263.

104 COSMOS.

of Africa, through South America westward, we shall find
in this direction that the distance of the nodes from one an-
other is about 8^° too great, which is a proof that the curve
of which we are here speaking is not one of the great circles.

According to the admirable and comprehensive determina-
tions which were made by Captain Elliot from 1846 to 1849,
between the meridians of Batavia and Ceylon, and which
coincide in a remarkable manner with those of Jules de
Blosseville (see page 65), it would appear that the magnetic
equator passes through the northern point of Borneo, and
almost due west into the northern point of Ceylon, in 9° 45''
N. lat. The curve of minimum total intensity runs almost
parallel to this part of the magnetic equator,* which enters
the western part of the continent of Africa, south of the Cape
of Gardafui. This important re-entering point of the curve
has been determined with great accuracy by Rochet cl'Heri-
court on his second Abyssinian expedition, from 1842 to
1845, and by the interesting discussion to which his magnet-
ic observations gave rise.j This point lies south of Gau-
bade, between Angolola and Angobar, the capital of the
kingdom of Schoa, in 10° 7^ N. lat, and in 41° 13' E.'long.
Tlie course of the mngnetic equatcr \n the interior of Africa,
from Angobar to the Gulf of Biafra, is as thoroughly unex-
plored as that in the interior of South America, east of the
chain of the Andes, and south of the geographical equator.
Both these continental districts are nearly of equal extent,
measured from east to west, each extending over a space of
about 80° of longitude, so that we are still entirely ignorant
of the magnetic condition of nearly one quarter of the earth's
circumference. My own observations of inclination and in-
tensity for the whole of the interior of South America, from
Cumana to the llio Negro, as well as from Cartagena de In-
dias to Quito, refer only to the tropical zone north of the
geographical equator, while those which I made in the south-
ern hemisphere, from Quito as far as Lima, were limited to
the district lying near the western coast.

The translation of the African node toward the west from
1825 to 1837, which we have already indicated, has been
confirmed on the eastern coasts of Africa by a comparison
of the inclination-observations made by Panton, in the year
1776, with those of Rochet d'He'ricourt. The latter ob-
server found the magnetic equator much nearer the Straits

* Elliot, in the Phil. Transact, for 1851, pt. i., p. 287-331.
t Duperrey, in the Comptes rendus, t. xxii., 1846, p. 804-806.

MAGNETIC INCLINATION. 105

of Bab-el-Mandeb, namely, 1° south of the island of Soco-
tora, in 8° 40^ N. lat. There was, therefore, an alteration
of 1° 27'' lat. for 49 years, while the corresponding altera-
tion in the longitude wtis determined by Arago and Duper-
rey to have been 10° from east to west. The direction of
the secular variation of the nodes of the magnetic equator on
the eastern coasts of Africa, toward the Indian Ocean, was
precisely similar to that on the western coast. The quanti-
ty of the motion must, however, be ascertained from much
more accurate results than we at present possess.

The periodicity of the alterations of the magnetic inclina-
tion, whose existence had been noticed at a much earlier pe-
riod, has only been established with certainty and thorough
completeness within the last twelve years, since the erection
of British magnetic stations in both hemispheres. Arago,'to
whom the theory of magnetism is so largely indebted, had
indeed recognized, in the autumn of 1827, " that the dip was
greater at 9 A.M. than at 6 P.M. ; while the intensity of "
the magnetic force, when measured by the oscillations of a
horizontal needle, attained its minimum in the first, and its
maximum in the second period."* In the British magnetic

* In a letter from Arago to myself, dated Mayence, 13th of Decem-
ber, 1827, he writes as follows : "I have definitely proved during the
late Auroraj Boreales, which have been seen at Paris, that this phe-
nomenon is always accompanied by. a variation in the position of tlie
horizontal and dipping needles, as well as in intensity. The changes
of inclination have amounted to T or 8'. To effect this change, after
allowing for every change of intensity, the horizontal needle must
oscillate more or less rapidly, according to the time at which the ob-
servation is made, but in correcting the results by calculating the
immediate effects of the inclination there still remained a sensible
variation of intensity. On repeating by a new method the diurnal
observation of inclination, on which I was engaged during your late
visit to Paris, I found a regular variation, not for the means but for
each day, which \^ greater in the morning at nine than in the even-
ing at six. You are aware that the intensity, measured with the hori-
zontal needle, is, on the contrary, at its miniiman at the first period,
while it attains its maximum between six and seven in the evening.
The total variation being very small, one might suppose that it was
merely du# to a change of inclination ; and, indeed, the greatest por-
tion of the apparent variation of intensity depends upon the dim-nal
alteration of the horizontal component, but, when every correction
has been made, there still remains a small quantity as an indication
of a real vai-iation of intensity. ^^ In another letter, which Arago wrote
to me from Paris on the 20th of March, 1829, shortly before my Sibe-
rian expedition, he expressed himself as follows : " I am not surprised
that you should have found it difficult to recognize the diurnal change
of inclination, of which I have already spoken to you, in the winter
months, for it is only during the warmer portions of the year that this

E2

106 COSMOS.

stations this opposition and the periodicity of the horary va-
riation in the dip have been firmly established by several
thousand regularly prosecuted observations, which have all
been submitted to a careful discussion since 1840. The
present would seem the most fitting place to notice the facts
that have been obtained as materials on which to base a
general theory of terrestrial magnetism. It must, however,
first be observed, that if we consider the periodical varia-
tions which are recognized in the three elements of terrestrial
magnetism, we must, with Sabine, distinguish, in the turn-
ing hours at which the maxima or minima occur, two great-
er, and therefore more important, extremes, and other slight
variations, which seem to be intercalated among the others,
as it were, and which are for the most part of an irregular
character. The recurring movements of the horizontal and

variation is sufficiently sensible to be observed with a lens. I would
still insist upon the fact thaj; changes of inclination are not sufficient
to explain the change of intensity, deduced from the observation of a
horizontal needle. An augmentation of temperature, all other cir-
cumstances remaining the same, retards the oscillations of the nee-
dles. In the evening the temperature of my horizontal.needle is al-
ways highe?- than in the morning ; hence the needle juust on that account
make fewer oscillations in a given time in the evening than in the
morning; in fact, it oscillates more frequently than we can account
for by the change of inclination, and hence there must be a real aug-
vientaiion of intensity from morning till evening in the ten-estrial mag-
netic force." Later and more numerous observations at Greenwich,
Berlin, St. Petersburg, Toronto, and Hobarton, have confirmed Ara-
go's assertion (in 1827) that the horizontal intensity was greater in
the evening than toward morning. At Greenwich the principal max-
imum of the horizontal force was about 6 P.M., the principal minimum
about 10 A.M., or at noon ; at Schulzendorf, near Berlin, the maxi-
mum falls at 8 P.M., the minimum at 9 A.M. ; at St. Petersburg the
maximum falls at 8 P.M., the minimum at llh. 20m. A.M. ; at To-
ronto the maximum falls at 4 P.M., the minimum at 11 A.M. The
time is always reckoned according to the true tira# of the respective
places (Airy, Magn. Observ. at Greenwich for 1845, p. 13; for 1846,
p. 102; for 1847, p. 241 ; Riess and Moser, in Poggend., ^rzna/ew, bd.
xix., 1830, s. 175 ; Kupffer, Compte rendu Annuel de P Observatoire Cen-
trale Magn. de St. Petersb., 1852, p. 28 ; and Sabine, Magn. Observ.
at Toronto, vol. i., 1840-1842, p. xhi.). The turning hAurs at the
Cape of Good Hope and at St. Helena, where the horizontal force is
the weakest in the evening, seem to be singularly at variance, and
almost the very opposite of one another (Sabine, Magn. Observ. at the
Cape of Good Hope, p. xl., at St. Helena, p. 40). Such, however, is
not the case further eastward, in other parts of the great southern
hemisphere. "The principal feature in the diurnal change of the
horizontal force at Hobarton is the decrease of force in the forenoon,
and its subsequent increase in the afternoon" (Sabine, Magn, Obs. at
Hobarton^ vol. i., p. liv., vol. ii., p. xliii.).

MAGNETIC INCLINATION. 107

(tipping needles, as well as the variation in the intensity of
the total force, consequently present principal and secondary
maxima or minima, and generally some of either type, which
therefore constitutes a double progression with four turning
hours (the ordinary case), and a simple progression with two
turning hours, that is to say, with a single maximum and a
single minimum. Thus, for instance, in Van Diemen's Land,
the intensity or total force exhibits a simple progression, com-
bined with' a double progression of the inclination, while at
one part of the northarn hemisphere, which corresponds ex-
actly witU the position of Hobarton, namely, Toronto, in
Canada, both the elements of intensity and inclination ex-
hibit a double progression.* At the Cape of Good Hope
there is only one maximum and one minimum of inclination.
The horary periodical variations of the magnetic dip are as
follows :

I. Northern Hemisphere.

Greenwich : Maxim. 9 A.M. ; minim. 3 P.M. (Airy, Oh-
serv. in 1845, p. 21; in 1846, p. 113; in 1847, p. 247).
Inclin. in the last-named year, about 9 A.M., on an average
08° 59^ 3'^; but at 3 P.M. it was 68° 58^ 6'^. In the
monthly variation the maximum falls between April and
June, and the minimum between October and December.

Paris: Maxim. 9 A.M.; minim. 6 P.M. This simple
progression from Paris and Greenwich is repeated at the
Cape of Good Hope.

St. Petersburg : Maxim. 8 A.M. ; minim. 10 P.M. Va-
riation of the inclination the same as at Paris, Greenwich,
and Pekin ; less in the cold months, and the maxima more
closely dependent on time than the minima.

Toronto : Principal maxim. 10 A.M. ; principal minim.
4 P.M. ; secondary maxim. 10 P.M. ; secondary minim. 6
A.M. (Sabine, Tor., 1840-1842, vol. i., p. Ixi.)

n. Southern Hemisphere.

Hobarton, Van Diemen's Land : Principal minim. 6 A.M. ;
principal maxim. 11*30 A.M.; secondary minim. 5 P.M.;
secondary maxim. 10 P.M. (Sabine, Hob., vol. i., p. Ixvii.).
The inclination is greater in the summer, when the sun is in
the southern zodiacal signs, 70° 36^-74; it is smaller in win-
ter, when the sqn is in the northern s!gns, 70° 34^*66. The
annual mean taken from the observations of six years gives
* Sabine, Tloharton. \6\. i., p. Ixvii., Ixix.

108 COSMOS.

70° 36'-01. (Sabine, Hob., vol. ii., p. xliv.)- Moreover, the
intensity at Hobarton is greater from October to February
than from April to August, p. xlvi.

Cape of Good Hope: Simple progression, the minim,
being Oh. 34m. P.M. ; maxim. 8h. 34m. P.M., with an ex-
ceedingly small intermediate variation between 7 and 9 A.M.
(Sabine, Cape Obs., 1841-1850, p. liii.).

The phenomena of the turning hours of the maximum of
the inclinations expressed in the time of the place fall with
remarkable regularity between 8 and 10 A.M. for places in
the northern hemisphere, such as Toronto, Pai'is, Green-
wich, and St. Petersburg, while in like manner the minima
of the turning hours all fall in the afternoon or evening, al-
though not within equally narrow limits (at 4, 6, and 10
P.M.). It is so much the more remarkable, that in the
course of very accurate observations made at Greenwich
during five years there was one year, 1845, in which the
epochs of the maxima and minima were reversed. The an-
nual mean of the inclinations was for 9 A.M. ; 68° 56^-8,
and for 3 P.M. : 68° 58'-l.

When we compare together the stations of Toronto and
Hobarton, which exhibit a corresponding geographical posi-
tion on either side of the equator, we find that there is at
Hobarton a great difference in the turning hours of the prin-
cipal minimum of inclination (at 4 o'clock in the afternoon
and 6 o'clock in the morning), although such is not the case
in the turning hours of the principal maximum (10 and
1 1 .30 A.M.). The period of the principal minimum (6 A.M.)
at Hobarton coincides with that of the secondary minimum
at Toronto. The principal and secondary maxima occur at
both places at the same hours, between 10 and 11.30 A.M.
and 10 P.M. The four turning hours of the inclination occur
almost precisely the same at Toronto as at Hobarton, only
in a reversed order (4 or 5 P.M., 10 P.M., 6 A.M., and 10 or
11.30 A.M.). This complicated effect of the internal terres-
trial force is very remarkable. If, on the other hand, we
compare Hobarton and Toronto in respect to the order in
which the turning hours of the alterations of intensity and
inclination occur, we shall find that at the former place in
the southern hemisphere the minimum of the intensity fol-
lows only two hours ^fter the principal minimum of the in-
clination, while the delay in the maximum amounts to six
hours ; while in the northern hemisphere, at Toronto, the
minimum of intensity precedes the principal maximum of

MAGNETIC INCLINATION. 109

inclination by eight hours, while the maximum of intensity
differs only by two hours from the minimum of inclination.*

The periodicity of inclination at the Cape of Good Hope
does not coincide with that at Hobarton, which lies in the
same hemisphere, nor with any one point of the northern
hemisphere. The minimum of inclination is indeed reached
at an hour at which the needle at Hobarton has very nearly
reached the maximum.

For the determination of the secular variation of the in-
clination it is necessary to have a series of observations that
have not only been conducted with extreme accuracy, but
which have likewise extended over long intervals of time.
Thus, for instance, we can not go with certainty as far back
as the time of Cook's voyages, for although in his third ex-
pedition the poles were always reversed, we frequently ob-
serve differences of 40^ to 55^ in the observations of this
great navigator and of Bayley on the Pacific Ocean, a dis-
crepancy which may very probably be referred to the imper-
fect construction of the magnetic needle at that time, and to
the obstacles which then prevented its free* motion. For
London we scarcely like to go further back than Sabine's
observation of August, 1821, which, compared with the ad-
mirable determination made by himself. Sir James Ross and
Fox in May, 1838, yielded an annual decrease of 2^-73,
while Lloyd with equally accurate instruments, but in a
shorter interval of time, obtained at Dublin the very accord-
ant result of 2^-38.1 At Paris, where the annual diminution
of inclination is likewise on the decrease, this diminution is
greater than in London. The very ingenious methods sug-
gested by Coulomb for determining the dip had, indeed, led
their inventor to incorrect results. The first observation
which was made with one of Le Noir's perfect instruments
at the Paris Observatory belongs to the year 1798. At that
time I found, after often repeating the experiments conjoint-
ly with the Chevalier Borda, 69° 51^; in the year 1810, in
conjunction with Arago, I found 68° 50^-2 ; and in the year
1826, with Mathieu, 67° 56^-7. In the year 1841 Arago
found 67° 9^, and in the year 1851 Laugier and Mauvais

* Total intensity at Hobarton, max. 5h. 30m. P.M., min. 8h. 30m.
A.M.; at Toronto, principal max. 6 P.M., principal min. 2 A.M.,
secondary max. 8 A.M., secondary min. 10 A.M. See Sabine, To-
ronto, vol. i., p. Ixi., Ixii., and Hobarton, vol. i,, p. Ixviii.

t Sabine, Report on the Isoclinal and Isodynamic Lines in the British
Islands, 1839, p. 61-63.

110 COSMOS.

found 66° 35^ — all these observers adopting similar methods
and using similar instruments. This entire period, which
extends over more than half a century (from 1798 to 1851),
gives a mean annual diminution of the inclination at Paris
of 3^*69. The intermediate periods stood as follows :

From 1798 to 1810 at 5'-08 I From 1826 to 1841 at 3'-13
" 1810 to 1826 " 3'-37 I " 1841 to 1851 " 3'-40 ^ '

The decrease between 1810 and 1826 has been strikingly
though gradually retarded ; for an observation which Gay-
Lussac made with extreme care (69° 12^), after his return in
1806 from Berlin, whither he had accompanied me after our
Italian expedition, gave an annual diminution of 4^-87 since
1798. The nearer the node of the magnetic equator ap-
proaches to the meridian of Paris in its secular progression
from east to west, the slower seems to be the decrease, rang-
ing in half a century from about 5^-08 to 3^-40. Shortly
before my Siberian expedition in April, 1829, I laid before
the Academy of Berlin a memoir, in which I had compared
together the different points observed by myself, and which,
I believe I may venture to say, had all been obtained with
equal care* Sabine, more than twenty-five years after me,
measured the inclination and intensity of the magnetic force
at the Havana, which, in respect to these equinoctial regions,
affords a very considerable interval of time, while he also de-
termined the variation of two important elements. Han-
steen, in 1831, gave the result of his investigations of the an-
nual variation of the dip in both hemispheres,! in a very ad-

* Humboldt, in Pogf]^end., Annalen, bd. xv., s. 319-336, bd. xix., s.
357-391 ; and in the Voyage aux Regions Equinox., t. iii., p, 616-625.

t Hansteen, Ueber jdhrlicke Verdnderung der Inclination, in Pog-
gend., Ann., bd. xxi., s. 403-429. Compare also, on the influence of
the progression of the nodes of the magnetic equator. Sir David Brew-
ster, Treatise on Magnetism, p. 247. As the great number of observa-
tions made at different stations have opened an almost inexhaustible
field of inquiry in this department of special investigation, we are
constantly meeting with new complications in our search for the laws
by which these forces are controlled. Thus, for instance, in the course
of a series of successive years we see that the dip passes in one of the
turning hours — that of the maximum from a deci-ease to an absolute
increase, while in the turning hour of the minimum the progressive
annual decrease continued the same. Thus, at Greenwich, the mag-
netic inclination in the maximum hour (9 A.M.) decreased in the
years 1844 and 1845, while it increased at the same hour from 1845
to 1846, and continued in the turning hour of the minimum (3 P.M.)
to decrease from 1844 to 1846 (Airv, Magn. Observ. at Greenwich,
1846, p. 113).

MAGNETIC OBSERVATIONS. Ill

mirable work, which is of a more comprehensive nature
than my own.

Although Sir Edward Belcher's observations for the year
1838, when compared with those I made in 1803 (see p. 73),
along the western coast of America, between Lima, Guaya-
quil, and Acapulco, indicate considerable alterations in the
inclination (and the longer the intermediate period the great-
er is the value of the results), the secular variation of the dip
at other points of the Pacific has been found to be strikingly
slow. At Otaheite, Bayley foynd, in 1773, 29° 43^ ; and
Fitzroy, in 1835, 30° IV ; while Captain Belcher, in 1840,
again found 30° 17^; and hence the mean annual variation
scarcely amounted, in the course of sixty-seven years, to
0^-51.=^ A very careful observer, Sawelieff, found in North-
ern Asia, twenty-two years after my visit to those regions,
in a journey which he made from Casan to the shores of the
Caspian Sea, that the incHnation to the north and south of
the parallel of 50° had varied very irregularly.!

Humboldt. Sawelieff.

1829. 1851.

Casan 68° 26-7 68° 30-8

Saratow •... 64M0'-9 64° 48'-7

Sarepta 62° 15'-9 (i2° 39 -6

Astrachau 59° 58-3 60° 27-9

For the Cape of Good Hope we now possess an extended
series of observations, which, if we do not go further back
than from Sir James Ross and Du Petit Thouars (1840) to
Vancouver (1791), may be regarded as of a very satisfactory
nature in respect to the variation of the inclination for near-
ly half a century.f

The solution of the question whether the elevation of the
soil does in itself exert a perceptible influence on magnetic
dip ^nd intensity,§ was made the subject of very careful in-
vestigation during my mountain journeys in the chain of the
Andes, in the Ural, and Altai. I have already observed, in

* Phil. Transact, for 1841, pt. i., p. 35.

t Compare Sawelieff, in the Bulletin Physico-M'athematique de VAcad.
Imp. de St. Petersb., t. x., No. 219, with Humboldt, Asie Centr.^ t. iii.,
p. 440.

X Sabine, Magn. Observ. at the Cape of Good Hope, vol. i., p. Ixv.
If we may trust to the observations made by Lacaille for the year 1751 ,
who, indeed, always reversed the poles, but who made his observations
with a needle which did not move freely, it follows that there has been
an increase in the incHnation at the Cape of Good Hope of 3°'08 in
eighty-nine years !

^ Arago, in the Annuaire du Bureau des Long, pour 1825, p. 285-288.

112 COSMOS.

the section on Magnetic Intensity, how very few localities
were able to afford any certainty as to this question, because
the distance between the points to be compared together
must be so small as to leave no ground for suspecting that
the difference found in the inclination may be a consequence
of the elevation of the soil, instead of the result of the curv-
ature of the isodynamic and isoclinal lines, or of some great
peculiarity in the composition of the rocks. I will limit
myself to the four results which I thought, at the time they
were obtained, showed more decisively than could be done
by observations of intensity the influence exerted by eleva-
tion in diminishing the dip of the needle.

The Silla de Cai^acas, which rises almost vertically above
La Guayra, and 8638 feet above the level of the sea, south
of the coast, but in its immediate vicinity, and north of the
town of Caracas, yielded 'the inclination of 41°-90; La
Guayra elevation 10 feet, inclination 42°-20 ; the town of
Caracas, height above the shores of the Rio Guayre, 2648
feet, inclination 42°-95. (Humboldt, Voy. aux Reg, Equi-
nox., t. i., p. 612.)

Santa Fe de Bogota: elevation 8735 feet, inclination
27°*15 ; the chapel of Nuestra Senora de Guadalupe, built
upon the projecting edge of a rock, elevation 10,794 feet,
inclination 26°-80.

Popayan: elevation 5825 feet, inclination 23°*25 ; mount-
ainous village of Purace on the declivity of the volcano, ele-
vation 8671 feet, inchnation 21°-80; summit of the volcano
of Purace, elevation 14,548 feet, inclination 20°*30.

Quito: elevation 9541, inclination 14°*85 ; San Antonio
de Lulumbamba, where the geographical equator intersects
the torrid valley, elevation of the bottom of the valley 8153
feet, inclination 16°*02. (All the above-named inclini^tions
have been expressed in decimal parts of a degree.)

It might, perhaps, be deemed unnecessary, considering
the extent of the relative distances and the influence of the
neighboring kinds of rock, for me to enter fully into the
details of the following observations : the Hospice of St.
Gotthard, 7087 feet, inclination 66° 12^; compared with
Airolo, elevation 3727 feet, inclination 66° 54^ ; and Altorf,
inclination 66° 55^ ; or to notice the appare'ntly contradict-
ory data yielded by Lans le Bourg, inclination 66° 9'', the
Hospice of Mont Cenis, 6676 feet, inclination 66° 22', and
Turin 754 feet, inclination 66° 3'; or by Naples, Portici,
and the margin of the crater of Vesuvius ; or by the summit

MAGNETIC OBSERVATIONS. 113

of the Great Milischauer (Phonolith), inclination 67° 53^-5,
Teplitz inclination 67° 19^-5, and Prague inclination 66°
47^*6.* Simultaneously with the series of admirable com-
parative observations published with the fullest details of
the horizontal intensity, which were made in 1844 by Bra-
vais, in conjunction with Martins and Lepileur, and com-
pared at thirty-five stations, including the summits of Mont
Blanc (15,783 feet), of the Great St. Bernard (8364 feet),
and of the Faulhorn (8712 feet), the above-named physicists
made a series of inclination experiments on the grand plateau
of Mont Blanc (12,893 feet), and at Chamouni (3421 feet).
Although the compariscjn of these results showed that the
elevation of the soil exerted an influence in diminishing the
magnetic inclination, observations made at the Faulhorn and
at Brienz (1870 feet in elevation) showed the opposite result
of the inclination increasing with the height. The dilFerent
investigations on horizontal intensity and inclination failed
to yield any satisfactory solution of the problem. (Bravais,
Sar VIntensite du Magnetisme Terrestre en Frojice, en Suisse,
et en Savoie, in the Annales de Chimie et de Physique, 3eme
serie, t. xviii., 1846, p. 225.) In a manuscript report by
Borda of his expedition to the Canary Islands in the year
1776, which is preserved at Paris in the Depot de la Marine,
and which I have been enabled to consult through the oblig-
ing courtesy of Admiral Rosily, I have discovered that Borda
was the first who made an attempt to investigate the influ-
ence of a great elevation on the inclination. He found that
the inclination was 1° 15^ greater at the summit of the Peak
of TenerifFe than in the harbor of Santa Cruz, owing un-
doubtedly to the local attractions of the lava, as I have oft-
en observed on Vesuvius and different American volcanoes.
(Humboldt, Voy. aux Hegions Equinox., t. i., p. 116, 277,
288.)

In order to try whether the deep interior portions of the
body of the earth influence magnetic inclination in the same
manner as elevations above the surface, I instituted an ex-
periment during my stay at Freiberg, in July, 1828, with all
the care that I could bestow upon it, and with a constant

* I would again repeat that all the European observations of incli-
nation which have been given in this page have been reckoned accord-
ing to the division of the circle into 360 parts, and it is only in those
observations of inclination which I made myself before the month of
June, 1804, in the New Continent, that the centesimal division of the
arc has been adhered to ( Voy. aux Regions Equinox., t. iii., p. 615-623).

114 COSMOS.

inversion of the poles ; when I found, after very careful in-
vestigation, that the neighboring rock, which was composed
■of gneiss, exerted no action on the magnetic needle. The
depth below the surface was 854 feet, and the difference be-
tween the inclination of the subterranean parts of the mine
and those points which lay immediately above it, and even
with the surface, was only 2^*06 ; but, considering the care
with which my experiments were made, I am inclined to
think, from the results given for each needle, as recorded in
the accompanying note,* that ' the inclination is greater in
the Churprinz mine than on the surface of the mountain.
It would be very desirable if opportunities were to present
themselves, in cases where there is evidence that the rock
has not exerted any local influence on the magnet, for care-
fully repeating my experiments in mines, in which, like those
of Valenciana, near Guanaxuato, in Mexico, the vertical
depth is 1686 feet; or in English coal mines nearly 1900
feet deep ; or in the now-closed shaft at Kuttenberg, in Bo-
hemia, 3778 feet in depth.j

After a violent earthquake at Cumana, on the 4th of
November, 1799, I found that the inclination was dimin-
ished 0°-90, or nearly a whole degree. The circumstances

* In the Churprinz mine at Freiberg, in the mountains of Saxony,
the subterranean point was 133^ fathoms deep, and was observed with
Freiesleben and Reich at 2^ P.M. (temperature of the mine being
60°-08 F.). The dipping-needle A showed 67° 87'-4, the needle B
67° 32'- 7, the mean of both needles in the mine was 67° 35'-05, In
the open air, at a point of the surface which lies immediately above
the point of subterranean observation, the needle A stood at 11 A.M.
at 67° 33'-87, and the needle B at 67° 32'-12. The mean of both
needles in the upper station was 67° 32'-99, the temperature of the
air being 60°-44 F., and the difference between the xipper and lower
result 2'-06. The needle A, which, as the stronger of the two, in-
spired me with most confidence, gave even 3'-53, while the influence
of the depth remained almost inappreciable when the needle B only
was used (Humboldt, in Poggend., AnnaL, bd. xv., s. 326). I have
already described in detail, and elucidated by examples, in Asie
Centr., t. iii., p. 465-467, the uniform method which I have always
employed in reading the azimuth circle in order to find the magnetic
meridian by corresponding inclinations, or by the perpendicular posi-
tion of the needle ; as also to find the inclination itself on the vertical
circle by reversing the bearings of the needle and by taking the read-
ings at both points, before and after the poles had been reversed. Tl:e
position of the two needles has, in each case, been read off sixteen
times, in order to obtain a mean result. Where so small an amount
has to be determined, it is necessary to enter fully- into the individual
details of the observation.

t Cosmos, vol. i., p. 157.

MAGNETIC OBSERVATIONS. 115

under which I obtained this result, and which I have else-
where fully described,* aiFord no sufficient ground for the sus-
picion of an error in the observation. Shortly' after my ar-
rival at Cumana I found that the inclination was 43°*53. A
few days before the earthquake I was induced to begin a
long series of carefully-conducted observations in the harbor
of Cumana, in consequence of having accidentally noticed a
statement in an otherwise valuable Spanish work, Mendoza's
Tratado de Navegacion, t. ii., p. 72, according to which it
was erroneously asserted that the hourly and monthly alter-
ations of inclination were greater than those of variation. I
found, between the 1st and 2d of November, that the inclina-
• tion exhibited very steadily the mean value of 43°-6o. The
instrument remained untouched and properly leveled on the
same spot, and on the 7th of November, and therefore three
days after the great earthquake, and when the instrument
had again been adjusted, it yielded 42°-75. The intensity
of the force, measured by vertical oscillations, was not
changed. I expected that the inclination would, perhaps,
gradually return to its former position, but it remained sta-
tionary. In September, 1800, in an expedition of more than
2000 geographical miles on the waters and along the shores
of the* Orinoco and the Rio Negro, the same instrument,
which was one of Borda's, which I had constantly carried
with me, yielded 42°-80, showing, therefore, the same dip as
before my journey. As mechanical disturbances and elec-
trical shocks excite polarity in soft iron by altering its mo-
lecular condition, we might suspect a connection between the
influences of the direction of magnetic currents and the di-
rection of earthquakes ; but carefully as I observed this phe-
nomenon, of whose objective reality I did not entertain a
doubt in 1799, I have never on any other occasion, in the
many earthquakes which I experienced in the course of three
years at a subsequent period in South America, noticed any
sudden change of the inclination which I could ascribe to
these terrestrial convulsions, however different were the di-
rections in which the undulations of the strata were propa-
gated. A very accurate and experienced observer, Erman,
likewise found that after an earthquake at Lake Baikal, on
the 8tli of March, 1828, there was no disturbance in the
declinatienf and its periodic changes.

* Humboldt, Voy. aux Regions Equinox., t. i., p. 515-517.
t Erman, Reise urn die Erde, bd. ii., s. 180.

116 . COSMOS.

Declination.

We have already referred to the historical facts of the
earliest recognition of those phenomena which depend upon
the third element of terrestrial magnetism, namely, declina-
tion. The Chinese, as early as the 12th century of our era,
were not only well acquainted with the fact of the varia-
tion of a horizontal magnetic needle (suspended by a cot-
ton thread) from the geographical meridian, but they also
knew how to determine the amount of this variation. The
intercourse which the Chinese carried on with the Malays
and Indians, and the latter with Arab and Moorish pilots,
led to the extensive use of the mariner's 'compass among the
Genoese, Majorcans, and Catalans, in the basin of the Med-
iterranean, on theVest coast of Africa, and in high northern
latitudes ; while the maps, which were published as early as
1436, even give the variation for different parts of the sea.*
The geographical position of a line of no variation, on which
the needle turns to the true north — the pole of the axis of
the earth — was determined by Columbus on the 13th of
September, 1492, and it did not escape his notice that the
knowledge of the magnetic declination might serve in the de-
termination of geographical longitudes. I have elsewhere
shown, from the Admiral's log, that when he was uncertain
of the ship's reckoning, he endeavored, on his second voyage,
April, 1496, to ascertain his position by observations of dec-
lination.! The horary changes of variation, which were sim-
ply recognized as certain facts by Hellibrand and Father
Tachard, at Louvo,'^in Siam, were circumstantially and al-
most conclusively observed by Graham in 1722. Celsius
was the first who made use of these observations to institute
simultaneous measurements at two widely remote points.J

* See page 53 ; Petrus Peregrine informs a friend that he found
the variation in Italy was 5° east in 1269.

t Humboldt, Examen. Grit, de tllist. de la Geogr., t. iii., p. 29, 36,
38, 44-51. Although Herrera {Dec.^ i., p. 23) says that Columbus had
remarked that the magnetic variation was not the same by day and by
night, it does not justify us in ascribing to this great discoverer a
knowledge of the horary variation. The actual journal of the admiral,
which has been published by Navarrete, informs ns that from the 17th
to the 30th of September, 1492, Columbus had reduced every thing to
a so-called "unequal movement" of the polar star and the pointers
(Guat'das), Examen Gnt., t. iii., p. 56-59.

t See pages 61, 70. The first-printed observations for London are
those by Graham, in the Philos. Transact, for 1724 awe? 1725, vol.
xxxiii., p. 96-107 {An Account of Observations made of the Horizontal

MAGNETIC VARIATION. 117

•

Passing to the consideration of the phenomena observed
in the variation of the magnetic needle, we must first notice
its alterations in respect to the different hours of the night
and day, the different seasons of the year, and the mean
annual values; next, in respect to the influence which the
extraordinary, although periodically recurring disturbances,
and the magnetic position, north or south of the equ^or,
exert on these alterations ; and, finally, in respect to the dif-
ferent lines passing through the terrestrial points at which
the variation is equal, or even null. These linear relations
are certainly most important in respect to the direct prac-
tical application of their results to the ship's reckoning, and
to navigation generally ; but all the cosmical phenomena of
magnetism, among which we must place those extraordinary
and most mysterious disturbances which often act simultane-
ously at very remote distances (magnetic storms), are so in-
timately connected with one another, that no single one of
them can be neglected in our attempt gradually to complete
the mathematical theory of terrestrial magnetism.

In the middle latitudes, throughout the whole northern
magnetic hemisphere (the terrestrial spheroid being assumed
to be divided through the magnetic equator), the north end
of the magnetic needle — that is to say, the end which points
toward the north pole — is most closely in the direction of
that pole about 8h. 15m. A.M. The needle moves from east
to west from this hour till about Ih. 45m. P.H., at which
time it attains its most westerly position. This motion
westward is general, and occurs at all places in the northern
hemisphere, whether they have a western variation — as the
whole of Europe, Pekin, Nertschinsk, and Toronto — or an
eastern variation, like Kasan, Sitka (in Russian Ameri-
caj, Washington, Marmato (New Granada), and Payta, on
the Peruvian coast.* From this most westerly point, at

Needle at London, 1722-1723, by Mr. George Graham). The change
of the variation depends "neither upon heat nor cold, dry or moist
air. The variation is greatest between 12 and 4 in the afternoon, and
the least at 6 or 7 in the evening." These, however, are not the true
turning hours.

* Proofs of this are afforded by numerous observations of George
Fuss and Kowanko ; at the observatory in the Greek convent at Pekin ;
by Anikin at Nertschinsk ; by Buchanan Riddell at Toronto, in Cana-
da (all these being places of western variation) ; by Kupffer and Si-
monoff at Kasan ; by Wrangle, notwithstanding the many disturb-
ances from the Aurora Borealis at Sitka, on the northwest coast of
America ; by Gilliss at Washington ; by Boussingault at Marmato, in
South America ; and by Duperrey at Payta, on the Peruvian shores

118 COSMOS.

Ih. 45m. P.M., the magnetic needle continues to retrograde
toward the east throughout the whole of the afternoon and
a portion of the night till midnight, or 1 A.M., while it often
makes a short pause about 6 P.M. In the night there is
again a slight movement toward the west, until the minimum
or eastern position is reached at 8h. 15m. A.M, This noc-
turnal period, which was formerly entirely overlooked, since
a gradual and uninterrupted retrogression toward the east
between Ih. 45m. P.M. and 8h. 15m. A.M. was assumed,
had already been carefully studied by me at Rome, when I
was engaged with Gay-Lussac in observing the horary
changes of variation with one of Prony's magnetic tele-
scopes. As the needle is generally unsteady as long as the
sun is below the horizon, the small nocturnal motion west-
ward is more seldom and less distinctly manifested. At
those occasions when this motion was clearly discernible, I
never saw it accompanied by any restlessness of the needle.
The needle, during this small western period, passes quietly
from point to point of the dial, exactly in the same manner
as in the reliable diurnal period, between 8h. 15m. A.M.
and Ih. 45m. P.M., and very differently from the manner in
"which it moves during the occurrence of the phenomenon
which I have named a magnetic storm. It is very remark-
able that when the needle changes its continuous western
motion into an eastern movement, or conversely, it does not
continue unchanged for any length of time, but it turns
round almost suddenly, more especially by day, at the above-
named periods, 8h. 15m. A.M. and Ih. 45m. P.M. The slight
motion westward does not commonly occur until after mid-
night and toward the early morning. On the other hand, it
has been observed at Berlin, and during the subterranean
observations at Freiberg, as well as at Greenwich, Makefs-

of the Pacific (all these being places with an eastern variation). I
would here observe that the mean declination was 2° 15' 42" west at
Pekin (Dec, 1831) (Poggend., Annalen, bd. xxxiv., s. 54); 4° 7' 44"
west at Nertschinsk (Sept., -1832) (Poggend., Op. cit., s. 61); 1° 33'
west at Toronto (November, 1847) (see Observ. at the Magnetical and
Meteorological Observatory at Toronto, vol. i., p. 11 ; and Sabine, in the
Phil Transact, for 1851, pt. ii., p. 636), 2° 21' east at Kasan (August,
1828) (KupfFer, Simonoff, and Erman, lleise urn die Erde, bd. ii., s.
532); 28° 16' east at Sitka (November, 1829) (Erman, 0;>. cit., s.
546); 6° 33' east at Marmato (August, 1828) (Humboldt, in Poggend.,
Annalen, bd. xv., s. 331); 8" 56' east at Payta (August, 1823) (Dii-
perrey, in the Connaissance des Tejnps pour 1 828, p. 252). At Tiflis
the declination was westerly from 7 A.M. till 2 P.M. (Pan'ot, ReisQ
zum Ararat, 1834, th. ii., s. 58).

MAGNETIC VARIATION. 119

ton in Scotland, Washington, and Toronto, soon after 10 or
11 P.M.

The four movements of the needle, which I recognized in
1805,* have been represented in the admirable collection of
observations made at Greenwich in the years 1845, 1846,
and 1847, as the results of many thousand horary observa-
tions in the following four turning points,! namely, the first

* See extracts from a letter, which I addressed to Karsten, from
Rome, June 22, 1805, "On four motions of the magnetic needle,
constituting, as it were, four periods of magnetic ebbing and flowing,
analogous to the barometrical periods." This communication was
printed in Hansteen's Magnetismus der Ercle, 1819, s. 459. On the
long-disregarded nocturnal alterations of variation, see Faraday, On
the Night Episode, § 3012-3021.

t Airy, Magnetic and Meteorological Observations made at Greenwich
(Results, 1845, p. G; 1846, p. 94; 1847, p. 236).- The close correspond-'
ence between the earliest results of the nocturnal and diurnal turning
hours, and those which were obtained four years later, in the admi-
rable observatories at Greenwich and at Toronto, in Canada, is clearly
shown by the investigation made by my old friend Enke, the distin-
guished director of the observatory at Berlin, between the correspond-
ing observations of Berlin and Breslau. He wrote as follows on the
11th of October, 1836: "In reference to the nocturnal maximum, or
the inflection of the curve of horary A^ariation, I do not think that
there can be a doubt, as, indeed, Dove has also shown from the Frei-
berg observations for 1830 (Poggend., Ann., bd. xix., s. 373). Graph-
ical representations are preferable to numerical tables for affording a
correct insight into this phenomenon. In the former great irregular-
ities at once attract the attention, and enable the observer to draw a
line of average ; while in the latter the eye is frequently deceived, and
individual and striking irregularities are mistaken for a true maximum
or minimum. The periods seem to fall regularly at -the following
turning hours :

Tlie greatest eastern declination falls at 8 A.M., 1 maximiim E.

The greatest ^restern declination falls at. 1 P.M., 1 minimum E.

Tlie secondary or lesser eastern maximum falls at . 10 P.M., 11 maximum E.

The secondary or lesser western minimum falls at . 4 A.M., 11 minimum E.

The secondary or.lesser minimum (the nocturnal elongation westward)
falls, more correctly speaking, between 3 and 5 A.M., sometimes nearer
the one hour, and sometimes nearer the other." I need scarcely ob-
serve that the periods which Enke and myself designate as the eastern
minima (the principal and the secondary minimum at 4 A.M.) are
named western maxima in the registers of the English and American
stations, which were established in 1840, and consequently our eas^erw
maxima (8 A.M. and 10 P.M.) would, in accordance with the same
form of expression, be converted into western minima. In order, there-
fore, to give a representation of the horary motion of the needle in its
general character and analogy in the northern hemisphere, I will em-
ploy the terms adopted by Sabine, beginning with the period of the
greatest loestern elongation, reckoned according to the mean time of the
place :

120 COSMOS.

minimum at 8 A.M. ; the first maximum at 2 P.M. ; the
second minimum at 12 P.M. or 2 A.M. ; and the second
maximum at 2 A.M. or 4 A.M. I must here content my-
self with merely giving the mean conditions, drawing atten-
tion to the fact that the morning principal minimum of 8h.

Freiberg, 1829. Breslau, 1836. Greenwich, 1846-47.

Maximum 1 P.M. 1 P.M. 2 P.M.

Minimum 1 A.M. 10 P. M. 12 P. jM.

Maximum 4 A.M. 4 A. M. 4 A.M.

Minimum 8 A.M. 8 A.M. 8 AM.

Makerston, 1842-43. Toronto, 1845-47. Washington, 1840-42.

. Maximum Oh. 40in. 1 1'. M. 2 P. ^i.

Minimum 10 P. M. 10 P.M. 10 P. M.

Maodmum 2h. 15m. A. M. 2 A. M. 2 A. M.

Minimum 7h. 15ra. A. M. 8 A. M. 8 A. M.

The different seasons exhibited some striking differences at Green-
wich. In the year 1847 there was only one maximum (2 P.M.) and
one minimum (12 night) during the winter ; in the summer there was
a double progression, but the secondary minimum occurred at 2 A.M.
instead of 4 A.M. (p. 236). The greatest western elongation (princi-
pal maximum) remained stationary at 2 P.M. in winter as well as in
summer, but the smaller or secondary minimum fell, in 1846, as usual
(p. 94), at about 8 A.M. in the summer, and in winter about 12 at
night. The mean winter western elongation continued, without inter-
mission, throughout the whole year between midnight and 2 P.M. (see
also for 1845, p. 5). We owe the erection of the observatory at Mak-
erston, Roxburghshire, in Scotland, to the generous scientific zeal of
Sir Thomas Brisbane (see John Allan Broun, Obs. in Mognetism and
Meteorology made at Makerston in 1843, p. 221-227). On the horary
diurnal and noctui'nal observations of St. Petersburg, see Kupffer,
Compte-renda Meteor, et Mag. a Mr. de Brock en 1851, p. 17. Sabine,
in his admirable and ingeniously combined graphic representation of
the curve of horary declination at Toronto (Phil. Transact, for 1851,
pt. ii., plate 27), shows that there is a singular period of rest (from 9
to 11 JP.M.) occurring before the small nocturnal western motion, which
begins about 11 P.M. and continues till about 3 A.M. "We find," he
observes, " alternate progression and retrogression at Toronto twice in
the 24 hours. In two of the eight quarters (1841 and 1842) the infe-
rior degree of regularity during the night occasions the occurrence of
a trijile maximum and minimum ; in the remaining quarters the turning
hours are the same as those of the mean of the two years." {Obs,
made at the Magn. and Meteor. Observatory at Toronto, in Canada, vol.
i., p. xiv., xxiv., 183-191, and 228; and Unusual Magn. Uisturbances,
pt. i., p: vi.) For the very complete observations made at Washing"
ton, see Gilliss, Magn. and Meteor. Observations made at Washington,
p. 325 {General Law). Compare with these Bache, Observ. at the
Magn. and Meteor. Observatory at the Girard College, Philadelphia, made
in the years 1840 to 1845 (3 volumes, containing 3212 quarto pages),
vol. i., p. 709; vol. ii., p. 1285; vol. iii., p. 2167, 2702. Notwith-
standing the vicinity of these two places (Philadelphia lying only 1° 4'
north, and 0° 7' 33" east of Washington), I find a diffei'ence in tho
lesser periods of the western secondary maximum and secondary min-
imum. The former falls about Ih. 30m., and the latter about 2h. 15m.
earlier at Philadelphia,

M^^NETIC VARIATION. 121

i? sitrv changed in our northern zone by the earlier or later
time of sunrise. At the two solstitial periods and the three
equinoxes, at which, conjointly with Oltmanns, I watched
the horary variations for five to six consecutive days and
nights, I found that the eastern turning point remained fixed
between 7h. 45m. A.M. and 8h. 15m. A.M. both in summer
and in winter, and was only very slightly anticipated by the
earlier period at which the sun rose.*

In the high northern latitudes near the Arctic circle, and
between the latter and the pole of the earth's rotation, the
regularity of the horary declination has not yet been very
clearly recognized, although there has been no deficiency in
the number of very carefully-conducted observations regard-
ing this point. The local action of the rocks and the fre-
quency of the disturbing action of the polar light, either in
the immediate vicinity or at a distance, made Lottin hesi-
tate ia drawing definite conclusions in reference to these
turning hours, from his own great and careful labors, which
were carried on during the French scientific expedition of
Lilloise in 1836, or from the earlier results that had been
obtained with much care and accuracy by Lowenorn in
A 786. It would appear that at Keikjavik, in Iceland, 64°
8^ lat., as well as at Godthaab, on the coast of Greenland,
according to observations made by the missionary Genge,
the minimum of the western variation fell almost as in the
middle latitudes at about 9 or 10 A.M., while the maximum
did not appear to occur before 9 or 10 P.M.f Farther to

* Examples of the slightly earlier occurrence of the turning hours
are given by Lieutenant Gilliss, in his Magn. Ohserv. of Washington,
p. 328. At Makerston, in Scotland (55° 35' N. lat.), variations are
observed in the secondary minimum, which occurs about 9 A.M. in
the first three and the last four months of the year, and about 7"A.M.
in the remaining five months (from April till August), the reverse be-
ing the case at Berlin and Greenwich (Allan Broun, Observ. made at
Makerston, p. 225). The idea of heat exerting an influence on the
regular changes of the horary variation, whose minimum falls in the
morning near the time of the minimum of the temperature, as the
maximum very nearly coincides with maximum heat, is most distinct-
ly contradicted by the nocturnal motions of the needle, constituting
the secondary minimum and secondary maximum. "There are two
maxima and two minima of variation in the twenty-four hours, but only
one minimum and one maximum of temperature" (Relshuber, in Pog-
gend., Annalen der Physik und Chemie, bd. 85, 1852, s. 416). On the
normal motion of the magnetic needle in Northern Germany, see Dove,
Poggend., Annalen, bd. xix., s. 364-374.

t Voy. en Islande et en Gtoenland, execute on 1835 et 1836, siir la
Corv. la Recherche; Physique (1838), p. 214-225, 358-867.

Vol. v.— F

122 COSMOS. .

the north, at Hammerfest, in Finmark, 70° 40' lat., Sabine
found that the motion of the needle was tolerably regular,
as in the south of Norway and Germany,* the western min-
imum being at 9 A.M. and the western maximum^ at Ih.
•30m. P.M. ; he found it, however, different at Spitzbergen,
in 79° 50' lat., where the above-named turning hours fell
at 6 and at 7h. 30m. A.M. In reference to the Arctic polar
Archipelago we possess an admirable series of observations,
made during Captain Parry's third voyage in 1825, by Lieu-
tenants Foster and James Boss, at Port Bowen, on the east-
ern coast of Prince Regent's Inlet, 73° 14' N. lat, which
were extended over a period of five months. Although the
needle passed twice in the course of twenty-four hours through
that meridian, which was regarded as the mean magnetic
meridian of the place, and although no Aurora Borealis was
visible for fully two months (during the whole of April and
May), the periods of the principal elongations varied from
four to six hours, and from January to May the means of
the maxima and minima of th% western variation differed by
only one hour ! The quantity of the declination rose in in-
dividual days from 1° 30' to 6° or 7°, while at the turn-
ing periods it hardly reaches as many minutes.f Not only
within the Arctic circle, but also in the equatorial regions —
as, for instance, at Bombay, 18° 56' lat. — a great complica-
tion is observable in the horary periods of magnetic varia-
tion. These periods may be grouped into two principal
classes, which present great differences between April and
October on the one hand, and between October and Decem-
ber on the other, and these are again divided into two sub-
periods, which are very far from being accurately determ-
ined.f ,, .. . .^^

* Sabine, Account of the Pendulum Experiments^ 1825, p. 500.

t See Barlow's " Report of the Observations at Port Bowen," in
the Edinb. New Philos. Journal, vol. ii., 1827, p. 347.

X Professor Orlebar, of Oxford, former superintendent of the Mag-
netic Observatory of the Island of Colaba, erected at the expense of
the East India Company, has endeavored to elucidate the complica-
ted laws of the changes of declination in the sub-periods {Observations
made at the Magn. and Meteor. Observatory at Bombay in 1845, Results,
p. 2-7). It is singular to find that the position of the needle during
the first period from April to October (western min. 7h. 30m. A.M.,
max. Oh. 30m. P.M. ; min. 5h. 30m,, max. 7 P.M.) coincides so close-
ly with that of Central Europe. The month of October is a transition
period, as-'the amount of diurnal variation scarcely amounts to two
minutes in November and December. Notwithstanding that this sta-
tion is situated 8° from the magnetic equator, there is no obvious reg-

MAGNETIC VARIATION. 123

Europeans could not have learned, from their own expe-
rience, the direction of the magnetic needle in the southern
hemisphere before the second half of the 15th century, when
they may have obtained an imperfect knowledge of it from
the adventurous expeditions of Diego Cam with Martin Be-
haim, and Bartholomew Diaz, and Vasco de Gama. The
Chinese, who, as early as the 3d century of our era, as well
as the inhabitants of Corea and the Japanese Islands, had
guided their course by the compass at sea, no less than by
land, are said, according to the "testimony of their earliest
writers, to have ascribed great importance to the south di-
rection of the magnetic needle, and this was probably main-
ly dependent on the circumstance that their navigation was
entirely directed to the south and southwest. During these
southern voyages, it had not escaped their notice that the
magnetic needle, according to whose direction they steered
their course, did not point accurately to the south pole. We
even know, from one of their determinations, the amount*
of the variation toward the southeast, which prevailed dur-
ing the 12th century. The application and farther diffusion
of such nautical aids favored the very ancient intercourse of
the Chinese and Indians with Java, and to a still greater
extent the voyages of the Malay races and their colonization
of the island of Madagascar.!

Although, judging from the present very northern position
of the magnetic equator, it is probable that the town of
Louvo, in Siam, was very near the extremity of the northern
magnetic hemisphere, when the missionary father, Guy Ta-
chard, first observed the horary alterations of the magnetic
variation at that place in the year 1682, it must be remem-
bered that accurate observations of the horary declination in
the southern magnetic hemisphere were not made for fully a
century later. John Macdonald watched the course of the

ularity in the turning hours. Eveiy where in nature, where various
causes of disturbances act upon a phenomenon of motion at recurring
periods (whose duration, however, is still unknown to us), the law by
which these disturbances are brought about often remains for a long
time unexplained, in consequence of the perturbing causes either re-
ciprocally neutralizing or intensifying one another.

* See my Exavien Crit. de IHist. de la Geogr., t. iii., p. 34-37.
The most ancient notice of the variation given by Keutsungchy, a
writer belonging to the beginning of the 12th century, was east ^
south. Klaproth's Lettre sur V invention de la Boussole, p. 68.

t On the ancient intercourse of the Chinese with Java, according to
statements of Fahian in the Fo-kue-si, see Wilhelm von Humboldt,
Ueber die Kawi Sprache^ bd. i., b. 16.

124 COSMOS.

needle during the years 1794 and 1795 in Fort Marlborough,
on the southwestern coast of Sumatra, as well as at St.
Helena.* The results which were then obtained drew the
attention of physicists to the great decrease in the quantity
of the daily alterations of variation in the lower latitudes.
The elongation scarcely amounted to three or four minutes.
A more comprehensive and a deeper insight into this phe-
nomenon was obtained through the scientific expeditions of
Freycinet and Duperrey, but the erection of magnetic sta-
tions at three important points of the southern magnetic
hemisphere — at Hobarton in Van Diemen's Land, at St.
Helena, and at the Cape of Good Hope (where for the last
ten years horary observations have been carried on for the
registration of the alterations of the three elements of terres-
trial magnetism in accordance with one uniform method) —
afforded us the first general and systematic results. In
the middle latitudes of the southern magnetic hemisphere the
needle moves in a totally opposite direction from that which
it follows in the northern ; for while in the south the needle
that is pointed southward turns from east to w^est between
morning and noon, the northern point of the needle exhibits
a direction from west to east.

Sabine, to whom we are indebted for an elaborate revision
of all these variations, has arranged the horary observations
that were carried on for five years at Hobarton (42° 53^ S.
lat., variation 9° 57^ east) and Toronto (43° 39^ N. lat., va-
riation 1° 33' west), so that we can draw a distinction be-
tween the periods from October to February, and from April

* Phil Transact, for 1795, p. 340-349, /or 1798, p. 397. The re-
sult which Macdonald himself draws from his observations at Fort
Marlborough (situated above the town of Bencoolen, in Sumatra, 3°
47' S. lat.), and according to which the eastern elongation was on the
increase from 7 A.M. to 5 P.M., does not appear to me to be entirely
justified. No regular observation was made between noon and 8, 4,
or 5 P.M. ; and it seems probable, from some scattered observations
made at different times from the normal hours, that the turning hours
between the eastern and western elongation fall, as early as 2 P.M.,
precisely the same as at Hobarton. We are in possession of declina-
tion observations made by Macdonald during 23 months (from June,
1794, to June, 1796), and from these I perceive that the eastern vari-
ation increases at all times of the year between 7h. 30m. A.M. till
noon, the needle moving steadily from west to east during that period.
There is here no trace of the type of the northern hemisphere (Toronto),
which was observable at Singapore from May till September ; and yet
Fort Marlborough lies in almost the same meridian, although to the
south of the geographical equator, and only 5'= 4' distant from Singa-
pore.

MAGNETIC VARIATION. 125

to August, since the intermediate months of March and Sep-
tember present, as it were, phenomena of transition. At
Hobarton the extremity of the needle, which points north-
ward, exhibits two eastern and two western maxima of elon-
gation,* so that in the period of the year from October to
February it moves eastward from 8 or 9 o'clock A.M. till 2
P.M., and then from 2 till 11 P.M., somewhat to the west;
from 11 P.M. to 3 A.M. it again turns eastward, and from
3 to 8 A.M. it goes back to the west. In the period between
April and August the eastern turning hours are later, oc-
curring at 3 P.M. and 4 A.M. ; while the w^estern turning
hours fall earlier, namely, at 10 A.M. and at 11 P.M. In
the northern magnetic hemisphere the motion of the needle
westward from 8 A.M. till 1 P.M. is greater in the summer
than in the winter; while in the southern magnetic hemis-
phere, where the motion has an opposite direction between
the above-named turning hours, the quantity of the elonga-
tion is greater when the sun is in the southern than when it
is in the northern sign§.

The question which I discussed seven years ago in the
Picture of Nature,f whether there may not be a region of
the earth, probably between the geographical and magnetic
equators, in which there is no horary variation (before the
return of the northern extremity of the needle to an oppo-
site direction of variation in the same hours), is one which,
it would seem, from recent experiments, and more especially
since Sabine's ingenious discussions of the observations made
at Singapore (1<^ 17' N. lat.), at St. Helena (15° 56^ S. lat.),
and at the Cape of Good Hope (33° 56^ S. lat), must be an-
swered in the negative. No point has hitherto been discov-
ered at which the needle does not exhibit a horary motion,
and since the erection of magnetic stations the important and
very unexpected fact has been evolved that there are places
in the southern magnetic hemisphere at which the horary
variations of the dipping-needle alternately participate in the
phenomena (types) of both hemispheres. The island of St.
Helena lies very near the line of weakest magnetic intensity,
in a region where this line divaricates very widely from the
geographical equator and from the line of no inclination.

* vSabine, Magn. Observ. macU at Hobarton, vol. i. (1841 and 1842),
p. XXXV.; 2, 148; vol. ii. (1843-1845), p. iii.-xxxv., 172-344. See
also Sabine, Obs. made at St. Helena^ and in Phil, Transact, for 1847,
pt. i., 55, pi. iv., and Phil. Transact, for 1851, pt. ii., p. 36, pi. xxvii.

t Cosmos, vol. i., p. 183.

126 COSMOS.

At St. Helena the movement of the end of the needle, which
points to the nortli, is entirely opposite, in the months from
May to September, from the direction which it follows in the
analogous hours from October to February, it has been
found after five years' horary observations, that during the
winter of the southern hemisphere, in the above-named peri-
ods of the year, while the sun is in the northern signs, the
northern point of the needle has the greatest eastern varia-
tion at 7 A.M., from which hour, as in the middle latitudes
of Europe and North America, it moves westward till 10
A.M., and remains very nearly stationary until 2 P.M. At
other parts of the year, on the other hand, namely, from Oc-
tober till February (which constitutes the summer of the
southern hemisphere, and when the sun is in the southern
signs, and therefore nearest to the earth), the greatest west-
ern elongation of the needle falls about 8 A.M., showing a
movement from west to east until noon, precisely in accord-
ance with the type of Hobarton (42^ 53^ S. lat.), and of oth-
er districts of the middle parts of th% southern hemisphere.
At the time of the equinoxes, or soon afterward, as, for in-
stance, in March and April, as well as in September and Oc-
tober, the course of the needle fluctuates on individual days,
showing periods of transition from one type to another, from
that of the northern to that of the southern hemisphere.*

Singapore lies a little to the north of the geographical
equator, between the latter and the magnetic equator, which,
according to Elliot, coincides almost exactly with the curve
of lowest intensity. According to the observations which

* Sabine, Observations mads at the Magn. and Meteo7\ Observatory at
St. Helena in 1840-1845, vol. i., p. 30 ; and in the Phil. Transact, for
1847, pt. i., p. 51-56, pi. iii. The regularity of this opposition in the
two divisions of the year, the first occurring between May and Septem-
ber (type of the middle latitudes in the northern hemisphere), and the
next between October and February (type of the middle latitudes in
the southern hemisphere), is graphically and strikingly manifested
when we separately compare the form and inflections of the curve of
horary variation in the portions of the day intervening between 2 P.M.
and 10 A.M., between 10 A.M. and 4 P.M., and between 4 P.M. and
2 A.M. Every curve above the line which indicates the mean decli-
nation has an almost similar one corresponding to it below it (vol. i.,
pi. iv,, the curves A A and BB). This opposition is perceptible even
in the rvDCturnal periods, and it is still more remarkable that, while the
type of St. Helena and of the Cape of Good Hope is found to be that
belonging to the noi-thern hemisphere, the same earlier occurrence of
the turning hours which is observed in Canada (Toronto) is noticed in
the same months at these two southern points. Sabine, Obscj'v. at
Hobarton, vol. i., p. xxxvi.

MAGNETIC VARIATION. 127

were made at Singapore every two hours during the years
1841 and 1842, Sabine again finds the St. Helena types in
the motion of the needle from May to August, and from
November to February; the same occurs at the Cape of
Good Hope, which is 34° distant from the geographical and
still more remote from the magnetic equator, and where the
inclination is 53° south and the sun never reaches the ze-
nith.* We possess the published horary observations made
at the Cape for six years, from May to September, according
to which, almost precisely as at St. Helena, the needle moves
westward till llh. 30m. A.M. from its extreme eastern po-
sition (7h. 30m. A.M.), while from October to March it
moves eastward from 8h. 30m. A.M. to Ih. 30m. and 2 P.M.
The discovery of this well-attested, but still unexplained and
obscure phenomenon, has more especially proved the import-
ance of observations continued uninterruptedly from hour to
hour for many years. Disturbartces which, as we shall soon
have occasion to show, have the power of diverting the nee-
dle either to the eastward or westward for a length of time,
would render > the isolated observations of travelers uncer-
tain.

By means of extended navigation and the application of
the compass to geodetic surveys, it was very early noticed
that at certain times the magnetic needle exhibited an ex-

* PhiL Transact, for 1847, pt. i,, p. 52, 57; and Sabine, Observa-
tions mack at the Magn. and Meteor. Observatory at the Cape of Good
Hope, 1841-1840, vol. i., p. xii.-xxiii., pi. iii. ' See also Faraday's in-
genious views regarding the causes of those phenomena, which depend
upon the alternations of the seasons, in \i\& Experiments on Atmospheric
Magnetism, § 3027-3068, and on the analogies with St. Petersburg, §
3017. It would appear that the singular type of magnetic declination,
varying with the seasons, which prevails at the Cape of Good Hope,
St. Helena, and Singapore, has been noticed on the southern shores
of the Red Sea by the careful observer D' Abbadie (Airy, On the Present
State of the Science of Terrestrial Magnetism, 1850, p. 2). "It results
from the present position of the four points of maximum of intensity
at the surface of the earth," observes Sabine, "that the important
curve of the relatively, but not absolutely, weakest intensity in the
Southern Atlantic Ocean should incline away from the vicinity of St.
Helena, in the direction of the southern extremity of Africa. The as-
tronomico-geographical position of this southern extremity, where the
sun remains throughout the whole year north of the' zenith, affords a
principal ground of objection against De la Rive's thermal explanation
(Annales de Chimie et de Physique, t. xxv., 1849, p. 310) of the phenom-
enon of St. Helena here referred to, which, although it seems at first
sight apparently abnormal, is nevertheless entirely in accordance with
established law, and is found to occur at other points." See Sabine,
in the Proceedings of the Royal Society, 1849, p. 821.

128 COSMOS.

traordinary disturbance in its direction, which was frequent-
ly connected with a vibratory, trembling, and fluctuating
motion. It became customary to ascribe this phenomenon
to some special condition of the needle itself, and this was
characteristically designated by French sailors Vaffolement de
raiguille, and it was recommended that une aiguille affolee
should be again more strongly magnetized. Halley was cer-
tainly the first who inferred that polar light was a magnetic
phenomenon — a statement* which he made on the occasion
of his being invited by the Royal Society of London to ex-
plain the great meteor of the 6th of March, 1716, which was
seen in every part of England. He says " that the^ meteor
is analogous with the phenomenon which Gassendi first des-
ignated in 1621 by the name o^ Aurora Borealis'^ Although,
in his voyages for the determination of the line of variation,
he advanced as far south as 52°, yet we learn, from his own
confession, that he had never seen a northern or southern
polar light before the year 1716, although the latter, as I
can testify, is visible in the middle of the tropical zone of
Peru. Halley, therefore, does not appear, fromihis own ob-
servation, to have been aware of the restlessness of the nee-
dle, or of the extraordinary distiirbimces and fluctuations
which it exhibits at tiie periods of visible or invisible north-

* Halley, Accozmt of the late stirprising A'ppearance of-Lightsin the
Air, in the Phil. Transact., vol. xxix., 1714-1716, No. 347, p. 422-
428. Halley's explanation of the Aurora Borealis is unfortunately
connected with the fantastic hypothesis which had been enounced by
him twenty-five years earlier, in the Phil. Transact, for 1693, vol.
xvii., No. 195, p. 563, according to which there was a luminous fluid
in the hollow terrestrial sphere lying between the outer shell which we
inhabit and the inner denser nucleus, Avhich is also inhabited by hu-
man beings. These are his words: "In order fo make that inner
globe capable of being inhabited, there might not improbably be con-
tained some luminous medium between the balls, so as to make a per-
petual day below." Since the outer shell of the earth's crust is far
less thick in the region of the poles of rotation (owing to the compres-
sion produced at those parts) than at the equator, the inner luminous
fluid (that is, the magnetic fluid), seeks at certain periods, more espe-
cially at the times of the equinoxes, to find itself a passage in the less
thick polar regions through the fissures of rocks. The emanation of
this fluid is, according to Falley, the phenomenon of the northern
light. When iron filings are strewn over a spheroidal magnet (a te-
rella), they serve to show the direction of the luminous colored rays of
the Aurora. "As each one sees his own rainbow, so also the corona
appears to every observer to be at a different point" (p. 424). Regard-
ing the geognostic dreams of an intellectual investigator, who display-
ed |uch profound knowledge in all his magnetic and astronomical la-
bors, see Cosmos, vol. i., p. 171.

MAGNETIC DISTURBANCES. 129

em or southern polar lights. OlavHiorter and Celsius at Up-
sala were the first who, in the year 1741, and therefore be-
fore Halley's death, confirmed, by a long series of measure-
ments and determinations, the connection, which he had mere-
ly conjectured to exist between the appearance of the Aurora
Borealis and a disturbance in the normal course .of the nee-
dle. This meritorious investigation led them to enter into
an arrangement for carrying on systematic observations si-
multaneously with Graham in London, while the extraordi-
nary disturbances of variation, observed on the appearance
of the Aurora, were made subjects of special investigation
by Wargentin, Canton, and Wilke.

The observations which I had the opportunity of making,
conjointly with Gay-Lussac, in 1805, on the Monte Pincio
at Rome, and more especially the investigations suggested by
these observations, and which I prosecuted conjointly with
Oltmanns during the equinoctial and solstitial periods of the
years 1806 and 1807, in a large isolated garden at Berlin,
by means of one of Prony's magnetic telescopes, and of a
distant tablet-signal, which admitted, of being well illumina-
ted by lamp-light, showed me that this element of terrestrial
activity (which acts powerfully at certain epochs, and not
merely locally, and which has been comprehended under the
general name of extraordinary disturbances) is worthy, on
account of its complicated nature, of being made the subject
of continuous observation. The arrangement of the signal
and the cross-wires in the telescope, which was suspended in
one instance to a silken thread, and in another to a metallic
wire, and attached to a bar magnet inclosed in a large glass
case, enabled the observer to read off to 8^^ in the arc. As
this method of observation allowed of the room in which the
telescope and the attached bar magnet stood being left unil-
luminated by night, all suspicion of the action of currents of
air was removed, and those disturbances avoided which oth-
erwise are apt to arise from the illumination of the scale
in variation compasses, provided with microscopes, however
perfect they may otherwise be. In accordance with the opin-
ion then expressed by me, that *' a continuous, uninterrupt-
ed hourly and half-hourly observation (Observatio Perpetua)
of several days and nights was greatly to be preferred to
isolated observations extending over many months," we con-
tinued our investigations for five, seven, and even eleven
days and nights consecutively,* during the equinoctial and

* When greatly fatigued by observing for many consecutive niglits,

130

COSMOS.

solstitial periods — the importance of such observations at
these times being admitted by all recent observers. We soon
perceived that, in order to study the peculiar physical char-
acter of these anomalous disturbances, it was not sufficient
to determine the amount of the alteration of the variation,
but that the numerical degree of disturbance of the nee-
dle must be appended to each observation by obtaining the
measured elongation of the oscillations. In the ordinary
horary course of the needle, it was found to be so quiet that
in 1500 results, deduced from 6000 observations, made from
the middle of May, 1806, to the end of June, 1807, the oa-
cillation generally fluctuated only from one half of a gradua-
ted interval to the other half, amounting therefore only to V
12^^; in individual cases, and often when the weather was
very stormy and much rain was falling, the needle appeared
to be either perfectly stationary, or to vary only 0*2 or 0*3
of a graduated interval, that is to say, about 2V^ or 28^^.
But on the occurrence of a magnetic storm, whose final and
strongest manifestation is the Aurora Borealis, the oscilla-
tions were either in some cases only IV and in others 38^
in the arc, each one being completed in from 1^ to 3 seconds
of time. Frequently, on account of the magnitude and in-
equality of the oscillations, which far exceeded the scale
parts of the tablet in the direction of one or both of its sides,
it was not possible to make any observation.* This, for in-

Professor Oltmanns and myself were occasionally relieved by very
trustworthy obsen^ers; as, for instance, by Mampel, the geographer
Friesen, the skillful mechanician Nathan Mendelssohn, and our great
geognosist, Leopold von Buch. It has always afforded me pleasure
to record the names of those who have kindly assisted me in my
labors.

* The month of September, 1806, was singularly rich in great mag-
netic disturbances. By way of illustration, I will give the following
extracts from my journal :

1806, from 4h. 36m. A.M. till 5h. 43m. A.M.

fi

Sept.

m

M

5_4.

2 5

ZA

an

If

27

28

4h. 40m.

' '' 7h. 2m. "

3h. 33m.

' " 6h. 27m. *'

3h. 4ra.

' " 6h. 2m. "

2h. 22m.

' " 4h. 30m. "

2h. 12m.

' " 4h. 3m. "

Ih. 5om.

' " 5h. 27m. *'

Oh. 3m.

' " Ih. 22m. "

The disturbance last referred to was very small, and was succeeded
by the greatest quiet, which continued throughout the whole night
and until the following noon.

f§ Sept., 1806, from lOh. 20m. P.M. till llh. 32m. P.M.

MAGNETIC DISTURBANCES.

131

stance, was the case for long and uninterrupted periods dur-
ing the night of the 24th September, 1806, lasting on the
first occasion from 2h. Om. to oh, o2m., and next from 3h.
57m. to 5h. 4m. A.M.

In general, during unusual or larger magnetic disturb-
ances (magnetic storms), the mean of the arc of the oscilla-
tions exhibited an increase either westward or eastward, al-
though with irregular rapiditj, but in a few cases extraor-
dinary fluctuations were also observed, even when the vari-
ation was not irregularly increased or decreased, and when
the mean of the oscillations did not exceed the limits apper-
taining to the normal position of the needle at the given
time. We saw, after a relatively long rest, sudden motions
of very unequal intensity, describing arcs of from 6^ to 15^,
either alternating with one another or abnormally inter-
mixed, after which the needle would become suddenly sta-
tionary. At night this mixture of total quiescence and vio-
lent perturbation, without any progression to either side, was
very striking.* One special modification of the motion, which

This was a small disturbance, which was succeeded by great calm
until 5h. 6m. A.M. ^ ^^p*-, 1806, about 2Ii. 46ra. A.M. a great but
short magnetic storm, followed by perfect calm. Another equally
great magnetic disturbance about 4h. 30m. A.M.

The great storm of |^| September had been preceded by a still
greater disturbance from 7h. 8m. till 9h. 11m. P.M. In the following
winter months there was only a very small number of storms, and
these could not be compared with the disturbances during the au-
tumnal equinox. I apply the term great storm to a condition in which
the needle makes oscillations of from 20 to 38 minutes, or passes be-
yond all the scale parts of the segment, or when it is impossible to
make any observation. In small storms the needle makes irregular
oscillations of from five to eight minutes.

* Arago, during the ten years in which he continued to make care-
ful observations at Paris (till 1829), never noticed any oscillations
without a change in the variation. He wrote to me as follows, in the
course of that year : " I have communicated to the Academy the re-
sults of our simultaneous observations. I am surprised to notice the
oscillations which the dipping-needle occasionally exhibited at Berlin
during the observations of 1806, 1807, and of 1828-1829, even when
the mean declination was not changed. Here (at Paris) we never ex-
perience any thing of the kind. The only time at which the needle
exhibits violent oscillations is on the occurrence of an Aurora Borealis,
and when its absolute direction has been considerably disturbed, and
even then the disturbances of direction are most frequmihj unaccom-
panied by any oscillatory movement." The condition here described
is, however, entirely opposite to the phenomena which were observed
a,t Toronto (43° 91' N. lat.) during the years 1840 and 1841, and which
nrrespond accurately with those manifested at Berlin. The observ-
ftrs at Toronto have paid so much attention to the nature of the mo-

132 COSMOS.

I must not pass without notice, consisted in the very rare
occurrence of a vertical motion, a kind of tilting motion, an
alteration of the inclination of the northern point of the
needle, which was continued for a period of from fifteen to
twenty minutes, accompanied by either a very moderate de-
gree of horizontal vibration or by the entire absence of this
movement. In the careful enumeration of all the secondary
conditions which are recorded in the registers of the English
observatories, I have only met with three references to " con-
stant vertical motion, the needle oscillating vertically,"* and
these three instances occurred in Van Diemen's Land.

The periods of the occurrence of the greater magnetic
storms fell, according to the mean of my observations in
Berlin, about three hours after midnight, and generally ceased
about 5 A.M. We observed lesser disturbances during the
daytime, as, for instance, between 5 and 7 P.M., and fre-
quently on the same days of September, during which vio-
lent storms occurred after midnight, when, owing to the
magnitude and rapidity of the oscillations, it was impossible
to read them off or to estimate the means of their elonga-
tion. I soon became so convinced of the occurrence of mag-
netic storms in groups during several nights consecutively,

tion that they indicate whether the vibrations and shocks are "strong"
or "slight," and characterize the disturbances in accordance with defin-
ite and uniform subdivisions of the scale, following a fixed and uni-
form nomenclature. Sabine, Days of Umisual Magn. Disturbances,
vol. i., pt. i., p. 4G. Six groups of successive days (146 in all) are
given from the two above-named years in Canada, which were marked
by very strong shocks, without any perceptible change in the horary
declination. Such groups (see Op. cit., p. 47, 54, 74, 88, 95, 101) are
designated as " Times of Observations at Toronto, at which the magnetom-
eters loere disturbed, but the mean readings were not materially changed^''
The changes of variation were also nearly ahvays accompanied by
strong vibrations at Toronto during the frequent Auroras Boreales ; in
some cases these vibrations were so strong as entirely to prevent the
observations from being read off. We learn, therefore, from these
phenomena, whose further investigation we can not too strongly rec-
ommend, that although momentary changes of decHnation which dis-
turb the needle may often be followed by great and definite changes
of variation (Younghusband, Unusual Distwbances, pt. ii., p. x.), the
size of the arc of vibration in no respect agrees with the amount of
the alteration in the declination ; that in very inconsiderable changes
of variation the vibrations may be very strong, while the progressive
motion of the needle toward a western or eastern declination may be
rapid and considerable, independently of any vibration ; and further,
that these processes of magnetic activity assume a special and different
character at different places.

* Unusual Disturb. ., vol. i., pt. i., p. 69, 101.

MAGNETIC DISTURBANCES, 133

that I acquainted the Academy at Berlin with the peculiar
nature of these extraordinary disturbances, and even invited
my friends to visit me at predetermined hours, at which I
hoped they might have an opportunity of witnessing this
phenomenon ; and, in general, I was not deceived in my an-
ticipations.* KupfFer, during his travels in the Caucasus in
1829, and at a later period, Kreil, in the course of the valu-
able observations which he made at Prague, were both en-
abled to confirm the recurrence of magnetic storms at the
same hours, f

The observations which I was enabled to make during the
year 1806 at the equinoctial and solstitial periods, in refer-
ence to the extraordinary disturbances in the variation, have
become one of the most important acquisitions to the theory
of terrestrial magnetism, since the erection of magnetic sta-
tions in the different British colonies (from 1838 to 1840),
through the accumulation of a rich harvest of materials,
which have been most skillfully elaborated by General Sa-
bine. In the results of both hemispheres this talented observ-
er has separated magnetic disturbances, according to diurnal
and nocturnal hours, according to different seasons of the year,
and a^ording to their deviations eastward or westward. At
Toronto and Hobarton the disturbances were twice as fre-
quent and strong by night as by day,{ and the same was the
case in the oldest observations at Berlin ; exactly the re-
verse of what was found in from 2600 to 3000 disturbances

* This was at the end of September, 3806. This fact, which was
published in Poggendorflfs Annalen der Physik, bd. xv. (April, 1829),
s. 330, was noticed in the following terms : " The older horary observ-
ations, which I made conjointly with Oltmanns, had the advantage
that at that period (1806 and 1807) none of a similar kind had been
prosecuted either in France or in England. They gave the nocturnal
maxima and minima; they also showed how remarkable magnetic
storms could be recognized, which it is often impossible to record,
owing to the intensity of the vibrations, and which occur for many
nights consecutively at the same time, although no influence of mete-
orological relations has hitherto been recognized as the inducing cause
of the phenomena." The earliest record of a certain periodicity of
extraordinary disturbances was not, therefore, noticed for the first
time in the year 1839. Report of the Fifteenth Meeting of the British
Association at Cambridge, 1845, pt. ii., p. 12.

t Knpffer, Voyage au Mont Elbruz dans le Caucase, 1829, p. 108.
"Irregular deviations often recur at the same hour and for several
days consecutively."

X Sabine, Unusual Disturb., vol. i., pt. i., p. xxi. ; and Younghus-
band, On Periodical Laws in the larger Magnetic Disturbances, in the
Phil. Transact, for 1853, pt. i., p. 173.

134 COSMOS.

at the Cape of Good Hope, and more especially at the island
of St. Helena, according to the elaborate investigation of Cap-
tain Younghusband. At Toronto the principal disturbances
generally occurred in the period from midnight to 5 A.M.;
it was only occasionally that they were observed as early as
from 10 P.M. to midnight, and consequently they predomin-
ated by night at Toronto, as well as at Hobarton. After
having made a very careful and ingenious investigation of
the 3940 disturbances at Toronto, and the 3470 disturbances
at Hobarton, which were included in the cycle of six years
(from 1843 to 1848), of which the disturbed variations con-
stituted the ninth and tenth parts, Sabine was enabled to
draw .the conclusion* that " the disturbances belong to a
special kind of periodically recurring variations, which fol-
low recognizable laws, depend upon the position of the sun
in the ecliptic and upon the daily rotation of the earth round
its axis, and, further, ought no longer to be designated as irreg-
ular motions, since we may distinguish in them, in addition to
a special local type, processes which affect the whole earth."
In those years in which the disturbances were more frequent
at Toronto, they occurred in almost equal numbers in the
southern hemisphere at Hobarton. At the first-naAed of
these places these disturbances were, on the whole, doubly as
frequent in the summer — namely, from April to September
— as in the winter months, from October to March. The
greatest number fell in the month of September, in the same
manner as at the autumn equinox in my Berlin observations
of 1806.t They are more rare in the winter months in all
places ; at Toronto they occur less frequently from Novem-

* Sabine, in the Phil Transact, for 1851, pt. i., p. 125-127. *' The
diurnal variation observed is, in fact, constituted by two variations
superposed upon each other, having different laws, and bearing differ-
ent proportions to each other in different parts of the globe. At trop-
ical stations the influence of what have been hitherto called the irreg-
ular disturbances (^magnetic storms) is comparatively feeble ; but it is
otherwise at stations situated as are Toronto (Canada) and Hobarton
(Van Diemen's Island), where their influence is both really and pro-
portionally ^greater, and amounts to a clearly recognizable part of the
whole diurnal variation." We find here, in the complicated efl'ect of
simultaneous but different causes of motion, the same condition which
has been so admirably demonstrated by Poisson in his theory of waves
(Annales de Chimie et de Physique, t. vii., 1817, p. 293). "Waves of
different kinds rrfay cross each other in the water as in the air, where
the smaller movements are superposed upon each other." See La-
mont's conjectures regai-ding the compound etfect of a polar and an
equatorial loave, in Poggeud., Annalen, bd. Ixxxiv., s. 583.

t See p. 130.

^MAGNETIC DISTURBANCES. 135

ber till February, and at Hobarton from May till August.
At St. Helena and at the Cape of Good Hope the periods at
which the sun crosses the equator are characterized, accord-
ing to Younghusband, by a very decided frequency in the
disturbances.

The most important point, and one which was also first
noticed by Sabine in reference to this phenomenon, is the
regularity with which, in both hemispheres, the disturbances
occasion an augmentation in the eastern or western varia-
tion. At Toronto, where the declination is slightly west-
ward (1° 33^), the progression eastward in the summer, that
is, from June till September, preponderated over the pro-
gression westward during the winter (from December till
April), the ratio being 411:290. In like manner, in Van
Diemen's Land, taking into account the local seasons of the
year, the winter months (from May till August) are charac-
terized by a strikingly diminished frequency of magnetic
storms.* The co-ordination of the observations obtained in
the course of six years at the two opposite stations, Toronto
and Hobarton, led Sabine to the remarkable result that,
from 1843 to 1848, there was in both hemispheres not only
an increase in the number of the disturbances, but also (even
when, in order to determine the normal annual mean of the
daily variation, 3469 storms were excluded from the calcu-
lation) that the amount of total variation from this mean
gradually progressed during the above-named five years from
7^*65 to 10^*o8. This increase was simultaneously percepti-
ble, not only in the amplitude of the declination, but also in
the inclination and in the total terrestrial force. This result
acquired additional importance from the confirmation and
generalization afforded to it by Lamont's complete treatise
(September, 1851) "regarding a decennial period, which is
perceptible in the daily motion of the magnetic needle."
According to the observations made at Gottingen, Munich,
and Kremsmunster,t the mean amplitude of the daily dec-

* Sabine, in the Phil. Transact, for 1852, pt. ii., p. 110 (Younghus-
band, Op. cit., p. 169).

t According to Lament and Relshuber, the magnetic period is ten
years four months, so that the amount of the mean of the diurnal mo-
tion of the needle increases regularly for five years, and decreases for
the same length of time; on which account the winter motion (the
amplitude of decHnation) is always twice as small as the summer mo-
tion (see Lament, Jahresbericht der Sternwarte zu Miinchen fiir 1852,
s. 54-60). The director of the Observatory at Berne, Rudolph Wolf,
finds, by a much more comprehensive series of operations, that the

136 COSMOS.

lination attained its minimum between 1843 and 1844, and
its maximum from 1848 to 1849. After the declination has
thus increased for five years, it again diminishes for a period
of equal length, as is proved by a series of exact horary ob-
servations, which go back as far as to a maximum in 1786^.*
In order to discover a general cause for such a periodicity in
all three elements of telluric magnetism, we are disposed to
refer it to cosmical influences. Such a connection is indeed
appreciable, according to Sabine's conjecture, in the altera-
tions which take place in the photosphere, that is to say, in
the luminous gaseous envelopes of the dark body of the sun-t
According to the investigations which were made throughout
a long series of years by Schwabe, the period of the greatest
and smallest frequency of the solar spots entirely coincides
with that which has been discovered in magnetic variations.
Sabine first drew attention to this coincidence in a memoir
which he laid before the Royal Society of London^ in March,
1852. "There can be no doubt," says Schwabe, in the re-
marks with which he has enriched the astronomical portion
of the present work, "that, at least from the year 1826 to
1850, there has been a recurring period of about ten years in
the appearance of the sun's spots, whose maxima fell in the
years 1828, 1837, and 1848, and the minima in the years
1833 and 1843."t The important influence exerted by the
sun's body, as a mass, upon terrestrial magnetism is con-
firmed by Sabine in the ingenious observation that the period
at which the intensity of the magnetic force is greatest, and
the direction of the needle most near to the vertical line,
falls, in both hemispheres, between the months of October
and February ; that is to say, precisely at the time when the
earth is nearest to the sun, and moves in its orbit with the?
greatest velocity.§

I have already treated, in the Picture of Nature, || of the

period of magnetic declination which coincides with the frequency of
the solar spots must be estimated at H*l years.

* See page 74.

t Sabine, in the Phil. Transact, for 1852, pt. i., p. 103, 121. See
the observations made in July, 1852, by Rudolph Wolf, above referred
to in page 76 of the present volume ; also the very similar conjectures
of Gautier, which were published very nearly at the same time in the
Bibliotheque Universelle de Geneve, t. xx., p. 189.

X Cosmos, vol. iv., p. 85-87.

§ Sabine, in the Phil. Transact, for 1850, pt. i., p. 216. Faraday,
Exper. Researches on Electricity, 1851, p. 56, 73, 76, § 2891, 2949,
2958.

II Cosmos, vol. i., p. 191 ; Poggend., Annalen, bd. xv., s. 334, 335;

MAGNETIC DISTURBANCES. 137

simultaneity of many magnetic storms, which are transmit-
ted for thousands of miles, and indeed almost round the en-
tire circumference of the earth, as on the 25th of September,
1841, whfen they were simultaneously manifested in Canada,
Bohemia, the Cape of Good Hope, Van Diemen's Land, and
Macao ; and I have also given examples of those cases in
which the perturbations were of a more local kind, passing
from Sicily to Upsala, but not from Upsala farther north in
the direction of Alten and Lapland. Li the simultaneous
observations of declination which were instituted by Arago
and myself in 1829 at Berlin, Paris, Freiberg, St. Petersburg,
Casan, and Nikolajew, with the same Gambey's instruments,
individual perturbations of a marked character were not
transmitted from Berlin as far as Paris, and not on any one
occasion to the mine at Freiberg, where Reich was making
a series of subterranean observations on the magnet. Great
variations and disturbances of the needle simultaneously with
the occurrence of the Aurora Borealis at Toronto certainly
occasioned magnetic storms in Kerguelen's Land, but not at
Hobarton. * When we consider the capacity for penetrating
through all intervening bodies, which distinguishes the mag-
netic force, as well as the force of gravity inherent in all
matter, it is certainly very difficult to form a clear concep-
tion of the obstacles which may prevent its transmission
through the interior of the earth. These obstacles are anal-
ogous to those which we observe in sound-waves, or in the
waves of commotion in earthquakes, in which certain spots
which are situated near one another never experience 'the
shocks simultaneously.* Is it possible that certain magnet-
ic intersecting lines may by their intervention oppose all fur-
ther transmission *?

We have here described the regular and the apparently ir-
regular motions presented by horizontally-suspended needles.
If by an examination of the normal-recurring motion of the
needle we have been enabled, from the mean numbers of the
e^j^remes of the horary variations, to ascertain the direction

Sabine, Unusual Disturb., vol. i., pt. i., p. xiv.-xviii. ; where tables are
given of the simultaneous storms at Toronto, Prap;ue, and Van Die-
men's Land. On those days in which the magnetic storms were the
most marked in Canada (as, for instance, on the 22d of March, the
10th of May, the Gth of August, and the 25th of September, 1841),
the same phenomena were observed in the southern hemisphere in
Australia. See also Edward Belcher, in the Phil, Transact, for 1843,
p. 133.

* Cosmo Sy vol. i., p. 212.

138

COSMOS.

of the magnetic meridian, in which the needle has vibrated
equally to either side, from one solstice to another, the com-
parison of the angles which the magnetic meridian describes
at different parallels with the geographical meridian has led,
in the first place, to the knowledge of lines of variation of
strikingly heterogeneous value (Andrea Bianco in 1436, and
Alonzo de Santa Cruz, cosmographer to the Emperor Charles
v., even attempted to lay down these lines upon charts);
and, more recently, to the successful generalization of isogonic
curves, lines of equal variation, which British seamen have
long been in the habit of gratefully designating by the his-
torical name of llalley's lines. Among the variously curved
and differently arranged closed systems of isogonic lines,
which are sometimes almost parallel, and more rarely re-en-
ter themselves so as to form oval closed systems, the great-
est attention, in a physical point of view, is due to those
lines on which the variation is null, and on both sides of
which variations of opposite denominations prevail, which in-
crease unequally with the distance.* I have already else-
where shown how the first discovery made by Columbus, on
the 13th of September, 1492, of a line of no variation in the
Atlantic Ocean, gave an impetus to the study of terrestrial
magnetism, which, however, continued for two centuries and
a half to be directed solely to the discovery of better meth-
ods for obtaining the ship's reckoning.

However much the higher scientific education of mariners
in recent times, and the improvement of instruments and
methods of observation, have extended our knowledge of in-
dividual portions of lines of no variation in Northern Asia,
in the Indian Archipelago, and the Atlantic Ocean, we have
still to regret that in this department of our knowledge,
where the necessity of cosmical elucidation is strongly felt,
the progress has been tardy and the results deficient in gen-
eralization. I am not ignorant that a large number of ob
servations of accidental crossings of lines of no variation have
been noted down in the logs of various ships, but we are ^
ficient in a comparison and co-ordination of the materials,
which can not acquire any importance in reference to this
object or in respect to the position of the magnetic equator,
until individual ships shall be dispatched to different seas
for the sole purpose of uninterruptedly following these lines
throughout their course. Without a simultaneity in the ob-

- * Op. cit., vol. i., p. 187-189; vol. ii., p. 657-659, and p. 64-60 of
the present volume.

LINES OF NO VARIATION. 139

servations, we can have no history of terrestrial magnetism.
I here merely reiterate a regret which I have often previous-
ly expressed.*

* At very different periods, once in 1809, in my Recucil d'Observ.
Astron., vol. i., p. 368, and again in 1839, when, in a letter addressed
to the Earl of Minto, then First Lord of the Admiralty, a few days
before the departure of Sir James Ross on his Antarctic expedition, I
endeavored more fully to develop the importance of the proposition
advanced in the text (see Report of the Committee of Physics and Me-
teor, of the Royal Soc. relative to the Antarctic Exped., 1840, p. 88-91).
" In order to follow the indications of the magnetic equator or those
of the lines of no variation, the ship's course must be made to cross the
lines 0 at very small distances, the bearings being changed each time
that observations of inclination or of declination show that the ship
has deviated from these points. I am well aware that, in accordance
with the comprehensive views of the true basis for a general theory of
terrestiial magnetism, which we owe to Gauss, a thorowgh knowledge
of the horizontal intensity, and the choice of the points at which the
three elements of declination, inclination, and total intensity have all
been simultaneously measured, suffice for finding the value of - (Gauss,
§ 4 and 27), and that these are the essential points for future investi-
gations ; but the sum total of the small local attractions, the requii'e-
ments of steering ships, the ordinary corrections of the compass, and
the safety of navigation, continue to impart special importance to the
knowledge of the position, and to the movements of the periodic trans-
lation oi lines of no variation. I here plead the cause of these various
requirements, which are intimately connected with the interests of
physical geography." Many years must still pass before seamen can
be enabled to guide the ship's course by charts of variation, construct-
ed in accordance with the theory of terrestrial magnetism (Sabine, in
the Phil. Transact, for 1849, pt. ii., p. 204), and the wholly objective
view directed to actual observation, which I would here advocate,
would, if it led to periodically-repeated determinations, and conse-
quently to expeditions prosecuted simultaneously by land and sea, in
accordance with some preconcerted plan, give the double advantage
of, in the first place, yielding a direct practical application, and afford-
ing us a correct knowledge of the annual progressive movement of
these lines : and, secondly, of supplying many new data for the fur-
ther development of the theory enounced' by Gauss (Gauss, § 25;.
It would, moreover, greatly facilitate the accurate determination of
the pj'ogression of the two lines of no inclination and no variation, if
landmarks could be established at those points where the lines enter
or leave continents at stated intervals ; as, for instance, in the years

1850, 1875, 1900 In expeditions of this kind, which would be

similar to those undertaken by Halley, many isoclinal and isogonic
systems would necessarily be intersected before the lines of no decli-
nation and no inclination could be reached, and by this means the hor-
izontal and total intensities might be measured along the coasts, so
that several objects would thus be simultaneously attained. The views
which I have here expressed are, I am happy to find, supported by a
very great authority in nautical questions, viz.. Sir James Ross. (See
his Voyage in the Southern and Antarctic Regions, vol. i., p. 105.)

140

COSMOS.

According to the facts which we already generally know
concerning the position of lines of no variation, it would ap-
pear that, instead of the four meridian systems which were
believed at the end of the 16th century to extend from pole
to pole,* there are probably three very differently formed
systems of this kind, if by this name we designate those
groups in which the line of variation does not stand in any
direct connection with any other line of the same kind, or
can not, in accordance with the present state of our knowl-
edge, be regarded as the continuation of any other line. Of
these three systems, which we will separately describe, the
middle, or Atlantic, is limited to a single line of no varia-
tion, inclining from SS.E. to NN.W., between the parallels
of 65° south and 67° north latitude. The second system,
which lies fully 150° farther east, occupying the whole of
Asia and Australia, is the most extended and most compli-
cated of all, if we merely take into account the points at
which the line of no variation intersects the geographical
equator. This system rises and falls in a remarkable man-
ner, exhibiting one curvature directed southward and anoth-
er northward; indeed it is so strongly curved at its north-
eastern extremity that the line of no variation forms an el-
lipse, surrounding those lines which rapidly increase in vari-
ation from without inward. The most westerly and the
most easterly portions of this Asiatic curve of no variation
incline, like the Atlantic line, from south to north, and in
the space between the Caspian Sea and Lapland even from
SS.E. to NN.W. The third system, that of the Pacific,
which has been least investigated, is the smallest of all, and,
lying entirely to the south of the geographical equator, forms
almost a closed oval of concentric lines, whose variation is
opposite -to that whicfh we observe in the northeastern part
of the Asiatic system, and decreases from without inward.
If we base our opinion upon the magnetic declination ob-
served on the coast, we find that the African continent! only

* Acosta, Ilistoria de las Indias, 1590, lib. i., cap. 17. I have al-
ready considered the question whether the opinion of Dutch naviga-
toi's regarding the existence of four lines of no variation may not,
through the difFei-ences between Bond and Beckborrow, have had some
influence on Halley's theory of four magnetic poles (Cosmos, vol. ii.,
p. 280).

t In the interior of Africa, the isogonic line of 22° 15' W. is espe-
cially deserving of careful cosmical investigation, as being the inter-
mediate line between very different systems, and as proceeding (ac-
cording to the theoretical views of Gauss) from the Eastern Indian

LINES OF NO VARIATION. 141

presents lines which exhibit a western variation of from 6°
to 29° ; for, according to Purchas, the Atlantic line of no
variation left the southern point of Africa (the Cape of Good
Hope) in the year 1605, inclining farther from east to west.
The possibility that we may discover in some part of Cen-
tral Africa an oval group of concentric lines of variation de-
creasing to 0°, and which is similar to that of the Pacific,
can neither be asserted nor denied on any sure grounds.

The Atlantic portion of the American curve of no varia-
tion was accurately determined in both hemispheres for the
year 1840, by the admirable investigations of General Sa-
bine, who employed 1480 observations, and duly took into
account the secular changes. It passes in the meridian of
70° S. lat.,and about 19° W. long.,* in a NN.W. direction,
to about 3° east of Cook's Sandwich Land, and to about 9°
30^ east of South Georgia ; it then approaches the Brazilian
coast, which it enters at Cape Frio 2° east of Rio Janeiro,
and traverses the southern part of the New Continent no
farther than 0° 36^ S. lat., where it again leaves it some-
what to the east of Gran Para, near Cape Tigioca, on the
Rio do Para, one of the secondary outlets of the Amazon,
crossing the geographical equator in 47° 44^ W. long., then
skirting along the coast of Guiana at a distance of eighty-
eight geographical miles as far as 5° N. lat, and afterward
following the arc of the small Antilles as far as the parallel
of 18°, and, finally, touching the shore of North Carolina
near Cape Lookout, southeast of Cape Hatteras, in 34° 50^
N. lat., 74° 8^ W. long. In the interior of North America,
the curve follows a northwestern direction as far as 41° 30^
N. lat., 77° 38^ W. long., toward Pittsburgh, Meadville, and
Lake Erie. We may conjecture that it has advanced very
nearly half a degree farther west since 1840.

The Australo- Asiatic curve of no variation (if, according
to Erman, we consider the part which rises suddenly from
Kasan to Archangel and Russian Lapland as identical with

Ocean, straight across Africa, on to Newfoundland. The very com-
prehensive plan of the African expedition, conducted by Richardson,
Barth, and Overweg, under the orders of the British government, may
probably lead to the solution of such magnetic problems.

* Sir James Ross intersected the curve of no variation in 61° 30' S.
lat. and 27° 10' W. long. ( Voyage to the Southern Seas, vol. ii., p.
357). Captain Crozier found the variation in March, 1843, 1° 38' in
70° 43' S. lat. and 21° 28' W. long., and he was therefore very near
the line of no variation. See Sabine, On the Magn. Declination in the
Atlantic Ocean for 1840, in the Phil. Transact, for 1849, pt. ii., p. 233.

142 COSMOS.

the part in the sea of Molucca and Japan) can scarcely be
followed as far as 62° in the southern hemisphere. This
starting-point lies farther west from Van Diem en's Land
than had hitherto been conjectured, and the three points at
which Sir James Ross crossed the curve of no variation, on
his Antarctic voyage of discovery in 1840 and 1841,* are
all situated in the parallels of 62°, 54° -30, and 46°, be-
tween 133° and 135° 40^ E. long., and therefore mostly in
a meridian-like direction running from south to north. In
its further course, the curve crosses Western Australia from
the southern coast of Nuyts' Land, about 10° W. of Ade-
laide, to the northern coast, near Vansittart River and
Mount Cockburn, from whence it enters the sea of the In-
dian Archipelago in a region of the world in which the in-
clination, declination, total intensity, and the maximum and
minimum of the horizontal force were investigated by Cap-
tain Elliot, from 1846 to 1848, with, more care than has
been done in any other portion of the globe. Here the line
passes south of Flores and through the interior of the small
Sandal-wood Island,! i" a direct east and west direction,
from about 120° 30^ to 93° 30" E. long., as had been ac-
curately demonstrated sixteen years before l3y Barlow. From
the last-named meridian it ascends toward the northwest in
9° 30" S. lat., judging by the position in which Elliot fol-
lowed the curve of 1° east variation to Madras. We are
not able here to decide definitely whether, crossing the
equator in about the meridian of Ceylon, it enters the con-
tinent of Asia between the Gulf of Cambay and Guzurat, or
farther west in the Bay of Muscat,J and whether, therefore,
it is identical^ with the <:urve of no variation, which appears

* Sir James Eoss, Op. ciL, vol. i., p. 104, 310, 317.

t Elliot, in the Phil. Transact, for 1851, pt. i., p. 331, pi. xiii. The
long and narrow small island from which we obtain the sandal-wood
{tschendana, Malay and Java; tschandana, Sanscrit; ./sa?2c?e/, Arab),

X According to Barlow, and the chart of Lines of Magnetic Declina-
tions computed according to the theory of Mr. Gauss, in the Report of
the Committee for the Antarctic Expedition, 1840. According to Bar-
low, the line of no variation proceeding from Australia enters the
Asiatic Continent at the Bay of Cambay, but turns immediately to the
northeast, across Thibet and China, near Thaiwan (Formosa), from
whence it enters the Sea of Japan. According to Gauss, the Aus-
tralian line ascends merely through Persia, past Nishnei-Novgorod to
Lapland. This great geometrician regards the Japan and Philippine
line of no variation, as well as the closed oval group in Eastern Asia,
as entirely independent of the line belonging to Australia, the Indian
Ocean, Western Asia, and Lapland.

ij I have already elsewhere spoken of this identity, which is based

MAGNETIC VARIATION. 143

to advance southward from the basin of the Caspian Sea ;
or whether, as Erman maintains, it may not curve to the
eastward, and, rising between Borneo and Malacca, reach
the Sea of Japan,* and penetrate into Eastern Asia through
the Gulf of Ochotsk. It is much to be lamented that, not-
withstanding the frequent voyages made to and from India-,
Australia,\he Philippines, and the northeast coasts of Asia,
a vast accumulation of materials should remain buried and
unheeded in various ships' logs, which might otherwise lead
to general views, by which we might be enabled to connect
Southern Asia with the more thoroughly explored parts of
Northern Asia, and thus to solve questions which were start-
ed as early as 1840. In order, therefore, not to blend to-
gether known facts with uncertain hypotheses, I will limit
myself to the consideration of the Siberian portion of the
Asiatic continent, as far as it has been explored in a souther-
ly direction to the parallel of 45° by Erman, Hansteen, Due,
Kupffer, Fuss, and myself In no other part of the earth
has so extended a range of magnetic lines been accessible to
us in continental regions ; and the importance which Euro-
pean and Asiatic Russia presents in this respect was ingen-
iously conjectured even before the time of Leibnitz.f

upon my own declination observations in the Caspian Sea, at Uralsk
on the Jaik, and in the Steppe of Elton Lake (Asie Centrale, t. iii., p.
458-461).

* Adolf Erman's Map of the Magnetic Declination, 1827-1830.
Elliot's chart shows, however, most distinctly that the Australian curve
of no variation does not intersect Java, but runs parallel with, and at
a distance of 1° 30' latitude from the southern coast. Since, accord-
ing to Erman, although not according to Gauss, the Australian line
of no variation between Malacca and Borneo enters the Continent
through the Japanese Sea, proceeding to the closed oval group of
Eastern Asia, on the northern coast of the Sea of Ochotsk (59° 30'
N. lat.), and again descends through Malacca, the ascending line can
only be 11° distant from the descending curve; and accoi-ding to this
graphical representation, the Western Asiatic line of no variation
(from the Caspian Sea to Russian Lapland) would be the shortest and
most direct prolongation of the part descending froni north to south.

t I drew attention as early as 1843 to the fact, which I had ascer-
tained from documents preserved in the Archives of Moscow and
Hanover {Asie Centrale, t. iii., p. 469-476), that Leibnitz, who con-
structed the first plan of a French expedition to Egypt, was also the
first who endeavored to profit by the relations which the czar, Peter
the Great, had established with Germany in 1712, by using his influ-
ence to secure the prosecution of observations for "determining the
position of the lines of variation and inclination, and for insuring that
these observations should be repeated at certain definite epochs" in
different parts of the Russian empire, whose superficies exceed those

144 COSMOS.

In order to follow the usual direction of Siberian expedi-,
tions from west to east, and starting from Europe, we will
begin with the northern part of the Caspian Sea. Here, in
the small island of Birutschikassa, in Astracan, on Lake El-
ton, in the Kirghis steppe, and at Uralsk, on the Jaik, be-
tween 45° 43" and Sl'^ 12' N. lat., and 46° 37' and 51° 2V
E. long., the variation fluctuates from 0° 10' ea^t to 0° 37'
west.* Farther northward, this line of no variation inclines
somewhat more toward the northwest, passing near Nishnei-
Novgorod.f In the year 1828 it passed between Osablikowo
and Doskino in the parallel of 56° N. lat. and 43° east long.
It becomes elongated in the direction of Russian Lapland be-
tween Archangel and Kola, or more accurately, according to
Hansteen (1830), between Umba and Ponoi.f It is not un-
til we have passed over nearly two thirds of the greatest
breadth of Northern Asia, advancing eastward to the lati-
tudes of from 50° and 60° (a district in which at present the
variation is entirely easterly), that we reach the line of no
variation, which in the northeastern part of the Lake of Bai-
kal rises to a point west of Wiluisk, which reaches the lati-

of the portions of the moon visible to us. In a letter addressed to the
czar, discovered by Pertz, Leibnitz describes a small hand-globe, or
ierrella, which is still preserved at Hanover, and on which he had rep-
resented the curve at which the variation is null (his linea magnet'ica
primaria). Leibnitz maintains that there is only one line of vo varia-
tion, which divides the terrestrial sphere into two almost equal parts,
and has four puncta Jlexits contrarii, or sinuosities, where the curves
are changed from convex to concave. From the Cape de Verd it
passes in lat. 36° toward the eastern shores of North America, after
which it directs its course through the South Pacific to Eastern Asia
and New Holland. This line is a closed one, and, passing near both
poles, it approaches closer to the southern than the northern pole; at
the latter the declination must be 25° west, and at the former only
5°. The motion of this important curve must have been directed to-
ward the north pole at the beginning of the 18th century. The varia-
tion must have ranged between 0° and 15° east over a great portion
of the Atlantic Ocean, the whole of the Pacific, Japan, a part of
China, and New Holland. "As the czar's private physician, Donelli,
is dead, it would be advisable to supply his place by some one else,
who will be disposed to administer very little medicine, but who may
be able to give sound scientific advice regarding determinations of
magnetic declination and inclination." These hitherto un-
noticed letters of Leibnitz certainly do not express any special theo-
retical views.

* See my Magnetic Observations, in Asie Centrale, t. iii., p. 460,
t Erman, Astron. und Magnet. Beohachtungen {Reise um die Erde<,
abth. ii., bd. 2, s. 532.

X Hansteen, in Poggend., Ann., bd. xxi., s. 371.

MAGNETIC VARIATION. 145

tude of 68°, iiuthe meridian of Jakutsk 129° 50' E. long.,
forming at this point the outer shell of the eastern group of
oval concentric lines of variation, to which we have frequent-
ly referred, again sinking in the direction of Ochotsk in 143°
10' E. long., intersecting the arc of the Kurile Islands, and
penetrating into the southern part of the Japanese Sea. All
the curves of from 5° to 15° eastern variation which occupy
the space between the lines of no variation in Western and
Eastern Asia have their concavities turned northward. The
maximum of their curvature falls, according to Erman, in
80° E. long., and almost in one meridian between Omsk and
Tomsk, and are therefore not very different from the merid-
ian of the southern extremity of the peninsula of Hindos-
tan. The axis major of the closed oval group extends 28°
of latitude as far as Corea.

A similar configuration, although on a still larger scale,
is exhibited in the Pacific. The closed curves here form
an oval between 20° N. lat. and 42° S. lat. The axis ma-
jor lies in 130° W. long. That which most especially dis-
tinguishes this singular group (the greater portion of which
belongs to the southern hemisphere, and exclusively to the
sea) from the continent of Eastern Asia is, as has been al-
ready observed, the relative succession in the value of the
curves of variation. In the former the eastern variation di-
minishes, while in the latter the western variation increases
the farther we penetrate into the interior of the oval. The
variation in the interior of this closed group in the southern
hemisphere amounts, however, as far as we know, only to
from 8° to 5°. Is it likely that there is a ring of southern
variation within the oval, or that we should again meet with
western variation farther to the interior of this closed line of
no variation ?

Curves of no variation, like all magnetic lines, have their
own history, which, however, does not as yet, unfortunately,
date further back than two centuries. Scattered notices may
indeed be met with as early even as in the 14th and 15th
centuries ; and here, again, Hansteen has the great merit of
having collected and carefully compared together all the va-
rious data. It would appear that the northern magnetic pole
is moving from west to east, and the southern magnetic pole
from east to west ; accurate observations show us, however,
that the different parts of the isogonic curves are progressing
very irregularly, and that where they were pai-allel they are
losing their parallelism ; and, lastly, that the domain of the
Vol. v.— G

146

COSMOS.

declination of one denomination, that is to s%y, east or west
declination, is enlarging and contracting in very diiFerent di-
rections in contiguous parts of the earth. Tlie lines of no
variation in Western Asia and in the Atlantic are advancing
from east to west, the former line having crossed Tobolsk in
1716 ; while in 1761, in Chappe's time, it crossed Jekather-
inenburg and subsequently Kasan ; and in 1829 it was found
to have passed between Osablikowo and Doskino, not far
from Nishnei-Novgorod, and consequently had advanced 24Q
45'' westward in the course of 113 years. Is the line of the
Azores, which Christopher Columbus determined on the 13th
of September, 1492, the same which, according to the ob-
servations of Davis and Keeling, in 1607, passed through the
Cape of Good Hope 1* and is it identical with the one which
we designate as the Western Atlantic, and which passes from
the mouth of the River Amazon to the sea-coast of North
Carolina ? If it be, we are led to ask, What has become of
the line of no variation which passed in 1600 through Kon-
igsberg, in 1620 (?) through Copenhagen, from 1657 to 1662
through London, and which did not, according to Picard,.
reach Paris, notwithstanding its more eastern longitude, un-
til 1666, passing through Lisbon somewhat before 1668 !t
Those points of the earth at which no secular progres.sion
has been observed for long periods of time are especially
worthy of our notice. Sir John Herschel has already drawn
attention to a corresponding long period of cessation in Ja-
maica,$ while Euler§ and BarlowJ refer to a similar condi-
tion in Southern Australia.

Polar Light.
We have now treated fully of the three elements of terres-
trial magnetism in the three principal types of its manifesta-
tion— namely. Intensity, Inclination, and Declination — in ref-

* Sabine, Magn. and Meteor. Observ. at the Cape of Good Hope, vol.
i., p. Ix.

f In judging of the approximate epochs of the crossing of the line of
no variation, and in endeavoring to decide upon the claim of no prior-
ity in this respect, we must bear in mind Jiow readily an error of 1°
may have been made with the instruments and methods then in use.

J Cosmos, vol. i., p. 181.

§ Euler, in the M^m. de VAcad. de Berlin, 1757, p.- 176.

II Barlow, in the Phil. Transact, for 1833, pt. ii., p. 671. Great un-
certainty prevails regarding the older magnetic observations of St. Pe-
tersburg during the first half of the 18th century. The variation seems
to have been always 3° 15' or 3° 30' from 1726 to 1772 ! Ilajisteen,
Magnetismus der Erde, s. 7, p. 143.

POLAR LIGHT. 147

erence to the movements which depend upon geographical
relations of place, and diurnal and annual periods. The ex-
traordinary disturbances which were first observed in the dip
are, as Halley conjectured, and as Dufay and Hiorter recog-
nized, in part forerunners, and in part accompaniments of
the magnetic- polar light. I have already fully treated, in
the Picture of Nature, of the peculiarities of this luminous
process, which is often so remarkable for the brilliant dis-
play of colors with which it is accompanied ; and more re-
cent observations have, in general, accorded with the views
which I formerly expressed. ''The Aurora Borealis has
not been described merely as an external cause of a disturb-
ance in the equilibrium of the distribution of terrestrial mag-
netism, but rather as an increased manifestation of telluric
activity, amounting even to a luminous phenomenon, exhib-
ited on the one hand by the restless oscillation of the needle,
and on the other by the polar luminosity of the heavens."
The polar light appears, in accordance with this view, to be
a kind of silent discharge or shock as the termination o*f a
magnetic storm, very much in the same manner as in the
electric shock the disturbed equilibrium of the electricity is
renewed by a development of light by lightning, accompa-
nied by pealing thunder. The reiteration of a definite hy-
pothesis in the case of a complicated and mysterious phenom-
enon has, at all events, the advantage of giving rise with a
view to its refutation to more persisteht and careful observa-
tions of the individual processes.*

Dwelling only on the purely objective description of these
processes, which are mainly based upon the materials yielded
by the beautiful and unique series of observations, which
were continued without intermission for eight months (1838,
1839) — during the sojourn of the distinguished physicists,
Lottin, Bravais, and Siljestrom — in the most northern parts
of Scandinavia,! we will first direct our attention to the so-

* Cosmos, vol. i., p. 193-203; and Dove, in Poggend., Annakn, bd.
xix., s. 388.

t The able narrative of Lottin, Bravais, Lilliehook, and Siljestrom,
who observed the phenomena of the northern light from the 19th of
September, 1838, till the 8th of April, 1839, at Bossekop (69° 58' N.
lat.), in Finmark, and at Jupvig (70° 6' N. lat.), was published in the
fourth section of Voyages en Scandinavie, en Laponie, au Sjntzberg et
aux Feroes, sur la Carvette, la Recherche {Aurores Boreales). To these
observations are appended important results obtained by the English
superintendent of the copper mines at Kalfiord (69° 56' N. lat.), p.
401-435.

148 COSMOS.

called black segment of the aurora, which rises gradually on
the horizon like a dark wall of clouds.* The blackness is
not, as Argelahder observes, a mere result of contrast, since
it is occasionally visible before it is bounded by the brightly-
illuminated arch. It must be a process effected within some
part of the atmosphere, for nothing has hitherto shown that
the obscuration is owing to any material blending. The
smallest stars are visible through the telescope in this black
segment, as well as in the colored illuminated portions of the
fully-developed aurora. In northern latitudes the black seg-
ment is seen far less frequently than in more southern re-
gions. It has even been found entirely absent in these last-
named latitudes in the months of February and March, when
the aurora was frequent in bright clear weather ; and Keil-
hau did not once observe it during the whole of a winter
which he spent at Tahvig, in Lapland. Argelander has
shown, by accurate determination of the altitudes of stars,
that no part of the polar light exerts any influence on these
altitudes. Beyond the segment there appear, although rare-
ly, black rays, which Hansteen and I have often watchedf
during their ascent ; blended with these appear round black
patches, or spots, inclosed by luminous spaces. The latter
phenomena have been made a special subject of investigation
by Siljestrom.J The central portion of the corona of the au-
rora (which, owing to the effect of linear perspective, corre-
sponds at its highest point with the magnetic inclination of
the place) is also usually of a very deep black color. Bra-
vais regards this blackness and the black rays as the effect
of optical illusions of contrast. Several luminous arches are

* See the work above referred to (p. 437-444) for a description of
the Segment obscure de VAurore Boreale.

t Schweigger's Jahrhuch der Chemie und Physik, 1826, bd. xvi., s,
198, and bd. xviii., s. 364. The dark segment and the incontestable
rising of black rays or bands, in which the luminous process is annihi-
lated (by interference?) reminds us of Quet's Recherches sur VElectro-
chimie dans le vide, and of RuhmkorfTs dehcate experiments, in which
in a vacuum the positive metallic balls glowed with red Hght, while
the negative balls showed a violet light, and the strongly luminous
parallel strata of rays were regularly separated from one another by
perfectly dark strata. "The light which is diffused between the
terminal knobs of the two electric conductors divides into numerous
parallel bands, which are separated by alternate obscure and perfectly
distinct strata." Comptes 7-endus de tAcad. des Sc, t. xxxv., 1852,
p. 949.

X Voyages en Scandinavie (Aurores Bor.), p. 558. On the corona
and bands of the northern light, see the admirable investigations of
Bravais, p. 602-514.

POLAR LIGHT. 149

frequently simultaneously present; in some rare cases a»
many as seven or nine are seen advancing toward the zenith
parallel to one another ; while in other cases they are alto-
gether absent. The bundles of rays and columns of light aS'
sume the most varied forms, appearing either in the shape of
curves, wreathed festoons and hooks, or resembling waving
pennants or sails.*

In the higher latitudes " the prevailing color of the polar
light is usually white, while it presents a milky hue when
the aurora is of faint intensity. When the colors brighten,
they assume a yellow tinge ; the middle of the broad ray be-
comes golden yellow, while both the edges are marked by
separate bands of red and green. When the radiation ex-
tends in narrow bands, the red is seen above the green.
When the aurora moves sideways from left to right, or from
right to left, the red appears invariably in the direction to-
ward which the ray is advancing, and the green remains be-
hind it." It is only in very rare cases that either one of the
complementary colors, green or red, has been seen alone.
Blue is never seen, while dark red, such as is presented by
the reflection of a great fire, is so rarely observed in the north
that Siljestrom noticed it only on one occasion.! The lu-
minous intensity of the aurora never even in Finmark quite
equals that of the full moon.

The probable connection which, according to my views,
exists between the polar light and the formation of very
srnall and delicate fleecy clouds (whose parallel and equiva-
lent rows follow the direction of the magnetic meridian), has
met with many advocates in recent times. It still remains
a doubtful question, however,| whether, as the northern trav-
elers, Thienemann and Admiral Wrangel believe, these par-
allel fleecy clouds are the substratum of the polar light, or
whether they are not rather, as has been conjectured by
Franklin, Richardson, and myself, the effect of a meteoro-

* Op. cit.j p. 35, 37, 45, 67, 481 {^^Draperie ondulante, flamme dhin
navire de guerre deployee horizontalement et ayitee par le vent, crochets,
fragments dares et de guiriandes).^' M. Bevalet, the distinguished
artist to the expedition, has given an interesting collection of the
many varied forms assumed by this phenomenon.

t See Voy, en Scandinavie (Aur. Boreal), p. 523-528, 557.

X Cosmos, vol. i., p. 200; see also FrankHn, Narrative of a Journey
to the Shores of the Polar Sea in 1819-1822, p. 51)7 ; and Kamtz, Lehr-
buch der Meteorolojie, bd. iii. (1836), s. 488-490. The earliest con-
jectures advanced in relation to the connection between the northern
light and the formation of clouds are probably those of Erobesius. (See
Aurorce Borealis spectacula, Helmst, 1739, p. 139.)

150 COSMOS.

logical process generated by and accompanying the magnetic
storm. The regular coincidence in respect to direction be-
tween the very fine cirrous clouds (polar bands) and the mag-
netic declination, together with the turning of the points of
convergence, were made the subjects of my most careful ob-
servation on the Mexican plateau in 1803, and in Northern
Asia in 1829. When the last-named phenomenon is com-
plete, the two apparent points of convergence do not remain
stationary, the one in the jiortheast and the other in the south-
west (in the direction of the line which connects together the
highest points of the arch of the polar light, which is lumin-
ous at night), but move by degrees toward the east and west.*
A precisely similar turning, or translation of the line, which
in the true aurora connects the highest points of the lumin-
ous arch, while its bases (the points of support by which it
rests on the horizon) change in the azimuth and move from
east-west toward north-south, has been several times observed
with much accuracy in Finmark.* These clouds, arranged

* I will give a single example from my MS. journal of my Siberian
journey: "I spent the whole of the night of the 5-Gth of August
(1829), separated from my traveling companions, in the open air, at
the Cossack outpost of Krasnajazarki, the most eastern station on the
Irtisch, on the boundary of the Chinese Dzungarei, and hence a place
whose astronomical determination was of considerable importance.
The night was extremely clear. In the eastern sky polar bands of
cirrous clouds were suddenly formed before midnight (which I have
recorded as ' de petits moutons egalement espaces, distrihues en bandQs
paralleles et polaires).' Greatest altitude 35°. The northern point of
convergence is moving slowly toward the east. They disappear with-
out reaching the zenith f and a few minutes afterward precisely simi-
lar cirrous bands are formed in the northeast, which move during a
part of the night, and almost till sunrise, regularly northward 70° E.
An unusually large number of falling stars and colored rings round
the moon throughout the night. No trace of a true aurora. Some
rain falling from speckled feathery masses of clouds. At noon on the
6th of August the sky was clear, polar bands were again formed, pass-
ing from N.N.E. to S.S.W., where they remained immovable, without
altering the azimuth, as I had so often seen in Quito and Mexico."
(The magnetic variation in the Altai is easterly.)

t Bravais, who, contraiy to my own experience, almost invariably
observed that the masses of cirrous clouds at Bossekop were directed,
like the Aurora Borealis, at right angles to the magnetic meridian
(Voyages en Scandinavie, Fhenomene de translation dans les pieds de I'arc
des Aurores Boreales, p. o3t-537), describes with his accustomed ex-
actitude the turnings or rotations of the true arch of the Aurora Borea-
. lis, p. 27, 92, 122, 487. Sir James Ross has. likewise observed in the
southern hemisphere similar progressive alterations of the arch of the
aurora (a progression in the southern lights from W.N.W. — E.S.E. to
N.N.E. — S.S.W.), Voyage in the Southern and Antartic Regions, vol. i..

POLAR LIGHT. 151

in the form of polar bands, correspond, according to the above
developed views, in respect to position, with the luminous
columns or bundles of rays which ascend in the true aurora
toward the zenith from the arch, which is generally inclined
in an east and west direction ; and they can not, therefore,
be confounded with those arches of which one was distinctly
seen by Parry in bright daylight after the occurrence of a
northern light. This phenomenon occurred in England on
the 3d of September, 1827, when columns of light were seen
shooting up from the luminous arch even by day.*

It has frequently been asserted that a continuous evolution
of light prevails in the sky immediately around the northern
magnetic pole. Bravais, who continued to prosecute his ob-
servations uninterruptedly for 200 nights, during which he
accurately described 152 aurorse, certainly asserts that nights
in which no northern lights are seen are altogether excep-
tional ; but he has sometimes found, even when the atmos-
phere was perfectly clear, and the view of the horizon was
wholly uninterrupted, that not a trace of polar light could
be observed throughout the whole night, or else that the
magnetic storm did not begin to be apparent until a very late
hour. The greatest absolute number of northern hghts ap-
pears to occur toward the close of the month of September ;
and as March, when compared with February and April,
seems to exhibit a relatively frequent occurrence of the phe-
nomenon, we are here led, as in the case of other magnetic
phenomena, to conjecture some connection with the period
of the equinoxes. To the northern lights which have been
seen in Peru, and to the southern lights which have been vis-
ible in Scotland, we may add a colored aurora, which was
observed for more than two hours continuously by Lafond in
the Candide, on the 14th of January, 1831, south of New
Holland, in latitude 45°.t

The accompaniment of sound in the aurora has been as
definitely denied by the French physicists and Siljestrom at

p. 311. An absence of all color seems to be a frequent characteristic
of southern lights, vol. i., p. 266; vol. ii., p. 209. Regarding the ab-
sence of the northern light in some nights in Lapland, see Bravais,
Op. cit., p. 545.

* Cosmos, vol. i., p. 197. The arch of the aurora seen in bright
daylight reminds us, by the intensity of its light, of the nuclei and
tails of the comets of 1843 and 1847, which were recognized in the
immediate vicinity of the sun in North America, Parma, and London.
Op. cit., vol. i., p. 85 ; vol. iii., p. 543.

t Comptes rendus de I' Acad, des Sciences, t. iv., 1837, p. 589.

152 COSMOS.

Bossekop* as by Thienemann, Parry, Franklin, Richardson,
Wrangel, and Anjou. Bravais estimated the altitude of the
phenomenon to be fully 51,307 toises (or 52 geographical
miles), while an otherwise very careful observer, Farquhar-
son, considers that it scarcely amounts to 4000 feet. The
data on which all these determinations are based are very
uncertain, and are rendered less trustworthy by optical illu-
sions, as well as by erroneous conjectures regarding the posi-
tive identity of the luminous arch seen simultaneously at two
remote points. There is, however, no doubt whatever of the
inflaence of the northern light on declination, inclination,
horizontal and total intensity, and consequently on all the
elements of terrestrial magnetism, although this influence is
exerted very unequally in the different phases of this great
phenomenon, and on the different elements of the force. The
most complete investigations of the subject were those made
in Lapland by the able physicists Siljestrom and Bravaisf
(in 1^38-1839), and the Canadian observations at Toronto
(1840-184:1), which have been most ably discussed by Sa-
bine.f In the preconcerted simultaneous observations which
were made by us at Berlin (in the Mendel ssohn-Bartholdy
Garden), at Freiberg below the purface of the earth, at St
Petersburg, Kasan, and Nikolajew, we found that the mag-
netic variation was affected at all these places by the Aurora
Borealis, which was visible at Alford, in Aberdeenshire (57°
15' N. lat.), on the night of the 19-20th of December, 1829.
At some of these stations, at which the other elements of
terrestrial magnetism could be noted, the magnetic intensity
and inclination were affected no less than the variation.§
During the beautiful aurora which Professor Forbes ob-

* Voyages en Scandinavie, en Laponie, etc. (Aurores Bor^aks), p. 559;
and Martin's Tj-ad. de la Meteorologie de Kaemtz, p. 460. In refer-
ence to the conjectured elevation of the northern light, see Bravais,
0}^. clL, p. 549, 559.

t Op, ciL, p. 462.

X Sabine, Unusual Magnet. Disturbances, pt. i., p. xviii., xxii., 3,
54.

§ Dove, in Poggend., Ann., bd. xx., s. 333-341. The unequal influ-
ence which an aurora exerts on the dipping-needle at points of the
earth's surface, which lie in very different meridians, may in many
cases lead to the local determination of the active cause, since the
manifestation of the luminous magnetic storm does not by any means
always origiifcate in the magnetic pole itself; while, moreover, as Ar-
gelander maintained and as Bravais has confirmed, the summit of the
luminous arch is in some cases as much as 11° from the magnetic me-
ridian.

TERRESTRIAL MAGNETISM. 153

served at Edinburgh on the 21st of March, 1833, the inclin-
ation was strikingly small in the mines at Freiberg, while
the variation was so much disturbed that the angles could
scarcely be read off. The decrease in the total intensity of
the magnetic force, which has been observed to coincide with
the increasing energy of the luminosity of the northern light,
is a phenomenon which is worthy of special attention. The
measurements which I made in conjunction with Oltmanns
at Berlin during a, brilliant aurora on the 20th of Decemni
ber, 1806,* and which are printed in Hansteen's " Unter-
suchungen iiber den Magnetism us der Erde," were confirmed
by Sabine and the French physicists in Lapland in 1838.f

While in this careful development of the present condition
of our positive knowledge of the phenomena of terrestrial
magnetism, I have necessarily limited myself to a mere ob-
jective representation of that which did not even admit of
being elucidated by merely theoretical views, based only
upon induction and analogy ; I have likewise purposely ab-
stained in the present work from entering into any of those
geognostic hypotheses in which the direction of extensive

* "On the 20th of December, 1806, the heavens were of an azure
blue, with not a trace of clouds. Toward 10 P.M. a reddish-yellow
luminous arch appeared in the NN.W., through which I could distin-
guish stars of the 7th magnitude in the night telescope. I found the
azimuth of this point by means of a Lyrse, which was almost directly
under the highest point of the arch. It was somewhat farther west
than the vertical plane of the magnetic variation. The aurora, which
was directed NN.W., caused the north pole of the needle to be de-
flected, for, instead of progressing westward like the azimuth of the
arch, the needle moved back toward the east. The changes in the
magnetic declination, which generally amount to from 2' 27" to 3' in
the nights of this month, increased progressively and without any great
oscillation to 26' 28" during the northern light. The variation was
the smallest about 9h. 12m., when the aurora was the most intense.
We found that the horizontal force amounted to 1' 37"-73 for 21 vi-
brations during the continuance of the aurora, while at^h. 50m. A.M.,
and consequently long after the disappearance of the aurora, which
had entirely vanished by 2h. 10m. A.M., it was 1' 37 '"17 for the same
number of vibrations. The temperature of the room, in which the
vibrations of the small needle were measured, was in the first case
37°'76 F., and in the second 37°-04 F. The intensity was, therefore,
slightly diminished during the continuance of the northern light.
The moon presented no colored rings." From my magnetic journal,
see Hansteen, s. 459.

t Sabine, On Days of Unusual Magn. Disturbances, pt. i., p. xviii.
"M. Bravais concludes from the observations made in Lapland that
the horizontal intensity diminishes when the phenomenon of the Au-
rora Borealis is at its maximum" (Martins, p. 461).

G2

154 COSMOS.

mountain chains and of stratified mountain masses is con-
sidered in relation to its dependence upon the direction of
magnetic lines, more especially the isoclinal and isodynamic
systems. I am far from denying the influence of all cosmical
primary forces — dynamic and chemical forces — as well as of
magnetic and electrical currents on the formation of crystal-
line rocks and the filling up of veins;* but owing to the
progressive movement of all magnetic lines and their conse-
#quent change of form, their present position can teach us
nothing in reference to the* direction in primeval ages of
mountain chains, which have been upheaved at very differ-
ent epochs, or to the consolidation of the earth's crust, from
which heat was being radiated during the process of its
hardening.

Of a different order, not referring generally to- terrestrial
magnetism, but merely to very partial local relations, are
those geognostic phenomena which have been designated by
the name of the magnetism! of mountain masses. These
phenomena engaged much of my attention before my Amer-
ican expedition, at a time when I was occupied in examin-
ing the magnetic serpentine rock of the Haidberg mountain,
in Franconia, in 1796, and then gave occasion in Germany
to a considerable amount of literary dissension, which, how-
ever, was of a very harmless nature. They present a num-
ber of problems, which are by no means incapable of solu-
tion, but which have been much neglected in recent times,
and only very imperfectly investigated both as regards ob-
servation and experiment. The force of this magnetism of
rocks may be tested for the determination of the increase of
magnetic intensity by means of pendulum experiments, and
by the deflection of the needle in broken-off fragments of
hornblende and chloritic schists, serpentine, syenite, dolerite,
basalt, melaphyre, and trachyte. We may in this manner
decide, by a comparison of the specific gravity, by the rins-
ing of finely pulverized masses, and by the application of the
microscope, whether the intensity of the polarity may not
depend in various ways upon the relative position, rather
than upon the quantity, of the granules of magnetic iron

* Delesse, Sur rassociation des mineraux dans les roches qui ont
un pouvoir magnetique eleve, in the Comptes rendus de I'Acad. des Sc,
t. xxxi., 1850, p. 806 ; and Annales des Mines, 4eme Serie, t. xv.
(1849), p. 130.

t Reich, Ueher Gebirgs-und Gesteins-Magnetismus, in Poggend.,
Ann.y bd. Ixvii., s. 36.

TERRESTRIAL MAGNETISM. 155

and protoxyd of iron intermixed in the mass. More im-
portant, however, in a cosmical point of view, is the question
which I long since suggested in reference to the Haidberg
mountain, wiiether there exist entire mountain ranges in
which opposite polarities are found to occur on opposite de-
clivities of the mass.* An accurate astronomical determin-

* This question was made the subject of lively discussion when, in
the year 1796, at the time tha'Pl fulfilled the duties of superintend-
ent of the mining operations in the Fichtelgebijrge, in Franconia, I dis-
covered the remarkable magnetic serpentine mountain (the Haidberg)
near Gefress, which had the property at some points of causing the
needle to be deflected at a distance of even 23 feet {Intelligenz-Blatt
der Allgem. Jenaer Litteratur-Zeitung, Dec, 1796, No. 169, s. 1447,
and Mdrz, 1797, No. 38, s. 323-326 ; Gren's Neues JourrOil der Physik,
bd. iv., 1797, s. 136; Annales de C/dmie, t. xxii., p. 47). I had thought
that the magnetic axes of the mountain were diametrically opposed to
the terrestrial poles ; but according to the investigations of Bischoff
and Goldfuss, in 1816 (Beschrelbung des Fichtelgebirges, bd. i., s. 176),
it would appear that they discovered magnetic poles, which penetrated
through the Haidberg and presented opposite poles on the opposite
declivities of the mountain, while the directions of the axes were not
the same as I had given them. The Haidberg consists of dull green
serpentine, which partially merges into chloritic and hornblende schists.
At the village of Voysaco, in the chain of the Andes of Pasto, we saw ,
the needle deflected by fragments of porphyritic clay, while on the
ascent to Chimborazo groups of columnar masses of trachyte disturbed
the motion of the needle at a distance of three feet. It struck me as a
very remarkable fact that I should have found in the black and red
obsidians of Quinche, north of Quito, as well as in the gray obsidian of
the Cerro de la Navajas of Mexico, large fragments with distinct poles.
The large collective magnetic mountains in the Ural chain, as Blago-
dat, near Kuschwa, Wyssokaja Gora, at Nishne Tagilsk, and Katsch-
kanar, near Nishne Turinsk, have all broken forth from augitic or
rather uralitic porphyry. In the great magnetic mountain of Blago-
dat, which I investigated with Gustav Rose, in our Siberian expedi-
tion in 1829, the combined effect of the polarity of the individual parts
did not, indeed, appear to have produced any determined and recog-
nizable magnetic axes. In close vicinity to one another lie irregular-
ly mixed opposite poles. A similar observation had previously been
made by Erman {Reise urn die Erde, bd. i., s. 362). On the degree of
inteinsity of the polar force in serpentine, basaltic, and trachytic rock,
compared with the quantity of magnetic iron and protoxyd of iron,
intermixed with these rocks, as well as on the influ^ceof the contact
of the air in developing polarity, which had already been maintained
by Gmelin and Gibbs, see the numerous and very admirable experi-
ments of Zaddach, in his Beobachtungen Uber die Magnetische Polaritat
des Basaltes und der Trachytischen Gesteine, 1851, s. 56, 65-78, 95. A
comparison of many basaltic quarries, made with a view of ascertain-
ing the polarity of individual columns which have stood isolated for a
long period, and an examination of the sides of these columns which
have been recently brought in contact with the outer air in conse-
quence of the removal from individual masses of a certain depth of

156

COSMOS.

ation of the position of such magnetic axes of a mountain
would be of the greatest interest, if it could be ascertained,
after considerable periods of time, that the three variable
elements of the total force of terrestrial magnetism caused
either an alteration in the direction of the axes, or that such
small systems of magnetic forces were at least apparently
independent of these influences.

earth, have led Dr. Zaddach to haza?d the conjecture (see s. 74, 80)
that the polar property, which always appears to be manifested with
the greatest intensity in rocks to which the air has been freely admit-
ted, and which are intersected by open fissures, " diffuses itself from
without inward, and generally from above downward." Gmelin ex-
presses himself as follows in respect to the great magnetic mountain,
Ulu-utasse-T^au, in the country of the Baschkiri, near the Jaik : " The
sides which are exposed to the open air exhibit the most intense mag-
netic force, while those which lie under ground are much weaker"
(Meise durch Siberien, 1740-1743, bd. iv., s. 345). My distinguished
teacher, Werner, in describing the magnetic iron of Sweden, in his
lectures, also spoke of " the influence which contact with the atmos-
phere might have, although not by means of an increased oxydation,
in rendering the polar and attracting force more intense." It is as-
serted by Colonel Gibbs, in reference to the magnetic iron mines at
Succassuny, in New Jersey, that " the ore raised from the bottom of
the mine has no magnetism at first, but acquires it after it has been
some time exposed to the influence of the atmosphere" {On the connec-
tion of Magnetism and Light, in Silliman's American Journal of Science y
vol. i., 1819, p. 89). Such an assertion as this ought assuredly to stim-
ulate observers to make careful and exact investigations ! Wh^n I
drew attention in the text (see page 154) to the fact that it was not
only the quantity of the small particles of iron which were intermixed
in the stone, but also their relative distribution (their position), which
acted as the resultant upon the intensity of the polar force, I consid-
ered the small particles to be so many small magnets. See the new
views regarding this subject in a treatise by Melloni, read by that dis-
tinguished physicist before the Royal Academy at Naples, in the month
of January, 1853 (Esperienze intorno al Magnetismo delle Rocche, Mem.
i., Sulla Polarita). The popular notion which has been so long cur-
rent, more especially on the shores of the Mediterranean, that if a
magnetic rod be rubbed with an onion, or brought in contact with the
emanations of the plant, the directive force will be diminished, while
a compass thus treated would mislead the steersman, is mentioned in
Prodi Diadochi Paraphrasis Ptolem., libri iv,, de Siderum affectionibus^
1635, p. 20 (Del^nbre, Hist, de V Astronomie Ancienne, t. ii., p. 545).
It is difficult to conceive what could have given occasion to so singular
a popular error.

VULCANICITY. 157

II.

REACTION OP THE INTEEIOR OF THE EARTH UPON ITS SURE ACE-,
MANIFESTING ITSELF:— a. MERELY DYNAMICALLY, BY TREMU-
LOUS UNDULATIONS (EARTHQUAKES): b. BY THE HIGH TEM-
PERATURE OF MINERAL SPRINGS, AND BY THE DIFFERENCE OF

. THE INTERMIXED SALTS AND GASES (THERMAL SPRINGS) ; c. BY
THE OUTBREAK OF ELASTIC FLUIDS, SOMETIMES ACCOMPANIED
BY PHENOMENA OF SPONTANEOUS IGNITION (GAS AND MUD VOL-
CANOES, BURNING NAPHTHA SPRINGS, SALSES); d. BY THE
GRAND AND MIGHTY ACTIONS OF TRUE VOLCANOES, WHICH
(WHEN THEY HAVE A PERMANENT CONNECTION WITH THE AT-
MOSPHERE BY FISSURES AND CRATERS) THROW UP FUSED EARTH
FROxM THE DEPTHS OF THE INTERIOR, PARTLY ONLY IN THE
FORM OF RED-HOT CINDERS, BUT PARTLY SUBMITTED TO VARY-
ING PROCESSES OF CRYSTALLINE ROCK FORMATION, POURED OUT
IN LONG, NARROW STREAMS.

In order to maintain, in accordance with the fundamental
plan of this work, the co-ordination of telluric phenomena —
the co-operation of a single system of impelling forces — in
the descriptive representation, we must here remind the
reader how, starting from the general properties of matter,
and the three principal directions of its activity (attraction,
vibrations producing light and heat, and electro-magnetic pro-
cesses), we have in the first, section taken into consideration
the size, form, and density of our planet, its internal diffusion
of heat and of magnetism, in their effects of intensity, dip,
and variation, changing in accordance with definite laws.
The directions of the activity of matter just mentioned are
nearly allied* manifestations of one and the same primitive
force. They occur in a condition of the greatest independ-
ence of all diflTerences of matter, in gravitation and molecular
attraction. We have at the same time represented our planet
in its cosmical relation to the central body of its system, be-
cause the internal jmmitive heat, which is probably produced
by the condensation of a rotating nebular ring, is modified
by the action of the sun (insolation). With the same view,
the periodical action of the solar spots (th^t is to say, the
frequency or rarity of the apertures in the solar envelopes)
upon terrestrial magnetism has been referred to, in accord-
ance with the most recent hypotheses.

The second section of this volume is devoted to the entire-
ty of those telluric phenomena which are to be ascribed to
the constantly active reaction of the interior of the earth upon
* Cosmos, vol. iii., p. 34.

158 COSMOS.

its su7'face.* To this entirety I give the general name of
Vulcanism or Vulcanicity ; and I regard it as advantageous
to avoid the separation of that which is causally connected,
and differs only in the strength of the manifestation of force
and the complication of physical processes. By taking this
general view, small and apparently unimportant phenomena
acquire a greater significance. The unscientific observer
who comes for the first time upon the basin of a thermal
spring and sees gases capable of extinguishing light rising
in it, or who wanders among rows of changeable cones of
mud volcanoes scarcely exceeding himself in height, never
dreams that in the calm space occupied by the latter erup-
tions of fire to the height of many thousand feet have often
taken place ; and that one and the same internal force pro-
duces colossal craters of elevation — nay, even the mighty,
desolating, lava-pouring volcanoes of ^i^tna and the Peak of
Teyde, and the cinder-erupting Cotopaxi and Tunguragua.

Among the multifarious, mutually intensifying phenomena
of the reaction of the interior of the earth upon its external
crust, I first of all separate those the essential character of
which is purely dynamical, namely, that of movement or
tremulous undulations in the solid strata of the earth ; a
volcanic activity which is not necessarily accompanied by any
chemical changes of matter, or by the expulsion or produc-
tion of any thing of a material nature. In the other phe-
nomena of the reaction of the interior upon the exterior of
the earth — in gas and mud volcanoes, burning springs and salses,
and in the large burning mountains to which the name of vol-
cano was first, and for a long time exclusively applied, the
production of something of a material nature (gaseous or
solid), and processes of decomposition and gas evolution,
such as the formation of rocks from particles arranged in a
crystalline form, are never wanting. When most fully gen-
eralized, these are the distinctive characters of the volcanic
vital activity of our planet. In so far as this activity is to be
ascribed, in great measure, to the high temperature of the
innermost strata of the earth, it becomes probable that all
cosmical bodies which have become conglomerated with an
enormous evolution of heat, and passed from a state of vapor
to a solid condition, must present analogous phenomena.
The little that we know of the form of the moon's surface
appears to indicate this.f Upheaval and plastic activity in

* Cos77jos, vol. i., p. 202-204.

t Cosmos, vol. iii., p. 44 ; iv., p. 104, 151, 154-156.

vuLCANicrrY. 159

the production of crystalline rock from a fused mass are con-
ceivable even in a sphere which is regarded as destitute of
both air and water.

The genetic connection of the classes of volcanic phenom-
ena here referred to is indicated by the numerous traces of
the simultaneousness of the simpler and weaker with stronger
and more complex effects, and the accompanying transitions
of the one into the other. The arrangement of the mate-
rials in the representation selected by me is justified by such
a consideration. The increased magnetic activity of our
planet, the seat of which, however, is not to be sought in
the fused mass of the interior (even though, according to
Lenz and Riess, iron in the fused state may be capable of
conducting an electrical or galvanic current), produces evolu-
tion of light in the magnetic poles of the earth, or at least
usually in their vicinity. We, concluded the first section of
the volume on telluric phenomena with the luminosity of the
earth. This phenomenon of a luminous vibration of the
ether hj magnetic forces is immediately followed by that class
of volcanic agencies which, in their essential nature, act pure-
ly dynamically, exactly like the magnetic force — causing
movement and vibrations in the solid ground, but neither
producing nor changing any thing of a material nature.
Secondary and unessential phenomena (the ascent of flames
during the earthquake, and eruptions of water and evolu-
tions of gas* following it) remind one of the action of ther-
mal springs and salses. Eruptions of flame, visible at a dis-
tance of many miles, and masses of rock, torn from their
deep seats and hurled about,! are presented by the salses,
which thus, as it were, prepare us for the magnificent phe-
nomena of the true volcanoes; which again, between their
distant epochs of eruption, like the salses, only exhale aque-
ous vapor and gases from their fissures : so remarkable and
instructive are the analogies which are presented in various
stages by the gradations of Vulcanism.

* Cosmos, vol. i., p. 217.

t Cosmos, vol. i., p. 225. Compare Bertrand-Geslin, " Sur les roches
lancces par le Volcan de boue du Monte Zibio pres du bourg de Sassuolo,"
in Humboldt, Voyage aux Regions Equinoxiales du Nouveau Continent
(Relation Historique), t. ill., p. 566.

160

COSMOS.

a. Earthquakes.
(Amplification of the Picture of Nature, Cosmos, vol. i., p. 204-217.)

Since the appearance in the first volume of this work
(1845) of the general representation of the phenomena of
earthquakes, the obscurity in which the seat and causes of
these phenomena are involved has but little diminished ; but
the excellent works ^ of Mallet (1846) and Hopkins (1847)
have thrown some light upon the nature of concussions, the
connection of apparently distinct eiFects, and the separation
of chemical and physical processes, which may accompany
it or occur simultaneously with it. Here, as elsewhere, a
mathematical mode of treatment, such as that adopted by
Poisson, may have a beneficial effect. The analogies between
the oscillations of solid bodies and the sound-waves in the or-
dinary atmos'phere, to which Thomas Young f had already
called attention, are peculiarly adapted to lead to simpler
and more satisfactory views in theoretical considerations upon
the dynamics of earthquakes.

Displacement, commotion, elevation, and formation of fissures
indicate the essential character of the phenomenon. We
have to distinguish the efficient force which, as the imjndse,
gives rise to the vibration ; and the nature, propagation, in-
crease, or diminution of the commotion. In the Picture of
Nature I have described what is especially manifested to the
senses ; what I had myself the opportunity of observing for
so many years on the sea, on the sea-bottom of the plains
{Llanos), and at elevations of eight to fifteen thousand feet ;
on the margin of the craters of active volcanoes, and in re-
gions of granite and mica schist, twelve hundred geograph-
ical miles from any eruptions of fire ; in districts where at
certain periods the inhabitants take no more notice of the

* Robert Mallet, in the Transactions of the Royal Insh Academy^.
vol. xxi. (1848), p. 51-113, and First Report on the f^acis of Earth-
quake Phenomena, in the Report of the Meeting of the British Associa-
tion, 1850, p. 1-89; also Maniial of Scientific Inquiry for the Use of the
British Navy, 1849, p. 196-223. William Hopkins, On the Geological
Theories of Elevation and Earthquakes, in the Report of the British As-
sociation for 1847, p. 33-92. The rigorous criticism to which Mr. Mal-
let has subjected my previous work, in his very valuable memoirs {Insh
Transactions, p. 99-101, and Meeting of the British Association at Edin-
burgh, p. 209), has been repeatedly made use of by me.

t Thomas Young, Lectures on Natural Philosophy, 1807j vol. i.. p.
717.

EARTHQUAKES. 161

number of earthquakes than we in Europe of that of the
showers of rain, and where Bonpland and I were compelled
to dismount, from the restiveness of our mules, because the
earth shook in a forest for 15 to 18 minutes without intermis-
sion. Bj such long custom, as Boussingault subsequently
experienced even in a still higher degree, one becomes fitted
for quiet and careful observation, and also for collecting
varying evidence with critical care on the spot, nay, even
for examining under what conditions the mighty changes of
the surface of the earth, the fresh traces of which one recog-
nizes, have taken place. Although five years had already
elapsed since the terrible earthquake of Riobamba, which,
on the 4th of February, 1797, destroyed upward of 30,000
people in a few minutes,* we nevertheless saw the formerly-
advancing cone of the Moya f which rose out of the earth,
and witnessed the employment of this combustible substance
for co(Jking in the huts of the Indians. I might describe the
results of alterations of the ground from this catastrophe,
which, although on a large* scale, were exactly analogous to
those presented by the famous earthquake of Calabria (Feb-
ruary, 1783), and were long considered to have been repre-
sented in an incorrect and exaggerated manner, because they
could not be explained in accordance with hagjM|pr-formed
theories.

By carefully separating, as we have already indicated,
the investigation of that which gives the impulse to the vi-
bration, from that of the nature and propagation of the
waves of commotion, we distinguish two classes of problems
of very unequal accessibility. The former, in the present
state of our knowledge, can lead to no generally satisfactory
results, as is the case with so many problems in which we
wish to ascend to primary causes. Nevertheless, while we
are endeavoring to discover laws in that which is submitted
to actual observation, it is of great cosmical interest that we
should bear constantly in mind the various genetic explana-
tions which have hitherto been put forward as probable.
As with all vulcanicity, the greater part of these refer,
under various modifications, to the high temperature and
chemical nature of the fused interior of the earth ; one of

* I follow the statistical account communicated to me by the Cor-
regidor of Tacunga in 1802. It rose to a loss of 30,000—34,000 peo-
ple, but some twenty years later the number of those killed immedi-
ately was reduced by about one third.

t Cosmos, vol. i., p. 213.

162 COSMOS,

the most recent explanations of earthquakes in trachytic re-
gions is the ivisult of geognostic suppositions regarding the
want of cohesion in rocky masst3S raised by volcanic action.
The following summary furnishes a more exact but very
brief indication of the variety of views as to the nature of
the first impulse to the commotion :

The nucleus of the earth is supposed to be in a state of
igneous fluidity, as the consequence of every planetary
process of formation from a gaseous material, by evolu-
tion of heat during the transition from fluidity to solidity.
The external strata were first cooled by radiation, and
were the first to become consolidated. The commotion is
occasioned by an unequal ascent of elastic vapors, formed
(at the limit between the fluid and solid parts) either from
the fused terrestrial mass alone or from the penetration
of sea-water into higher strata of rock, nearer to the sur-
face of the earth, the sudden opening of fissures, and by
the sudden ascent of vapors produced in the hotter and
consequently more elastic depljis. The attraction of the
moon and sun* on the fluid, fused surface of the nucleus

* Hopkins has expressed doubts as to the action npon the fused
"subjacent fluid confined into internal lakes, ^^ SLt tho. Meeting of the
British Ai/tMation for 1847 (p. 57), as Mallet has also done with re-
gard to '*nie subteiTaneous lava tidal wave, moving the solid crust
above it," at the British Association Meeting for 1850 (p. 20). Poisson
also, with whom I have often spoken regarding^he hypothesis of the
subterranean ebb and flow caused by the sun and moon, considers the
impulse, which he does not deny, to be inconsiderable, "as in the open
sea the effect scarcely amounts to 14 inches." Ampere, on the other
hand, says: "Those who admit the fluidity of the internal nucleus of
the earth do not appear to have sufficiently considered the action which
would be exercised by the moon upon this enormous liquid mass — an
action from which would result tides analogous to those of our seas,
but far more terrible, both from their extent and from the density of
the liquid. It is difficult to conceive how the envelope of the earth
should be able to resist the incessant action of a sort of hydraulic
ram(?) of 1400 leagues in length" (Ampere, Theorie de la Terre, in
Revue des deux Mondes, July, 1833, p. 148). If the interior of the
earth be fluid, which in general can not be doubted, as, notwithstand-
ing the enormous pressure, the particles are still displaceable, then the
same conditions are fulfilled in the interior of the earth that give rise
on the surface to the ocean tides ; and the tide-producing force will
constantly become weaker in approaching the centre, as the difference
of the distances of every two opposite points, considered in their rela-
tion to the attracting bodies, constantly becomes less in receding from
the surface, and the" force depends exclusively upon the difference of
the distances. If the solid crust of the earth opposes a resistance to
this effort, the interior of the earth will only exert a pressure against
its crust at these points ; as my astronomical friend, Dr. Brunnow,

EARTHQUAKES. 163

^f iM earth may also be regarded as the subsidiary action
of a hon-telluric cause, by which .an increased pressure
must be j^roduced, either immediately against a solid, su-
perimposed rocky arch ; or indirectly, when the solid mass
is separated, in subterranean basins, from the fused, fluid
mass by eiastic vapors. «

The nucleus of our planet is supposed to consist of un-
oxydized ma«ses,'the metalloids of the alkalies and earths.
Volcanic activity is excited in the nucleus by the access
of water and air. Volcanoes certainly pour forth a great
quantity of aqueous vapor into the atmosphere ; but the
assumption of the penetration of water into the volcanic
focus is attended with much difficulty, considering the op-
posing pressure* of the external column of water and of

expresses himself, no more tide will be produced than if the ocean had
an indestructible covering o/ ice. The thickness of the solid, unfused
crust of the earth is calculated from the fusing points of the different
kinds of rock, and the law of the increase of heat from the surface
into the depths of the earth. I have already (Cosmos, vol. i., p. 45)
justified the assumption that at somewhat more than twenty geograph-
ical miles (21-/ij, 25 English) below ihe surface a heat capable of melt-
ing granite prevails. Nearly the same number (45,000 metres =24
geographical miles) was named by Elie de Beaumont (Geologie, edited
by Vogt, 1846, vol. i., p. 32) as the thickness of the solid crust of the
earth. Moreover, according to the ingenious experiments of Bischof
on the fusion of vaiious minerals, of which the importance to the prog-
ress of geology is so great, the thickness of the unfused strata of the
earth is between 122,590 and 136,448 feet, or, on the average, 21i geo-
graphical (24i^ English) miles; see BiscLof, Wdrmelehre des Innern
unsers Erdkorpers, p. 286 and 271. This renders it the more remark-
able to me to find that, with the assumption cf a definite limit between
the solid and fused parts, and not of a gradual transition, Hopkins,
from the fundamental principles of his speculative geology, establishes
the result that "the thickness of the solid shell can not be less than
about one fourth or one fifth (?) of the radius of its external surface"
{Meeting of British Association, 1847, p. 51). Cordier's earliest sup-
position Avas only 56 geographical (72 English) miles, without correc-
tion, which is dependent upon the increased pressare of the strata at
great depths, and the hypsometrical form of the surface. The thick-
ness of the solid part of the crust of the earth is probably very un-
equal.

* Gay-Lussac, Reflexions sur les Volcans, in the Annaks de Chi-
mie et de Physique, tome xxii., 1823, p. 418 and 426. The author,
Vvho, in company with Leopold von Buch and myself, observed the
great eruption of lava from Vesuvius in September, 1805, has the
merit of having submitted the chemical hypotheses ta a strict criti-
cism. He seeks for the cause of volcanic phenomena iri a "very en-
ergetic and still unsatisfied affinity between the substances, which a
fortuitous contact permits them to obey ;" in general, he favors the
hypothesis of Davy and Ampere, which is now given up, ** supposing

164 COSMOS.

the internal lava; and the deficiency, or, at all events,
very rare occurrence of burning hydrogen gas during the
eruption (which the formation of hydrochlorio acid,* am-
monia, and sulphureted hydrogen certainly does not suffi-
ciently replace), has led the celebrated originator of this
hypothesis to abandon it of his own accord, f

According to a third view, that of the highly-endowed
South American traveler, BoussingauHt, a deficiency of co-
herence in the trachytic and doleritic masses which form
the elevated volcanoes of the chain of the Andes, is regard-
ed as a primary cause of many earthquakes of very great
extent. The colossal cones and dome-like summits of the
Cordilleras, according to this view, have by no means been
elevated in a soft and semi-fluid state, but have been thrown
up and piled on one another when perfectly hardened, in
the form of enormous, sharp-edged fragments. In an ele-
vation and piling of this description, large interstices and
cavities have necessarily been produced ; so that by sud-
den sinking, and by the fall of solid masses which are too
weakly supported, shocks are produced. {

that the radicals of silica, alumina, lime, and iron are combined with
chlorine in the interior of the earth," and the penetration of sea-wa-
ter does not appear to him to be improbable under certain conditions
(p. 419, 420, 423, and 426). Upon the difficulty of a theory founded
upon the penetration of water, see Hopkins, JSrif.Jiapoc. jR(^jo., 1847, p. 38.

* According to the beautiful analyses made by Boussingault on the
margins of five craters (Tolima, Purace, Pasto, Tuqueras, and Cum-
bal), hydrochloric acid is entirely wanting in the vapors poured forth
by the South American volcanoes, but not in those of Italy (Annales
de Chimie, tome Hi., 1833, p. 7 and 23).

t Cosmos, vol. i., p. 236. While Davy, in the most distinct man-
ner, gave up the opinion that volcanic eruptions are a consequence of
the contact of the metalloid bases with water and air, he still assert-
ed that the presence of oxydizable metalloids in the interior of the
earth might be a co-operating cause in volcanic processes already com-
menced.

X Boussingault says : "I attribute most of the earthquakes in the
Cordillera of the Andes to falls produced in the interior of these
mountains by the subsidence which takes place, and which is a conse-
quence of their elevation. The mass which constitutes these gigantic
ridges has not been raised in a soft state ; the elevation did not take
place until after, the solidification of the rocks. I assume, therefore,
that the elevated masses of the Andes are composed of fragments
heaped upon each other. The consolidation of the fragments could
not be so stable from the beginning as that there should be no
settlements after the elevation, or that there should be no inte-
rior movements in the fragmentary masses" (Boussingault, >$«?• les
Tremhlemens de Terre des Andes, in Annales de Chimie et de Phy-
sique, tome Iviii., 1835, p. 84-86). In the description of his mem-

EARTHQUAKES. 165

The efects of the impulse, and waves of commotion, may be
reduced to simple mechanical theories with more distinctness
than is furnished by the consideration of the nature of the
first impulse, which indeed may be regarded as heterogene-
ous. As already observed, this part of our knowledge has
advanced essentially in very recent times. The earth-waves
have been represented in their progress and their propaga-
tion through rocks of different density and elasticity ;* the
causes of the rapidity of propagation, and its diminution by
the refraction, reflection, and interferencef of the oscillations,
have been mathematically investigated. Attempts have been
made to reduce to a rectilinearf standard the apparently

orable ascent of Chimborazo (Ascension au Chimhorazo le 16 Dec.^
1831, be. cit., p. 176), he says again: "Like Cotopaxi, Antisana, Tun-
guragua, and the volcanoes in general which project from the plateaux
of the Andes, the mass of Chimborazo is formed by the accumulation
of trachytic debris, heaped together without any order. These frag-
ments, often of enormous volume, have been elevated in the solid
state by elastic fluids which have broken out through the points of
least resistance ; their angles are always sharp." The cause of earth-
quakes here indicated is the same as that which Hopkins calls "a
shock produced by the falUng of the roof of a subterranean cavity," in
his "Analytical Theory of Volcanic Phenomena" (^Brit. Assoc. Report,
1847, p. 82).

* Mallet, Dynamics of Earthquakes, p. 74, 80, and 82 ; Hopkins,
B7it. Assoc. Report, 1847, p. 74-82. All that we know of the waves
of commotion and oscillations in solid bodies shows the untenability
of the older theories as to the facilitation of the propagation of the
movement by a series of cavities. Cavities can only act a secondary
part in the earthquake, as spaces for the accumulation of vapors and
condensed gases. "The earth, so many centuries old," says Gay-
Lussac, very beautifully (Ann. de Chimie et de Phys., tome xxii., 1823,
p.* 428), "still preserves an internal force, which raises mountains (in
the oxydized crust), overturns cities, and agitates the entire mass.
Most mountains, in issuing from the bosom of the earth, must have
left vast cavities, which have remained empty, at least unless they
have been filled with water (and gaseous fluids). It is certainly in-
correct for Deluc and many geologists to make use of these empty
spaces, which they imagine produced into long galleries, for the propa-
gation of earthquakes to a distance. These phenomena, so grand and
terrible, are very powerful sonorous waves, excited in the solid mass
of the earth by some commotion, which propagates itself therein with
the same velocity as sound. The movement of a carriage over the
pavement shakes the vastest edifices, and communicates itself through
considerable masses, as in the deep quarries below the city of Paris."

t Upon phenomena of interference in the earth-waves, analogous to
those of the waves of sound, see Cosmos, vol. i., p. 215 ; and Hum-
boldt, Kieinere Schriften, bd. i., p. 379.

X Mallet on vorticose shocks and cases of twisting, in Brit. Assoc.
Report, 1850, p. 33 and 49, and in the Admiralty Manual, 1849, p. 213
(see Cosmos, vol. i., p. 204).

166 COSMOS.

circling (rotatory) shocks of which the obelisks before the
monastery of San Bruno, in the small town of Stephano del
Bosco (Calabria, 1783), furnished such a well-known ex-
ample. Air, water, and earth-waves follow the same laws
which are recognized by the theory of motion, at all events
in space ; but the earth- waves are accompanied, in their de-
structive action, by phenomena which remain more obscure
in their nature, and belong to the class of physical processes.
As such we have to mention — discharges of elastic vapors,
and of gases ; or, as in the small, moving Moyacones of Pel-
ileo, grit-like mixtures of pyroxene crystals, carbon, and in-
fusorial animalcules with silicious shields. These wandering
cones have overthrown a great number of Indian huts.*

In the general Delineation of Nature many facts are nar-
rated concerning the great catastrophe of Eiobamba (4th of
February, 1797), which were collected on the spot from the
lips of the survivors, with the most earnest endeavors after
historic truth. Some of them are analogous to the occur-
rences in the great earthquake of Calabria in the year 1783;
others are new, and especially characterized by the mine-like
manifestation of force frvm below upward. The earthquake
itself was neither accompanied nor announced by any subter-
ranean noise. A prodigious explosion, still indicated by the
simple name of el gran ruido, was not perceived until 18 or
20 minutes afterward, and only under the two cities of Quito
and Ibarra, far removed from Tacunga, Hambato, and the
principal scene of the destruction. There is no other event
in the troubled destinies of the human race by which in a
few minutes, and in sparingly-peopled mountain lands, ^o
many thousands at once may be overtaken by death, as by
the production and passage of a few earth-waves, accom-
panied by phenomena of cleavage !

In the earthquake of Riobamba, of which the celebrated
Valencian botanist, Don Jose Cavanilles, gave the earliest
account, the following phenomena are deserving of special
attention: Fissures which alternately opened and closed
again, so that men saved themselves by extending both arms
in order to prevent their sinking ; the disappearance of en-
tire caravans of riders or loaded mules (recuas), some of
which disappeared through transverse fissures suddenly open-

* The Moyacones were seen by Boussingault nineteen years after I
saw them. " Muddy eruptions, consequences of the earthquake, like
the eruptions of the Moya of Felileo, which have buried entire vil-
lages" (Ann. de Chirn. et de Fhys.^ t. Iviii., p. 81).

EARTHQUAKES. 167

ing in their path, while others, flying back, escaped the dan-
ger ; such violent oscillations (non-simultaneous elevation
and depression) of neighboring portions of the ground, that
people standing upon the choir of a church at a height of
more than 12 feet got upon the pavement of the street with-
out falling ; the sinking of massive houses,* in which the
inhabitants could open inner doors, and for two whole days,
before they were released by excavations, passed uninjured
from room to room, procured lights, fed upon supplies acci-
dentally discovered, and disputed with each other regarding
the probability of their rescue ; and the disappearance of
such great masses of stones and building materials. Old
Riobamba contained churches and monasteries among houses
of several stories ; and yet, when I took the plan of the de-
stroyed city, I only found in the ruins heaps of stone of eight
to ten feet in height. In the southwestern part of Old Kio-
bamba (the former Barrio de Sigchuguaicu) a mine-like ex-
plosion, the effect of a force from below upward, was dis-
tinctly perceptible. On the Cerro de la Culca^ a liill of some
hundred feet in height, which rises above the Cerro de Cum-
bicarca, situated to the north of it, there lies stony rubbish
mixed with human bones. Translatory movements, in a hori'
zontal direction, by which avenues of trees become displaced
without being uprooted, or fragments of cultivated ground
of very different kinds mutually cfisplace each other, have
occurred repeatedly in Quito, as well as in Calabria. A
still more remarkable and complicated phenomenon is the
discovery of utensils belonging to one house in the ruins of
another at a great distance — a circumstance which has given
rise to lawsuits. Is it, as the natives believe, a sinking fol-
lowed by an eruption? or, notwithstanding the distance, a
mere projection"? As, in nature, every thing is repeated
when similar conditions again occur, we must, by not con-
cealing even what is still imperfectly observed, call the atten-
tion of future observers to special phenomena.

According to my observations, it must not be forgotten

* Upon the displacement of buildings and plantations during the
earthquake o^ Calabria, see Lyell's Principles of Geology^ vol. i,, p.
484-491. Upon escapes in fissures during the great earthquake of
Riobamba, see my Relation Historique^ tome ii., p. 642. As a re-
markable example of the closing of a fissure, it must be mentioned
that, according to Scacchi's report, during the celebrated earthquake
(in the summer of 1851) in the Neapolitan province of Basilicata, a
hen was found caught by both feet in the street pavement in Barile,
near Melfi.

168 COSMOS.

that, besides the commotion of solid parts as earth-waves,
very different forces — as, for instance, physical forces, emana-
tions of gas and vapor — also assist in most cases in the pro-
duction ofjissures. When in the undulatory movement the
extreme limit of the elasticity of matter set in motion (accord-
ing to the difference of the rocks or the looser strata) is ex-
ceeded, and separation takes place, tense elastic fluid may
break out through the fissures, bringing substances of various
kinds from the interior to the surface, and giving rise again,
by their eruption, to translatory movements. Among these
phenomena which only accompany the primitive commotion
(the earthquake) are the elevation of the undoubtedly wan-
dering cone of the Moya, and probably also the transporta-
tion of objects upon the surface of the earth.* When large
clefts are formed, and these only close again at their upper
parts, the production of permanent -subterranean cavities may
not only become the cause of new earthquakes, as, according
to Boussingault's supposition, imperfectly-supported masses
become detached in course of time and fall, producing com-
motions, but we may also imagine it possible that the circles
of commotion are enlarged thereby, and that in the new earth-
quake the clefts opened in the previous one enable elastic
fluids to act in places to which they could not otherwise
have obtained access. It is, therefore, an accompanying
phenomenon, and not th% strength of the wave commotion,
which has once passed through the solid parts of the earth,
that gives rise to the gradual and very important, but too
little considered enlargement of the circle of commotionj^

Volcanic activities, of which the earthquake is one of the
lower grades, almost always include at the same time move-
ment and the physical production of matter. In the Delin-
eation of Nature we have already repeatedly indicated that
water and hot vapors, carbonic acid gas and other ?nofettes,

* Cosmos^ vol. i., p. 206. Hopkins has very correctly shown theo-
retically that the fissures produced by earthquakes are very instruct-
ive as regards the formation of veins and the. phenomenon of dis-
location, the more recent vein displacing the older formations. But
long before Phillips (in his "Theorie der Gauge," 1791), Werner
showed the comparative ages of the displacing penetr9,ting vein and
of the disrupted penetrated rock (see British Assoc, Report, 1847,
p. 62).

t Upon the simultaneous commotion of the tertiary limestone of
Cumana and Maniquarez since the great earthquake of Cumana, on
the 14th December, 1796, see Humboldt's Relation Ilistorique, tome i.,
p. 314 ; Cosmos, vol. i., p. 212 ; and Mallet, Brit. Assoc. Report, 1850,
p. .28.

EARTHQUAKES. 1G9

black smoke (as was the case for several days in the rock of
Alvidras, during the earthquake of Lisbon, on the 1st No-
vember, 1755), flames of fire, sand, mud, and moyas, mixed
with charcoal, rise from fissures at a distance from all vol-
canoes. The acute geognosist, Abich, has proved the con-
nection which exists in the Persian Ghilan between the
thermal springs of Sarcin (5051 feet), on the road from Ar-
debil 4o Tabriz, and the earthquakes which frequently visit
the elevated districts in every second year. In October,
1848, an undulatory movement of the earth, which lasted
for a whole hour, compelled the inhabitants of Ardebil t©
abandon the town ; and the temperature of the springs,
which is between 44° and 46° C. (=111° — 115° F.), rose
immediately to a most painful scalding heat, and continued
so for a whole month.* As Abich says, nowhere, perhaps,
upon the face of the earth is "the intimate connection of
fissure-producing earthquakes, with the phenomena of mud
volcanoes, of salses, of combustible gases penetrating through
the perforated soil, and of petroleum springs, more distinctly
expressed or more clearly recognizable, than in the south-
eastern extremity of the Caucasus, between Schemacha,
Baku, and Sallian. It is the part of the great Aralo-Cas-
pian basin, in which the earth is most frequently shaken." f
I was myself struck with the remarkable fact that in North-
ern Asia the circle of commotion, the centre of which ap-
pears to be in the vicinity of Lake Baikal, extends west-
ward only to the eastern borders of the Eussian Altai, as
far as the silver mines of Riddersk, the trachytic rock of
Kruglaia Sopka, and the hot springs of Rachmanowka and
Arachan, but not to the Ural chain. Further, toward the

* Abich, on Daghestan, Schagdagh, and Ghilan, in Poggend., An-
nalen, bd. Ixxvi., 1849, p. 157. The salt spring in a well near Sas-
scndorf, in Westphalia (in the district of Amsberg), also increased
about I2 per cent, in amount of saline matter, in consequence of tlie
widely-extended earthquake of the 29th July, 1846, the centre of
commotion of which is placed at St. Goar, on the Rhine ; this was
probably because other fissures of supply had opened (Noggerath, Das
Erdbeben im Rheingebiete vom 29 Juli, 1846, p. 14). According to
Cliarpentier's observation, the tempei*ature of the sulphureous spring
of Lavey (above St. Maurice, on the bank of the Rhone) rose from
87" -8 to 97°-3 F. during the Swiss earthquake of the 25th August,
1851.

t At Schemacha (elevation 2393 feet), one of the numerous mete-
orological stations founded by Prince Wftronzow, in the Caucasus,
under Abich 's directions, eighteen earthquakes were recorded by the
observer in the journal in 1848 alone.
Vol. v.— H

170 • ' COSMOS.

south, on the other side of the parallel of 45° N.^ in the
chain of the Thianschan (IVIountains of Heaven), there ap-
pears a zone of volcanic activity directed from east to west,,
with every kind of manifestation. It extends not only from
the fire district (Ho-tscheu) in Turfan, through the small
chain of Asferah to Baku, and thence over Ararat into Asia
Minor ; but it is believed that it may be traced, oscillating
between the parallels of 38° and 40° N.,. through thg vol-
canic basin of the Mediterranean as far as Lisbon and the
Azores. I have elsewhere* treated in detail of this import-
ant subject of volcanic geography. In Greece also, which
has suffered from earthquakes more than any other part of
Europe (Curtius, PeloponnesoSy i., s. 42-46), it appears that
an immense number of thermal springs, some still flowing,,

* See Asie Centrale, tome i., p. 324-329, and tome ii., p. 108-120;
and especially my Carte des Montagues et Volcans de tAsie, compared
with the geognostic maps of the Caucasus, and of the plateau of Ar-
menia by Abich, and the map of Asia Minor (Argaens) by Peter
Tschichatschef, 1853 (Rose, Reise nachdem Ural, Altai, und Kaspischem
Meere, bd. ii., p. 576 and 597). In Asie Centrale we find: "From
Tourfan, situated upon the southern slope of the Thianchan, to the
Archipelago of the Azores, there are 120 degrees of longitude. This
is probably the longest and most regular band of volcanic reactions, os-
cillating slightly between 38° and 40° of latitude, which exists upon
the face of the earth ; it greatly surpasses in extent the volcanic band
of the Cordillera of the Andes, in South America. I insist the more
upon this singular line of ridges, of elevations, of fissures, and of
propagations of commotions, which comprises a third of the circum-
fei^ence of a parallel of latitude, because some small accidents of sur-
face, the unequal elevation and the breadth of the ridges, or linear
elevations, as well as the interruption caused by the sea-basins (Aralo-
Caspian, Mediterranean, and Atlantic basins), tend to mark ,the gi*eat
features of the geological constitution of the globe. (This bold sketch
of a regularly prolonged line of commotion by no means excludes
other lines in the direction of which the movements may also be
propagated.)" As the city of Khotan and the district south of the
Thianschan has been the most ancient and celebrated seat of Bud-
dhism, the Buddhistic literature was occupied very early and earnestly
with the causes of earthquakes (see Foe-koue-ki, ou Relation des Roy-
mmes Bouddiques, translated by M. Abel Remusat, p. 217). By the
followers of Sakhyamuni eight of these causes are adduced, among
which a revolving wheel of steel, hung with reliques ('sarira, signify-
ing body in Sanscrit), plays a principal part — a mechanical explana-
tion of a dynamic phenomenon, scarcely more absurd than many of
our geological and magnetic myths, which have but recently become
antiquated ! According to a statement of Klaproth's, priests, and es-
pecially begging monks (Bhikchous), have the power of causing the
earth to tremble and of setting the subterranean wheel in motion.
The travels of Fahian, the author of the Foe-koue-ki, date about the
commencement of the fifth century.

EARTHQUAKES. 171

others already lost, have broken out with earth-shocks. A
similar thermic connection is indicated in the remarkable
book of Johannes Lydus upon earthquakes {De Ostentis, cap.
liv., p. 189, Hase). The great natural phenomenon of the
destruction of Helice and Bura, in Achaia (373 B.C. ; Cos-
mos, vol. iv., p. 188), gave rise in an especial manner to hy-
potheses regarding the causal connection of volcanic activ-
ity. With Aristotle originated the curious theory of the
force of the winds collecting in the cavities of the depths of
the earth {Meteor., ii., p. 368). By the part which they
have taken in the early destruction of the monuments of the
most flourishing period of the arts, the unhappy frequency
of earthquakes in Greece and Southern Italy has exercised
the most pernicious influence upon all the studies which have
been directed to the evolution of the Greek and Roman civ-
ilization at various epochs. Egyptian monuments also, for
example that of a colossal Memnon (27 years B.C.), have
suffered from earthquakes, which, as Letronne has proved,
have been by no means so rare as was supposed in the val-
ley of the Nile {Les Statues Vocales de Memnon, 1833, p.
23-27, 255).

The physical changes here referred to, as induced by earth-
quakes by the production of fissures, render it the more re-
markable that so many warm mineral springs retain their
composition and temperature unchanged for centuries, and
therefore must flow from fissures 'which appear to have un-
dergone no alteration either vertically or laterally. The
establishment of communications with higher strata would
have produced a diminution, and that with lower ones an
increase of heat.

When the great eruption of the volcano of Conseguina (in
Nicaragua) took place, on the 23d of January, 1835, the
subterranean noise* {los ruidos suhterraneos) was heard at the
same time on the island of Jamaica and on the plateau of
Bogota, 8740 feet above the sea, at a greater distance than
from Algiers to London. I have also elsewhere observed,
that in the eruptions of the volcano on the island of Saint
Vincent, on the 30th of April, 1812, at two o'clock in the
morning, a noise like the report of cannons was heard with-
out any sensible concussion of the earth over a space of
160,000 geographical square miles.j It is very remarkable

* Acosta, Viajes cientijicos a los Andes ecuatoriales, 1849, p. 56.
t Cosmos, \o\. i., p. 208-210; Humboldt, Relation Historique, t. iv.,
chap. 14, p. 31-38. Some sagacious theoretical observations by Mai-

172 COSMOS.

that when earthquakes are combined with noises, which is
by no means constantly the case, the strength of the latter
does not at all increase in proportion to that of the former.
The most singular and mysterious phenomenon of subter-
ranean sound is undoubtedly that of the bramidos de Gua-
naxuato, which lasted from the 9th of January to the middle
of February, 1784, regarding which I was the first to col-
lect trustworthy details from the lips of living witnesses,
and from official records {Cosmos, vol. i., p. 209).

The rapidity of the propagation of the earthquake upon
the surface of the earth must, from its nature, be modified in
many ways by the variable densities of the solid rocky strata
(granite and gneiss, basalt and trachytic porphyry, Jurassic
limestone and gypsum), as well as by that of the alluvial
soil, through which the wave of commotion passes. It
would, however, be desirable to ascertain once for all with
certainty what are the extreme limits between which the
velocities vary. It is probable that the more violent com-
motions by no means always possess the greatest velocity.
The measurements, moreover, do not always relate to the
same direction which the waves of commotion have followed.
Exact mathematical determinations are much wanted, and
it is only at a very recent period that a result has been ob-
tained with great exactitude and care from the Rhenish
earthquake of the 29th of July, 1846,. by Julius Schmidt,
assistant at the Observatory of Bonn. In the earthquake
just mentioned the velocity of propagation was 14,956 geo-
graphical miles in a minute, that is, 1466 feet in the second.
This velocity certainly exceeds that of the waves of sound in
the air ; but if the propagation of sound in water is at the
rate of 5016 feet, as stated by Colladon and Sturm, and in
cast-iron tubes 11,393 feet, according to Biot, the result
found for the earthquake appears very weak. For the
earthquake of Lisbon, on the 1st of November, 1755, Schmidt
(working from less accurate data) found the velocity between
the coasts of Portugal and Holstein to be more than five
times as great as that observed on the Rhine, on the 29th of
July, 1846. Thus, for Lisbon and Gliickstadt (a distance

let upon sonorous waves in the earth and sonorous wares in the air
occur in the Brit. Assoc. Report, 1850, p. 41-46, and in the Admiral-
ty Manual, 1849, p. 201 and 217. The animals which in tropical
countries are disquieted by the slightest commotions of the earth
sooner than man are, according to my experience, fowls, pigs, dogs,
asses, and crocodiles (Caymans) ; the latter suddenly quit the bottom
of the rivers.

EARTHQUAKES. 173

of 1348 English miles), the velocity obtained was 89*26
miles in a minute, or 7953 feet in a second ; which, how-
ever, is still 3438 feet less than in cast iron.*

Concussions of the earth and sudden eruptions of fire from
volcanoes which have been long in repose, whether these
merely emit cinders, or, like intermittent springs, pour forth
fused, fluid earths in streams of lava, have certainly a single,
common causal connection in the high temperature of the in-
terior of our planet ; but one of these phenomena is usually
manifested quite independently of the other. Thus, in the
chain of the Andes in its linear extension, violent earth-
quakes shake districts in which unextinguished, often indeed
active, volcanoes exist without the latter being perceptibly
excited. During the great catastrophe of Riobamba, the
volcanoes of Tunguraliua and Cotopaxi — the fbrmer in the
immediate vicinity, and the latter rather farther off — re-
mained perfectly quiet. On the other hand, volcanoes have
presented violent and long-continued eruptions without any
earthquake being perceived in their vicinity, either previous-
ly or simultaneously. In fact, the most destructive earth-
quakes recorded in history, and which have passed through
many thousand square miles, if we may judge from what is
observable at the surface, stand in no connection with the

* Julius Schmidt, in No.!Ji;geratli, Ueher das Erdhehen vom 29 JuU^
18-1:6, s. 28-37. With the velocity stated in the text, the earthquake
of Lisbon would have passed round the equatorial circumference of
the earth in about 45 hours. Michell {Phil. Transact, vol. i., pt. ii.,
p. 572) found for the same earthquake of the 1st November, 1755, a
velocity of only 50 English miles in a minute — that is, instead of 7956,
only 4444 feet in a second. The inexactitude of the older observa-
tions and difference in the direction of propagation may conduce to
this result. Upon the connection of Neptune Avith earthquakes, at
which I have glanced in the text (p. 181), a passage of Proclus, in the
commentary to Plato's Cratylus, throws a remarkable light. " The
middle one of the three deities, Poseidon, is the cause of movement
in all things, even in the immovable. As the originator of movement
he is called 'Evvoacyaiog ; to him, of those who shared the empire of
Saturn, fell the middle lot, the easily-moved sea" (Creuzer, SymhoUh
und Mi/thologie, th. iii,, 1842, s. 260). As the Atlantis of Solon and
the Lyctonia, which, according to my idea, was nearly allied to it, are
geological myths, both the lands destroyed by earthquakes are regard-
ed as standing under the dominion of Neptune, and set in opposition
to the Saturnian continents. According to Herodotus (lib, ii., c. 43
et 50), Neptune was a Libyan deity, and unknown in Egypt. Upon
these circumstances — the disappearance of the Libyan lake Tritonis
by earthquake — and the idea of the great rarity of earthquakes in tha
valley of the Nile, see my Examen Critique de la Gcographie, t. i., p.
171 and 179.

174 COSMOS.

activity of volcanoes. They have lately been called Plu-
tonic, in opposition to the true Volcanic earthquakes, which
are usually limited to smaller districts. In respect of the
more general views of Vulcanicity, this nomenclature is,
however, inadmissible. By far the greater part of the earth-
quakes upon our planet must be called Plutonic.

That which is capable of exciting earth-shocks is every
where under our feet; and the consideration that nearly
three fourths of the earth's surface are covered by the sea
(with the exception of some scattered islands), and without
any permanent communication between the interior and the
atmosphere, that is to say, without active volcanoes, contra-
dicts the erroneous but widely disseminated belief that all
earthquakes are to be ascribed to the eruption of some dis-
tant volcano: Earthquakes on continents are certainly prop-
agated along the sea-bottom from the shores, and give rise to
the terrible sea-waves, of which such memorable examples
were furnished by the earthquakes of Lisbon, Callao de Lima,
and Chili. When, on the contrary, the earthquakes start
from the sea-bottom itself, from the realm of Poseidon, the
earth-shaker (aeialxOcov, tctV7jOLx6(j^v), and are not accompa-
nied by upheaval of islands (as in the ephemeral existence
of the island of Sabrina or Julia), an unusual rolling and
swelling of the waves may still be observed at points where
the navigator would feel no shock. The inhabitants of the
desert Peruvian coasts have often called my attention to a
phenomenon of this kind. Even in the harbor of Callao, and
near the opposite island of San Lorenzo, I have seen wave
upon wave suddenly rising up in the course of a few hours
to more than 10 or 15 feet, in perfectly still nights, and in
this otherwise so thoroughly peaceful part of the South Sea.
That such a phenomenon might have been the consequence
of a storm which had raged far off upon the open sea, was
by no means to be supposed in these latitudes.

To commence from those commotions^ which are limited
to the smallest space, and evidently owe their origin to the
activity of a volcano, I may mention, in the first place, how,
when sitting at night in the crater of Vesuvius, at the foot
of a small cone of eruption, with my chronometer in my
hand (this was after the great earthquake of Naples, on the
2Cth of July, 1805, and the eruption of lava which took
place seventeen days subsequently), I felt a concussion of the
soil of the crater very regularly every 20 or 25 seconds, im-
mediately before each eruption of red-hot cinders. The cin-

EARTHQUAKES. 175

ders, thrown up to a height of 50 — 60 feet, fell back partly
into the orifice of eruption, while a part of them covered the
walls of the cone. The regularity of such a phenomenon
renders its observation free from danger. The constantly-
repeated small earthquake was quite imperceptible beyond
the crater — even in the Atrio del Cavallo and in the Her-
mitage Del Salvatore. The periodicity of the concussion
shows that it was dependent upon a determinate degree of
tension which the vapors must attain, to enable them to
break through the fused mass in the interior of the cone of
cindei'S. In the case just described no concussions were felt
on the declivity of the ashy cone of Vesuvius, and in an ex-
actly analogous but far grander phenomenon, on the ash-
cone of the volcano of Sangai, which rises to a height of
17,006 feet to the southeast of the city of Quito, no trem-
bling of the eartli* was felt by a very distinguished observer,
M. Wisse, when (in December, 1849) he approached within a
thousand feet of the summit and crater, although no less than
267 explosions (eruptions of cinders) were counted in an hour.

A second, and infinitely more important kind of earth-
quake, is the veiy frequent one which usually accompanies
or precedes great eruptions of volcanoes — whether the vol-
canoes, like ours in Europe, pour forth streams of lava ; or,
like Cotopaxi, Pichincha, and Tunguragua of the Andes,
only throw out calcined masses, ashes and vapors. For
earthquakes of this kind the volcanoes are especially to be
regarded as safety-valves, as indicated even by Strabo's ex-
pression concerning the fissure pouring out lava near Le-
lante, in Euboea. The earthquakes cease when the great
eruption has taken place.

Most widelyt distributed, however, are the ravages of the

* The explosions of the Sangai, or Volcan de Macas, took place on
an average every 13"*4; see Wisse, Comptes rendus de VAcad. des
Sciences, tome xxxvi., 1853, p. 720. As an example of commotions
confined within the narrowest limits, I might also have cited the re-
port of Count Larderel upon the lagoons in Tuscany, The vapors
containing boron or boracic acid give notice of their existence, and of
their approaching eruption at fissures, by shaking the surrounding
rocks (Larderel, Sur les etablissements industriels de la production
d'acide boracique en Toscane, 1852, p. 15).

t I am glad that I am able to cite an important authority in con-
firmation of the views that I have endeavored to develop in the text.
" In the Andes the oscillation of the soil, due to a volcanic eruption,
is, so to speak, local, while an earthquake, which, at all events in ap-
pearance, is not connected with any volcanic eruption, is propagated
to incredible distances. In this case it has been remarked that the

176 COSMOS.

waves of commotion, which pass sometimes through com-
pletely non-trachytic, non-volcanic countries, and sometimes
through trachytic, volcanic regions, jvithout exerting any
influence upon tlie neighboring volcanoes. This is a third
group of phenomena, and is that which most convincingly
indicates the existence of a general cause, lying in the ther-
mic nature of the interior of our planet. To this third
group also belongs the phenomenon sometimes, though rare-
ly, met with in non-volcanic lands, but little disturbed by
earthquakes, of a trembling of the soil within the most nar-
row limits, continued uninterruptedly for months together,
so as to give rise to apprehensions of an elevation and for-
mation of an active volcano. This was the case in the Pied-
montese valleys of Pelis and Clusson, as well as in the vi-
cinity of Pignerol, in April and May, 1805, and also in the
spring of 1829 in Murcia, between Orihuela and the sea-
shore, upon a space of scarcely sixteen square miles. When
the cultivated surface of JoruUo, upon the western declivity
of the plateau of Mechoacan, in the interior of Mexico, was
shaken uninterruptedly for 90 days, the volcano rose with
many thousand cones of 5 — 7 feet in height {los hornitos) sur-
rounding it, and poured forth a sliort but vast stream of
lava. In Piedmont and Spain, on the contrary, the concus-
sions of the earth gradually ceased, without the production
of any other phenomenon.

I have considered it expedient to enumerate the perfectly
distinct kinds of manifestation of the same volcanic activity
(the reaction of the interior of the earth upon its surface), in
order to guide the observer, and bring together materials
which may lead to fruitful results with regard to the causal
connection of the phenomena. Sometimes the volcanic ac-
tivity embraces at one time or within short periods so large
a portion of the earth, that the commotions of the soil excited
may be ascribed simultaneously to many causes related to
each other. The years 1796 and 1811 present particularly
memorable examples* of such a grouping of the phenomena.

shocks followed in preference the direction of the chains of mountains,
and were principally felt in Alpine districts. The frequency of the
movements in the soil of the Andes, and the little coincidence ob-
served between these movements and volcanic eruptions, must neces-
sarily lead us to suppose that in most cases they are occasioned by a
cause independent of volcanoes'''' (Boussingault, Annales de Chiinie et de
Physique, t. Iviii., 1835, p. 83).

* The great phenomena of 1796 and 1797, and 1811 and 1812, oc-
curi^d in the following order:

THERMAL SPRINGS. 177

b. Thermal Springs.

(Amplification of the Eepresentation of Nature, Cosmos, vol. i., p.
219-224.)

As a consequence of th'e vital activity of the interior of
our planet, evidenced in irregularly repeated and often fear-
fully destructive phenomena, we have described the earth-
quake. In this there prevails a volcanic power, which in

27th of September, 179G. Eruption of the volcano of the island of
Guadaloupe, in the Leeward Islands, after a repose of many
years ;

November, 1796, The volcano on the plateau of Pasto, between
the small rivers Guaytara and Juanambu, became ignited and be-
gan to smoke permanently ;

14th of December, 1796. Earthquake and destruction of the city
of Cumana ;

4th of February, 1797. Earthquake and destruction of Eiobamba.
On the same morning the columns of smoke of the volcano of
Pasto, at a distance of at least 200 geographical miles from Eio-
bamba, disappeared suddenly, and never reappeared; no com-
motion was felt in its vicinity.

30th of January, 1811. First appearance of the island of Sabrina,
in the group of the Azores, near the island of St. Michael. The
elevation preceded the eruption of fire, as in the case of the little
Kameni (Santorin) and that of the volcano of Jorullo. After an
eruption of cinders, lasting for six days, the island rose to a
height of 320 feet above the surface of the sea. It was the third
appearance and disappearance of the island nearly at the same
point, at intervals of 91 and 92 years.

May, 1811. More than 200 shocks of earthquake on ttie island of
St. Vincent up to April, 1812.

December, 1811. Innumerable shocks in the river-valleys of the

Ohio, Mississippi, and Arkansas, up to 1813. Between New

Madrid, Little Prairie, and La Saline, to the north of Cincin-

■ nati, the earthquakes occurred almost every hour for months

together.

December, 1811. A single shock in Caraccas.

26th of March, 1812. Earthquake and destruction of the town of
Caraccas. The circle of commotion extended over Santa Marta,
the town of Honda, and the elevated plateau of Bogota, to a dis-
tance of 540 miles from Caraccas. The motion continued until
the middle of the year 1813.

30th of April, 1812. Eruption of the volcano of St. Vincent; and
on the same day, about two o'clock in the morning, a fearful sub-
terranean noise, like the roar of artillery, was heard at the same
time and with equal distinctness on the shores of Caraccas, in
the Llanos of Calabazo and of the Rio Apure, without being ac-
companied by any concussion of the earth (see ante, p. 171). The
subterranean noise was also heard upon the island of St. Vin-
cent, but (and this is very remarkable) it was stronger at some
distance upon the sea.

IT 2

178 COSMOS.

its essential nature only acts dynamically, producing move-
ment and commotion ; but when it is favored at particular
points by the fulfillment of subsidiary conditions, it is capa-
ble of bringing to the surface material products, although not
of generating them like true volcanoes. Just as water, va-
pors, petroleum, mixtures of gases, or pasty masses (mud and
moya) are thrown out, through fissures suddenly opened in
earthquakes sometimes of short duration, so do liquid and
aerial fluids flow permanently from the bosom of the earth
through the universally diffused net-work of communicating
fissures. The brief and impetuous eruptive phenomena are
here placed beside the great peaceful spring-system of the
crust of the earth, which beneficently refreshes and supports
organic life. For thousands of years it returns to organized
nature the moisture which has been drawn from the atmos-
phere by falling rain. Analogous phenomena are mutually
illustrative in the eternal economy of nature ; and wherever
an attempt is made at the generalization of ideas, the inti-
mate concatenation of that which is recognized as allied
must not remain unnoticed.

The widely-disseminated classification of springs into cold
and hot, which appears so natural in ordinaiy conversation,
has but a very indefinite foundation when reduced to nu-
merical data of temperature. If the temperature of springs
be compared with the internal heat of man (found, with ther-
mo-electrical apparatus, to be 98° — 98°'6 F., according to
Brechet and Becquerel), the degree of the thermometer at
which a fluid is called cold, warm, or hot, when in contact
with parts of the human body, is very different according to
individual sensations. No absolute degree of temperature
can be established above which a spring should be designated
warm. The proposition to call a spring cold in any climatic
zone, when its average annual temperature does not exceed
the average annual temperature of the air in the same zone,
at least presents a scientific exactitude, by affording a com-
parison of definite numbers. It has the advantage of lead-
ing to considerations upon the different origin of springs, as
the ascertained agreement of their temperature with the an-
nual temperature of the air is recognized directly in unchange-
able springs ; and in changeable ones, as iias been shown by
Wahlenberg and Erman the elder, in the averages of the sum-
mer and winter months. But in accordance with the crite-
rion here indicated, a spring in one zone must be denomin-
ated warm, which hardly attains the seventh or eighth part

THERMAL SPRINGS. 179

of temperature of one which in another zone, near the equa-
tor, will be called cold. I may mention the differences be-
tween the average temperature of St. Petersburg (38°-12 F.)
and of the shores of the Orinoco. The purest spring water
which I drank in the vicinity of the cataracts of Atures*
and Maypures (81°-14 F.) or in the forest of Atabapo, had
a temperature of more than 79° F. ; even the temperature
of the great rivers in tropical South America corresponds
with the high degrees of heat of such coldf springs. •

* Humboldt, Voyage aux Regions JSquinoxiales, t. ii., p. 376.

t For the sake of comparing the temperature of springs where they
break forth directly from the earth, with that of large rivers flowing
through open channels, I here bring together the following average
numbers from my journals :

Eio Apure, lat. 71°; temperature, 81°.

Orinoco, between 4° and 8° of latitude; 81°-5— 85°-3.

Springs in the forest, near the cataract of Maypures, breaking forth
from the granite, 82°.

Cassiquiare, the branch of the Upper Orinoco, which forms the union
with the Amazon; only 75°-7.

Rio Negro, above San Carlos (scarcely 1° 53' to the north of the
equator) ; only 74° '8.

Rio Atabapo, 79° -2 (lat. 3° 50').

Orinoco, near the entrance of the Atabapo, 82°.

Rio Grande de la Magdalena (lat. 5° 12' to 9° 56'), 79° 9'.

Amazon, 5"^ 31' S. latitude, opposite to the Pongo of Rentema
(Provincia Jaen de Bracaraoros), scarcely 1300 feet above the
South Sea, only 72° -5.

The great mass of water of the Orinoco consequently approaches
the average temperature of the air of the vicinity. During great in-
undations of the savannas, the yellowish-brown waters, which smell
of sulphureted hydrogen, acquire a temperature of 92°-8 ; this I found
to be the temperature in the Lagartero, to the east of Guayaquil, which
swarmed with crocodiles. The soil there becomes heated, as in shallow-
rivers, by the warmth produced in it by the sun's rays falling upon it.
With regard to the multifarious causes of the low temperature of the
water of the Rio Negro, which is of a coffee-hrown color by reflected
light, and of the white waters of the Cassiquiare (a constantly clouded
sky, the quantity of rain, the evaporation from the dense forests, and
the want of hot sandy tracts upon the banks), see my river voyage, in
the Relation Historique, t. ii., p. 463 and 509. In the Rio Guanca-
bamba or Chamaya, which falls into the Amazon, near the Pongo de
Rentema, I found the temperature of the water to be only 67°*6, as its
waters come with prodigious swiftness from the elevated Lake Simi-
cocha, on the Cordillera. On my voyage of 52 days up the River Mag-
dalena, from Mahates to Honda, I perceived most distinctly, from
numerous observations, that a rise in the level of the water was indi-
cated for hours previously by a diminution of the temperature of the
river. The refrigeration of the stream occurred before the cold mount-
ain waters from the Paramos, near the source, came down. Heat and
water move, so to speak, in opposite directions and with very unequal

180 COSMOS.

The breaking out of springs, effected by multifarious
causes of pressure and by the communication of fissures con-
taining water, is such a universal phenomenon of the sur-
face of the earth, that waters flow forth at some points fronn
the most elevated mountain strata, and at others from the
bottom of the sea. In the first quarter of this century nu-
merous results were collected by Leopold von Buch, Wahlen-
berg, and myself, Avith regard to the temperature of springs
and the diffusion of heat in the interior of the earth in
both hemispheres, from 12° S. latitude to 71° N * The
springs which have an unchangeable temperature were care-
fully separated from those which vary with the seasons ;
and Leopold von Buch ascertained the powerful influence of
the distribution of rain in the course of the year ; that is
to say, the influence of the proportion between the relative
abundance of winter and summer rain upon the temperature
of the variable springs, which, as regards number, are the
most widely distributed. More recently! some very ingen-

velocities. When the water near Badillas rose suddenly, the tempera-
ture fell long before from 80°*6 to 74:°-'d. As, during the night, when
one is established iiix)n a low sandy islet, or upon the bank, with bag
and baggage, a rapid rise of the river may be dangerous, the discov-
ery of a prognostic of the approaching rise (the avenida) is of some
importance.

* Leopold von Buch, Physioalische Beschreihung der canarischen In-
seln, s. 8 ; Foggendorf, Annalen, bd. xii., s. 403 ; Bihliotheque Britan-
niqice, Sciences et Arts, t. xix., 1802, p. 2G3 ; Wahlenberg, De Veget. et
dim. in Helvetia Septentrionali Observatis, p. Ixxviii. and Ixxxiv. ;
Wahlenberg, Flora Carpathica, p. xciv., and in Gilbert's Annalen, bd.
xli., s. 115 ; Humboldt, in the 3Ie?n. de la Soc. d' Arcueil, t. iii. (1817),
p. 699.

t De Gasparin, in the Bihliotheque Univ. Sciences et Arts, t. xxxviii.,
1828, p. 51, 113, and 261; Mem de la Soc. Centrale d' Agricidture,
1820, p. 178 ; Schouw, Tableau du Cliinat et de la Vegetation de Vltalie,
vol. i,, 1839, p. 133-195 ; Thurmann, Sar la temperature des sources dc
la chaine du Jura, comparce a celle des sources de la plaine Suisse, des
Alpes et des Vosges, in the Annuaire Meteorologique de la France, ] 850,
p. 258-268. As regards the frequency of the summer and autumn
rains, De Gasparin divides Europe into two strongly-contrasted regions.
Valuable materials are contained in Kamtz, Lehrhuch der Meteor o log ie,
bd. i., s. 448-506. According to Dove (Poggend., Annalen, bd. xxxv., s.
376) in Italy, " at places to the north of which a chain of mountains
is situated, the maxima of the curves of monthly quantities of rain
fall in March and September ; and where the mountains lie to the
south, in April and October." The totality of the proportions of rain
in the temperate zones may be comprehended under the following
general point of view : " The period of winter rain in the borders of
the tropics constantly divides, the farther we depart from these, into
two maxima united by slighter falls, and these again unite into a

THERMAL SPRINGS. 181

ious comparative observations by De Gasparin, Schouw, and
Thurmann have thrown considerable light, in a geographical
and hjpsoraetrical point of view, in accordance with latitude
and elevation, upon this influence. Wahlenberg asserted
that in very high latitudes the average temperature of vari-
able springs is rather higher than that of the atmosphere ;
he sought the cause of this, not in the dryness of a very cold
atmosphere and in the less abundant winter rain caused
thereby, but in the snowy covering diminishing the radiation
of heat from the soil. In those parts of the plain of Noa(Ji-
crn Asia in which a perpetual icy stratum, or at least a
frozen alluvial soil, mixed with fragments of ice, is found at
a depth of a* few feet,* the temperature of springs can only
be employed with great caution for the investigation of
Kupffer's important theory of the isogeotherraal lines. A
two-fold radiation of heat is then produced in the upper
stratum of the earth : one upward toward the atmosphere,
and another downward toward the icy stratum. A long se-
ries of valuable observations made by my friend and com-
panion, Gustav Kose, during our Siberian expedition in the
heat of summer (often in springs still surrounded by ice), be-
tween the Irtysch, the Obi, and the Caspian Sea, revealed a
great complication of local disturbances. Those which pre-
sent themselves from perfectly different causes in the tropic-
al zone, in places where mountain springs burst forth upon
vast elevated plateaux, eight or ten thousand feet above the
sea (Micuipampa, Quito, Bogota), or in narrow, isolated
mountain peaks many thousand feet higher, not only include
a far greater part of the surface of the earth, Ijut also lead
to the consideration of analogous thermic conditions in the
mountainous countries of the temperate zones.

In this important subject it is above all things necessary
to separate the cycle of actual observations from the theoret-
ical conclusions which are founded upon them. What we
seek, expressed in the most general way, is of a triple nature
— the distribution of heat in the crust of the earth which is
accessible to us, in the aqueous covering (the ocean) and in
the atmosphere. In the two envelopes of the body of the
iearth, the liquid and gaseous, an opposite alteration of tem-

summer maximum in Germany; where, therefore, a temporary want
of rain ceases altogether." See the section "Geothermik," in the
excellent Lehrhuch de?- Geognosie, by Naumann, bd. i. (1850), s.
41-73.

* See above, p. 47.

182

COSMOS.

perature (diminution and increase in the superposed strata)
prevails in a vertical direction. In the solid parts of the
body of the earth the temperature increases with the depth ;
the alteration is in the same direction, although in a very
different proportion, as in the aerial ocean, the shallows and
rocks of which are formed by the elevated plateaux and mul-
tiform mountain peaks. We are most exactly acquainted
by direct experiments with the distribution of heat in the
atmosphere — geographically by local determination in lati-
tiMe and longitude, and in accordance with hypsometric re-
lations in proportion to the vertical elevation above the sur-
face of the sea ; but in both cases almost exclusively in close
contact with the solid and fluid parts of the surface of our
planet. Scientific and systematically arranged investigations
by aerostatic voyages in the free aerial ocean, beyond the
near action of the earth, are still very rare, and therefore
but little adapted to furnish the numerical data of average
conditions which are so necessary. Upon the decrease of
heat in the depths of the ocean observations are not want-
ing; but currents, which bring in water of different lati-
tudes, depths, and densities, prevent the attainment of gen-
eral results, almost to a greater extent than currents in the
atmosphere. We have here touched preliminarily upon the
thermic conditions of the envelopes of our planet, which will
be treated of in detail hereafter, in order to consider the in-
fluence of the vertical distribution of heat in the solid crust
of the earth, and the system of the geo-isothermic lines, not
in too isolated a condition, but as a part of the all-penetrat-
ing motion of heat, a truly cosmical activity.

Instructive as are, in many respects, observations upon
the unequal diminution of temperature of springs which do
not vary with the seasons as the height of their point of
emergence increases — still the local law of such a diminish-
ing temperature of springs can not be regarded, as is often
done, as a universal geothermic law. If we were certain
that waters flowed unmixed in a horizontal stratum of great
extent, we might certainly suppose that they have gradually
acquired the temperature of the solid ground, but in the
great net-work of fissures of elevated masses this case can
rarely occur. Colder and more elevated waters mix w^ith
the lower ones. Our mining operations, inc(5nsiderable as
may be the depth to which they attain, are very instructive
in this respect ; but we should only obtain a direct knowl-
edge of the isogeothermal lines if thermometers were buried,

THERMAL SPRINGS. 183

according to Boussingault's method,* to a depth below that
affected by the influences of the changes of temperature of
the neighboring atmosphere, and at very different elevations
above the sea. From the forty-fifth degree of latitude to the
parts of the tropical regions in the vicinity of the equator,
the depth at which the stratum of invariable temperature
commences diminishes from 60 to 1^ or 2 feet. Burying
the geothermometer at a small depth, in order to obtain a
knowledge of the average temperature of the earth, is there-
fore readily practicable only between the tropics or in the
sub-tropical zone. The excellent expedient of Artesian
wells, which have indicated an increase of heat of 1° F. for
every 54 to 58 feet in absolute depths of from 745 to 2345
feet, has hitherto only been afforded to the physicist in dis-
tricts not much more than 1600 feet above the level of the
sea.f I have visited silver mines in the chain of the Andes,
6° 45^ south of the equator, at an elevation of nearly 13,200
feet, and found the temperature of the water penetrating
through the fissures of the limestone to be 52°.*3 F.j: The
waters which were heated in the baths of the Inca Tupac
Yupanqui, upon the ridge of the Andes {Paso del Assiiay),
probably come from springs of the Ladera de Cadlud, where
I have traced their course, near which the old Peruvian
causeway also ran, barometrically to an elevation of 15,526
feet (almost that of Mont Blanc).§ These are the highest
points at which I could observe spring water in South Amer-
ica. In Europe the brothers Schlagintweit have found gal-
lery-water in the gold mine in the Eastern Alps at a height
of 9442 feet, and fbund that the temperature of small springs
near the opening of the gallery is only 33°*4 F.,|| at a dis-
tance from any snow or glacier ice. The highest limits of
springs are very different according to geographical latitude,
the elevation of the snow line and the relation of the highest
peaks to the mountain ridges and plateaux.

If the radius of our planet were to be increased by the
height of the Himalaya at the Kintschindjunga, and therefore
uniformly over the whole surface by 28,175 feet (4-34 En-
glish miles), with this small increase of only ^-ou^^ ^^ *^^®

* See Cosmos, vol. i., p. 221, and vol. v., p. 42.

f See above, p. 39.

j Mina de Gaudalupe, one of the Minas de Chota, I. c. sup., p. 41.

§ Humboldt, Views of Nature, p. 393.

11 Mine on the Great Fleuss in the Moll Valley of the Tauern ; see
Hermann and Adolph Schlagintweit, Untersuchungen uber die phjsika-
lische Geographie der Alpen, 1850, s. 242-273.

184 COSMOS.

radius, the heat in the surface, cooled by radiation, would be
(according to Fourier's analytical theory) almost the same
as it now is in the upper crust of the earth. But if individ-
ual parts of the surface raise themselves in mountain chains
and narrow peaks, like rocks upon the bottom of the aerial
ocean, a diminution of heat takes place in the interior of the
elevated strata, and this is modified by contact with strata
of air of different temperature, by the capacity for heat and
conductive power of heterogeneous kinds of rocks, by the
sun's action on the forest-clad summits and declivities, by the
greater and less radiation of the mountains in accordance
with their form (relief), their massiveness, or their conical
and pyramidal narrowness. The special elevations of the
region of clouds, the snow and ice coverings at various ele-
vations of the snow line, and the frequency of the cool cur-
rents of air coming down the steep declivities at particular
times of the day, alter the effect of the terrestrial radiation.
In proportion as the towering cones of the summits become
cooled, a weak current of heat tending toward, but never
reaching an equilibrium, sets in from below upward. The
recognition of so many factors acting upon 'the vertical dis-
tribution of heat leads to well-founded presumptions regard-
ing the connection of complicated local phenomena, but not
to direct numerical determinations. In the mountain springs
(and the higher ones, being important to the chamois-hunt-
er, are carefully sought) there so often remains the doubt
that they are mixed with waters, which by sinking down in-
troduce the colder temperature of higher strata, or by ascend-
ing introduce the warmer temperature of lower strata. From
nineteen springs observed by Wahlenberg, Kamtz draws the
conclusion that in the Alps we must rise from 9C0 to 1023
feet in order to see the temperature of the springs sink 1° C.
(1°*8 F.). A greater number of observations, selected with
more care by Hermann and Adolph Schlagintweit, in the
eastern Carinthian Alps and in the western Swiss Alps, on
the Monte Kosa, give only 767 feet. According to the great
work* of these excellent observers, " the decrease of the tem-
perature of springs is certainly somewhat more gradual than
that of the average annual temperature of the air, which in
the Alps amounts to about 320 feet for 1° F. The springs
tliere are, in general, warmer than the average temperature
of the air at the same level ; and the difference between the
temperature of the air and springs increases with the eleva-
* Monte Hosa, 1853, chap, vi., s. 212-225.

THERMAL SPRIGS. 185

tion. The temperature of the soil is not the same at equal
elevations in the entire range of the Alps as the isothermal
surfaces, which unite the points of the same average temper-
ature of springs, rise higher above the level of the sea, inde-
pcndently of the injluence of latitude, in proportion to the av-
erage convexity of the surrounding soil ; perfectly in accord-
ance with the laws of the distribution of heat in a solid body
of varying thickness, with which the relief (the mass-eleva-
tion) of the Alps may be compared.'*

In the chain of the Andes, and indeed iil those volcanic
parts of it which present the greatest elevations, the burying
of thermometers may, in particular cases, lead to deceptive
results by the influence of local circumstances. Fronsi the
opinion formerly held by me, that black, rocky ridges, visible
at a great distance, which penetrate the snowy region, are
not always indebted for their entire freedom from snow to
the steepness of their sides, but to other causes, I buried the
bulb of a thermometer only three inches deep in the sand,
which filled the fissure in a ridge on the Chimborazo at an
elevation of 18,290 feet, and therefore 3570 feet above the
summit of Mont Blanc. The thermometer permanently
showed 10°-5 F. above th|fe freezing-point, while the air was
only 4P'o F. above that point. The result of this observa-
tion is of some importance ; for even 2558 feet lower, at the
lower limit of perpetual snow of the volcano of Quito, ac-
cording to numerous observations collected by Boussingault
and myself, the average temperature of the atmosphere is
not higher than 34°-9 F. The ground temperature of 42°-5
must, therefore, be ascribed to the subterranean heat of the
doleritic mountain — I do not say of the entire mass, but to
the currents of air ascending in it from the depths. At the
foot of Chimborazo, at an elevation of 948G feet toward the
hamlet of Calpi, there is, moreover, a small crater of erup-
tion, Yana-Urcu, which, as indeed is shown by its black,
slag-like rock (augitic porphyry), appears to have been act-
ive in the middle of the fifteenth century.*

The aridity of the plain from which Chimborazo rises, and
the subterranean brook, which is heard rushing under the
volcanic hill (Yana-Urcu) just mentioned, have led Boussin-
gault and myselff at very different times to the idea that
the water which the enormous masses of snow produce daily
by melting at their lower limit sinks into the depths through

* Humboldt, Kldriere Schriften, bd. i., p. 139 and 147.
, f Humboldt, Op. cit., s. 140 and 203.

I

186 ^

COSMOS.

the fissures and chambers of the elevated volcano. These
waters perpetually produce a refrigeration in the strata
through which they run down. "Without them the whole
of the doleritic and trachytic mountains would acquire, even
at times when no near eruption is foretold, a still higher
temperature in their interior, from the volcanic source, per-
petually in action, although perhaps not lying at the same
depth in all latitudes. Thus, in the varying struggle of the
causes of heat and cold, we have to assume a constant tide
of heat upward and downward in those places where conical
solid parts ascend into the atmosphere.

As regards the area which they occupy, however, mount-
ains* and elevated peaks form a very small phenomenoii in
the relief formation of continents ; and, moreover, nearly
two thirds of the entire surface of the earth is sea-bottom
(according to the present state of geographical discovery in
the polar regions of both hemispheres, we may assume the
proportion of sea and land to be in the ratio of 8 : 3). This
is directly in contact with aqueous strata, which, being
slightly salt, and depositing themselves in accordance with
the maximum of their density (at 38° -9), possess an icy cold-
ness. Exact observations by Louz and Du Petit-Thouars
have shown that within the tropics, where the temperature
of the surface of the ocean is 78°-8 to 80°*6, water of the
temperature of 36° "5 could be drawn up from a depth of
seven or eight hundred fathoms — phenomena which prove
the existence of under currents from the polar regions. The
consequences of this constant, sub-oceanic refrigeration of by
far the greater part of the crust of the earth deserve a degree
of attention which they have not hitherto received. Rocks
and islands of small size, which project, like cones, from the
sea-bottom above the surface of the water, and narrow isth-
muses, such as Panama and Darien, washed by great oceans,
must present a distribution of heat in their rocky strata dif-
ferent from that of parts of equal circumference and mass in
the interior of continents. In a very elevated mountainous
island, the submarine part is in contact with a fluid which
has an increasing temperature from below upward. But as
the strata pass into the atmosphere unmoistened by the sea,
they come in contact, under the influence of insolation and
free radiation of dark heat, with a gaseous fluid in which
the temperature diminislies with the elevation. Similar
thermic conditions of opposed decrease and increase of tem-
perature in a vertical direction are repeated between t^o

THERMAL SPRINGS. 187

large inland seas, the Caspian and Aral Sea, in the narrow
Ust-Urt, which separates them from each other. In order,
however, to clear up such complicated phenomena, the only
means to be employed are such as borings of great depth,
which lead directly to the knowledge of the internal heat of
the earth, and not merely observations of springs, or of the
temperature of the air in caves, which give just as uncertain
results as the air in the galleries and chambers of mines.

When a low plain is compared with a mountain chain or
plateau, rising boldly to a height of many thousand feet, the
law of the increase and diminution of temperature does not
depend simply upon the relative vertical elevation of two
points on the earth's surface (in the plain and on the sum-
mit of the mountain). If we should calculate from the sup-
position of a definite proportion in the change of tempera-
ture in a certain number of feet from the plain upward to
the summit, or from the summit downward to the stratum
in the interior of the mountain mass which lies at the same
level as the surface of the plain, we should in the one case
find the summit too cold, and in the other the stratum in
the interior of the mountain far too hot. The distribution
of heat in a gradually sloping mountain (an undulation of
the surface of the earth) is dependent, as has already been
remarked, upon form, mass, and conductibility ; upon inso-
lation, and radiation of heat toward the clear or cloudy
strata of the atmosphere ; and upon the contact and play of
the ascending and descending currents of air. According
to such assumptions, mountain springs must be very abund-
ant, even at very moderate elevations of four or five thou-
sand feet, where the temperature would exceed the average
temperature of the locality by 72 or 90 degrees; and how
would it be at the foot of mountains under the tropics,
which at an elevation of 14,900 feet are still free from per-
petual snow, and often exhibit no volcanic rock, but only
gneiss and mica schist!* The great mathematician, Fou-
rier, who had been much interested in the fact of the vol-
cano of Jorullo having been upheaved, in a plain where for
many thousands of square miles around no unusual terres-
trial heat was to be detected, occupied himself, at my re-

* I differ here from the opinion of one of my best friends, a pliys-
icist who has done excellent service as regards the distribution of tel-
luric heat. See, "upon the cause of the hot springs of Leuck and
Warmbrum," Bischof, Lehrhuch der Chemischen und Physikalischen Ge-
o%?e,bd.i.,s. 127-133.

188 COSMOS.

quest, in the very year before his death, with theoretical in-
vestigations upon the question, how in the elevation of
mountains and alterations in the surface of the earth, the
isothermal surfaces are brought into equilibrium with the
new form of the soil. The lateral radiation from strata
which lie in the same level, but are differently covered,
plays in this case a more important part than the direction
(inclination) of the cleavage planes of the rock, in cases
where stratification is observable.

I have already elsewhere mentioned* how the hot springs
in the environs of ancient Carthage, probably the thermal
springs of Pertusa {aquce calidce of Hammam-el-Enf), led
Bishop Patricius, the martyr, to the correct view of the
cause of the higher or lower temperature of the bubbling
waters. When the Proconsul Julius tried to confuse the
accused bishop by the mocking question, " Quo auctore fer-
vens haic aqua tantum ebulliatf Patricius set forth his the-
ory of the central heat, " which causes the fiery eruptions of
*JEtna and Vesuvius, and communicates more and more heat

* With regard to this passage, discovered by Bureau de la Malle,
see Cos7)ios, vol. i., p. 223, 224. "Est autem," says Saint Patricius,
" et supra firmamentum cash, et suhter terram ignis atque aqua ; et
quae supra terram est aqua, coacta in unum, appellationcm marium:
quffi vero infra, abyssorum suscepit; ex quibus ad generis humani
iisus in terram velut siphones quidam emittuntur et scaturiunt. Ex
iisdera quoque et thermae exsistunt : quarum quae ab igne absunt
longius, provida boni Dei erga nos mente, frigidiores ; quae vero/»ro-
plus admodum,7%ruewies fluunt. In quibusdam etiam locis et tepidoe
aquse reperiuntur, pro ut majore ab igne intervallo sunt disjunctae."
So run the words in the collection : Acta Primorum Martyrum, opera
et studio Theodorici Ruinart, ed. 2, Amstelaedami, 1713 fol., p. 655.
According to another report (A. S. Mazochii, in vctus murmoreum
sanctce. NeapoUtance Ecclesice Kalendarium commentarhis, vol. ii., Neap.
] 744, 4to, p. 385), Saint Patricius developed nearly the same theory
of telluric heat before the Proconsul Julius ; but at the conclusion of
his speech the cold hell is more distinctly indicated : " Nam quae lon-
gius ab igne subterraneo absunt, Dei optimi providentia frigidiores
erumpunt. At quce propiores igni sunt, ab eo fervefactae, intolerabili
calore praeditae promuntur foras. Sunt et alicubi tepidse, quippe non
parum sed longiuscule ab eo igne remotae. Atque ille infernus ignis
impiarum est animarum carnificina ; non secus ac subterraneus frigi-
dissimus gurges, in glaciei glebas concretus, qui Tartarus nuncupatur."
The Ai-abic name, Hammdm-el-Evf, signifies nose-baths, and is, as Tem-
ple has already remarked, derived from the form of a neighboring
promontory, and not from a favorable action exerted by this thermal
Avater upon diseases of the nose. The Arabic name has been various-
ly altered by reporters : Hammam I'Enf or Lif, Emmamelif (Peys-
sonel), la Mamelif (Desfontaines). See Gumprecht, Die Mineralquel-
len avfdem Fcstlande von Africa (1851), s. 140-144.

THERMAL SPRINGS. 189

to the springs, in proportion as they have a deeper origin."
With the learned bishop Plato's Pyriphlegethon was the hell
of sinners ; and as though he desired at the same time to re-
mind one of the cold hells of the Buddhists, an aqua gelidissi-
ma concrescens in glaciem is admitted, somewhat unphysically
and notwithstanding the depth, for the nunquam jiniendum
suppliciu?7i impiorum.

Among hot springs, those which, approaching the boiling
heat of water, attain a temperature of 194° F. are far more
rare than is usually supposed, in consequence of inexact ob-
servations ; least of all do they occur in the vicinity of still
active volcanoes. I was so fortunate, during my American
travels, as to investigate two of the most important of these
springs, both between the tropics. In Mexico, not far from
the rich silver mines of Guanaxuato, in 21° N. lat., and at
an elevation of about 6500 feet above the surface of the sea,
near Chichemequillo,* the Aguas de Comangillas burst forth
from a mountain of basalt and basaltic breccia. In Septem-
ber, 1803, I found their temperature to be 205°-5 F. This
mass of basalt has broken in the form of veins through a
columnar porphyry, which again rests upon a white syenite
rich in quartz. At a greater elevation, but not far from
this nearly boiling spring, near Los Joares, to the north of
Santa Rosa de la Sierra, snow falls from December to April
even at an elevation of 8700 feet, and the inhabitants pre-
pare ice the whole year round by radiation in artificial ba-
sins. On the road from Nueva Valencia, in the Valles de
Aragua, toward the harbor of Porto Cabello (in about 101°
of latitude), on the northern slope of the coast chain of Ven-
ezuela, I saw the aguas calientes de las Trincheras springing
from a stratified granite, which does not pass at all into
gneiss. I foundf the springs, in February, 1800, at 194°-5
F., while the Banos de Mariara, in the Valles de Aragua,
which belong to the gneiss, showed a temperature of 138°-7
F. Twenty-three years later, and again in the month of
February, Boussingault and Riverof found in the Mariara

* Humboldt, Essai Politique sur la Nouvelle Espagne^ ed. 2, t. iii.
(1827), p. 190. •

t Relation Ilistoriqne, t. ii., p. 98 ; Cosmos, vol. i., p. 222. The hot
springs of Carlsbad also originate in the granite (Leop. von Buch, in
Foggend., Anmlen, bd. xii., s. 230), just like the hot springs of Mo-
may, in Thibet, visited by Joseph Hooker, which break forth near
Changokhang, at an elevation of 16,000 feet above the sea, with a
temperature of 115° (IIi?nalai/an Journal, a'oI. ii., p. 133).

t Boussingault, " CoHsiderations sur les eaux thermales des Cor-

190 COSMOS.

exactly 147° -2 F. ; and in the Trincheras de Porto Cabello,
at a small elevation above the Caribbean Sea, in one basin
198^ F., in the other 206O-6 F. The temperature of these
hot springs had, therefore, risen unequally in the short inter-
val between these two periods — in Mariara about 8°'5 F.,
and in the Trincheras about 12°*1 F. Boussingault has
justly called attention to the fact that it was in the above-
mentioned interval that the fearful earthquake took place
which overwhelmed the city of Caraccas on the 26th of
March, 1812. The commotion -at the surfiice was, indeed,
not so strong in the vicinity of the Lake of Tacarigua (Nu-
eva Valencia) ; but in the interior of the earth, where elas-
tic vapors act upon fissures, may not a movement which
propagated itself so far and so powerfully readily alter the
net-work of fissures, and open deeper canals of supply ? Thb
hot waters of the Trincheras, rising from a granite formation,
are nearly pure, as they only contain traces of silicic acid, a
little sulphureted hydrogen and nitrogen ; after forming nu-
merous very picturesque cascades, surrounded by a luxuri-
ant vegetation, they constitute a river, the Rio de Aguas
calientes; and this, toward the coast, is full of large croco-
diles, to which the warmth, already considerably diminished,
is very suitable. In the most northern parts of India (30°
52^ N. lat.), and also from granite, issues the very hot well
of Jumnotri, which attains a temperature of 194° F., and,
as it presents this high temperature at an elevation of 10,850
feet, almost reaches the boiling point proper to this atmos-
pheric pressure.*

Among the intermittent hot springs, the Icelandic boiling
fountains, and of these especially the Great Geyser and
Strokkr, have justly attained the greatest celebrity. Ac-
cording to the admirable recent investigations of Bunsen,
Sartorius von Waltershausen, and Descloiseaux, the tem-
perature of the streams of water in both diminishes in a re-
markable manner from below upward. The Geyser possess-
es a truncated cone of 25 to 30 feet in height, formed by
horizontal layers of silicious sinter. ,In this cone there lies
a shallow basin of 52 feet in diameter, in the centre of which
the funnel of the boiling spring, one third of its diameter,
and surrounded by perpendicular walls, goes down to a

dill^res, in the Annaks de Chimie et de Physique, t. iii., 1833, p. 188-
190.

* Captain Newbold, " On the Temperature of the Wells and Riverff
in India and Egypt" {Phil. Transact, for 1845, pt. i., p. 127).

THERMAL SPRINGS. 191

depth of 75 feet. The temperature of the water, which
constantly fills the basin, is 180°. At very regular inter-
vals of one hour and 20 or 30 minutes the thunder below
proclaims the commencement of the eruption. The jets of
water, of 9 feet in thickness, of which about three large
ones follow one. another, attain a height of 100 and some-
times 150 feet. The temperature of the water ascending in
the funnel has been found to be 260° -6 at a depth of 72
feet a little while before the eruption, during the eruption
255^-5, and immediately after it 251°'6; at the surface of
the basin it is only 183° — 185°. The Strokkr, which is also
situated at the base of the Bjarnafell, has a smaller mass of
water than the Geyser. The sinter margin of its basin is
only a few inches in height and breadth. The eruptions
are more frequent than in the Geyser, but do not announce
themselves by subterranean thunder. In the Strokkr the
temperature during the eruption is 235° — 239° at a depth
of 42 feet, and almost 212° at the surface. The eruptions
of the intermittent boiling springs, and the slight changes in
the type of the phenomena, are perfectly independent of the
eruptions of Hecla, and were by no means disturbed by the
latter in the yea^s 1845 and 1846.* With his peculiar
acuteness in observation and discussion, Bunsen has refuted
the earlier hypotheses regarding the periodicity of the Gey-
ser eruptions (subterranean caldrons, which, as steam-boil-
ers, are filled sometimes with vapors and sometimes with wa-
ter). According to him the eruptions are caused by a por-
tion of the column of water, which has acquired a high tem-
perature at a lower point under great pressure of accumu-
lated vapors, being forced upward, and thus coming under
a pressure which does not correspond with its temperature.

* Sartorius von Waltershausen, Physisch-geograpMsche Sidzze von
Island, mit besonderer Riicksicht auf Vulkanische Erscheinungen, 184:7,
s. 128-132 ; Bunsen and Descloiseaux, in the Comptes rendus des St-
ances de VAcad. des Sciences, t. xxiii., 1846, p. 935; Bunsen, in the
Annalen der Cheinie und Pharmacie, bd. Ixii., 1847, s. 27-45. Lottin
and Robei-t had already found that the temperature of the jet of wa-
ter in the Geyser diminishes from below upward. Among the forty
silicious bubbling springs, which are situated in the vicinity of the
Great Geyser and Strokkr, one bears the name of the Little Geyser.
Its jet of water only rises 20 or 30 feet. The term boiling springs
(Kochbrunnen) is derived from the word Geyser, which is connected
with the Icelandic giosa (to boil). On the high land of Thibet also,
according to the report of Esoma de Koros, there is, near the Alpitie
Lake Mapham, a Geyser, which rises to the height of 12 feet.

192 COSMOS.

In this way " the Geysers are natural collectors of steam
power."

Of the hot springs a few approach nearly to absolute
purity; others contain solutions of 8 — 12 parts of solid or
gaseous matters. Among the former are the baths of Lux-
euil, PfefFer, and Gastein, the efficacy of which may appear
so mysterious on account of their purity.* As all springs
are fed principally by meteoric water, they contain nitrogen,
as Boussingault has proved in the very puref springs flowing
from the granite in Las Trincheras de Porto Cabello, and
BunsenJ in the Cornelius spring at Aix and in the Geyser
of Iceland. The organic matter dissolved in many springs
also contains nitrogen, and is even sometimes bituminous.
Until it was known, from the experiments of Gay-Lussac
and myself, that rain and snow water contain more oxygen
than the atmosphere (the former 10, and the latter at least
8 per cent, more), it appeared very remarkable that a gase-
ous mixture rich in oxygen could be evolved from the springs
of Nocera, in the Apennines. The analyses made by Gay-
Lussac during our stay at this mountain spring showed that
it only contained as much oxygen as might have been fur-
nished to it by atmospheric moisture.§ If^we be astonished
at the silicious deposits as a constructive material of which

* Trommsdorf finds in the springs of Gastein only 0-303 of sohd
constituents in 1000 parts; Lovvig, 0-201 in Pfeffer; and Longchamp
only 0-236 in Luxeuil; on the other hand, 0-478 were found in 1000
parts of common well-water in Berne; 5-459 in the Carlsbad bub^
bling spring ; and even 7-454 in Wiesbaden (Studer, Physikal. Geo-
graphie und Geologie, ed. 2, 1847, cap. i., s. 92).

t "The hot springs which gush from the granite of the Cordillera
of the coast (of Venezuela) are nearly pure ; they only contain a small
quantity of silica in solution, and hydrosulphuric acid gas, mixed
with a little nitrogen. Their composition is identical with that which
would result from the action of water upon sulphuret of silicon" {An-
nates de Chimie et de Physique, t. Hi., 1833, p. 189). Upon the great
quantity of nitrogen which is contained in the hot spring of Orense
(154° -4), see Maria Rubio, Tratado de las Fuentes Minerales de Es-
2mna, 1853, p. 331.

X Sartorius von Waltershausen, Shizze von Island, s. 125.

§ The distinguished chemist Morechini, of Home, had stated the
oxygen contained in the spring of Nocera (situated 2240 feet above
the sea) to be 0-40; Gay-Lussac (26th September, 1805) found the
exact quantity of oxygen to be only 0-299. We had previously found
0-31 of oxygen in meteoric waters (rain). Upon the nitrogen gas con-
tained in the acid springs of Neris and Bourbon I'Archambault, see the
works of Anglade and Longchamp (1834); and on carbonic acid
exhalations in general, see Bischof 's admirable investigations in his
Chemische Geologie, bd. i., s. 243-350.

THERMAL SPRINGS. 193

nature, as it were, artificially composes the apparatus of
Geysers, we must remember that silicic acid is also diffused
in many cold springs which contain a very small portion of
carbonic acid.

Acid springs and jets of carbonic acid gas, which were
long ascribed to deposits of coal and lignite, appear rather to
belong entirely to the processes of deep volcanic activity —
an activity which is universally disseminated, and therefore
does not exert itself merely in those places where volcanic
rocks testify to the existence of ancient local fiery eruptions.
In extinguished volcanoes jets of carbonic acid certainly re-
main longest after the Plutonic catastrophes; they follow
the stage of Solfatara activity ; but nevertheless waters im-
pregnated with carbonic acid, and of the most various tem-
peratures, burst forth from granite, gneiss, and old and new
floetz mountains. Acid springs become impregnated with
alkaline carbonates, and especially with carbonate of soda,
wherever water impregnated with carbonic acid acts upon
rocks containing alkaline silicates.* In the north of Ger-
Biany many of the cal'bonic acid springs and gaseous jetg
are particularly remarkable for the dislocation of the strata
about them, and for their eruption in circular valleys (Pyr-
mont, Driburg), which are usually completely closed. Fried-
rich Hoffman and Buckland have almost at the same time
very characteristically denominated such depressions valleys
of elevation {Erhehungs-Thdler).

In the springs to which the name of sulphurous waters is
given, the sulphur by no means constantly occurs combined
in the same way. In many, which contain no carbonate of
soda, sulphureted hydrogen is probably dissolved ; in others, *
for example in the sulphurous waters of Aix (the Kaiser,
Cornelius, Kose, and Quirinus springs), no sulphureted hy-
drogen is contained, according to the precise experiments of
Bunsen and Liebig, in the gases obtained by boiling the
waters without access of air ; indeed the Kaiserquelle alone
contains 0"31 per cent, of sulphureted hydrogen in gas bub-
bles which rise spontaneously from the springs.!

* Bunsen, in Poggendoiif's Annaien, bd. Ixxxiii., s. 257; Bischof,
Geohgie, bd. i., s. 271.

t Liebig and Bunsen, Untersuchung derAachener Schewefelquelkn, in •
the Annaien der Chemie und Pharrnacie, bd. Ixxix. (1851), s. 101. In
the chemical analyses of mineral waters, which contain sulphuret of
Bodium, carbonate of soda and sulphureted hydrogen are often stated
to occur, from an excess of carbonic acid being present in those waters.
Vol. v.— I

194

COSMOS.

A thermal spring which gives rise to an entire river of
vi^ater acidified by sulphur, the Vinegar River {Rio Vinagre),
called Fusambio by the aborigines, is a remarkable phenom-
enon to which I first called attention. The Rio Vinagre
rises at an elevation of about 10,660 feet on the northwest-
ern declivity of the volcano of Purace, at the foot of which
the city of Popayan is situated. It forms three picturesque
cascades,* of one of which I have given a representation,
falling over a steep trachytic wall probably 320 feet in per-
pendicular height. From the point where the small river
falls into the Cauca, this great river, for a distance of 2 — 3
miles (from 8 to 12 English miles) downward, as far as the
junctions of the Pindamon and Palace, contains no fish ;
which must be a great inconvenience to the inhabitants of
Popayan, who are strict observers of fasts! According to
Boussingault's subsequent analysis, the waters of the Pusam-
bio contain a great quantity of sulphureted hydrogen and
carbonic acid, with some sulphate of soda. Near the source,
Boussingault found the temperature to be 163°. The up-
per part of the Pusambio runs underground. Degenhardfc
(of Clausthal, in the Harz), whose early death has caused a
great loss to geognosy, discovered a hot spring in 1846 in
the Paramo de Ruiz, on the declivity of the volcano of the
same name, at the sources of the Rio Guali, and at an alti-
tude of 12,150 feet, in the water of which Boussingault
found three times as much sulphuric acid as in the Rio
Vinagre.

The equability of the temperature and chemical constitu-
tion of springs, as far as we can ascertain from reliable ob-
' servations, is far more remarkable than the instabilityf which

* One of these cascades is represented in my Vues des Cordilleres^
pi. XXX. On the analysis of the water of the Kio Vinagre, see Bous-
singault, in the Annales de Chimie et de Physique, 2e serie,t. lii., 1833,
p. 397, and Dumas, 3e serie, t. xviii., 1846, p. 503; on the spring in
the Paramo de Ruiz, see Joaquin Acosta, Viajes Cientificos d los Andes
Ecuatoriales, 1849, p. 89.

t The examples of alteration of temperature in the thermal springs
of Mariara and Las Trincheras lead to the question whether the Styx
water, whose source, so difficult of access, is situated in the wild
Aroanic Alps of Arcadia, near Nonacris, in the district of Pheneos,
has lost its pernicious qualities by alteration in the subterranean fis-
• sures of supply ? or whether the.waters of the Styx have only occasion-
ally been injurious to the wanderer by their icy coldness ? Perhaps
they are indebted for their evil reputation, which has been transmitted
to the present inhabitants of Arcadia, only to the awful wildness and
desolation of the neighborhood, and to the myth of their origin from

THERMAL SPRINGS. 195

has been occasionally detected. The hot spring waters,
which, during their long and tortuous course, take up such
a variety of constituents from the rocks with which they
are. in contact, and often carry them to places where they
are deficient in the strata through which the springs burst
forth, have also an action of a totally different nature. They
exert a transforming and at the same time a formative ac-
tivity, and in this respect they are of great geognostic im-
portance. Senarmont has shown with wonderful acuteness
how extremely probable it is that many vein-crevices (an-
cient courses of thermal waters) have been filled from below
upward by the deposition of the dissolved elements. By
changes of pressure and temperature, by internal electro-
chemical processes, and the specific attraction of the lateral
walls (the rock traversed), sometimes lamellar deposits, and
sometimes masses of concretion are produced in fissures and
vesicular cavities. In this way druses and porous amygda-
loids appear to have been sometimes formed. Where the
deposition of the veins has taken place in parallel zones, these
zones usually correspond with each other symmetrically in
their nature, both vertically and laterally. Senarmont has
succeeded in preparing a considerable number of minerals
artificially, by perfectly analogous synthetical methods.!

Tartarus. A young and learned philologist, Theodor Schwab, suc-
ceeded a few years ago, with great exertion, in penetrating to the
rocky wall from which the spring trickles down, exactly as described
by Homer, Hesiod, and Herodotus. He drank some of the water,
which was extremely cold, but very pure to the taste, without per-
ceiving any injurious effects (Schwab, Arkadieii, seine Natur und Ge-
schichte, 1852, s. 15-20). Among the ancients it was asserted that the
coldness of the water of the Styx burst all vessels except those made
of the hoof of an ass. The legends of the Styx are certainly very old,
but the report of the poisonous properties of its spring appears to have
been widely disseminated only in the time of Aristotle. According
to a statement of Antigonus of Carystus (Hist. Mirab., § 174), it was
contained very circumstantially in a book of Theophrastus, which has
been lost to us. The calumnious fable of the poisoning of Alexander
by. the water of the Styx, which Aristotle communicated to Cassander
by Antipater, was contradicted by Plutarch and Arrian, and dissem-
inated by Vitruvius, Justin, and Quintus Curtius, but without men-
tioning the Stagirite (Stahr, Aristotelia, th. i., 1830,' s. 137-140).
Pliny (xxx., 53) says, somewhat ambiguously: "Magna Aristotelis
infamia excogitatum." See Ernst Curtius, Peloponnesus (1851), bd.
i., s. 194-196, and 212; St. Croix, Examen Critique des Anciens His-
ioriens d' Alexandre, p. 496. A representation of the cascade of the
Styx, drawn from a distance, is contained in Fiedler's Meise durch
Griechenland, th. i., s. 400.
* " Very important metalliferous lodes, perhaps the greater num-

196 COSMOS.

One of my intimate friends, a highly endowed scientific
observer, will, I hope, before long publish a new and import-
ant work upon the conditions of temperature of springs, and
in it treat with great acumen and universality, by induction
from a long series of recent observations, upon the involved
phenomenon of disturbances. In the determinations of tem-
perature made by him in Germany (on the Rhine) and in
Italy (in the vicinity of Rome, in the Albanian mountains
and the Apennines) from the year 1845 to 1853, Eduard
Kallmann distinguishes : 1. Purely meteorological springs, the
average temperature of which is not increased by the internal
heat of the earth ; 2. Meteorologico-geological springs, which,
being independent of the distribution of rain, and warmer
than the air, only undergo such alterations of temperature as
are communicated to them by the soil through which tlky
flow out ; 3. Abiiomialbj cold springs, which bring down their
coldness from great elevations.* The more we have advanced

ber, appear to have been formed by solution, while the veins filled with
concretions of metal seem to be nothing but immense canals more or
less obstructed, and formerly traversed by incrusting thermal waters.
The formation of a great number of minerals which are met with in
these lodes does not always presuppose conditions or agents very far
removed from existing causes. The two principal elements of the most
widely-diffused thermal waters, the alkaline sulphurets and carbonates,
have enabled me to reproduce artificially, by very simple synthetic
methods, 29 distinct.mineral species, nearly all crystallized, belong-
ing to the native metals (native silver, copper, and arsenic), quartz,
specular iron, carbonates of iron, nickel, zinc, manganese, sulphate
of baryta, pyrites, malachite, copper pyrites, sulphuret of copper, red
arsenical and antimonial silver. . . . We approach as closely as pos-
sible to the processes of nature, if we succeed in reproducing minerals
in their conditions of possible association, by means of the most wide-
ly diffused natural chemical agents, and by imitating the phenomena
which we still see realized in the foci in which the mineral creation
has concentrated the remains of that activity which it formerly dis-
played with a very different energy." (H. de Senarmont, Sur la Forma-
tion des Mineravx par la Vole Humide, in the Annales de Chetnie et de
Physique^ 3e se'rie, t. xxxii., 1851, p. 234 ; see also Elie de Beaumont,
Sur les Emanations Volcaniques et Metallijeres, in the Bulletin de la
Societe Geologique de France, 2e se'rie, t. xv., p. 129.)

* "In order to ascertain the amount of variation of the average
temperature of springs from that of the air, Dr. Eduard Kallmann ob-
serv^ed at his former residence, Marienberg, near Boppard, on the
Rhine, the temperature of the air, the amount of rain, and the tem-
perature of seven springs for five years, from the 1st December, 1845,
to the 30th November, 1850 ; upon these observations he has founded
a new elaboration of the relative temperature of springs. In this in-
vestigation the springs with a perfectly constant temperature (the
purely geological springs) are excluded. On the other hand, all those

THERMAL SPRINGS. 197

of late years, by the successful employment of chemistry, in
the geognostic investigation of the formation and metamorph-

springs have been made the subject of investigation which undergo an
alteration in their temperature according to the seasons.

"The variable springs fall into two natural groups :

"1. Purely meteorological springs: that is to say, those whose aver-
age is demonstrably not elevated by the heat of the earth. In these
springs the amount of variation of the average from the aerial avenge
is dependent upon the distribution of the annual amount of rain through
the twelve months. These springs are on the average colder than the
air when the proportion of rain for the four cold months, from Decem-
ber to March, amounts to more than 33^^ per cent. ; they are on the
average warmer than the air when the proportion of rain for the four
warm months, from July to October, amounts to more than 33J per
cent. The negative or positive difference of the spring average from
the air average is larger in proportion to the excess of rain in the
above-mentioned cold or warm thirds of the year. Those springs in
which the difference of the average from that of the air is in accord-
ance with the law, that is to say, the largest possible by reason of the
distribution of rain in the year, are called purely meteorological springs
of undistorted average ; but those in which the amount of difference of
the average from the air average is diminished by the disturbing ac-
tion of the atmospheric heat during the seasons which are free from
rain are called purely meteorological springs of approxirnate average.
The approximation of the average to the aerial average is caused
either by the inclosure, especially by a channel at the lower extremity
of which the temperature of the spring was observed, or it is the con-
sequence of a superficial coui'se and the poverty of the feeders of the
spring. In each year the amount of difference of the average from
the aerial average is similar in all purely meteorological springs, but
it is smaller in the approximate than in the undistorted springs, and in-
deed is smaller in proportion as the disturbing action of the atmospher-
ic heat is greater. Of the springs of Marienberg four belong to the
group of purely meteorological springs ; of these four one is undis-
torted in its average, the three others are approximated in various de-
grees. In the first year of observation the portion of rain of the cold
third predominated, and all four springs were on the average colder
than the air. In the four following years of observation fhe rain of
the warm third predominated, and in these all the four springs had a
higher average temperature than the air; and the positive variation
of the average of tlie spring from that of the air was higher, the greater
the excess of rain in the warm third of one of the four years.

" The view put forward in the year 1825 by Leopold von Buch, that
the amount of variation of the average of springs from that of the air
must depend upon the distribution of rain in the seasons of the year,
has been shown to be perfectly correct by Hallmann, at least for his
place of obsei-vation, Marienbei'g, in the Rhenish Graywacke mount-
ains. The purely meteorological springs of undistorted average alone
have any value for scientific climatology ; these springs are to be
sought for every where, and to be distinguished on the one hand from
the purely meteorological springs with an approximate average, and
on the other from the meteorologico-geological springs.

"2. Meteorologico-geological springs: that is to say, those of which

198 COSMOS.

ic transformation of rocks, the greater importance has been
acquired for the consideration of the waters impregnated with
gases and salts which circulate in the interior of the earth,
and which, when they burst forth at the surface as thermal
springs, have already fulfilled the greater part of their forma-
tive, alterative, or destructive activity.

c. Vapor and Gas Springs, Salses, Mud Volcanoes, Naphtha
Fire.

(Amplification of the Picture of Nature, Cosmos, vol. i., p. 224-226.)

In the General Representation of Nature I have shown by
well-ascertained examples, which, however, have not been
sufficiently taken into consideration, how the salses in the
various stages through which they pass, from the first erup-

the average is demonstrably heightened by the heat of the earth.
Whatever the distribution of rain may be, these springs are in their
average warmer than the air all the year round (the alterations of
temperature which they exhibit in the course of the year are commu-
nicated to them by the soil through which they flow). The amount
by which the average of a meteorologico-geological spring exceeds the
atmospheric average depends upon the depth to which the meteoric
waters have sunk down into the interior of the earth, where the temper-
ature is constant, before they again make their appearance in the form
of a spring ; this amount, consequently, possesses no climatological in-
terest. The climatologist must, however, know these springs, in order
that he may not mistake them for purely meteorological springs. The
meteorologico-geological springs may also be approximated to the
aerial average by an inclosure or channel. The springs were observed
on particular fixed days, four or five times a month. The elevation
above the sea, both of the place where the temperature of the air was
observed and of the difierent springs, was carefully taken into ac-
count."

After the completion of the elaboration of his observations at Mari-
enberg, Dr. Kallmann passed the winter of 1852-1853 in Italy, and
found abnormally cold springs in the vicinity of ordinary ones. This is
the name he gives " to those springs which demonstrably bring down
cold from above. These springs are to be regai'ded as subterranean
drains of open lakes or subterranean accumulations of water situated
at a great elevation, from which the waters pour down very rapidly in
fissures and clefts, and break forth at the foot of the mountain or
chain of mountains in the form of springs. The idea of the abnorm-
ally cold springs is, therefore, as follows : They are too cold for the
elevation at which they come forth ; or, which indicates the conditions
better, they come forth at too low a part of the mountain for their low
temperature." These views, which are developed in the first volume
of Hallmann's Temperaturverhaknimen der Quellen, have been modified
by the author in his second volume (s. 181-183), because in every
meteorological spring, however superficial it may be, there must h6
some telluric heat.

SALSES. 199

tions accompanied by flames to the subsequent condition of
simple eruptions of mud, form, as it were, an intermediate
step between hot springs and true volcanoes, which throw
out fused earths, either in the form of disconnected cinders
or as newly-formed rocks, often arranged in many beds one
over the other. Like all transitions and intermediate steps,
both in organic and inorganic nature, the salses and mud
volcanoes deserve a more careful consideration than was be-
stowed upon them by the older geognosists, from the want
of special knowledge of the facts.

The salses and naphtha springs are sometimes arranged in
isolated close gi-oups — like the ]VI^calubi, near Girgenti, in
Sicily, which were mentioned even by Solinus; those near
Pietra Mala, Barigazzo, and on the Monte Zibio, not far
from Sassuolo, in the north of Italy ; or those near Turbaco,
in South America ; . sometimes they appear to be arranged
in narrow chains, and these are the most instructive and im-
portant. We have long known* as the outermost members

* Humboldt, Asie Centrale, t: ii., p. 58. Upon the reasons which
render it probable that the Caucasus, which for five-sevenths of its
length, between the Kasbegk and Elburuz, runs from E.S.E. to
W.N.W. in the mean parallel of 42° 50', is the continuation of the
volcanic fissure of the Asferah (Aktagh) and Thian-schan, see the work
cited above, p. 54-Gl. Both the Asferah and Thian-schan oscillate
between the parallels of 40|° and 43°. I regard the great Aralo-
Caspian depression, the surface of which, according to the accurate
measurements of Struve, exceeds the area of the whole of France by
nearly 107,520 geographical square miles (Op. cit, supra, p. 309-312),
as more ancient than the elevations of the Altai and Thian-schan.
The fissure of elevation of the last-mentioned mountain chain has not
been continued through the great depression. It is only to the west
of the Caspian Sea that we again meet with it, with some alteration
in its direction, as the chain of the Caucasus, but associated with tra-
chytic and volcanic phenomena. This geognostic connection has also
been recognized by Abich, and confirmed by valuable observations.
In a treatise on the connection of the Thian-schan with the Caucasus
by this great geognosist, which is in my possession, he says express-
ly : "The frequency and decided predominance of a system of paral-
lel dislocations and lines of elevation (nearly from east to west) dis-
tributed over the whole district (between the Black Sea and the Cas-
pian) brings the mean axial direction of the great latitudinal central
Asiatic mass elevations most distinctly westward from the Kosyurt
and Bolar systems to the Caucasian Isthmus. The mean direction
of the Caucasus, S.E. — N.W., is E.S.E. — W.N.W. in the central parts
of the mountain chain, and sometimes even exactly E. — W., as in
the Thian-schan. The lines of elevation which unite Ararat with the
trachytic mountains Dzerlydagh and Kargabassar near Erzeroum, and
in the southern parallels of which Mount Argaeus, Sepandagh, and
Sabalan are arranged, constitute the most decided expression of a

200

COSMOS.

of the Caucasus, in the northwest the mud volcanoes of Ta-
man, and in the southeast of the great mountain chain the
naphtha springs and naphtha fire of Baku and the Caspian

mean volcanic axial direction, that is to say, of the Thian-schan be-
ing prolonged westward through the Caucasus. Many other mountain
directions of Central Asia, however, also revert to "this remarkable
space, and stand, as elsewhere, in mutual relation to each other, so
as to form vast mountain nuclei and maxima of elevation." Pliny
(vi., 17) says : " Pers^e appellavere Caucasum montem Graucasim
(var. Graucasum, Groucasim, Grocasum), hoc est nive candidum;" in
which Bohlen thought the Sanscrit words kds, to shine, and gravan,
rock, were to be recognized (see my Asie Centrale, t. i., p. 109). As
Klausen says, in his investigations on the wanderings of lo {Rheinisches
Museum fur Philologiey JahrgViii., 1845, s. 298), if the name Graucasus
was corrupted into Caucasus, then a name "in which each of its first
syllables gave the Greeks the idea of burning might certainly charac-
terize a burning mountain, with which the history of the Fire-burner
(Fire-igniter, TrvpKatvg) would become readily and almost spontaneous-
ly associated." It can not be denied that mylhs sometimes originate
from names, but the production of so great and important a fable as
the Typhonico-caucasic can certainly not be derivable from the acci-
dental similarity of sound in the misunderstood name of a mountain.
There are better arguments, of which Klausen also mentions one.
From the actual association of Typhon and the Caucasus, and from
the express testimony of Pherecydes of Syros (in the time of the 58th
Olympiad), it is clear that the eastern cxriomity of the world was re-
garded as a volcanic mountain. According to one of the Scholia to
Apollonius (Scholia in Apoll. Rhod., ed. Schaefferi, 1813, v. 1210, p.
524), Pherecydes says, in the Theogony, "that Typhon, when pur-
sued, fled to the Caucasus, and that then the mountain burned (or
was set on fire) ; that from thence Typhon fled to Italy, when the isl-
and Pithecusa was thrown around (as it were, poured around) liim."
But Pithecusa is the island ^naria (now Ischia), upon which the Epo-
meus (Epopon) cast forth fire and lava, according to Julius Obsequens,
95 years before our era, then during the reigns of Titus and Diocle-
tian, and, lastly, in the year 1302, according to the statement of To-
lomeo Fiadoni of Lucca, who was at that time Prior of Santa Maria
Novella. "It is singular," as Boeckh, the profound student of antiq-
uity, writes to me, " that Pherecydes should make Typhon fly from
the Caucasus because it burned, as he himself is the originator of sub-
terraneous fire ; but that his residence upon the Caucasus rests upon
the occurrence of volcanic eruptions there, appears to me to be unde-
niable." Apollonius Rhodius {Argon., lib. ii., v. 1212-1217, ed. Beck),
in speaking of the birth of the Colchian Dragon, also places in the
Caucasus the rock of Typhon, on which the giant was struck by the
lightning of Jupiter. Although the lava-streams and crater-lakes of
the high land of Kely, the eruptions of Ararat and Elburuz, or tha
currents of obsidian and pumice-stone from the old craters of the Rio-
tandagh, may be placed in a pre-historic period, still the many hun-
dred flames which even now break forth from fissures in the Cauca-
sus, both from mountains of seven or eight thousand feet in height and
from broad plains, may have been a sufficient reason for regarding the
entire mountain district of the Caucasus as a Typhonic seat of fire.

SALSES. 201

peninsula, Apscheron. The magnitude and connection of
this phenomenon was, however, first discovered by Abich,
distinguished by his profound knowledge of this part of Asia.
According to him, the mud volcanoes and naphtha fires of
the Caucasus are arranged in a distinctly recognizable man-
ner in certain lines, which stand in unmistakable relation
w^ith the axes of elevation and the directions of dislocation
of the strata of rock. The greatest space, of nearly 4000
square miles, is occupied by genetically-connected mud vol-
canoes, naphtha emanations, and saline springs in the south-
eastern part of the Caucasus, in an isosceles triangle, the
base of which is the shore of the Caspian Sea, near Balacha-
ni (to the north of Baku), and one of the mouths of the Kur
(Araxes), near the hot springs of Sallian. The apex of such
a triangle is situated near the Schagdagh, in the elevated
valley of Kinalughi. There, at the boundary of a dolomitic
and slate formation, at an elevation of 8350 feet above the
Caspian Sea, close to the village of Kinalughi itself, break
forth the perpetual fires of the Schagdagh, which have never
been extinguished by meteorological occurrences. The cen-
tral axis of this triangle corresponds with the direction which,
the earthquakes^ so often experienced in Schamacha, upon the
banks of the Pyrsagat, appear constantly to follow. When
the northwestern direction just indicated is traced further, it
strikes upon the hot sulphurous springs of Akti, and then
becomes the line of strike of the principal crest of the Cau-
casus, where it rises up into the Kasbegk and bounds Daghes-
tan- The salses of the lower region, which are often regu-
larly arranged in series, gradually become more numerous
toward the shore of the Caspian, between Sallian, the mouth
of the Pyrsagat (near the island of Swinoi), and the penin-
sula of Apscheron. They present traces of repeated mud
eruptions in earlier times, and often bear at their summits
small cones, from which combustible and often spontaneous-
ly ignited gas is poured forth, and which are exactly similar
in form to the hoimitos of JoruUo, in Mexico. Considerable
eruptions of flame were particularly frequent between 1844
and 1849, at the Oudplidagh, Nahalath, and Turandagh.
Close to the mouth of the Pyrsagat, on the mud volcano
Toprachali, "black marly fragments, which at the first glance
might be confounded with dense basalt, and extremely fine-
grained doleritic rocks" are found (a proof of the exception-
al, greatly increased intensity of the subterranean heat). At
other points on the peninsula of Apscheron, Lenz found

12

202 COSMOS.

slag-like fragments as products of eruption; and during
the great eruption of flame of Backlichli (7th February,
1839), small hollow balls, like the so-called ashes of the true
volcanoes, were carried by the wind to a long distance.*

In the northwestern extremity, toward the Cimmerian
Bosphorus, are the mud volcanoes of the peninsula of Ta-
man, which form one group with those of Aklanisowka and
Jenikale, near Kertsch. One of the salses of Taman ex-
hibited an eruption of mud and gas on the 27th of P^ebru-
aiy, 1793, in which, after much subterranean noise, a col-
umn of fire half enveloped in black smoke (dense aqueous
vapor ?) rose to a height of several hundred feet. It is a re-
markable phenomenon, and instructive as regards the nature
of the Volcancitos de Turhaco^ that the gas of Taman, which
was tested in 1811 by Frederick Parrot and Engelhardt,
was not injlammable ; while the gas collected by Gobel in
the same place, twenty-three years later, burned,' from the
mouth of a glass tube, with a bluish flame, like all emana-
tions from the salses in the southeastern Caucasus, but also,
when carefully analyzed, contained in 100 parts 92-8 of car-
bureted hydrogen and 5 parts of carbonic oxyd gas.l

A phenomenon certainly nearly allied to these in its
origin, although different as regards the matter produced, is
presented by the eruptions of boracic acid vapors in the
Tuscan Maremma, known under the names of lagoni, fum-
marole, soffioni, and even volcani, near Possara, Castel Novo,
and Monte Cerboli. The vapors have an average tempera-
ture of 205° to 212°, and according to Pella, in certain
points, as much as 347°. They rise in p^rt directly from
clefts in the rocks, and partly from stagnant pools, in which
they throw up small cones of fluid clay. They are seen to
diffuse themselves in the air in whitish eddies. The boracic
acid, which is brought up by the aqueous vapors from the
bosom of the earth, can not be obtained when the vapors of
the soffioni are condensed in very wide and long tubes, but

* Humboldt, Asie Centrale, t. ii., p. 511 and 513. I have already
(t. ii., p. 201) called attention to, the fact that Edrisi does not men-
tion the fire of Baku, although it is described diffusely as a Nefala-
land, that is to say, rich in burning naphtha springs, by Massudi Coth-
beddin, two hundred years before, in the tenth century (see Frahn,
Ibn Fozlan, p. 245 ; and on the etymology of the Median word najih-
tha, Asiatic J(mmal, vol. xiii., p. 124).

t Compare Moritz von Engelhardt and F. Parrot, Rdse in die Kryvi
und den Kaukasus, 1815, th. 1., s. 71 ; with Gobel, Reise in die Steppen
des sUdlichen Russknds, 1838, th. i., s. 249-253, and th. ii.,s. 138-144.

SALSES. 203

becomes difFased -in the atmosphere in consequence of its
volatility. The acid is only procured in the beautiful estab-
lishments of Count Larderel, when the orifices of the soffioni
are covered directly by the fluid of the basin.* According
to Payen's excellent analysis, the gaseous emanations contain
0-57 of carbonic acid, 0-35 of nitrogen, and only 0*07 of
oxygen, and 0-001 of sulphuric acid. Where the boracic
acid vapors permeate the clefts of the rock they deposit sul-
phur. According to Sir Roderick Murchison's investiga-
tions, the rock is in part of a chalky nature, and in part an
eocene formation, containing nummulites — a macigno, which
is penetrated by the uncovered and elevated serpentinef of
the neighborhood (near Monte Rotondo). In this case, and
in the crater of Volcano, asks Bischof, do not hot aqueous
vapors act upon and decompose boracic minerals, such as
rocks rich in datolithe, axinite, or tourmalin ^J

In the variety and grandeur of the phenomena, the sys-
tem of soffioni in Iceland exceeds any thing that we are ac-
quainted with on the continent. Actual mud springs burst
forth in the fumarole-field of l^risuvek and Reykjalidh, from
small basins with crater-like margins in a bluish-gray clay.§
Here also the fissures of the springs may be traced in de-

* Pay en, De Vacii^e boracique des Suffioni de la Toscane, in the An-
nales de Chimie et de Physique, 3me serie, t. i., 1841, p. 247-255 : Bis-
chof, Chem. und Phjsik. Geologie, bd. i., s. 669-691 ; Etablissements in-
dustriels de Vacide boracique enToscane,hy the Count de Larderel, p. 8.

t Sir Roderick Irapey Murchison, On the Vents of hot Vopor in Tus-
vany, 1850, p. 7 (see also the earlier geognostic observations of Hoff-
mann, in Karsten's und Dechen's Archiv/Ur Mineral., bd. xiii., 1839,
s. 19). From old but trustworthy traditions, Targioni Tozzeti asserts
that some of these boracic acid springs which are constantly changing
their place of eruption were once seen to be luminous (ignited) at
night. In order to increase the geological interest of the observations
of Murchison andPkreto upon the volcanic relations of the serpentine
formation in Italy, I may here advert to the fact that the flame of the
Asiatic Chimcera (near the town of Deliktasch, the ancient Phaselis
in Lycia, on the west coast of the Gulf of Adalia), which has been
burning for several thousand years, also rises from a hill on the slope
of the Solimandagh, in which serpentine in position and blocks of
limestone have been found. Rather more to the south, on the small
island of Grambusa, the limestone is deposited upon dark-colored
serpentine. See the important work of Admiral Beaufort (Survey of
the Coasts of Caramania, 1818, p. 40 and 48), whose statements arc
confirmed by the specimens of rocks just brought home (May, 1854)
by a highly talented artist, Albrecht Berg (Pierre de TchihatchefF, Asie
Mineure, 1853, t. i., p. 407). J Bischof, o;?. cit., s. 682.

() Sartorius von Waltershausen, Physisch-geographische Skizze von
Island, 1847, s. 123 ; Bunsen " upon the processes of formation of the
volcanic rocks of Iceland," Poggend., Annalen, bd. Ixxxiii., s. 257*.

204 COSMOS.

terminate directions.* There is no portion of the earth,
where hot springs, salses, and gas eruptions occur, that has
been made the subject of such admirable and complete chem-
ical investigations as those on Iceland, which we owe to the
acute and persevering exertions of Bunsen. Nowhere, per-
haps, in such a great extent of country, or so near the sur-
face, is such a multifarious spectacle of chemical decomposi-
tions, conversions, and new formations to be witnessed.

Passing from Iceland to the neighboring American conti-
nent, we find in the State of New York, in the neighborhood
of Fredonia, not far from Lake Erie, a multitude of jets of
inflammable gas (carbureted hydrogen) breaking forth from
fissures in a basin of Devonian sandstone strata, and partly
employed for the purpose of illumination. Other springs
of inflammable gas, near Rushville, assume the form of mud
cones; and others, in the valley of the Ohio, in Virginia,
and on the Kentucky River, also contain chlorid of sodium,
and are there connected with weak naphtha springs. But
on the other side of the Caribbean Sea, on the north coast
of South America, 11^ mile%south-southeast from the har-
bor of Cartagena de Indias, near the pleasant village of Tur-
baco, a remarkable group of salses or mud volcanoes exhibits
phenomena which I was the first to describe.

In the neighborhood of Turbaco, where one enjoys a mag-
nificent view of the colossal snowy mountains {Sierras Neva-
das) of Santa Marta, on a desert spot in the midst of the
primeval forest, rise the Volcancitos, to the number of 18 or
20. The largest of the cones, which consist of blackish
gray loam, are from 19 to 23 feet in height, and probably
80 feet in diameter at the base. At the apex of each cone
is a circular orifice of 20 to 28 inches in diameter, surround-
ed by a small mud wall. The gas rushes up with great vio-
lence, as in Taman, forming bubbles, each of which, accord-
ing to my measurements in graduated vessels, contains 10 —
12 cubic inches. The upper part of the funnel is filled with
water, which rests upon a compact floor of mud. The
eruptions are not simultaneous in neighboring cones, but in
each one a certain regularity was observable in the periods
of the eruptions. Bonpland and I, standing on the outer-
most parts of the groups, counted pretty regularly five erup-
tions every two minutes. On bending down over the small
orifice of the crater a hollow sound is perceived in the in-
terior of the earth, far below the base of the cone, usually

* Waltershauseii, o/j. ctV,, s. 118.

SALSES. 205

•

twenty seconds lefore each eruption. A very thin burning
wax taper was instantly extinguished in the gas, which was
twice collected with great care ; this was also the case with
a glowing chip of the wood Bomhax Ceiba. The gas could
not be ignited. Lime-water was not rendered turbid by it ;
no absorption took place. When tested for oxygen with
nitrous acid gas, this gas showed no trace of the former in
one experiment ; in a second case, when the gas of the Vol-
cancitos had been confined for many hours in a bell glass
with water, it exhibited rather more than one hundredth of
oxygen, which had probably been evolved from the water
and accidentally intermixed.

From these analytical results I then declared, perhaps not
very incorrectly, that the gas of the Volcancitos of Turbaco
was nitrogen gas, which might be mixed with a small quan-
tity of hydrogen. At the same time, I expressed my regret
in my journal that, in the state of chemistry at that time
(April, 1801), no means were known by which, in a mix-
ture of nitrogen and hydrogen gases, the numerical propor-
tions of the mixture might be determined. The expedient,
by the employment of which three thousandths of hydrogen
may be detected in a. gaseous mixture, was only discovered
by Gay-Lussac and myself four years afterward.* During
the half century that has elapsed since my residence in Tur-
baco, and my astronomical survey of the Magdalena lliver,
no traveler had occupiedhimself scientifically with the small
mud volcanoes just described, until, at the end of December,
1850, my friend Joaquin Acosta,| so well versed in modern

* Humboldt and Gay-Lussac, Memoire sur ^analyse de Vair atmos-
phcrique in the Journal de Physique, par Latnctherie, t. Ix., p. 151 (see
my Khinere Schriften, bd. i., s. 346).

t "It is with emotion that I have just visited a place which you
made known fifty years ago. The appearance of the small volcanoes
of Turbaco is such as you have described ; there is the same luxuri-
ance of vegetation, the same form of cones of clay, and the same ejec-
tion of liquid and muddy matter ; nothing has changed, unless it be
the nature of the gas which is evolved. I had with me, in accordance
with the advice of our mutual friend, M. Boussingault, all that was
necessary for the chemical analysis of the gaseous emanations, and
even for making a freezing mixture for the purpose of condensing the
aqueous vapor, as the doubt had been expressed to me that nitrogen
might have been confounded with this vapor. But this apparatus Avas
by no means necessary. As soon as I arrived at the Volcancitos, the
distinct odor of bitumen set me in the right course; I commenced by
lighting the gas upon the very orifice of each small crater. Even now
one sees on the surface of the liquid, which rises intermittently, a
delicate film of petroleum. The gas collected hums away entirely^

206

COSMOS.

geognosy and chemistry, made the remarkable observation
that at present "the cones diffuse a bituminous odor" (of
which no trace existed in my time) ; " that some petroleum
floats upon the surface of the water in the small orifices, and
that the gas pouring out may be ignited upon every mud-
cone of Turbaco." Does this, asks Acosta, indicate an al-
teration of the phenomena brought about by internal pro-
cesses, or simply an error in the earlier experiments? I
would admit the latter freely, if I had not preserved the leaf
of the journal on which the experiments were recorded in
detail,* on the very morning on which they were made. I

without any residue of nitrogen (?), and without depositing sulphur
(when in contact with the atmosphere). Thus the nature of the phe-
nomenon has completely changed since your journey, unless we admit an
error of observation, justified by the less advanced state of experi-
mental chemistry at that period. I no longer doubt that the great
eruption of Galera Zamba, which illuminated the country in a radius
of 100 kilometres (62 miles), is a salses-like phenomenon, developed
on a great scale, since there exist hundreds of little cones, vomiting
saline clay, upon a surface of 400 square leagues. I propose examin-
ing the gaseous products of the cones of Tubara, which are the inost
distant salses from your Volcancitos of Turbaco. From the powerful
manifestations which have caused the disappearance of a part of the
peninsula of Galera Zamba, now become an island, and from the ap-
pearance of a new island raised from the bottom of the sea in 1848,
and which has since disappeared, I am led to think that it is near
Galera Zamba, to the west of the delta of the Rio Magdalena, that
the principal focus of the phenomenon of salses in the province of
Carthagena is situated" (from a letter from Colonel Acosta to A. von
Humboldt, Turbaco, 21st December, 1850). See also Mosquera, il/g-
moria politica sabre la Nueva Granada, 1852, p. 73; and Lionel Gis-
borne, The Isthmus of Darien, p. 48.

* During the whole of my American expedition I always adhered
strictly to the advice of Vauquelin, under whom I worked for some
time before my voyage : to write down and preserve the details of ev-
ery experiment on the same day. From my journals of the 17th and
18th April, 1801, I here copy the following: "As, therefore, the gas
showed scarcely 0*01 of oxygen from experiments with phosphorus
and nitrous acid gas, and not 0*02 of carbonic acid with lime-water,
the question is, what are the other 97 hundredths? I supposed, first
of all, carbureted and sulphureted hydrogen ; but no sulphur is de-
posited on the margins of the small craters in contact with the atmos-
phere, and nc odor of sulphureted hydrogen was to be perceived.
The problematical part might appear to be pure nitrogen, for, as above
mentioned, nothing was ignited by a burning taper; but I know, from
the time of my analyses of fire-damp, that a light hydrogen gas, free
from any carbonic acid, which merely stood at the top of a gallery,
did not ignite, but extinguished the pit candles, while the latter
burned clearly in deep places, when the air was considerably mixed
with nitrogen gas. The residue of the gas of the Volcancitos is,
therefore, probably to be regarded as nitrogen, with a portion of by-

SALSES. 207

find nothing in them that could make me at all doubtful now ;
and the observation already referred to (from Parrot's Re-
ports), that " the gas of the mud volcanoes of the peninsula
of Taman in 1811 had the property of preventing combus-
tion, as a glowing chip w^as extinguished in the gas, and
even the ascending bubbles, a foot in diameter, could not be
ignited at the moment of their bursting," while in 1834
. Gobel saw readily inflammable gas burning with a bluish
flame at the same place — leads me to believe that the ema-
nations undergo chemical changes in different stages. Very
recently Mitscherlich has, at my request, determined the
limits of inflammability of artificially prepared mixtures of
nitrogen and hydrogen gases. It appeared that mixtures of
one part of hydrogen gas and three parts of nitrogen gas
not only took fire from a light, but also continued to burn.
When the quantity of nitrogen gas was increased, so that the
mixture consisted of one part of hydrogen and three and a
half parts of nitrogen, it was still inflammable, but did not
continue burning. Jt was only with a mixture of one part of

drogen gas, the quantitative amount of which we do not at present
know. Does the same carbonaceous schist that I saw farther west-
ward on the Rio Sinu, or marl and clay, lie below the Volcancitos ?
Does atmospheric air penetrate through narrow fissures into cavities
formed by water and become decomposed in contact with blackish
gray loam, as in the pits in the saline clay of Hallein and Berch-
tholdsgaden, where the chambers are filled with gases which extin-
guish Hghts ? or do the gases, streaming out tense and elastic, prevent
the penetration of atmospheric air ?" These questions were set down
by me in Turbaco 53 years ago. According to the most recent ob-
servations of M. Vauvert de Mean (1854), the inflammability of the
gas emitted has bben completely retained. The traveler brought with
him samples of the water which fills the small orifice of the craters
of the Volcancitos. In this Boussingault found in the litre : common
salt, 6*59 gr. ; carbonate of soda, 0-31; sulphate of soda, 0*20; and
also traces of borate of soda and iodine. In the mud which had fall-
en to the bottom, Ehrenberg, by a careful microscopiip examination,
found no calcareous parts or scoriaceous matter, but quartz granules
mixed with micaceous laminae, and many small crystalline prisms of
black Augite, such as often occurs in volcanic tufa ; no trace of Spon-
giolites or Poly gastric Infusoria, and nothing to indicate the vicinity
of the sea, but on the contrary many remains of* Dicotyledonous
plants and grasses, and sporangia of lichens, reminding one of the
constituents of the Moija of Felileo. While C. Sainte-Claire, Deville,
and George BoKnemann, in their beautiful analyses of the Macalube
di Terrapilata, found 0 99 of carbureted hydrogen in the gas emitted,
the gas which rises in the Agua Santa di Limosina, near Catanea,
gave. them, like Turbaco formerly, 0*98 of nitrogen, without a trace
of oxygen (Comptes rendus de VAcad. des Sciences, t. xliii., 1856, p.
361 and 366).

208 nosMos.

hydrogen and four parts of nitrogen gas that no ignition took
place. The gaseous emanations, which from their ready in-
liammability and the color of their flame are usually called
emanations of pure and carbureted hydrogen, need, therefore,
consist quantitatively only of one third part of one of the
last-mentioned gases. With mixtures of carbonic acid and
hydrogen, which occur more rarely, the limits of inflamma-
bility prove different again, on account of the capacity for
heat of the former. Acosta justly suggests .the question:
"Whether a tradition disseminated among the inhabitants
of Turbaco, descendants of the Indios dc Taruaco, according
to which the Volcancitos formerly all burned, and were con-
verted from Volcanes de fuego into Volcanes de agua, by be-
ing exorcised and sprinkled with holy water by a pious
monk,* may not refer to a condition which has now re-
turned ?" Single great eruptions of flames from mud vol-
canoes, which both before and since have been very inactive
(Taman, 1793; on the Caspian Sea, near Jokmali, 1827;
and near Baklichli, 1839; near Kuschtschy, 1846, also in
the Caucasus), present analogous examples.

The apparently unimportant phenomenon of the salses of
Turbaco has gained in geological interest by the terrible
eruption of flame, and the terrestrial changes which occurred
in 1839, more than 32 geographical miles to the NN.E. of
Cartagena de Indias, between this harbor and that of Saba-
nilla, not far from the mouth of the great Magdalena River.
The true central point of the phenomenon was the Cape
Galera Zamba, which projects 6 — 8 geographical miles into
the sea, in the form of a narrow peninsula. J"^or the knowl-
edge of this phenomenon we are also indebted to Colonel
Acosta, of whom science has unfortunately been deprived by
an early death. In the middle of the tongue of land there
stood a conical hill, from the crater of which smoke (vapors)
and gases soUietimes poured forth with such violence that,
boards and large pieces of wood which were thrown into it
were cast back again to a great distance. In the year 1839
the cone disappeared during a considerable eruption of fire,
and the entire peninsula of Galera Zamba became an island,

* Humboldt, Vues des CordiUeres et Monuments des peuples indigenes
de VAmerique, pi. xli., p. 239. The beautiful drawing of the Volcan-
citos de Turbaco, from which the copper-plate was engraved, was
made by my young fellow-traveler, Louis de Rieux. Upon the old
Taruaco in the first period of the Spanish Conquista, see Herrera, Dec.
i., p. 251.

SALSES. 209

separated from the continent by a channel of 30 feet in
depth. The surface of tlie sea continued in this peaceful
state until, on the 7th of October, 1848, at the place of the
previous breach, a second terrible eruption of flames* ap-
peared, without any perceptible earthquake in the vicinity,
lasted for several days, and was visible at a distance of from
40 to 50 miles. The salse only emitted gases, but no solid
matters. When the flames had disappeared the sea-bottom
was found to be raised into a small sandy islet, which how-
ever soon disappeared again. More than 50 volcancitos
(cones similar to those of Turbaco) now surround the sub-
marine gas volcano of Galera Zamba, to a distance of from
18 to 23 miles. In a geological point of view we may cer-
tainly regard this as the principal seat of the volcanic ac-
tivity which strives to place itself in contact with the atmos-
phere, over the whole of the low country from Turbaco to
beyond the delta of the Rio Grande de la Magdalena.

The uniformity of the phenomena which are presented in
the various stages of their activity, by the salses, mud vol-
canoes, and gas springs on the Italian peninsula, in the
Caucasus and in South America, is manifested in enormous
tracts of land in tlie Chinese empire. The art of man has
there from the most ancient periods known how to make use
of this treasure; nay, even led to the ingenious discovery
of the Chinese rope-boring, which has only of late become
known to Europeans. Borings of several thousand feet in
depth are produced by the most simple application of human
strength, or rather of the weight of man. I have elsewheref
treated in detail of this discovery, and also of the "fire
springs," Ho-tsing, and "fiery mountains," Ho-schan, of
Eastern Asia. They bore for water, brine springs, and in-
flammable gas, from the southwestern provinces, Yun-nan,
Kuang-si, and Szu-tschuan on the borders of Thibet, to the

* Lettre de M. Joaquin Acost-a a M. Elie de Beaumont, in the
Comptes rendus de I' Acad, des Sciences, t. xxix., 1849, p. 530-534.

t Humboldt, Asie Centrale, t. ii., p. 519-540; principally from ex-
tracts from Chinese works by Klaproth and Stanislas Julien. The
old Chinese rope-boring, which was repeatedly employed, and some-
times with advantage, in coal-pits in Belgium and Germany between
1830 and 1842, had been described (as Jobard has discovered) as early
as the 17th century, in the Relation of the Dutch embassador, Van
Hoorn ; but the most exact account of this method of boring the fire-
springs {Ho-tsirtg) is given by the French missionary, Imbert, who re-
sided so many years in Kia-ting-fu (see Annates de la Propagation de
la Foy, 1829, p. 369-381).

210 cos]5ios.

northern province Schan-si. When it has a reddish flame,
the gas often diffuses a bituminous odor ; it is- transferred
partly in portable and partly in lying bamboo tubes to re-
mote places, for use in salt-boiling, for heating the houses,
or for lighting the streets. In some rare cases supply of
carbureted hydrogen gas has been suddenly exhausted, or
stopped by earthquakes. Thus we know that a celebrated
Ho'tsing, situated to the southwest of the town of Khiung-
tscheu (latitude 50° 27^ longitude 101° 6^ East), which
was a salt spring burning with noise, was extinguished in
the 13th century, after it had illuminated the neighborhood
from the second century of our era. In the province of
Schan-si, which is so rich in coal, there are some ignited
carbonaceous strata. Fiery mountains {Ho-schan) are dis-
tributed over a great part of China. The flames often rise
to a great height, for example, in the mass of rock of tho
Py-kia-schan, at the foot of a mountain covered with perpet'
ual snow (lat. 31° 400, from long, open, inaccessible fissures.
a phenomenon which reminds us of the perpetual fire of the
Shagdagh mountain in the Caucasus.

On the island of Java, in the province of Samarang, at a
distance of about fourteen miles from the north coast, there
are salses similar to those of Turbaco and Galera Zamba.
Very variable hills of 25 to 30 feet in height throw out mud,
salt-water, and a singular mixture of hydrogen gas and car-
bonic acid* — a phenomenon which is not to be confounded
with the vast and destructive streams of mud which are
poured forth during the rare eruptions of the true, colossal
volcanoes of Java ( Gunung Kelut and Gunung Idjen). Some
mofette grottoes or sources of carbonic acid in Java are also
very celebrated, particularly in consequence of exaggerations
in the statements of some travelers, as also from their con-
nection with the myth of the Upas poison-tree, already men-
tioned by Sykes and Loudon. The most remarkable of the
six has been scientifically described by Junghuhn, the so-
called Vale of Death of the island (Fakaraman) in the mount-
ain Dieng, near Batur. It is a funnel-shaped sinking on the
declivity of a mountain, a depression in which the stratum
of carbonic acid emitted attains a very different height at

* According to Diard, Asie Centrale, t. ii., p. 515. Besides the mud
volcanoes of Damak and Surabaya, there are upon other islands of the
Indian Archipelago the mud volcanoes of Pulu-Semao, Pulu-Kam-
bing, and Pulu-Koti; see Junghuhn, Java, seine Gestalt undFflanzen-
decke, 1812, abth. iii., s. 830.

SALSES. 211

different seasons. Skeletons of wild hogs, tigers, and birds
are often found in it.* The poison-tree, ^oAoti (or better,
puhi) upas of the Malays (Antiatis toxicaria of the traveler
Leschenault de la Tour), with its harmless exhalations, has
nothing to do with these fatal actions.f

I conclude this section on the salses and steam and gas
springs with the description of an eruption of hot sulphur-
ous vapors, which may attract the interest of geognosists on
account of the kind of rock from which they are evolved.
During my delightful but somewhat fatiguing passage over
the central Cordillera of Quindiu (it took me 14 or 15 days
on foot, and sleeping constantly in the open air, to get over
the mountain crest of 11,500 feet from the valley of the Eio
Magdalena into the Cauca valley), when at the height of
6810 feet I visited the Azufral to the west of the station El
Moral. In a mica-schist of a rather dark color, which, re-
posing upon a gneiss containing garnets, surrounds, with
the latter, the elevated granite domes of La Ceja and La
Garita d^ Paramo, I saw hot sulphurous vapors flowing
out from the clefts of the rocks in a narrow valley (Que-
brada del Azufral). As they are mixed with sulphureted
hydrogen gas and much carbonic acid, a stupefying dizziness
is experienced on stooping down to measure the tempera-
ture, and remaining long in their vicinity. The tempera-
ture of the sulphurous vapors was 117°*7; that of the air
69° ; and that of the sulphurous brook, which is probably
cooled in the upper parts of its course by the snow-waters
of the volcano of Tolima, 84:°'6. The mica-schist, which
contains some pyrites, is permeated by numerous fragments
of sulphur. The sulphur prepared for sale is principally
obtained from an ochre-yellow loam, mixed with native sul-
phur and weathered mica-slate. The operatives (Mestizoes)
suffer from diseases of the eyes and muscular paralysis.

* Junghuhn, Op. cit, abth. i., s. 201, and abth. iii., s. 854-858.
The weaker suiFocating caves on Java are Gua-Upas and Gua-Galan
(the first word is the Sanscrit guhd, cave). As there can certainly be
no doubt that the Grotto del Cane, in the vicinity of the Lago di Ag-
nano, is the same that Pliny (ii., cap. 93) described nearly 18 centu-
ries ago, "in agro Puteolano," as "Charonea scrobis mortiferum
spiritum exhalans," we must certainly share in the surprise felt by
Scacchi {Memorie geol. suUa Cavipania, 1849, p. 48), that in a loose
soil, so often moved by earthquakes, so small a phenomenon (the sup-
ply of a small quantity of carbonic acid) can have remained unaltered
and undisturbed.

t Blume, Rumphia sive Comment, botanicce, t. i. (1835), p. 47-59.

212 COSMOS.

When Boussingault visited the Azufral de Quindiu, thirty
years after me (1831), the temperature of the vapors which
he analyzed* had so greatly diminished as to fall below that
of the open air (71°-6), namely to 66°— 68°. The same
excellent observer saw the trachytic rock of the neighboring
volcano of Tolima, breaking through the mica-schist, in the
Quebrada de Aguas calientes : just as I have very distinctly
seen the equally eruptive, black trachyte of the volcano of
Tunguragua covering a greenish mica-schist containing gar-
net near the rope bridge of Penipe. As sulphur has hither-
to been found in Europe, not in the primitive rocks, as they
were formerly called, but only in the tertiary limestone, in
gypsum, in conglomerates, and in true volcanic rocks, its
occurrence in the Azufral de Quindiu (4^° N. lat.) is the
more remarkable, as it is repeated to the south of the equa-
tor between Quito and Cuenca, on the northern slope of the
Paramo del Assuay. In the Azufral of the Cerro Cuello
(2° 13^ S. lat,), again in mica-schist, at an elevation of
7980 feet, I met with a vast bed of quartz,! in which the
sulphur is disseminated abundantly in scattered masses. At
the time of my journey the !ragments of sulphur measured
only 6 — 8 inches, but they were formerly found of as much
as 3 — 4 feet in diameter. Even a naphtha spring rises vis-
ibly from mica-schist in the sea-bottom in the Gulf of Cari-
aco, near Cumana. There the naphtha gives a yellow color
to the surface of the sea to a distance of more than a thou-
sand feet, and I found that its odor was diffused as far as the
interior of the peninsula of Araya.f

* Humboldt, Essai Geognostique sur h Gisement des Roches dans les
deux Hemispheres, 1823, p. 76 ; Boussingault, in the Annales de Chemie
et de Physique, t. Hi., 1833, p. 11.

t With regard to the elevation of Alausi (near Ticsan), on the Cerro
Cuello, see the " Nivellement barometrique, No. 206," in my Observ.
Astron., vol. i., p. 311.

X " The existence of a naphtha spring issuing at the bottom of -the
sea from a mica-schist, rich in garnets, and diffusing, according to
the expression of the historian of the Conquista, Oviedo, a " resinous,
aromatic, and medicinal liquid," is an extremely remarkable fact.
All those hitherto known belong to secondary mountains; and this
mode of stratification appeared to favor the idea that all the mineral
bitumens (Hatchett, Transact. Linnc^an Society, 1798, p. 129) were due
to the destruction of vegetable and animal mattei's, or to the ignition
of coal. The phenomenon of the Gulf of Cariaco acquires fresh im-
portance, if we bear in mind that the same so-called primitive stra-
tum contains subterranean fires, that the odor of petroleum is ex-
perienced frorfi time to time at the edge of ignited craters (for ex-
ample, in the eruption of Vesuvius in 1805, when the volcano threw

SALSES. 213

If we now cast a last glance at the kind of volcanic activ-
ity which manifests itself by the 'production of vapors and
gases, either with or without phenomena of combustion, Ave
find sometimes a great affinity, and sometimes a re*markable
difference in the matters escaping from fissures of the earth,
according as the high temperature of the interior, modifying
the action of the affinities, has acted upon homogeneous or
very composite materials. The matters which are driven to
the surface by this low degree of volcanic activity are : aque-
ous vapor in great quantity, chloryd of sodium, sulphur, car-
bureted and sulphureted hydrogen, carbonic acid and nitro-
gen ; naphtha (colorless or yellowish, or in the form of brown
petroleum) ; boracic acid and alumina from the mud volca-
noes. The great diversity of these matters, of which, how-
ever, some (common salt, sulphm'et^d hydrogen gas, and pe-
troleum) are almost always associated together, shows the
unsuitableness of the denomination salses, which originated
in Italy, where Spallanzani had the great merit of having
been the first to direct the attention of geognogists to this
phenomenon, which had been long regarded as so unimport-
ant, in the territory of Modena. The name vajwr (Md gas
springs is a better expression of the general idea. If many
of them, such as the Fumaroles, undoubtedly stand in rela-
tion to extinct volcanoes, and are even, as sources of carbon-
ic acid, peculiarly characteristic of a last stage of such vol-
canoes, others, on the contrary, appear to be quite independ-
ent of the true fiery mountains which vomit forth fused
earths. Then, as Abich has already shown in the Cauca-
sus, they follow definite directions in large tracts of country,
breaking out of fissures in rocks, both in the plains, even in
the deep basin of the Caspian Sea, and in mountain eleva-
tions of nearly 8500 feet. Like the true volcanoes, they
sometimes suddenly augment their apparently dormant ac-
tivity by the eruption of columns of fire, which spread ter-
ror all around. In both continents, in regions widely sep-
arated, they exhibit the same conditions following one upon

up scoriae), and that most of the very hot springs of South America
issue from granite (Las Trincheras, near Porto Cabello), gneiss, and
micaceous schist. More to the eastward of the meridian of Cu-
mana, in descending from the Sierra de Meapire, we first came to
the hollow ground (tierra hueca), which, during the great earthquakes
of 1766, threw up asphalt enveloped in viscous petroleum; and aft-
erward, beyond this ground, to an infinity of hydrosulphurous hot
springs (Humboldt, Halation Historique^ t. i., p. 136, 344, 347, and
447).

214 COSMOS.

the other; but no observation has hitherto justified us in
supposing that they are the forerunners of the formation of
true volcanoes vomiting lava and cinders. Their activity
is of anofher kind, perhaps originating at a smaller depth,
and caused by different chemical processes.

d. Volcanoes, according to the difference of their formation and
activity. — Action hy fismres and caldron-like depressions. —
Circumvallation of the craters of elevation. — Volcanic conical
and bell-shaped Mountains, with open or closed summits. —

Difference of the Hocks through which Volcanoes act.

/
(Amplification of the Eepresentation of Nature, Cosmos, vol. i., p.
228-248.)

Among the various specific manifestations of force in the
reaction of the interior of our planet upon its uppermost
strata, the mightiest is that presented by the true volcanoes ;
that is to say, those openings through -which, besides gases,
solid massQS of various materials are forced up from un-
measured depths to the surface, either in a state of igne-
ous fision, as lava streams, or in the form of cinders, or as
products of the finest trituration (ashes). If we regard the
words volcano and fery mountain as synonymous, in accord-
ance with the old usage of Speech, we thus, according to a
preconceived and very generally diffused opinion, attach to
the idea of volcanic phenomena the picture of an isolated
conical mountain, with a circular or oval orifice at the sum-
mit. Such views, however, lose their universality when the
observer has the opportunity of wandering through connect-
ed volcanic districts, occupying a surface of many thousand
square geographical miles ; for example, the entire central
part of the highlands of Mexico, between the Peak of Ori-
zaba, JoruUo, and the shores of the South Sea; or Central
America ; or the Cordilleras of New Granada and Quito,
between the Volcano of Purace, near Popayan, that of Pasto
and Chimborazo ; or the isthmian chain of the Caucasus, be-
tween the Kasbegk, Elburuz, and Ararat. In Lower Italy,
between the Phlegr^ean Fields of the main land of Campa-
nia, Sicily, and the islands of Lipari and Ponza, as also in
the Greek Islands, part of the intervening land has not been
elevated with the volcanoes, and part of it has been swallow-
ed by the sea.

In the above-mentioned great districts of America and
the Caucasus, masses of eruptions (true Trachytes, and not

VOLCANOES. 215

trachytic conglomerates ; streams of obsidian ; quarried
blocks of pumice-stone, and not pumice-bowlders transported
and deposited by water) make their appearance, seeming to
be quite independent of the mountains, which only rise at a
considerable distance. AVhy should not the surface have
been split in many directions during the progressive refriger-
ation of the upper strata of the earth by radiation of heat,
before the elevation of isolated mountains or mountain chains
had yet taken placet Why should not these fissures have
emitted masses in a state of igneous fusion, which have hard-
ened into rocks and eruptive stones (trachyfe, dolerite, mela-
phyre, raargarite, obsidian, and pumice)? A portion of
these trachytic or doleritic strata which have broken, out in
a viscid fluid state, as if from earth-springs,* and which
were originally deposited in a horizontal position, have,
during the subsequent elevation of volcanic cones and bell-
shaped mountains, b^n tilted into a position which by no
means belongs to the more recent lavas produced from ig-
neous mountains. Thus, to advert, in the first place, to a
very well known European example, in the Yal del Bove on
-^tna (a depression which cuts deeply into the interior of
the mountain), the declination of the strata of lava, which
alternate very regularly with masses of bowlders, is 25° to
30°, while, according to Elie de Beaumont's exact determ-
inations, the lava streams which cover the surface of ^tna,
and which have only flowed from it since its elevation in the
form of a mountain, only exhibit a declination of 3° to 5°
on an average of 30 streams. These conditions indicate the
existence of very ancient volcanic formations, which have
broken out from fissures, before the production of the vol-
cano as an igneous mountain. A remarkable phenomenon of
this kind is also presented to us by antiquity — a phenomenon
which manifested itself on Euboea, the modern Negropont,
in an extended plain, situated at a distance from all active
and extinct volcanoes. " The violent earthquakes, which
partially shook the island, did not cease until an abyss,
which had opened on the plain of Lelantus, threw up a
stream of glowing mud (lava)."t

* Cosmos, vol. i., p. 231.

t Strabo, i., p. 58, ed. Casaub. The epithet SiaTrvpoQ proves that
in this case mud volcanoes are not spoken of. Where Plato, in his
geognostic phantasies, alludes to these, mixing mythical matter with
observed facts, he says distinctly (in opposition to the phenomenon
described by Strabo) vypov rrrjXov iroTajxoi. Upon the denominations
'rrrikoQ and pya?, as volcanic emissions, I have treated on a former oc-

216

COSMOS.

If the oldest formations of eruptive rock (often perfectly
similar to the more recent lavas in its composition), which
also in part occupy veins, are to be ascribed to a previous
fissure of the deeply-shaken crust of the earth, as I have
long been inclined to think, both these fissures and the less
simple craters of elevation subsequently produced must be
regarded only as volcanic eruptive orijices, not as volcanoes
themselves. The principal character of these last consists
in a connection of the deep-seated focus with the atmosphere,
which is either permanent, or at least renewed from time to
time. For this plirpose the volcano requires a peculiar frame-
work ; for, as Seneca* says very appropriately, in a letter to
Lucilius, "ignis in ipso monte non aliraentum habet, sed
viam." The volcanic activity exerts, therefore, a formative
action by elevating the soil ; and not, as was at one time uni-
versally and exclusively supposed, a building action by the
accumulation of cinders, and new stra^ of lava, superposed
one upon the other. The resistance experienced In the canal
of eruption, by the masses in a state of igneous fluidity when
forced in excessive quantities toward the surface, gives rise to
the increase in the heaving force. A " vesicular inflation of
the soil" is produced, as is indicated by the regular outward
declination of the elevated strata. A mine-like explosion,
the bursting of the central and highest part of the convex
inflation of the soil, gives origin sometimes only to what
Leopold von Buch has called a crater of elevation,^ that is to

casion {Cosmos, vol. i., p. 237), and I shall only advert here to an-
other passage in Strabo (vi., p. 269), in which hardenii% lavff, called
TrrjXbg /xkXaQ, is most distinctly characterized. In the description of
JEtna Ave find : " The red-hot stream (pva^) in the act of solidifica-
tion converts the surface of the earth into stone to a considerable
depth, so that whoever wishes to uncover it must undertake the labor
of quarrying. For, as in the cratei-s, the stone is molten and then up-
heaved, the fluid streaming from the summit is a black excrementitious
mass (ttt/Xoc) falling down the mountain, which, afterward hardening,
becomes a millstone, and retains the same color that it had before."

* Cosmos, vol. i., p. 239.

t Leopold von Buch, On Basaltic Islands and Craters o/ Elevation,
in the Abhandl. der konig. Akad. der Wiss. zu Berlin, 1818-1819, s. 51 ;
and Physikalische Beschreibuuff der canarischen Inseln, 1825, s. 213, 262,
284, 313, 323, and 341. This work, which constitutes an era in the
profound knowledge of volcanic phenomena, is the fruit of a voyage
to Madeira and TenerifFe, from the beginning of April to the end of
October, 1815 ; but Naumann indicates with much justice, in his Lehr-
buch der Geognosie, that in the letters written in 1802 by Leopold von
Buch, from Auvergne (Gcognostische Beobachtung auf Reisen durch
Deutschland und Italien, bd. ii., s. 282), in reference to the description

CRATERS OF ELEVATION. 217

say, a crater-like, round or oval depression, bounded by a
circle of elevation, a ring-shaped wall, usually broken down
in places ; sometimes (when the frame-work of a permanent
volcano is to be completed) to a dome-shaped or conical
mountain in the middle of the crater of elevation. The
latter is then generally open at its summit, and on the bot-
tom of this opening (the crater of the permanent volcano)
rise transitory hills of eruption and hills of scoriae, small and
large cones of eruption, which, in Vesuvius, sometimes far
exceed the margins of the crater of the cone of elevation.
The signs of the first eruption, the old frame-work, are not,
however, always retained. The high wall of rock which sur-
rounds the inner circular wall (the crater of elevation) is not
recognizable, even in scattered detritus, on many of the larg-
est and most active volcanoes.

It is a great merit of modern times not only to have more
accurately investigated the peculiar conditions of the forma-
tion q£^volcanoes by h. careful comparison of those which are
widely separated from each other, but also to have intro-
duced more definite expressions into language, by which the
heterogeneous features of the general outline, as well as the
manifestations of volcanic activity, are distinguished. If we

of Mont d'Or, the theory of craters of elevation and their essential dif-
ference from the true volcanoes was already expressed. An instruct-
ive counterpart to the three craters of elevation of the Canary Islands
(on Gran Canaria, Teneriffe, and Palma) is furnished by the Azores.
The admirable maps of Captain Vidal, for the publication of which we
are indebted to the English Admiralty, elucidate the wonderful geog-
nostic-construction of these islands. On San Michael is situated the
enormous Caldeira das sete Cidades which was formed in the year 1444,
almost under Cabral's eyes, a crater of elevation which incloses two
lakes, the Lagoa grande and the Lagoa azul, at a height of 876 feet.
The Caldeira de Corvo, of which the dry part of the bottom is 1279
feet high, is almost of the same circumference. Nearly three times
this height are the craters of elevation of Fayal and Terceira. To the
same kind of eruptive phenomena belong the innumerable but ephem-
eral platforms which were visible only by day, in 1691, in the sea
around the island of San George, and in 1757 around San Michael.
The periodical inflation of the sea-bottom, scarcely four miles to the
west of the Caldeira das sete Cidades, producing a larger and some-
what more permanent island (Sabrina), has already been mentioned
(Cos7nos, vol. i., p. 242). Upon the crater of elevation of Astruni, in
the Phlegraean plains, and the trachytic mass driven up in its centre,
as an unopened bell-shaped hill, see Leopold von Buch, in Poggend.,
Annaleri, bd. xxxvii., s. 171 and 182. A fine crater of elevation is that
of Rocca Monfina, measured and figured in Abich's Geolog. JBeobacht.
iiber die Vulkan. Erschein. in Unter-und Mittel Italieuy 1841, bd. i., s.
113, taf. ii.

Vol. v.— K

218 COSMOS.

are not decidedly disinclined to all classifications, because in
the endeavor after generalization these always rest only upon
imperfect indications, we may conceive the bursting forth of
fused masses and solid matter, vapors and gases, in four dif-
ferent ways. Proceeding from the simple to the complex
phenomena, we may first mention eruptions from fissures,
not forming separate series of cones, but producing volcanic
rocks superlying each other, in a fused and viscid state;
secondly, eruptions through heaped-up cones, without any cir-
cumvallation, and yet emitting streams of lava, as was the
case for five years during the destruction of the island of
Lancerote, in the first half of the last century ; thirdly, cra-
ters of elevation, with upheaved strata, but without central
cones, emitting streams of lava only on the outside of the
circumvallation, never from the interior, which is soon closed
up with detritus; fourthly, closed hell-shaped mountains or
cones of elevation, open at the summit, either inclosed by a
circular wall, which is at least partially retained — aa||n the
Pic of TenerifFe, in Fpgo, and Rocca Monfina ; or entirely
without circumvallation or crater of elevation — as in Ice-
land,* in the Cordilleras of Quito, and the central parts of
Mexico. The open cones of elevation of this fourth class
maintain a permanent connection between the fiery interior
of the earth and the atmosphere, which is more or less effect-
ive at undetermined intervals of time. Of the dome-shaped
and bell-shaped trachytic and doleritic mountains which have
remained closed at the summit, there appear, according to
my observations, to be more than of the open cones, whether
active or extinct, and far more than of the true volcanoes.
Dome-shaped and bell-shaped mountains, such as Chimbora-
zo, Puy de D6me, Sarcouy, Rocca Monfina, and Vultur, give
the landscape a peculiar character, by which they contrast
pleasingly with the schistose peaks, or the serrated forms of
limestone.

In the tradition preserved to us so picturesquely by Ovid
regarding the great volcanic phenomenon of the peninsula of
Methone, the production of such a bell-shaped hnd unopen-
ed mountain is indicated with methodical clearness. " The
force of the winds imprisoned in dark caves of the earth, and
seeking in vain for an opening, drive up the heaving soil
(extentam tumefecit humum), as when one fills a bladder or
leather bag with air. By gradual hardening the high pro-

* Sartorius von Waltershausen, Physisch-geographische Skizze von
Island, 1847, s. 107.

CRATERS OF ELEVATION. 219

jecting eminence has retained the form of a hill." I have
already elsewhere adverted to the fact of how completely
different this Roman representation is from Aristotle's nar-
ration of the volcanic phenomenon upon Hiera, a newly-
formed ^olic (Liparian) island, in which " the subterranean,
mightily urging blast does indeed also raise a hill, but after-
ward breaks it up to pour forth a fiery shower of ashes."
The elevation is here clearly represented as preceding the
eruption of flame {Cosmos, vol. i., p. 241.). According to
Strabo, the elevated dome-like hill of Methana had also
opened in fiery eruptions, at the close of which an agreeable
odor was diffused in the night-time. It is very remarkable
that the latter was observed under exactly similar circum-
stances during the volcanic eruption of Santorin, in the au-
tumn of 1650, and was denominated " a consoling sign, that
God would not yet destroy his flock," in the penitential ser-
mon delivered and written shortly afterward by a monk.*

* It has been a much disputed point to what particular locality of
the plain of Troezeu, or the peninsula of Methana, the description of
the Roman poet may refer. My friend, Ludwig Ross, the great Greek
antiquarian and chorograph, who has had the advantage of many
travels, thinks that the immediate vicinity of Troezen presents no
locality which can be referred to as the bladder-like hills, and that,
by a poetic license, Ovid has removed the phenomenon described with
such truth to nature to the plain. " To the south of the peninsula of
Methana, and east of the plain of Troezen," writes Ross, " lies the
island Calauria, well known as the place where Demosthenes, being
pressed by the Macedonians, took poison in the temple of Neptune.
A narrow arm of the sea separates the limestone rocks of Calauria
from the coast ; from this arm of the sea (passage, Tropoc) the town
fend island take their present name. In the middle of the strait,
\ united with Calauria by a low causeway, probably of artificial origin,
lies a small conical islet, comparable in form to an egg cut through
the middle. It is volcanic throughout, consisting of grayish yellow
and yellowish red trachyte, mixed with eruptions of lava and scoriae,
and is almost entirely destitute of vegetation. Upon this islet stands
the present town of Poros, on the place of the ancient Calauria. The
formation of the islet is exactly similar to that of the more recent
volcanic islands in the. Bay of Thera (Santorin). In his animated
description, Ovid has jjrobably followed a Greek original or an old
tradition" (Ludw. Ross, in a letter to me dated November, 1845)
As a member of the French scientific expedition, Virlet has set up
the opinion that the volcanic upheaval may have been only a subse-
quent increase of the trachytic mass of the peninsula of Methana.
This increase occurs in the northwest extremity of the peninsula,
where the black burned rock, called Kammeni-petra, resembling the
Kammeni, near Santorin, betrays a more recent origin. Pausanias
communicates the tradition of the inhabitants of Methana, that, on
the north coast, before the now-celebrated sulphurous springs burst

220 COSMOS.

Does not this pleasant odor afibrd indications of naphtha?
The same thing is also referred to by Kotzebue, in his Rus-
sian voyage of discovery, in connection with an igneous
eruption (1804) of the volcanic island of Umnack, newly
elevated from the sea in the Aleutian Archipelago. During
the great eruption of Vesuvius, on the 12th August, 1805,
which I observed in company with Gay-Lussac, the latter
found a bituminous odor prevailing at times in the ignited
crater. I bring together these little-noticed facts, because
they contribute to confirm the close concatenation of all
manifestations of volcanic activity, the intimate connection
of the weak salses and naphtha springs with the true vol-
canoes.

Circumvallations, analogous to those of the craters of ele-
vation, also present themselves in rocks which are very dif-
ferent from trachyte, basalt, and porphyritic schists ; for ex-
ample, according to Elie de Beaumont's acute observation,
in the granite of the French Alps. The mountain mass of
Oisans, to which the highest* summit of France, Mont Pel-
-voux, near Briangon (12,905 feet), belongs, forms an amphi-
theatre of thirty-two geographical miles in circumference, in
the centre of which is situated the small village of La Be-
rarde. The steep walls of this circular space rise to a height
of more than 9600 feet. The circumvallation itself is gneiss;
all the interior is granite.f In the Swiss and Savoy Alps
the same formation presents itself repeatedly in small dimen-
sions. The Grand Plateau of Mont Blanc, in which Bravais

forth, fire rose out of the earth (see Curtius, Peloponnesos, bd. i., s. 42
and 46). On the " indescribable pleasant odor" which followed the
stinking sulphurous odor, near Santorin (September, 1650), see Ross,
Reisen aufden Griech. Inseln des agaischen Meeres, bd. i., s. 196. Upon
the odor of naphtha in the fumes of the lava of the Aleutian island
Umnack, which appeared in 1796, see Kotzebue's Entdeckungs-Reise,
bd. ii., s. 106, and Leopold de Buch, Description phys. des lies Cana-
ries, p. 458.

* The highest summit of the Pyrenees, that is, the Pic de Nethou
(the eastern and highest peak of the Maladetta or Malahita group), has
been twice measured trigonometrically ; its height, according to Ee-
boul, is 11,443 feet (3481 metres), and, according to Coraboeuf, 11,167
feet (3404 metres). It is, therefore, 1705 feet lower than Mont Pel-
voux, in the French Alps, near Brian9on. The next in height to the
Pic de Nethou, in the Pyrenees, are the Pic Posets or Erist, and of
the group of the Marbore', the Montperdu, and the Cylindre.

t Memoire pour servir a la Description Geologique de la France^ t. ii.,
p. 339. Upon "valleys of elevation" and " encircling ridges" in the
Silurian formation, see the admirable description of Sir Roderick Mur-
chison in " The Silurian System," pt. i., p. 427-442.

MAARS. 221

and Martins encamped for several days, is a closed amphi-
theatre with a nearly flat bottom, at an elevation of nearly
12,811 feet; from the midst of which the colossal pyramid
of the summit rises.* Tlie same upheaving forces produce
similar forms, although modified by the composition of the
different rocks. The annular and caldron-like valleys (val-
leys of elevation) described by Hoffman, Buckland, Murchi-
son, and Thurmann, in the sedimentary rocks of the north
of Germany, in Herefordshire, and the Jura mountains of
Porrentruy, are also connected with the phenomena here de-
scribed, as well as, although with a less degree of analogy,
some elevated plains of the Cordilleras inclosed on all sides
by mountain masses, in which are situated the towns of
Caxamarca (93G2 feet), Bogota (8729 ^eet), and Mexico
(7469 feet), and in the Himalayas the caldron-like valley of
Caschmir (5819 feet).

Less related to the craters of elevation than to the above
described simplest form of volcanic activity (the action from
mere fissures) are the numerous Maars among the extinct
volcanoes of the Eifel — caldron-like depressions in non-vol-
canic rock (Devonian slate), and surrounded by slightly ele-
vated margins, -formed by themselves. *' These are, as it
were, the funnels of mines, indications of mine-like erup-
tions," resembling the remarkable phenomenon described by
me of the human bones scattered upon the hill of La Culcaf
during the earthquake of Riobamba (4th February, 1797).
When single Maars, not situated at any great height, in the
Eifel, in Auvergne, or in Java, are filled with water, such
former craters of explosion may in this state be denominated
crateres-lacs ; but it seems to me that this term should not

* Bravais and Martins, Ohserv. fakes au Sommet et au Grand Pla-^m
teau du Mont Blanc, in the Annuaire MdMorol. de la France pour 1850,
p. 131.

t Cosmos, vol. v., p. 173. I have twice visited the volcanoes of the
Eifel, Avhen geognosy was in very different states of development, in
the autumn of 1794, and in August, 1845 ; the first time in the vicin-
ity of the Lake of Laach and the monastery there, which was then
still inhabited by monks ; the second time in the neighborhood of
Bertrich, the Mosenberg, and the adjacent Maars, but never for more
than a few days. As in the latter excursion I had the good fortune
to be able to accompany my intimate friend, the mining •surveyor.
Von Dechen, I have been enabled by many years' correspondence,
and the communication of important manuscript memoirs, to make
free use of the observations of this acute geognosist. I have often in-
dicated by quotation marks, as is my wont, what I have borrowed,
word for word, from his communications.

222 COSMOS.

be taken as a synonymous name for Maar, as small lakes
have been found by Abich and myself on the summits of the
highest volcanoes, on true cones of elevation in extinguished
craters ; for example, on the Mexican volcano of Toluca at
an elevation of 12,246 feet, and on the Caucasian Elburuz
at 19,717 feet. In the volcanoes of the Eifel we must- care-
fully distinguish from each other two kinds of volcanic ac-
tivity of very unequal age — the true volcanoes emitting
streams of lava, and the weaker eruptive phenomena of the
Maars. To the former belong the basaltic stream of lava,
rich in olivin, and cleft into upright columns, in the valley
of Uesbach, near Bertrich ;* the volcano of Gerolstein, which
is seated in a limestone containing dolomite, deposited in the
form of a basin in the Devonian gray wacke schists ; and the
long ridge of the Mosenberg (1753 feet above the sea), not
far from Bettenfeld, to the west of Manderscheid. The last-
named volcano has three craters, of which the first and sec-
ond, those furthest to the north, are perfectly round, and
covered with peat mosses ; while from the third and most
southernf crater there flows down a vast, reddish brown,
deep stream of lava, separated into a columnar form, toward
the valley of the little Kyll. It. is a remarkable phenome-
non, foreign to lava-producing volcanoes in general, that nei-
ther on the Mosenberg nor on the Gerolstein, nor in other
true volcanoes of the Eifel, are the lava eruptions visibly sur-
rounded at their origin by a trachytic rock, but, as far as
they are accessible to observation, proceed directly from the
Devonian strata. The surface of the Mosenberg does not at
all prove what is hidden in its depths. The scoriae contain-
ing augite, which by cohesion pass into basaltic streams,
contain small, calcined fragments of slate, but no trace of
•inclosed trachyte. Nor is the latter to be found inclosed in
the crater of the Rodderberg, notwithstanding that it lies
in the immediate vicinity of the Siebengebirge, the greatest
trachytic mass of the Rhine district.

"The Maars appear," as the mining surveyor Von De-

* H. von Dechen, Geognost. Uebersicht der Umgegend von Bad Ber-
trich, 1847, s. 11-51.

t Steni^cel, in Niiggerath, das Gebirge von Rheinland und WestphaleUy
bd. i., s, 79, taf. iii. See also C. von Oeynhausen's admirable expla-
nations of his jreognostic Map of the Lake of Laach, 1847, p. 34, 39,
and 42, including the Eifel and the basin of Neuwied. Upon th«
Maars, see Steininger, Geognostische Beschreibung der Eifel, 1853, s.
113. His earliest meritorious work, '■'■Die erloschenen Vulkane in d«r
Eifel und am Nieder-Rhein,'' belongs to the year 1820.

MAARS.

223

chen has ingeniously observed, " to belong in their formation
to about the same epoch as the eruption of the lava streams
of the true volcanoes. Both are situated in the vicinity of
deeply-cut valleys. The lava-producing volcanoes were de-
cidedly active at a time when the valleys had already at-
tained very nearly their present form ; and we also see the
most ancient lava streams of this district pouring down into
the valleys." The Maars are surrounded by fragments of
Devonian slates, and by heaps of gray sand and tufa mar-
gins. The Laacher lake, whether it be regarded as a large
Maar, or, with my old friend C. von Oeynhausen, as part of
a large caldron-like valley in the clay-slate (like the basin
of Wehr), exhibits some volcanic eruptions of scorice upon
the ridge surrounding it, as is the case on the Krufter Ofen,
the Veitskopf, and Laacher Kopf It is not, however, mere-
ly the entire want of lava streams, such as are to be ob-
served on the Canary Islands upon the outer margin of true
craters of elevation and in their immediate vicinity — it is
not the inconsiderable elevation of the ridge surrounding the
Maar, that distinguishes this from craters of elevation ; the
margins of the Maars are destitute of a regular stratifica-
tion of the rock, falling, in consequence of the upheaval, con-
stantly outward. The Maars sunk in the Devonian slate
appear, as has already been observed, like the craters of
mines, into which, after the violent explosion of hot gases
and vapors, the looser ejected masses {EapUli) have for the
most part fallen back. As examples I shall only mention
here the Immerather, the Pulvermaar, and the Meerfelder
Maar. In the centre of the first mentioned, the diy bottom
of which, at a depth of two hundred feet, is cultivated, are
situated the two villages of Ober- and Unter-Immerath.
Here, in the volcanic tufa of the vicinity, exactly as on the
Laacher lake, mixtures of feldspar and augite occur in sphe-
roids, in which particles of black and green glass are scat-
terecl. Similar spheroids of mica, hornblende, and augite,
full of vitrified portions, are also contained in the tufa veins
of the Pulvermaar near Gillenfeld, which, however, is en-
tirely converted into a deep lake. The regularly circular
Meerfelder Maar, covered partly with water and partly with
peat, is characterized geognostically by the proximity of the
three craters of the great Mosenberg, the most southern of
which has furnished a stream of lava. The Maar, however,
is situated 639 feet below the long ridge of the volcano, and
at its northern extremity, not in the axis of the series of

224 COSMOS.

craters, but more to the northwest. The average elevation
of the Maars of the Eifel above the surface of tlie sea falls
between 922 feet (Laacher lake ?) and 1588 feet (Mosbrucher
Maar).

As this is peculiarly the place in which to call attention
to the uniformity and agreement exhibited by volcanic ac-
tivity in its production of material results, in the most dif-
ferent forms of the outer frame-work (as Maars, as circum-
vallated craters of elevation, or cones opened at the sum-
mit), I may mention the remarkable abundance of crystal-
lized minerals which have been thrown out by the Maars in
their first explosion, and which still in part lie buried in the
tufas. In the environs of the Laacher lake this abundance
is certainly greatest ; biit other Maars also, for example the
Immerather, and the Meerfelder Maar, so rich in bombs of
olivin, contain fine crystallized masses. We may here men-
tion zircon, hauyne, leucite,* apatite, nosean, olivin, augite,
ryacolite, common feldspar (orthoclase), glassy feldspar (san-
idine), mica, sodalite, garnet, and titanic iron. If ^e num-
ber of beautifully crystallized minerals on Vesuvius be so
much greater (Scacchi counts 43 species), we must not for-
get that very few of them are ejcclod from the volcano, and
that the greater number belongs to the portion of the so-
called eruptive matters of Vesuvius, which, according to the

* Leucite (of tlie same kind from Vesiivius, from Rocca di Papa in
the Albanian mountains, from Viterbo, from the Rocca Monfina, ac-
cording to Pilla, sometimes of more than three inches in diameter,
and from the dolerite of the Kaiserstuhl, in the Breisgau) occurs also
"in position as leucite- rock in the Eifel, on the Burgberg, near Rie-
den. The tufa in the Eifel incloses large blocks of leucitophyre near
Boll and Weibern." I can not resist the temptation to borrow the
following important observation from a chemico-geognostic mei#oir
read by Mitscherlich a few weeks since before the Academy of Ber-
lin: " Aqueous vapors alone may have effected the eruptions of the
Eifel, but they would have divided olivin and augite into the finest
drops and powder if they had met with them in a fluid-state. With
the fundamental inass of the erupted matters fragments of the old,
broken-up rock are most intimately mixed, for example, on the Drei-
ser Weiher, and these are frequently caked together. The larger ol-
ivin masses and the masses of augite even usually occur surrounded
by a thick crust of this mixture ; a fragment of the old rock never oc-
curs in the olivin or augite ; both were consequently formed before
they reached the spot where the breaking up took place. Olivin and
augite had, therefore, separated from the iluid^ basaltic mass before
this met with an accumulation of water or a spring which caused its
expulsion." See also upon the bombs an older memoir by Leonard
Horner, in the Transactions of the Geological Society^ 2d serie^ vol.
iv., pt. 2, 1836, p. 467.

MAARS. 225

opinion of Leopold von Buch,* "are quite foreign to Vesu-
vius, and to be referred to a tufaceous covering diifused far
beyond Capua, which was upheaved by the rising cone of
Vesuvius, and has probably been produced by a deeply-seat-
ed submarine volcanic action."

Certain definite directions of the various phenojnena of
volcanic activity are unmistakable even in the EifeiT ^' The
eruptions producing lava streams of the Upper Eifel lie in
one fissure, nearly 32 English miles in length, from Bert-
rich to the Goldberg, near Ormond, directed from southeast
to northwest ; on the other hand, the Maars, from the Meer-
felder Maar to Mosbruch and the Laacher lake, follow a line
of direction from southwest to northeast. These two pri-
mary directions intersect each other in the three Maars of
Daun. In the neighborhood of the Laacher lake trachyte
is nowhere visible on the surface. The occurrence of this
rock below the surface is only indicated by the peculiar na-
ture of the perfectly feldspar-like pumice-stone of Laach,
and by the bombs of augite and feldspar thrown out. But
the trachytes of the Eifel, composed of feldspar and large
crystals of hornblende, are only visibly distributed among
basaltic mountains: as in the Sellberg (1893 feet), near
Quiddelbach ; in the rising ground of Struth, near Kelberg ;
and in the wall-like mountain chain of lieimerath, near
Boos."

Next to the Lipari and Ponza islands few parts of Europe
have probably produced a greater mass of pumice-stone than
this region of Germany, which, with a comparatively small el-
evation, presents such various forms of volcanic activity in its
Maars {crateres d' explosion), basaltic rocks, and lava-emitting
volcanoes. The principal mass of the pumice-stone is situ-
ated between Nieder Mendig and Sorge, Andernach and Rii-
benach ; the principal mass of the duckstein, or Trass (a very
recent conglomerate, deposited by water), lies in the valley
of Brohl, from its opening into the Rhine upward to Burg-
brohl, near Plaidt and Kruft. The Trass formation of the
Brohl valley contains, together with fragments of graywacke-
slate and pieces of wood, small fragments of pumice-stone,
differing in nothing from the pumice-stone which constitutes
the superficial covering of the region, and even that of the

* Leopold von Buch, in Po.Gjgend., Annalen, bd. xxxvii., s. 179. Ac-
cording to Scacchi, the eruptive matters belong to the first outbreak
of Vesuvius in the vear 79. Leonhard's Neues Jahrbuchjur Mineral, j
1853, s. 259.

Ji:2

226

COSMOS.

duckstein itself. Notwithstanding some analogies "which
the Cordilleras appear to present, I have always doubted
whether the Trass can be ascribed to eruptions of mud from
the lava-producing volcanoes of the Eifel. I rather suppose,
with H. von Dechen, that the pumice-stone was thrown out
dry, andthat the Trass was formed in the same way as oth-
er congWnerates. " Pumice-stone is foreign to the Sieben-
gebirge ; and the great pumice eruption of the Eifel, the
principal mass of which still lies above the loess (Trass) and
alternates therewith in particular parts, may, in accordance
with the presumption to which the local conditions lead,
have taken place in the valley of the Rhine, above Neuwied,
in the great Neuwied basin, perhaps near Urmits, on the left
bank of the Rhine. From the friability of the material, the
place of eruption may have disappeared without leaving any
traces by the subsequent action of the current of the Rhine.
In the entire tract of the Maars of the Eifel, as in that of
its volcanoes from Bertrich to Ormond, no pumice-stone is
found. That of the Laacher lake is limited to the rocks
upon its margin ; and on the other Maars the small frag-
ments of feldspathic rock, which lie in the volcanic sand and
tuff, do not pass into pumice."

We have already touched upon the relative antiquity of the
Maars and of the eruptions of the lava streams, which differ so
much from them, compared with that of the formation of the
valleys. " The trachyte of the Siebengebirge appears to be
much older than the valley formation, and even older than the
Rhenish brown coal. Its appearance has been independent
of the cutting of the valley of the Rhine, even if we should
ascribe this valley to the formation of a fissure. The forma-
tion of the valleys is more recent than the Rhenish brown
coal, and more recent than the Rhenish basalt ; but older
than the volcanic eruptions with lava streams, and older than
the great pumice eruption and the Trass. Basalt formations
decidedly extend to a more recent period than the formation
of trachyte, and the principal mass of the basalt is, therefore,
to be regarded as younger than the trachyte. In the pres-
ent declivities of the valley of the Rhine many basaltic groups
(the quarry of Unkel, Rolandseck, Godesberg) were only laid
bare by the opening of the valley, as up to that time they were
probably inclosed in the Devonian graywacke rocks."

The infusoria, whose universal diffusion, demonstrated by
Ehrenberg, upon the continents, in the greatest depths of the
sea, and in the upper strata of the atmosphere, is one of the

MAARS. 227

most brilliant discoveries of our time, have their principal
seat in the volcanic Eifel, in the Rapilli, Trass strata, and
pumice conglomerates. Organisms with silicious shields fill
the valley of Brohl and the eruptive matters of Hochsim-
mer ; sometimes, in the Trass, they are mixed with uncar-
bonized twigs of coniferse. According to Ehrenberg, the
whole of this microcosm is of fresh- water formation, and
marine Polythalamia^ only show themselves exceptionally
in the uppermost deposit of the friable, yellowish loess at the
foot and on the declivities of the Siebengebirge (indicating
its former brackish coast nature). ^

Is the phenomenon of Maars limited to "Western Germa-
ny ? Count Montlosier, who was acquainted with the Eifel
by personal observations in 1819, and who pronounces the
Mosenberg to be one of the finest volcanoes that he ever saw
(like Rozet), regards the Gouffre de Tazenat^ the Lac Pavin
and Lac de la Godivel, in Auvergne, as Maars or craters of
explosion. They are cut into very difierent kinds of rock —
in granite, basalt, and domite (trachytic rock), and surround-
ed at the margins with scorias and rapilli.f

The frame-works, which are built up by a more powerful
eruptive activity of volcanoes, by upheaval of the soil and
emission of lava, appear in at least six difierent forms, and
reappear with this variety in their forms in the most distant
zones <Jf the earth. Those who are born in volcanic districts,
among basaltic and trachytic mountains, are often genially
impressed in spots where the same forms greet them. Mount-
ain forms are among the most important determining elements
of the physiognomy of nature — they give the district either a
»

* Upon the antiquity of formation of the valley of the Rhine, see
H. von Dechen, Geognost. Beschreihung dcs SieLengebirges, in the Ver-
handl. des Naturhist. Vereins dcr Preuss. Wieinlande und Westphakns,
1852, s. 556-559. The infusoria of the Eifel are treated of by Ehren-
berg in the Monatsher. der Akad. de?- Wiss. zu Berlin, 1844, s. 337 ;
1845, s. 133 and 148 ; and 184G, s. 161-171. The Trass of Brohl,
which is filled with crumbs of pumice-stone containing infusoria,
forms hills of as much as 850 feet in height.

t See Rozet, in the Memoir es de la Socicte Geologique, 2m e serie, t.
i., p. 119. On the island of Java also, that wonderful seat of multi-
farious volcanic activity, there occur "craters without cones, as it were
flat volcanoes" (Junghuhn, Java, seine, Gestalt und Pflanzendecke, Lief,
vii., p. 640), between Gunung Salak and Perwakti, analogous to the
Maars as " craters of explosion." Destitute of any elevated margins,
they are situated partly in perfectly flat districts of the mountains,
have angular fragments of the burst rocky strata scattered around
them, and now only emit vapors and gases.

228 COSMOS.

cheerful, or a stern and magnificent character, according as
they are adorned with vegetation or surrounded by a dreary
barrenness. I have quite recently endeavored to bring to-
gether in a separate atlas a number of outlines of the Cordil-
leras of Quito and Mexico, sketched from my own drawings.
As basalt occurs sometimes in conical domes somewhat
rounded at the summit, sometimes in the form of closely-
arranged twin-mountains of unequal elevation, and some-
times in that of a long horizontal ridge bounded at each ex-
tremity by a more elevated dome, so we principally distin-
guish in trachyte the majestic dome form* (Chimborazo,
21,422 feet), not to be confounded with the form of the un-
opened but less massive bell-shaped mountains. The con-
ical form is most perfectlyt exhibited in Cotopaxi (18,877
feet), and next to this in Popocatepetl^: (1-7,727 feet), as seen
on the beautiful shores of the lake of Tezcuco, or from the
summit of the ancient Mexican step-pyramid of Cholula ;
and in the volcano of Orizaba§ (17,374 feet; according to
Ferrer, 17,879 feet). A strongly truncated conical form|| is
exhibited by the Nevado de Cayambe-Urcu (19,365 feet),
which is intersected by the equator, and by the volcano of
Tolima (18,129 feet), visible above the primeval forest at
the foot of the Paramo de Quindiu, near the little town of
Ibague.^ To the astonishment of geognosists an elongated
ridge is formed by the volcano of Pichincha (15,891 Teet), at
the less elevated extremity of which the broad, still ignited
crater** is situated.

Fallings of the walls of craters, induced by great natural
phenomena, or their rupture by mine-like explosion from the

* Humboldt, Umrisse von Vulkanen der Cordillerm von Quito una
Mexico, ein Beitrag zur Physiognomik der Natur, Tafel iv. (^Kleinere
Schrifien, bd. i., s. 133-205).

f Umrisse von Vulkanen, Tafel vi.

X Op. cit. sup., Tafel viii. (Kleinere Schri/ten, bd. i., s. 463-467). On
the topographical position of Popocatepetl (smoking mountain in the
Aztec language), near the (recumbent) White woman, Iztaccihuatl,
and its geographical relation to the western lake of Tezcuco and the
pyramid of Cholula situated to the eastward, see my Atlas Geogra-
phique et Physique de la JVouvelle Espagne, pi. 3.-

§ Umi-isse von Vulkanen., Tafel ix. ; the Star-mountain, in the Aztec
language Citlaltepetl; Kleinere Schriften, bd. i., s. 467-470, and my
Aths Geogr. et Phys. de la Nouvelle Espagne, pi. 17.

II Umrisse von Vtdkanen, Tafel ii.

^ Humboldt, Vnes des Cordillercs et Monumens des peuples indigenes
de VAmerique (fol.), pi. Ixii.

** Umrisse von Vulkanen, Tafel i. and x. {Kleinere Schriften, bd. i.,
s. 1-99).

TRUE VOLCANOES. ' 22b

depths of the interior, produce remarkable and contrasting
forms in conical mountains : such as the cleavage into dou-
ble pyramids of a more or less regular kind in the Carguai-
razo (15,667 feet), which suddenly fell in* on the night of
the 19th July, 1698, and in the still more beautiful pyra-
mids! of Ilinissa (17,438 feet); and a crenulation of the up-
per walls of the crater, in which two very similar peaks, op-
posite to each other, betray the previous primitive form (Ca-
pac-Urcu, Cerro del Altar, now only 17,456 feet in height).
Among the aborigines of the highlands of Quito, between
Chambo and Lican, between the mountains of Condorasto
and Cuvillan, the tradition has been universally preserved
that fourteen years before the invasion of Huayna Capac the
son of the Inca Tupac Yupanqui, and after eruptions which
lasted uninterruptedly for seven or eight years, ih6 summit
of the last-mentioned volcano fell in, and covered the entire
plateau, in which New Riobamba is situated, with pumice-
stone and volcanic ashes. The volcano, originally higher
than Chimborazo, was called, in the Inca or Quichua lan-
guage, capac, the king or prince of mountains {urcu\ because
the natives saw its summit rise to a greater height above the
lower snow-line than that of any other mountain of the
neighborhood.^ The great Ararat, the summit of which

* Umrisse von Vulkanen, Tafel iv.

t Ibid., Tafel iii. and vii.

i Long before the visit of Bouguer and La Condamine (1736) to the
plateau of Quito, long before any measurements of the mountains by
astronomers, the natives knew that Chimborazo was higher than any
other N#vado in that region. They had detected two lines of level
which remained almost exactly the same all the year round — that of
the lower limit of perpetual snow, and that of the elevation to which
a single, occasional snow-fall reached down. As in the equatorial re-
gion of Quito, the snow-line, as I have proved by measurements else-
where (Asie Centrale, t. iii., p. 255), only varies about 190 feet in eleva-
tion on six of the most colossal peaks ; and as this variation, as well as
smaller ones caused by local conditions, is imperceptible to the naked
eye when seen from a great distance (the height of the summit of
Mont Blanc is the same as that of the lower equatorial snow-limit),
this circumstance gives rise within the tropics to an appai-ently unin^
terrupted regularity of the snowy covering, that is to say, the form of
the snow-line. The pictorial representation of this horizontality is as-
tounding to the physicists who are only accustomed to the irregularity
of the snowy covering in the variable, so-called temperate zones. The
uniformity of elevation of the snow about Quito, and the knowledge of
the maximum of its oscillation, presents perpendicular bases of 15,777
feet above the surface of the. sea, and of 6396 feet above the plateau
in which the cities of Quit6, Hambato, and Nuevo Riobamba are situ-
ated; bases which, combined with very accurate measurements of

230 • COSMOS.

(17,084 feet) was reached by Friedrich Parrot in the year
1829, and by Abich and Chodzko in 1845 and 1850, forms,
like Chimborazo, an unopened dome. Its vast lava streams
have burst forth far below the snow-line. A more import-
ant character in the formation of Ararat is a lateral chasm,
the deeply-cut valley of Jacob, which may be compared with
the Val delBove of -<3Etna. In this, according to Abich's ob-
servation, the inner structure of the nucleus of the trachytic
dome-shaped mountain first becomes really visible, as this nu-
cleus and the upheaval of the whole of Ararat are much more
ancient than the lava streams.* The Kasbegk and Tschegem,
which have broken out upon the same principal Caucasian
mountain ridge (E.S.E. — W.N.W.) as the Elburuz (19,716
feet), are also cones without craters at their summits, while
the colossal Elburuz bears a crater-lake upon its summit.

As conical and dome-like forms are by far the most fre-
quent in all regions of the earth, the isolated occurrence of
the long ridge of the volcano of Pichincha, in the group of
volcanoes of Quito, becomes all the more remarkable. I
have occupied myself long and carefully with the study of
its structure, and, besides its profile view, founded upon nu-
merous angular measurements, have also published a topo-
graphical sketch of its transverse valleys. f Pichincha forms
a wall of black trachytic rock (composed of augite and oli-
goclase) more than nine miles in length, elevated upon a fis-
sure in the most western Cordilleras, near the South Sea,
but without the axis of the high mountain ridge coinciding

angles of elevation, may be employed for determining distance in
many topographical labors which are to be rapidly executed. The
second of the level lines here indicated, the horizontal, which bounds
the lower portion of a single occasional snow-fall, is decisive as to the
relative height of the mountain domes, which do not reach into the
region of perpetual snow. Of a long chain of such mountains, which
have been erroneously supposed to be of equal height, many are be-
low the temporary snow-line, and thus the snow-fall decides as to the
relative height. I have heard such considerations as these upon per-
petual and accidental snow limits from the mouths of rough country
people and herdsmen in the mountains of Quito, where the Sierras
Nevadas are often close together, although they are not connected by
the same line of perpetual snow. Grandeur of nature sharpens the
perceptive faculties in particular individuals among the colored abo-
rigines, even when they are on the lowest steps of civilization.

* Abich, Bulletin de la Soci^U de Geographie, 4me serie, t. i. (1851),
p. 517, with a very beautiful representation of the form of the old vol-
cano.

t Humboldt, Vues de Cordilllres, p. 295, pi. Ixi., and Atlas de la
Relat, Hist, du Voyage, pi. 27.

TRUE VOLCANOES. 231

in direction with that of the Cordillera. Upon the ridge of
the wall, the three domes, set up like castles, follow from
S.W. to N.E. : Cuntur-guachana, Guagua-Pichincha (the
child of the old volcano), and el Picacho de los Ladrillos.
The true volcano is called the Father, or the Old Man, Rucu-
Pichincha. It is the only part of the long mountain ridge
that reaches into the region of perpetual snow, and therefore
ris^es to an elevation which exceeds the dome of Guagua-
Pichincha, the child, by about 190 feet. Three tower-like
rocks surround the oval crater, which lie somewhat to the
southwest, and therefore beyond the axial direction of a wall
which is on the average 15,406 feet in height. In the spring
of 1802 I reached the eastern rocky tower accompanied only
by the Indian, Felipe Aldas. We stood there upon the ex-
treme margin of the crater, about 2451 feet above the bot-
tom of the ignited chasm. Sebastian Wisse, to whom the
physical sciences are indebted for so many interesting observ-
ations during his long residence in Quito, had the courage
to pass several nights in the year 1845 in a part of the cra-
ter where the thermometer fell toward sunrise to 28°. The
crater is divided into two portions by a rocky ridge, covered
with vitrified scoriae. The eastern portion lies more than a
thousand feet deeper than the western, and is now the real
seat of volcanic activity. Here a cone of eruption rises to a
height of 266 feet. It is surrounded by more than seventy
ignited fumaroles, emitting sulphurous vapors.* From this
circular eastern crater, the cooler parts of which are now
covered with tufts of rushy grasses, and a Pourretia with
Bromelia-like leaves, it is probable that the eruptions of fiery
scoriae, pumice, and ashes of Rucu-Pichincha took place in
1539, 1560, 1566, 1577, 1580, and 1660. The city of
Quito was then frequently enveloped in darkness for days to-
getTOr by the falling, dust-like rapilli.

To the rarer class of volcanic forms which constitute elon-
gated ridges belong, in the Old World, the Galungung, with
a large crater, in the western part of Java ;t the doleritic
mass of the Schiwelutsch, in Kamtschatka, a mountain
chain upon the ridge of which single domes rise to a height
of 10,170 feet \X Hecla, seen from the northwest side, in the
normal direction upon the principal and longitudinal fissure

* Kleinere Schriflen, bd. i., s. 61, 81, 83, and 88.
t Junghuhn, Eeise durch Java, 1845, s. 215, Tafel xx.
X See Adolf Erman's Eeise um die Erde, which is also very import-
ant in a geognostic point of view, bd. iii., s. 271 and 207.

232 COSMOS.

over which it has burst forth, as a broad mountain chain,
furnished with various small peaks. Since the last erup-
tions of 1845 and 1846, which yielded a lava stream of
eight geographical miles in length, and in some places more
than two miles in breadth, similar to the stream from ^tna
in 1669, five caldron-like craters lie in a row upon the ridge
of Hecla. As the principal fissure is directed N. 65° E., the
volcano, when seen from Selsundsfjall, that is from the south-
west side, and therefore in transverse section, appears as a
pointed conical mountain.*

If the forms of volcanoes are so remarkably different
(Cotopaxi and Pichincha) without any variation in the
matters thrown out, and jn the chemical processes taking
place in the depths of their interior, the relative position
of the cones of elevation is sometimes still more singular.
Upon the island of Luzon, ifi the group of the Philippines,
the still active volcano of Taal, the most destructive erup-
tion of which was that of the year 1754, rises in the midst
of a large lake inhabited by crocodiles (called the Laguna
de Bombon). The cone, which was ascended in Kotzebue's
voyage of discovery, has a crater-lake, from which again a
cone of eruption, with a second crater, rises. f This descrip-
tion reminds one involuntarily of Hanno's journal of his
voyage, in which an island is referred to, inclosing a small
lake, from the centre of which a second island rises. The
phenomenon is said to occur twice, once in the Gulf of the

* Sartorius von Waltershausen, Phydsch-geograpMsche Skizze von
Island, 1847, s. 107; and his Geognostischer Atlas von Island, 1853,
Tafel XV. and xvi.

t Otto von Kotzebue, Entdeckungs-Rdse in die Siidsee und in die
Berings-Strasse, 1815-1818, bd. iii., s. 68 ; Reise-Atlas von Choris, 1820,
Tafel 5 ; Vicomte d'Archiac, Histoire des P7-ogrts de la Gcologie,J^7,
t. i., p. 544 ; antt'Buzeta, Biccionario Geogr. estad. Historico de la^slas
Fil/pinas, t. ii. (Madrid, 1851), p. 436 and 470, 471, in which, however,
the double encircling of a crater in the crater-lake, mentioned alike
accurately and circumstantially by Delamare, in his letter to Arago
(November, 1842, Coinpies rendus de I' Acad, des Sciences, t. xvi., p. 756),
is not referred to. The great eruption in December, 1754 (a previous
and more violent one took place on the 24th September, 1716), de-
stroyed the old village of Taal, situated on the southwestern bank of
the lake, which was subsequently rebuilt at a greater distance from the
volcano. The small island of the lake upon which the volcano rises
is called Isla del Volcan. (Buzeta, lac. cit.) The absolute elevation
of the volcano of Taal is scarcely 895 feet. It is, therefore, like Cosi-
ma, one of the lowest. At the time of the American expedition of
Captain Wilkes (1842) it was in full activity. See United States Ex-
ploring Expedition, vol. v., p. 317.

TRUE VOLCANOES. 233

Western Horn, and again in the Bay of the Gorilla Apes,
on the West African coast.* Such particular descriptions
may be believed to rest upon actual observation of nature !

The smallest and greatest elevation of the points at which
the volcanic energy of the interior of the earth shows itself
permanently active at the surface is a hypsometric consider-
ation possessing that interest for the physical description of
the earth which belongs to all facts relating to the reaction
of the fluid interior of the planet upon its surface. The de-
gree of the upheaving forcej is certainly evidenced in the
height of volcanic conical mountains, but an opinion as to
the influence of comparative elevation upon the frequency
and violence of eruptions must be given with great caution.
Individual contrasts of the frequency and strength of simi-
lar actions in very high or very low volcanoes can not be
decisive in this case, and our knowledge of the many hun-
dred active volcanoes supposed to exist upon continents and
islands is still so exceedingly imperfect, that the only deci-
sive method, that of average numbers, is as yet misapplied.
But such average numbers, even if they should furnish the
definite result at what elevation of the cones a quicker re-
turn of the eruptions is manifested, would still leave room
for the doubt that the incalculable contingencies occurring in
the net- work of fissures, which may be stopped up with more
or less ease, may act together with the elevation ; that is to
say, the distance from tlie volcanic focus. The phenomenon
is consequently an uncertain one as regards its causal con-
nection.

Adhering cautiously to matters of fact, where the compli-
cation of the natural phenomena and the deficiency of his-
torical records as to the number of eruptions in the lapse of
ages have not yet allowed us to discover laws, I am content-
ed with establishing five groups for the comparative hypso-
metry of volcanoes, in which the classes of elevation are
characterized by a small but certain number of examples.
In these five groups I have only referred to conical mount-
ains rising isolated and furnished with still ignited craters,
and consequently to true and still active volcanoes, not to
unopened dome-shaped mountains, such as Chimborazo. All
cones of eruption which are dependent upon a neighboring
volcano, or which, when at a distance from the latter, as

* Humboldt, Examen Critique de I' Hist, de la Geogr., t. iii., p. 135;
Hannonis Periplus^ in Hudson's Geogr. Greed min., t. i., p. 45.
t Cosmos, vol. i. p. 229.

234 COSMOS.

upon the island of Lancerote, and in the Arso, on the Epo-
meus of Ischia, have preserved no permanent connection be-
tween the interior of the earth and the atmosphere, are here
excluded. According to the testimony of the most zealous
observer of the vulcanicitj of ^tna, Sartorius von Walters-
hausen, this volcano is surrounded by nearly 700 larger and
smaller cones of eruption. As the measured elevations of
the summits relate to the level of the sea, the present fluid
surface of the planet, it is of importance here to advert to
the fact that insular volcanoes — of which some (such as the
Javanese volcano Cosima,* at the entrance of the Straits of
Tsugar, described by Horner and Tilesius) do not project a
thousand feet, and others, such as the Peak of Teneriffe,t are
more thian 12,250 feet above the surface of the sea — have
raised themselves by volcanic forces above a sea-bottom,
which has often been found 20,000 feet, nay in one case
more than 45,838 feet, below the present surface of the ocean.
To avoid an error in the numerical proportions it must also
be mentioned that, although distinctions of the first and fourth
classes— volcanoes of 1000 and 18,000 feet (1066 and 19,188
English feet) — appear very considerable for volcanoes on con-
tinents, the ratios of these numbers are quite changed if
(from IVIitscherlich's experiments upon the melting point of
granite, and the not very probable hypothesis of the uniform
increase of heat in proportion to the depth in arithmetical
progression) we infer the upper limit of the fused interior of
the earth to be about 121,500 feet below the present sea-
level. Consiflering the tension of elastic vapors, which is
vastly increased by the stopping of volcanic fissures, the dif-
ferences of elevation of the volcanoes hitherto measured are
certainly not considerable enough to be regarded as a hinder-
ance to the elevation of the lava and other dense masses to
the height of the crater.

* For the position of this volcano, which is only exceeded in small-
ness by the volcano of Tanna, and that of 'the Mendana, see the fine
map of Japan by F. von Siebold, 1840.

t I do not mention here, with the Peak of TenerifFe, among the in-
sular volcanoes, that of Mauna-Roa, the conical form of which does
not agree with its name. In the language of the Sandwich Islanders,
mauna signifies mountain, and roa both long and 7nuch, Nor do I men-
tion Hawaii, upon the height of which there has so long been a dis-
pute, and which has been described as a trachytic dome not opened at
the summit. The celebrated crater Kiraueah (a lake of molten, boil-
ing lava) lies to the eastward, near the foot of the Mauna-Roa, accord-
ing to Wilkes, at an elevation of 3970 feet. See the excellent de-
scription in Charles Wilkes's Exploring Expedition, vol. iv., p 165-196.

TRUE VOLCANOES. 235

Hypsometi^ of Volcanoes.

First group, from 700 to 4000 Paris or 746 to 4264 English
feet in height.

The volcano of the Japanese island Cosima^ to the south of Jezo :

746 feet, according to Horner.
The volcano of the Liparian island Volcano : 1305 English feet, ac-
cording to F. Hoffmann.*
Gunung Api (signifying Fiery Mountain in the Malay language), the

volcano of the island of Banda: 1949 feet.
The volcano of Izalco,t in the state of San Salvador (in Central

America), which was first ascended in the year 1770, and which

is in a state of almost constant eruption: 2132 feet, according to

Squier.
Gunung Ringgit, the lowest volcano of Java : 2345 feet, according

to Junghuhn.J
Stromboli: 2958 feet, according to F. Hoffmann.
Vesuvius, the Rocca del Palo, on the highest northern margin of the

crater : the average of my two barometrical raeasurements§ of

1805 and 1822 gives 3997 feet.
The volcano of Jorullo, which broke out in the elevated plateau of

Mexicoll on the 29th Septembej, 1759 : 4266 feet.

Second group, from 4000 to 8000 Paris or 4264 to 8528 En-
glish feet in height.

Mont Pele, of Martinique : 4707 feet, according to Dupuget.
The Soufriere, of Guadaloupe : 4867 feet, according to C. Deville.
Gunung Lamongan, in the most eastern part of Java: 5341 feet, ac-
cording to Junghuhn.

* Letter from F. Hoffmann to Leopold von Buch, upon the Geog-
nostic Constitution of the Lipari Islands, in Poggend., Annalen, bd.
xxvi., 1832, s. 59. Volcano, 1268 feet, according to the recent meas-
urement of C. Sainte-Claire Deville, had violent eruptions of scorise
and ashes in the year 1444, at the endof the 16th century, in 1 731, 1739,
and 1771. Its fumaroles contain ammonia, borate of selenium, sul-
phuret of arsenic, phosphorus, and, according to Bornemann, traces
of iodine. The last three substances occur here for the first time
among volcanic products {Comptes rendus de I' Acad, des Sciences^
t. xliii., 1856, p. 683).

t Squier, in the tenth annual meeting of the American Association^
New Haven, 1850.

t See Franz Junghuhn's exceedingly instructive work, Java, seine
Gestalt und PJianzendecke, 1852, bd. i., s. 99. Ringgit has been near-
ly extinct since its fearful eruption in the year 1586, which cost the
lives of many thousand people. •

§ The summit of Vesuvius is, therefore, only 260 feet higher than
the Brocken.

II Humboldt, Vues des Cordilleres, pi. xliii., and Atlas Geogr. et
Physique, pi. 29.

236 COSMOS.

Gunung Tengger, which has the largest cratet* of all the volcanoes
of Java : height at the cone of eruption of Bromo, 7547 feet, ac-
cording to Junghuhn.

The volcano of Osorno (Chili) : 7550 feet, according to Fitzroy.

The volcano of Pico\ (Azores) : 7614: feef, according to Captain
Vidal.

The volcano of the island of Bourbon: 8002 feet, according to
Berth.

Third group, from 8000 to 12,000 Paris or 8528 to 12,792

English feet in height.

The volcano o^ Awatscha (peninsula of Kamtschatka), not to be
confounded^ with the rather more northern Strjeloschnaja Sopka,
which is usually called the volcano of AAvatscha by the English
navigators : 8912 feet, according to Erman.

The volcano of Antuco^ or Anto'io (Chili) : 8920 feet, according to
Domeyko.

The volcano of the island of Fogo\\ (Cape Verd Islands): 9154 feet,
according to Charles Deville.

* Junghuhn, Op. dt. sup., bd. i., s. 68 and 98.

t See my Relation Histonqtie., t. i., p. 93, especially with regard to
the distance at which the summit^ of the volcano of the island of Pico
has sometimes been seen. Ferrer's old measurement gave 7918 feet,
and therefore 304 feet more than the certainly more careful survey of
Captain Vidal in 1843.

X Erman, in his intei'esting geognostic description of the volcanoes
of the peninsula of Kamtschatka, gives the Awatschinskaja or Gore-
laja Sopka as 8912 feet, and the Strjeloschnaja Sopka, which is also
called Korjaskaja Sopka, as 11,822 feet (Reise, bd. iii., s. 494 and
540). See with regard to these two volcanoes, of which the former is
the most active, Leopold de Buch, Descr. Physique des lies Canaries,
p. 447-450. Erman's measurement of the volcano of Awatscha agrees
best with the earliest measurements of Mongez (8739) during the ex-
pedition of La Perouse (1787), and with the more recent one of Cap-
tain Beechy (9057 feet). Hofmann in Kotzebue's voyage, and Lenz
in Lutke's voyage, found only 8170 and 8214 feet; see Lutke, Voyage
autour du Monde, t. iii., p. 67-84. The admiral's measurement of the
Strjeloschnaja Sopka gave 11,222 feet.

§ See Pentland's table of elevations in Mrs. Somerville's Physical
Geography, vol. ii., p. 452 ; Sir Woodbine Parish, Buenos Ayres and
the Province of the Rio de la Plata, 1852, p. 343; Poppig, Reise in
Chile und Peru, bd. i., s. 411-434.

II Is it probable that the height of the sumnjit of this remarkable
volcano is gradually diminishing? A barometrical measurement by
Baldey, Vidal, and Mudge, in the year 1819, gave 2975 metres, or
9760 feet; while a very accurate and practiced observer, Sainte-Claire
Devillij, who has done such important service to the geognosy of vol-
canoes, only found 2790 metres, or 9154 feet, in the year 1842 ( Voy-
age aux lies Antilles et d, Vile de Fogo, p. 155). Captain King had a
little while before determined the height of the volcano of Fogo to be
only 2686 metres, or 8813 feet.

TRUE VOLCANOES. 237

The volcano of Schiwelutsch (Kamtschatka) : the northeastern sum-
mit 10,551 feet, according to Erman.*

uEtna:f according to Smyth, 10,871 feet.

Peak of Teneriffe: 12,161 feet, according to Charles Deville.J

The volcano Gunung Semeru, the highest of all mountains on the
island of Java : 12,237 feet, according to Junghuhn's barometrical
measurement.

The volcano jErebus, lat. 77° 32', the nearest to the south pole :§
12,366 feet, according to Sir James Ross.

The volcano Arc/cEus,\\ in Cappadocia, now Erdschisch-Dagh, south-
southeast of Kaisarieh: 12,603 feet, according to Peter von
TschichatschefF.

* Erman, Reise, bd. iii., s. 271, 275, and 297. The volcano Schi-
welutsch, like Pichincha, has a form which is rare among active vol-
canoes, namely, that of a long ridge (chrebet), upon which single domes
and crests (grebni) rise. Dome-shaped and conical mountains are
always indicated in the volcanic district of the peninsula by the name
sopJcL

f For an account of the remarkable agreement of the trigonomet-
rical with the barometrical measurement of Sir John Herschel, see
Cosmos, vol. i., p. 28.

I The barometrical measurement of Sainte-ClaireDeville ( Voy. aux
Antilles, p. 102-118), in the year 1842, gave 3706 metres, or 12,161
feet, nearly agreeing with the result (12,184 feet) of Borda's second
trigonometrical measurement in the year 1776, which I was enabled
to publish for the first time from the manuscript in the Depot de la
Marine (Humboldt, Voy. mix Regions Equinox., t. i., p. 116 and 275-
287). Borda's first trigonometrical measurement, undertaken in con-
junction with Pingrc in the year 1771, gave, instead of 12,184 feet,
only 11,142 feet. The cause of the error was the false reading of an
angle (33' instead of 53'), as was told mc by Borda himself, to whose
great personal kindness I was indebted for much useful advice before
ray voyage on the Orinoco.

§ I follow Pentland's estimate of 12,367 feet, especially because in
Sir James Ross's Voyage of Discovery in the Antarctic Regions, vol. i.,
p. 216, the height of the volcano, tlie eruptions of smoke and flame
from which were seen even in the daytime, is given in round numbers
at 12,400 feet.

II With regard to Argaeus, which Hamilton was the first to ascend
and measure barometrically (at 12,708 feet, or 3905 metres), see Peter
von Tschichatscheff', Asie Mineure (1853), t, i., p. 441-449, and 571.
In his excellent work (Researches in Asia Minor), William Hamilton
obtained, as the mean of one barometrical measurement and several
angles of elevation, 13,000 feet ; but if the height of Kaisarieh is 1000
feet less than he supposes, it would be only 12,000 feet. See Hamil-
ton, in Trans. Geolog. Society, vol. v., pt. 3, 1840, p. 596. Twvard the
southeast from Ar^jeus (Erdschisch-Dagh), in the great plain of Eregli,
numerous very small cones of eruption rise to the south of the village
of Karabunar and the mountain group Karadscha-Dagh. One of these,
furnished with a crater, has a singular shape like that of a ship, run-
ning out in front like a beak. This crater is situated in a salt Jake,
on the road from Karabunar to Eregli, at a distance of fully four miles

238 COSMOS.

Fourtn group, from 12,000 to 16,000 Paris or 12,792 io
17,056 English feet in height.

The volcano of Tuqueres* in the highlands of the provincia de los

Pastos : 12,824 feet, according to Boussingault.
The volcano of Pasfo.-f 13,453 feet, according to Boussingault.
The volcano Mauna-Roa :% 13,761 feet, according to Wilkes.
The volcano of Cumhal,^ in the provincia de los Pastos : 15,621 feet,

according to Boussingault.
The volcano Kliutscheivsk\\ (Kamtschatka) : 15,766 feet, according

to Erman.
The volcano Rucii-Pichincha : 15,926 feet, according to Humboldt's

barometrical measurements.
The volcano Tungurahua: 16,494 feet, according to a ti'igonomet-

rical measurement^ by Humboldt.

from the former place. The hill bears the same name (Tschichatscheff,
t. i., p. 455 ; William Hamilton, Researches in Asia Minor, vol. ii.,
p. 217).

* The height here given is properly that of the grass-green mount-
ain lake, Laguna verde, on the margin of which is situated the sol-
fatara examined by Boussingault (Acosta, Viajes Cientificos d los Andes
EcuatoHales, 1849, p. 75).

t Boussingault succeeded in reaching the crater, and determined
the altitude barometrically ; it agrees very nearly with that which I
made known approximately twenty-three years before, on my journey
from Popayan to Quito.

J The altitude of few volcanoes has been so over-estimated as that
of the Colossus of the Sandwich Islands. We see it gradually fall
from 18,410 feet (the estimate given in Cook's third voyage), 16,486
feet in King's, and 16,611 feet in Marchand's measurement, to 13,761
feet by Captain Wilkes, and 13,524 feet by Horner in Kotzebue's voy-
age. The grounds of the last-mentioned result were first made known
by Leopold von Buch in the Description Physique des Iks Canaries, p.
379. See Wilkes, Exploring Expedition, vol. iv., p. 111-162, The
eastern margin of the crater is only 13,442 feet. The assumption of
a greater height, considering the asserted freedom from snow of the
Mauna-Roa (lat. 19° 28'), would also be in contradiction to the result
that, according to my measurements in the Mexican' continent in the
same latitude, the limit of perpetual snow has been found at 14,775
feet (Humboldt, Voyage aux Regions Equinox., t. i., p. 97 ; Asie Cen-
trale, t. iii., p. 269 and 359).

§ The volcano rises to the west of the village of Cumbal, which is
itself situated 10,565 feet above the sea-level (Acosta, p. 76).

II I give the result of Erman's repeated measurements in Septem-
ber, 1829. The height of the margin of the crater is exposed to alter-
ations by frequent eruptions, for in August, 1828, measurements which
might inspire equal confidence gave an altitude of 16,033 feet. Com-
pare Erman's Physikalische Beohachtungen aifeiner Reise um die Erde,
bd. i., s. 400 and 419, with the historical account of the journey, bd.
iii., s. 358-360.

^ Bouguer and La Condamine, in the inscription at Quito, give
16,777 feet for Tungurahua before the great eruption of 1772, and

TRUE VOLCANOES. 239

The volcano of Purace* near Popayan: 17,010 feet, according to
Jose Caldas.

Fifth group, from 16,000 to more than 20,000 Fans or from
17,056 to 21,320 English feet in height.

The volcano Sangai/-, to the southwest of Quito: 17,128 feet, ac-
cording to Bouguer and La Condamine.f

The \o\cano Popocatepetl :X 17,729 feet, according to a trigonomet-
rical measurement by Humboldt.

The volcano of Orizaba:^ 17,783 feet, according to Ferrer.

Elias MountW (on the west coast of North America): 17,855 feet,
according to the measurements of Quadra and Galeano.

the earthquake of Riobamba (1797), which gave rise to great depres-
sions of mountains. In the year 1802 I found the summit of the
volcano trigonometrically to be only 16,494 feet.

* The barometrical measurement of the highest peak of the Volcan
de Purace by Francisco Jose Caldas, who, like my dear friend and
traveling companion, Carlos Montufar, fell a sacrifice to his love for
the independence and freedom of his country, is given by Acosta
{Vlajes Clentificos, p. 70) at 5184 metres (17,010 feet). I found the
height of the small crater, which emits sulphureous vapors with a
violent noise (Azufral del Boqueron), to be 14,427 feet ; Humboldt,
Recudl d Observ. Astronomiques et d^ Operations Trigonoinctriques, vol. i.,
p. 304.

t The Sangay is extremely remarkable from its uninterrupted ac-
tivity and its position, being removed somewhat to the eastward from
the eastern Cordillera of Quito, to the south of the Rio Pastaza, and
at a distance of 120 miles frqp the nearest coast of the Pacific — a
position which (like that of the volcanoes of the Celestial mountains
in Asia) by no means supports the theory according to which the east-
ern Cordilleras of Chili are free from volcanic eruptions on account
of their distance from the sea. The talented Darwin has not omitted
referring in detail to this old and widely diffused volcanic littoral
theory in the Geological Observations on South America, 1846, p. 185.

X I measured Popocatepetl, which is also called the Volcan Grande
de Mexico, in the plain of Tetimba, near the Indian village San Nico-
las de los Ranchos. It seems to me to be still uncertain which of the
two volcanoes, Popocatepetl or the peak of Orizaba, is the highest
(see Humboldt, Eeceuil d^Observ. Astron., vol. ii., p. 543).

§ The peak of Orizaba, clothed with perpetual snow, the geographic-
al position of which was quite erroneously indicated on all maps be-
fore my journey, notwithstanding the importance of this point for navi-
gation near the landing-place in Vera Cruz, was first measured trigo-
nometrically from the Encero by Ferrer, in 1796.. The measurement
gave 17,879 feet. I attempted a similar operation in a small plain
near Xalapa. I found only 17,375 feet, but the angles of elevation
were very small, and the base-line difficult to level. See Humboldt,
Essai Politique sur la Nouv. Espagne, 2me ed.,t. i., 1825, p. 166 ; Atlas
du Mexique (Carte des fausses positions), pi. x., and Kleinere Schriften,
bd. i., s. 468.

Ii Humboldt, Essai sur la Geograplde des Plantes, 1807, p. 153. The
elevation is uncertain, perhaps more than -jl^th too high.

240 COSMOS.

The volcano of Tolima:* 18,143 feet, according to a trigonomet-
rical measurement bj Humboldt.

The volcano of Arequipa:f 18,883 feet, according to a trigonomet-
rical measurement by Dolley.

* I measured the truncated cone of the volcano of Tolima, situated
at the northern extremity of the Paramo de Qiiindiu, in the Valle del
Carvajal, near the little town of Ibague, in the year 1802. The mount-
£lin is also seen at a great distance upon the plateau of Bogota. At
this distance Caldas obtained a tolerably approximate result (18,430
feet) by a somewhat complicated combination in the year 1806 ; Sema-
naiio dc la Neuva Granada, nueva edicion, aumentada por J. Acosta, 1849,
p. 349.

t The absolute altitude of the volcano ofArequipa has been so
variously stated that it becomes difficult to distinguish between mere
estimates and actual measurements. Dr. Thaddaus Hanke, of Prague,
the distinguished botanist of Malaspina's voyage round the world, as-
cended the volcano of Arequipa in the year 1796, and found at the
Bummit a cross which had been erected there twelve years before.
By a trigonometrical operation Hanke found the volcano to be 3180
toises (20,235 feet) above the sea. This altitude, which is far too
great, was probably the result of an en-oneous assumption of the ele-
vation of the town of Arequipa, in the vicinity of which the operation
was performed. Had Hanke been provided with a barometer, a bot-
anist entirely unpracticed in trigonometrical measurements would
certainly not have resorted to such means after ascending to the sum-
mit. The first who ascended the volcano after Hanke was Samuel
Curzon, from the United States of North America (Boston Philosophic-
alJournal, 1823, November, p. 168). In the year 1830 Pentland esti-
mated the altitude at 5600 metres (18,374 feet), and I have adopted
this number {Annuaire du Bureau des Longitudes, 1830, p. 325) for my
Carte Hypsometrique de la Cordillere des Andes, 1831. There is a satis-
factory agreement (within •^^th) between this and the trigonometrical
measurement of a French naval officer, M. Dolley, for which I was
indebted in 1826 to the kind communication of Captain Alphonse de
Moges in Paris. Dolley found the summit of the volcano ofArequipa
(trigonometrically)tobe 11,031 feet, and the summit of Charcani 11,860
feet above the plateau in which the town of Arequipa is situated. If
now we fix the town of Arequipa at 7841 feet, in accordance with the
barometrical measurements of Pentland and Eivero (Pentland, 7852
feet in the Table of Altitudes to the Physical Geography of Mrs. Som-
erville, 3d ed., vol. ii., p. 454; Eivero, in the Memorial de Ciencias
Natiirales, t. ii., Lima, 1828, p. 65 ; Meyen, Reise urn die Erde, Theil.
ii., 1835, s. 5), DoUey's trigonometrical operation will give for the
volcano of Arequipa 18,88i feet (2952 toises), and for the volcano
Charcani 19,702 feet (3082 toises). But Pentland's Table of Aki-
tudes, above cited, gives for the volcano of Arequipa 20,320 English
feet, 6190 metres (19,065 Paris feet) ; that is to say, 1945 feet more
than the detennination of 1830, and somewhat too identical with
Hanke's trigonometrical measurement in the year 1796 ! In opposi-
tion to this result the volcano is stated, in the Anales de la Universidad
de Chile, 1852, p. 221, only at 5600 metres, or 18,378 feet: consequent-
ly 590 metres lower ! A sad condition of hypsometry !

TRUE VOLCANOES. 241

The volcano Cotopaxi:* 18,881 feet, according to Bouguer.

The volcano Sahamaf (Bolivia): 22,354 feet, according to Pentland.

The volcano with which the fifth group ends is more than
twice as high as ^tna, and five times anij a half as high as
Vesuvius. The scale of volcanoes that I have suggested,
starting from the lowly Maars (mine-craters without a
raised frame-work, which have cast forth olivin bombs sur-
rounded by half-fused fragments of slate) and ascending to
the still burning Sahama 22,354 feet in height, has shown
us that there is no necessary connection between the maxi-
mum of elevation, the smaller amount of the volcanic activi-
ty and the nature of the visible species of rock. Observa-
tions confined to single countries may readily lead us to er-
roneous conclusions. For example, in the part of Mexico
which lies in the torrid zone, all the snow-covered mount-
ains— that is to say, the culminating points of the whole
country — are certainly volcanoes; and this is also usually
the case in the Cordilleras of Quito, if the dome-shaped
trachytic mountains, not opened at the summit (Chimborazo
and Corazon), are to be associated with volcanoes ; on the
other hand, in the eastern chain of the Bolivian Andes the
highest mountains are entirely non-volcanic. The Nevados

* Boussingault, accompanied by the talented Colonel Hall, has near-
ly reached the summit of Cotoimxi. He attained, according to bar-
ometrical measurement, to an ahitude of 574G metres, or 18,855 feet.
There was only a small space between him and the margin of the
crater, but the great looseness of the snow prevented his ascending
farther. Perhaps Bouguer's statement of altitude is rather too small,
as his complicated trigonometrical calculation depends upon the hy-
pothesis as to the elevation of the city of Quito.

t The Sahama, which Pentland (^Annuaire du Bureau des Longi-
tudes, 1830, p. 321) distinctly calls an active volcano, is situated, ac-
cording to his new map of the Vale of Titicaca (1848), to the east-
ward of Arica, in the western Cordillera. It is 928 feet higher than
Chimborazo, and the relative height of the lowest Japanese volcano
Cosima to the Sahama is as 1 to 30. I have hesitated in placing the
Chilian Aconcagua, which, stated by Fitzroy in 1835 at 23,204 feet,
is, according to Pentland's correction, 23,911 feet, and according to
the most recent measurement (1845) of Captain Kellet of the frigate
Herald, 23,004 feet, in the fifth group, because, from the contradictory
opinions of Miers ( Voyage to Chili, vol. i., p. 283) and Charles Dar-
win (Journal of Researches into the Geology and Natural History of the
Various Countries visited hy the Beagle, 2d ed., p. 291), it remains
doubtful whether this colossal mountain is a still ignited volcano.
Mrs. Somerville, Pentland, and Gilliss (Naval A sir. Exped., vol. i., p.
126) also deny its activity. Darwin says : "I was surprised at hear-
ing that the Aconcagua was in action the same night (15th January,
1835), because this mountain most rarely shows any sign of action."
Vol. v.— L

242 cosmos:

of Sorata (21,292 feet) and Illimani (21,15ci- feet) consist
of graywacke schists, which are penetrated by porphyritic
masses,* in which (as a proof of this penetration) fragments
of schist are inclosed. In the eastern Cordillera of Quito,
south of the parallel of 1° 35^, the high summits (Condoras-
to, Cuvillan, and the Collanes) lying opposite to the tra-
chytes, and also entering the region of perpetual snow, are
also mica-slate and fire-stone. According to our present
knowledge of the mineralogical nature of the most elevated
parts of the Himalaya, which we owe to the meritorious la-
bors of B. H. Hodgson, Jacquemont, Joseph Dalton Hooker,
Thomson, and Henry Strachey, the primary rocks, as they
were formerly called, granite, gneiss, and mica-slate, appear
to be visible here also, although there are no trachytic for-
mations. In Bolivia, Pentland has found fossil shells in the
Silurian schists on the Nevado de Antacaua, 17,482 feet
above the sea, between La Paz and Potosi. .The enormous
height to which, from the testimony of the fossils collected
by Abich from Daghestan, and by myself from the Peruvian
Cordilleras (between Guambos and Montan), the chalk for-
mation is elevated, reminds us very vividly that non-volcan-
ic sedimentary strata, full of organic remains, and not to be
confounded with volcanic tufaceous strata, show themselves
in places where for a long distance around melaphyres,
trachytes, dolerites, and other pyroxenic rocks, which we
regard as the seat of the upheaving, urging forces, remain
concealed in the depths. In what immeasurable tracts of
the Cordilleras and the districts bordering them upon the
east is no trace of any granitic formation visible !

The frequency of the eruptions of a volcano appearing to
depend, as I have already repeatedly observed, upon multifa-

* These penetrating porphyritic masses show themselves in peculiar
vastness near the Illimani, in Cenipampa (15,949 feet) and Totora-
pampa (13,709 feet) ; and a quartzose porphyry containing mica, and
inclosing garnets, and at the same time angular fragments of silicious
schist, forms the superior dome of the celebrated argentiferous Cerro
de Potosi (Pentland in MSS. of 1832). The Illimani, which Pent-
land estimated first at 7315 (23,973 feet), and afterward at 6445
(21,139 feet) metres, has also been, since 1847, the object of a care-
ful measurement by the engineer Pissis, who, on the occasion of his
great trigonometrical survey of the Llanura de Bolivia, found the Illi-
mani to be on the average 6509 metres (21,349 feet) in height, by
three triangles between Calamarca and La Paz : this only differs
about 64 metres (210 feet) from Pentland's last determination. See
Investigaciones Sohre la Altitud de los Andes, in the Anales de Chih,
1852, p. 217 and 22l.

TRUE VOLCANOES. 243

rioua and very complicated causes, no general law can safely
be established with regard to the relation of the absolute ele-
vation to the frequency and degree of the renewal of combus-
tion. If in a small group the comparison of Stromboli, Ve-
suvius, and ^tna may mislead us into the belief that tho
number of eruptions is in an inverse ratio to the elevation
of the volcanoes, other facts stand in direct contradiction
to this proposition. Sartorius von Waltershausen, who has
done such good service to our knowledge of ^tna, remarks
that, on the average furnished by the last few centuries, an
eruption of this volcano is to be expected every six years,
while in Iceland, where no part of the island is really secure
from destruction by submarine fire, the eruptions of Hecla,
which is 5756 feet lower, are only observed every 70 or 80
years.* The group of volcanoes of Quito presents a Itill more
remarkable contrast. The volcano of Sangay, 17,000 feet
in height, is far more active than the little conical mountain,
Stromboli (2958 feet) ; it is of all known volcanoes the one
which exhibits, every quarter of an hour, the greatest quanti-
ty of fiery, widely-luminous eruptions of scoriae. Instead of
losing ourselves in hypotheses upon* the causal relations of
inaccessible phenomena, we will rather dwell here upon the
consideration of six points of the surface of the earth, which
are peculiarly important and instructive in the history of vol-
canic activity — Stromboli, the Lycian Chimocra, the old vol-
cano oi*Masaya, the very recent one of Izalco, the volcano
Fvgo on the Cape Verd Islands, and the colossal Sangay.

The Chimoira in Lycia, and Stromboli, the ancient Stron-
gyle, are the two igneous manifestations of volcanic activity,
the historic proof of whose permanence extends the furthest
back. The conical hill of Stromboli, a doleritic rock, is twice
the height of the island of Volcano (Hiera, Thermessa), the
last great eruption of which occuiTcd in the year«1775. The
uninterrupted activity of Stromboli is compared by Strabo
and Pliny with that of the island of Lipari, the ancient Me-
ligunis ; but they ascribe to " its flame," that is, its erupted
scoriae, "a greater purity and luminosity, with less heat."!

* Sartorius von "Waltershausen, Skizze von Island, s. 103 and JOT.

f Strabo, hb. vi., p. 276, ed. Casaubon ; FHny, Hist. Nat., iii., 9 :
* ' Strongyle, quae a Lipara liquidiore flamma tantum differt ; e cujus
fumo quinam flaturi sint venti, in triduo praedicere incolaj traduntur."
See also Urlichs, Vindicice Pliniance, 1853, Fasc. i., p; 39. The volcano
of Lipara (in the northeastern part of the island), once so active, appears
to me to have been either the Monte Campo Bianco or the Monte di
Capo Castagno. (See Hoffmann, in Poggend., Ann.,hd. xxvi., s. 49-54.)

244 COSMOS.

The number and form of the small fiery chasms are very va-
riable. Spallanzani's description of the bottom of the cra-
ter, which was long regarded as exaggerated, has been com-
pletely confirmed by an experienced geognosist, Friedrich
Hoffmann, and also very recently by an acute naturalist,
A. de Quatrefages. One of the incandescent chasms has an
opening of only 20 feet in diameter; it resembles the pit
of a blast-furnace, and the ascent and overflow of the fluid
lava are seen in it every hour, froin a position on the margin
of the crater. The ancient, permanent eruptions of Strombo-
li still sometimes serve for the guidance of the mariner, and,
as among the Greeks and Romans, afford uncertain predic-
tions of the w^eather, by the observation of the direction of
the flame and of the ascending column of vapor. Polybius,
who displays a singularly exact knowledge of the state of the
crater, connects the multifarious signs of an approaching
change of wind with the myth of the earliest sojourn of ^o-
lus upon Strongyle, and still more with observations upon the
then violent fire upon Volcano (the " holy island of Hephaes-
tos "). The frequency of the igneous phenomena has of late
exhibited some irregularity. The activity of Stromboli, like
that of iEtna, according to Sartorius von Waltershausen, is
greatest in November and the winter months. It is some-
times interrupted by isolated intervals of rest ; but these, as
we learn from the experience of centuries, are of very short
duration.

The Chimcera in Lycia, which has been so admirably de-
scribed by Admiral Beaufort, and to which I have twice re-
ferred,* is no volcano, but a perpetual burning spring — a gas

* Cosmos; vol. i., p. 223, and vol. v., p. 203. Albert Berg, who had
previously pubhshed an artistic work, Physiognomie der Tropisclien
Vegetation von Siidamerika, visited the Lycian Chimera, near Delik-
tasch and Yanartaseh, from Rhodes and the Gulf of Myra, in 1853.
(The Turkish word tdsch signifies stone, as dagh and.tdgh signify mount-
ain; deliktasch signifies perforated stone, from the Turkish delik, a
hole.) The traveler first saw the serpentine rocks near Adrasan, while
Beaufort met with the dark-colored serpentine deposited upon lime-
stone, and perhaps deposited in it, even near the island Garabusa (not
Grambusa), to the south of Cape Chelidonia. "Near the ruins of the
ancient temple of Vulcan rise the remains of a Christian church in the
later Byzantine style : the remains of the nave and of two side chap-
els. In the fore-court, situated to the east, the flame breaks out of a
fireplace-like opening about two feet broad and one foot high in the
serpentine rock. » It rises to a height of three or four feet, and (as a
naphtha spring ?) diffuses a pleasant odor, which is perceptible to a
distance of forty paces. Near this large flame, and without the chim-
ney-like opening, numerous very small, constantly ignited, lambent

TRUE VOLCANOES. 245

spring always ignited by the volcanic activity of the interior
of the earth. It was visited a few months ago by a talented
artist, Albert Berg, for the purpose of making a picturesque
survey of this locality, celebrated even in periods of high an-
tiquity (sinpe the times of Ctesias and Scylax of Caryanda),
and of collecting the rocks from which the Chimeera breaks
fdrth. The descriptions of Beaufort, Professor Edward
Forbes, and Lieutenant Spratt in the " Travels in Lycia," are
completely confirmed. An eruptive mass of serpentine rock
penetrates the dense limestone in a ravine, which ascends
from southeast to northwest. At the northwestern extrem-
ity of this ravine the serpentine rock is cut off, or perhaps
only concealed, by a curved ridge of limestone rocks. The
fragments brought home are partly green and fresh, partly
brown and in a weathered state. In both serpentines diallage
is clearly recognizable.

The volcano of Masaya* the fame of which was already

flames make their appearance from subordinate fissures. The rock
which is in contact with the flame is much blackened, and the soot
deposited is collected to alleviate smarting of the eyelids, and espe-
cially for coloring the eyebrows. At a distance of three paces from
the flame of the Chimera the heat which it diff'uses is scarcely endura-
ble. A piece of dry wood ignites when it is held in the opening and
brought near the flame without touching it. Where the old mined
walls lean against the rock, gas also pours forth from the interstices of
the stones of the masonry, and this, probably from its being of a lower
temperature or diiferently composed, does not ignite spontaneously,
but whenerer it is brought in contact with a light. Eight feet below
the great flame in the interior of the ruins there is a r^ii^d opening,
six feet in depth, but only three feet wide, which was probably arched
over formerly, as a spring of water bi-eaks out in it in the wet seasons,
near a fissure over whjch a small flame plays." (From the traveler's
manuscripts.) On a plan of the locality. Berg shows the geographical
relations of the aUuvial strata, of the (tertiary ?) limestone, and of the
serpentine rocks.

* The oldest and most important notice of the volcano of Masaya
is contained in a manuscript of Oviedo's, first edited fourteen years ago
by the meritorious historical compiler, Ternaux-Compans — Historia do
Nicaragua (cap. v. to x.), see p. 115-197. The French translation
forms one volume of the Voyages, Relations et Memoires Originaux pour
servir a tllistoire et a la Dccouverte de V Amerique. See also Lopez de
Gomara, Historia General de las Indias (Zaragoza, 1553), fol. ex., b ; and
among the most recent works, Squier, Nicaragim, its People, Scenery,
and Monuments, 1853, vol. i., p. 21 1-223, and vol. ii., p. 17. So wide-
ly-famed was the incessantly active volcano of Masaya, that a special
monograph of this mountain exists in the royal library at Madrid, un-
der the title of Entrada y Desmbrimiento del Volcan de Masaya, que estd
en la Prov. de Nicaragua, fecha por Juan Sanchez del Portero. The au-
thor was one of those who let themselves down into the crater in the

246 COSMOS.

widely spread in the beginning of the 16th century under
the name of el Injiemo de Masaya, and gave occasion for re-
ports to the Emperor Charles V., is situated between the
two lakes of Nicaragua and Managua, to the southwest of the
charming Indian village of Nindiri. For centuries together it
presented the same rare phenomenon that we have described
in the volcano of Stromboli. From the margin of the crater
the waves of fluid lava, set in motion by vapors, were seen
rising and falling in the incandescent chasm. The Spanish
historian, Gonzalez Fernando de Oviedo, first ascended the
Masaya in July,* 1529, and made comparisons with Vesuvius,
which he had previously visited (1501), in the suite of the
Queen of Naples as her xefe de guai^daropa. The name 3Ia-
scuja belongs to the Chorotega language of Nicaragua, and
signifies burning mountain. The volcano, surrounded by a
wide lava-field {mal-pays), which it has probably itself pro-
duced, was at that time reckoned among the mountain group
of the " nine burning Maribios." In its ordinary condition,
says Oviedo, the surface of the lava, upon which black scoriaa
float, stands several hundred feet below the margin of the
crater; but sometimes the ebullition is suddenly so great
that the lava nearly reaches the upper margin. The per-
petual luminous phenomenon, as Oviedo definitely and acute-
ly states, is not caused by an actual flame,* but by vapors
illuminated from below. It is said to have been of such in-
tensity that on the road from the volcano toward Granada,

wonderful e»editions of the Dominican monk, Fray Bias de Inesta
(Oviedo, liisf! de Nicaragua, p. 141).

* In the French translation of Ternaux-Compans (the Spanish
original has never been published), we find, at p. 123 and 132 : " It
can not, however, be said precisely that a flame Issues from the crater,
but a smoke as hot as fire ; it is not seen from far during the day, but
is well seen at night. The volcano gives as much light as the moon a
few days before it is at the full." This old observation upon the prob-
lematical mode of illumination of a crater, and the strata of air lying
above it, is not without importance, on account of the doubt, so often
raised in recent times, as to the disengagement of hydrogen gas from
the craters of volcanoes. Although in the ordinary condition here in-
dicated the Hell of Masaya did not throw out scoriae or ashes (Gomara
adds, cosa que hazen otros volcanes\ it has nevertheless sometimes had
true eruptions of lava ; the last of which probably occurred in the year
1670, Since that date the volcano has been quite extinct, after a per-
petual luminosity had been observed for 140 years. Stephens, who as-
cended it in 1840, found no perceptible trace of ignition. Upon the
Chorotega language, the signification of the word Masaya, and the
Maribios, see Buschmann's ingenious ethnographical researches, Ueher
die Aztekischen Ortsnamen, s. 130, 140, and 171.

TRUE VOLCANOES. 247.

at ^ distance of more than three leagues, the illumination of
the district was almost equal to that of the full moon.

Eight years after Oviedo, the volcano was ascended by
the Dominican monk, Fray Bias del Castillo, who enter-
tained the absurd opinion that the fluid lava in the crater
was liquid gold, and associated himself with an equally ava-
ricious Flemish Franciscan, Fray Juan de Gandavo. The
pair availing themselves of the credulity of the Spanish set-
tlers, established a joint-stock company to obtain the metal
at the common cost. They themselves, Oviedo adds sa^c-
ally, declared that as ecclesiastics they were free from any
pecuniary contributions. The report upon the execution of
this bold undertaking, which was sent to the Bishop of Cas-
tilla del Oro, Thomas de Yerlenga, by Fray Bias del Cas-
tillo (the same person who is denominated Fray Bias de In-
esta in the writings of Gomara, Benzoni, and Herrera), was
only made known (in 1840) by the discovery of Oviedo's
work upon Nicaragua* Fray Bias, who had previously
served on board ship as a sailor, proposed to' imitate the
method of hanging upon ropes over the sea, by which the
natives of the Canary Islands collect the coloring matter
of the Orchil {Lichen Boccella) on precipitous rocks. For
months together all sorts of preparations were made, in
order to let down a beam of more than thirty feet in length,
by means of a windlass and crane, so that it might project
over the deep abyss. The Dominican, his head covered
with an iron helmet and a crucifix in his hand, was let
down with three other members of the association ; they re-
mained for a whole night in this part of the solid crater bot-
tom, from which they made vain attempts to dip out the
supposed liquid gold with earthen vessels, placed in an iron
pot. Not to frighten the shareholders, they agreed^ that

* " The three companions agi'ced to say that they had found great
riches ; and Fray Bias, whom I had known as an ambitious man,
gives, in his relation, the oath which he and his associates tcok upon
the Gospel, to persist forever in their opinion that the volcano con-
tained gold and silver in a state of fusion !" (Oviedo, Descr. de Nic-
aragua, cap. X., p. 18G and 196). The Cronista de las Indias is, how-
ever, very indignant (cap. 5) that Fray Bias narrated that "Oviedo
had begged the Hell \)f Masaya from the emperor as his armorial
bearings." Such a geognostic memento would certainly not have
been in opposition to the heraldic customs of the period, for the cour-
ageous Diego de Ordaz, who boasted of having reached the crater of
the Popocatepetl when Cortez first penetrated into the valley of Mex-
ico, bore this volcano as an heraldic Astinction, as did Oviedo the

,^48 COSMOS.

when they were drawn up again they should say that they
had found great riches, and that the Injierno of Masaya de-
served in future to be called el Paraiso del Masaya, The op-
eration was afterward repeated several times, until the Gov-
ernor of the neighboring city of Granada conceived some sus-
picion of the deceit, or perhaps of a fraud upon the revenue,
and forbade any " further descents on ropes into the crater."
This took place in the summer of 1538 ; but in 1551 Juan
Alvarez, the Dean of the Chapter of Leon, again received
frdii Madrid the naive permission " to open the volcano and
procure the gold that it contained." Such was the popular
credulity of the 16th century ! But even in Naples, in the
year 1822, Monticelli and Covelli were obliged to prove, by
chemical analysis, that the ashes thrown out from Vesuvius
on the 28th October contained no gold !*

The volcano of Izalco, situated on the west coast of Cen-
tral America, 32. miles northward from San Salvador, and
eastward from the harbor of Sonsonate, broke out 1 1 years
after the vokano of Jorullo, deep in the interior of Mexico.
Both eruptions took place in a cultivated plain, and after the
prevalence of earthquakes and subterranean noises (bramidos)
for several months. A conical L!ll rose in the Llano de
Izalco, and with it simultaneously an eruption of lava poured
from its summit on the 23d February, 1770. It still remains
undecided how much is to be attributed, in the rapidly-in-
creasing height, to the upheaval of the soil, and how much
to the accumulation of erupted scoriae, ashes, and tufa masses ;
only this much is certain, that since the first eruption the
new volcano, instead of soon becoming extinguished, like Jo-
rullo, has remained uninterruptedly active, and often serves
as a beacon-light for mariners near the landing-place in the
Bay of Acajutla. Four fiery eruptions are counted in an
hour, and the great regularity of the phenomenon has aston-
ished its few accurate observers.! ThS violence of the erup-
tions was variable, but not the time of their occurrence.
The elevation which the volcano of Izalco has now attained
since the last eruption of 1825 is calculated at about 1600
feet, nearly the same as the elevation of Jorullo above the
original cultivated plain, but almost four times that of the

constellation of the Southern Cross, and earliest of all Columbus
{Exam, crlt., t. iv., p. 235-240), a fraorment of a map of the Antilles.

* Humboldt, Views of Nature, p. 368.

t Squier, Nicaragua, its People and Monuments, vol. ii., p. 104. (John
Bailey, Centi-al America^ 185% p. 75.)

TRUE VOLCANOES. 249

crater of elevation (Monte Nuovo) in the Phlegrsean Fields,
to which Scacchi* ascribes a height of 432 feet from accu-
rate measurement. Tlie permanent activity of the volcano
of Izaleo, which was long considered as a safety-valve for
the neighborhood of San Salvador, did not, however, pre-
serve the town from complete destruction on Easter eve in
this year (1854).

One of tlie Cape Verd Islands, which rises between S. Jago
and Brava, early received from the Portuguese the name of
Ilha do Fogo, because, like Stromboli, it produced fire unin-
terruptedly from 1680 to 1713. After a long repose, the
volcano of this island resumed its activity in the summer of
the year 1798, soon after the last lateral eruption of the
Peak ofTeneriffe in the crater of Chahorra, which is errone-
ously denominated the volcano of Chahorra, as if it were a
distinct mountain.

The most active of the South American volcanoes, and
indeed of all those which I have here specially indicated, is
the Sangay, which is also called the Volcan de Macas, be-
cause the remains of this ancient city, so populous in the
early period of the Conquista, are situated upon the Rio
Upano, only 28 geographical miles to the south of it. The
colossal mountain, 17,128 feet in height, has risen on the
eastern declivity of the eastern Cordillera, between two sys-
tems of tributaries of the Amazons, those of the Pastaza and
the Upano. The grand and unequaled fiery phenomenon
which it now exhibits appears only to have commenced in
the year 1728. During the astronomical measurements of
degrees by Bouguer and La Condamine (1738 to 1740), the
Sangay served as a perpetual fire signal.j In the year
1802, I myself heard its thunder for months together, espe-
cially in the early morning, in Chillo, the pleasant country
seat of the Marquis de Selvalegre near Quito, as half a cen-
tury previously Don Jorge Juan had perceived the ronquidos
del Sangay, somewhat further 'toward the northeast, near
Pintac, at the foot of the Antisana.| In the years 1842

* Memorie geohgiche sulla Campania, 1849, p. 61. I found the
height of the volcano of Jorullo to be 1G82 feet above the plain in
which it rose, and 4266 feet above the sea-level.

t La Condamine, Journal du Voyage a V Equateur, p. 163 ; and in
the Mesure de Trois Degrcs de la Meridienne de V Hemisphere Austral.^
p. 56.

X In the country house of the Marquis of Selvalegre, the father of
my unfortunate companion and friend, Don Carlos' Montufar, one was
often inclined to ascribe the bramidos, which resembled the discharge

L2

i

250 COSMOS.

and 1843, when the eruptions were associated with most
noise, the latter was heard most distinctly not only in the
harbor of Guayaquil, but also'further to the south along the
coast of the Pacific Ocean, as far as Payta and San Buena-
ventura, at a distance equal to that of Berlin from Basle,
the Pyrenees from Fontainebleau, or London from Aber-
deen. Although, since the commencement of the present
century, the volcanoes of Mexico, New Granada, Quito,
Bolivia, and Chili have been visited by some geognosists,
the Sangay, which exceeds the Tungurahua in elevation, has
unfortunately remained entirely neglected, in consequence of

of a distant battery of heavy artillery, aijd which with the same wind^
the same clearness of the atmosphere, and the same temperature,
were so extremely unequal in their intensity, not to the Sangay, but
to the Guacamayo, a mountain forty miles nearer, at the foot of which
a road leads from Quito, over the hacienda de Antisana to the plairr
of Archidona and the Rio Napo. (See my special map of the prov-
ince Quixos, No. 23 of my Atlas Geogr. et JPhys. de VAmerique, ]814-
1834.) Don Jorge Juan, who heard the Sangay thundering when
closer to it than I have been, says decidedly that the bramidos, which he
calls ronquidos del Volcan {Relacion del Viage a la America Meridional,
pt. i., t. 2, p. 569), and perceived in Pintac, a few miles from the
hacienda de Chillo, belong to the Sangay or Volcan de Macas, whose
voice, if I may make use of the expression, is very characteristic.
This voice appeared to the Spanish astronomer to be peculiarly harsh,
for which reason he calls it a snore (nn ronquido) rather than a roar
(bramido). The very disagreeable noise of the volcano Pichincha,
which I have frequently heard at night in the city of Quito, without
its being followed by any earthquake, has something of a clear rattling
sound, as though chains were rattled and masses of glass were falling
upon each other. On the Sangay, Wisse describes the noise to be
sometimes like rolling thunder, sometimes distinct and sharp, as if
one were in the vicinity of platoon firing. Payta and San Buenaven-
tura (in the Choco), where the bramidos of the Sangay, that is to
say, its roaring, were heard, are distant from the summit of the vol-
cano, in a southwestern direction, 252 and 348 geographical miles.
(See Carte de la Prov. Du Choco, and Cai'te hypsometrique des Cordil-
leres, Nos. 23 and 3 of my Atlas Geogr. et Physique.) Thus, in this
mighty spectacle of nature, reckojaing in the Tungurahua and the Co-
topaxi, which is nearer to Quito, and the roar of which I heard in
February, 1803, in the Pacific Ocean (^Kleinere Schriften, bd. i., s.
384), the voices of four volcanoes are perceived at adjacent points.
The ancients also mention " the difference of the noise," emitted at
different times on the iEolian islands by the same fiery chasm
{Straho, lib. vi., p. 276). During the great eruption (23d January,
1835) of the volcano of Conseguina, which is situated on the coast of
the Pacific, at the entrance of the Bay of Fonseca, in Central Ameri-
ca, the subterranean propagation of the sound was so great, that it
was most distinctly perceived on the plateau of Bogota, at a distance
equal to that froiflSffitna to Hamburg (Acosta, Viajes Cieatificos de M.
Boussingault a los~Andes, 1849, s. 56).

TRUE VOLCANOES. 251

its solitary position, at a distance from all roads of commu-
nication. It was only in December, 1849, that an adven-
turous and highly infornied traveler, Sebastian Wisse, after
a sojourn of five years on the chain of the Andes, ascended
it, and nearly reached the extreme summit of the snow-
covered, precipitous cone. He not only made an accu-
rate chronometric determination of the wonderful frequency
of the eruptions, but also investigated the nature of the
trachyte which, confined to such a limited space, breaks
through the gneiss. As has already been remarked,* 267
eruptions were counted in one hour, each lasting on an
average 13 ''''•4, and, which is very remarkable, unaccom-
panied by any concussion perceptible on the ashy cone. The
erupted matter, enveloped in much smoke, sometimes of a
gray and sometimes of an orange color, is principally a mix-
ture of black ashes and rapilli, but it also consists partly of
cinders, which rise perpendicularly, are of a globular form
and a diameter of 15 or 16 inches. In one of the more vio-
lent eruptions, however, Wisse counted only fifty or sixty
red-hot stones as being simultaneously thrown out. They
usually fall back again into the crater, but sometimes they
cover its upper margin, or, visible by their luminosity at a
distance, glide down at night upon a portion of the cone,
which, when seen from a great way off, probably gave origin
to the erroneous notion of La Condamine, "that there was
an effusion of burning sulphur and bitumen." The stones
rise singly one after the other, so that some of them are fall-
ing down while others have only just left the crater. By
an exact determination of time, the visible space of falling
(calculated, therefore, to the margin of the crater) was ascer-
tained to be on the average only 786 feet. On ^tna, ac-
cording to the measurements of Sartorius von Waltershau-
sen and the astronomer D. Christian Peters, the ejected
stones attain an elevation of as much as 2665 feet above
the walls of the crater. Gemellaro's estimates during the
eruption of ^tna in 1832 gave even three times this eleva-
tion ! The black, erupted ashes form layers of three or four
hundred feet in thickness upon the declivities of the Sangay
for a circle of nearly fourteen miles in circumference. The
color of the ashes and rapilli gives the upper part of the
cone a fearfully ^tern character. We must here again call
attention to the colossal size of this volcano, which is six
times greater than that of Stromboli, as this consideration is
* Cosmos, see page 175.

252 COSMOS.

strongly in opposition to the absolute belief that the lower
volcanoes always have the most frequent eruptions.

The grouping of volcanoes is of more importance than
their form and elevation, because it relates to the great geo-
logical phenomenon of upheaval upon fissures. These groups,
whether, according to Leopold von Buch, they rise in lines,
or, united around a central volcano, indicate the parts of
the crust of the earth, where the eruption of the fused in-
terior has found the least resistance, in consequence either
of the reduced thickness of the rocky strata, of their natural
structure, or of their having been originally fissured. Three
degrees of latitude are occupied by the space in which the
volcanic energy is formidably manifested in ^tna, in the
-5^olian islands, in Vesuvius, and the parched land (the Phle-
grjean Fields) from Puteoli (Dicasarchia) to Cumse, and as far
as the fire-vomiting Epopeus on Ischia, the Tyrrhenian isl-
and of Apes, ^naria. Such a connection of analogous
phenomena could not escape the notice of the Greeks. Stra-
bo says: "The whole sea, commencing from Cumas, as far
as Sicily, is penetrated by fire, and has in its depths certain
conduits communicating with each other and with the conti-
nent.* In such a (combustible) nature, as all describe it, ap-

* See Strabo, lib. v., p. 248, Casaubon : ix^L KOi?uag Tivug ; and Iib«
vi., p. 276. Upon a double mode of production of islands the geogra-
pher of Amasia expresses himself (vi., p. 258) with much geological
acumen. "Some islands," says he (and he names them), "are frag-
ments of the main land; others have proceeded from the sea, as still
happens. For the islands of the high sea (those which lie far out in
the sea) were probably upheaved from the depths ; while, on the con-
trary, it is more reasonable to consider those situated at promontories,
and' separated by a strait, as torn from the main land." The small
grou)) of the Pithecusai consists of Ischia, originally called ^naria,
and Procida (Prochyta). The reason why this group was considered
to be an ancient habitation of apes, why the Greeks and the Italian
Tyrrhenians, consequently Etruscans, gave it such a name (apes were
called dpifj.oL, in the Tyrrhenian ; Strabo, lib. xiii., p. 626), remains
veiy obscure, and is perhaps connected with the myth, according to
which the old inhabitants were transforme'd into apes by Jupiter. The
name of the apes, uptfiot, might relate to Arima, or Arimer, of Homer
{Iliad, ii., 783) and Hesiod {Tlieog., v., 301). The words elv 'Apijuoig
of Homer are contracted into one word in some codices, and in this
contracted form we find the name in the Roman writers (Virgil,
yEneid, ix., 716; Ovid, Metamorph., xiv., 88). Pliny (Hist. Nat.,
iii., 5) even says decidedly : "^naria, Homero liiarime dicta, Grsecis

Pithecusa." The Homeric country of the Arimer, Typhon's

resting-place, was sought, even in ancient times in Cilicia, Mysia,
l^ydia, in the volcanic Pithecusse, at the crater Puteolanus, and in the
Phrygian Phlegraja, beneath which Typhon once lay, and even in the

TRUE VOLCANOES. 253

pear, not only JEtna, but also the districts around Dicaearchia
and Naples, and around Baias and Fithecusa;" and from
this arose the fable that Typhon lay under Sicily, and that,
when he turned himself, flames and water burst forth, nay
sometimes even small islands with boiling water. "Fre-.
quently, between Strongyle and Lipara (in this wide dis-
trict), flames have been seen bursting forth at the surface of
the sea, the fire opening itself a passage out of the cavities in
the depths and pressing upward with force." According to
Pindar,* the body of Typhon is of such extent that " Sicily

Katakekaumene. That apes should have lived within historical times
upon Ischia, at such a distance from the African coast, is the more
improbable, because, as I have already observed elsewhere, the an-
cient presence of the apes upon the Rock of Gibraltar does not appear
to be proved, since Edrisi (in the 12th century) and other Arabian
geographers, who describe the Straits of Hercules in such detail, do
not mention them. Pliny also denies the apes of iEnaria,but derives
the name of the Pithecusae in a most improbable manner from TrtOog,
dolium {ajiglinis doliorum). " It appears to me," says Bockh, " to be
the main point in this investigation that Inarima is a name of the
Pithecusae, produced by learned interpretation and fiction, just as
Corcyra became Scheria; and that ^neas was probably only con-
nected with the Pithecusae (JEneae insulae) by the Romans, who find
their progenitors every where in these regions. NEevius also testifies
to their connection with -ZEneas in the first book of the Punic War."
* Pind., Pyth., i., 31. See Straho, v., p. 245 and 248, and xiii.,
p. G27. We have already observed {Cosmos, vol. v., p. 200, that Ty-
phon fled from the Caucasus to Lower Italj^ as though the myth would
indicate that the volcanic eruptions in the latter country w*jre of less
antiquity than those upon the Caucasian Isthmus. The consideration
of mythical views in popular belief can not be separated either from
the geography or the history of volcanoes. The two often reciprocally
illustrate each other. That which was regarded upon, the surface of
the earth as the mightiest of moving forces (Aristotle, MeteoroL, ii.,
8, 3), the wind, the inclosed pnenma, was recognized as the universal
cause of vulcanicity (of fire-vomiting mountains and earthquakes).
Aristotle's contemplation of nature was founded upon the mutual ac-
tion of the external and the internal subterranean air, upon a theory
of transpiration, upon differences of heat and cold, moisture and dry-
ness (Aristotle, Meteor., ii,, 8, 1, 25, 31, and ii., 9, 2). The greater
the mass of the wind inclosed " in subterranean and submarine pas-
sages," and the more it is obstructed in its natural, essential property
of moving far and quickly, the more violent are the eruptions. "Vis
fera ventorum, caecis inclusa cavernis" (Ovid, Metamoi-ph., xv., 299).
Between the wind and the fire there is a peculiar relation. (To nvp
brav /jETu TTvev/iaToc; y, yiverai (j)?ib^ koI (peperai raxeoc; Aristotle,
Met.eoroL, ii,, 8, 3. — nal yap to Tzvp olov TrvevfiaroQ rig (pvoig ; Theo-
phrastus, De Jgne, § 80, p. 715.) The wind {pneuma) suddenly set
free from the clouds, sends the consuming and widely luminous light-
ning flash (jrpriaTTJp). "In the Phlegrsea, the Katakekaumene of
Lydia," says Strabo (lib. xiii., p. G28), " three chasms, fully forty

254 COSMOS.

and the sea-girt heights above Cumae (called Phlegra, or the
burned field) lie upon the shaggy breast of the monster."

Thus Typhon (the raging Enceladus) was, in the popular
fancy of the Greeks, the mythical symbol of the unknown
cause of volcanic phenomena lying deep in the interior of
the earth. By the position and the space which he occupied
were indicated the limitation and the co-operation of partic-
ular volcanic systems. In the fanciful geological picture of
the interior of the earth, in the great contemplation of the
universe which Plato establishes in the Phasdo (p. 112-114),
this co-operation is still more boldly extended to all volcanic
systems. The lava streams derive their materials from the
Pyriphlegethon, which, " after it has repeatedly rolled around
beneath the earth," pours itself into Tartarus. Plato says
expressly that the fire-vomiting mountains, wherever such
occur upon the earth, blow upward small portions from the
Pyriphlegethon (" ovrog d' earlv 6v enovoiid^ovot UvptcpXe-
yedovra, ov teal ol pvaKsg dnoandaixara dvacpvaojoiv, ottt}
dv Tvx^Oi rrjg y^c")* This expression (p. 113 B.) of the
expulsion with violence refers, to a certain extent, to the
moving force of the previously -inclosed wind, then sud-
denly breaking through, upon which the Stagirite after-
ward, in the Meteorology, founded his entire theory of vul-
canicity.

According to these ancient views, the linear arrangement
of volcanoes is more distinctly characterized in the consider-
ation of the entire body of the earth than their grouping
around a central volcano. The serial arrangement is most

stadia from each other, are still shown, which are called the wind-
bags ; above them lie rough hills, which are probably piled up by the
red-hot masses blown up." He had already stated (lib. i., p. 57) "that
between the Cyclades (Thera and Therasia) flames of fire burst forth
from the sea for four days together, so^hat the whole sea boiled and
burned; and an island composed of calcined masses was gradually
raised as if by a lever." All these well-described phenomena are
ascribed to the compressed wind, acting like elastic vapors. Ancient
physical science troubled itself but little "about the peculiar essentials
of material bodies ; it was dynamic, and depended on the measure of
the moving force. We find the opinion that the increasing heat of
the planet with the depth is the cause of volcanoes and earthquakes,
first expressed toward the close of the third century by a Christian
bishop in Africa under Diocletian (Cosmos, vol. v., p. 188). The Pyri-
phlegethon of Plato, as a stream of fire circulating in the interior of the
earth, nourishes all lava-giving volcanoes, as we have already men-
tioned in the text. In the earliest presentiments of humanity, in a
narrow circle of ideas, lie the germs of that which we now think we
may explain under the form of other symbols.

TRUE VOLCANOES. 255

remarkable in those places where it depends upon the situa-
tion and extension of fissures, which, usually parallel to each
other, pass through great tracts of country in a linear direc-
tion (like Cordilleras). Thus, to mention only the most im-
portant series of closely-approximated volcanoes, we find in
the new continent those of Central America^ with their ap-
pendages in Mexico; those of New Granada and Quito, of
Peru, Bolivia, and Chili; in the old continent the Sunda Isl-
ands (the Indian Archipelago, especially Java), the peninsu-
la of Kamtschatka and its continuation in the Kurile Islands,
and the Aleutian Islands, which bound the nearly-closed Beh-
ring's Sea on the south. We shall dwell upon some of the
principal groups ; individual details, by being brought to-
gether, lead us to the causes of phenomena. *

The linear volcanoes of Central America, according to the
older denominations the volcanoes of Costa Rica, Nicaragua,
San Salvador, and Guatemala, extend from the volcano Tur-
rialva, near Cartago, to the volcano of Soconusco, over six
degrees of latitude, between 10° 9^ and 16° 2^, in a line the
general direction of which is from S.E. to N.W., and which,
with the few curvatures which it undergoes, has a length of
540 geographical miles. This length is about equal to the
distance from Vesuvius to Prague. The most closely-ap-
proximated of them, as if they had broken out upon one and
the same fissure only 64 miles in length, are the eight volca-
noes situated between the Laguna de Managua and the Bay
of Fonseca, between the volcano of Momotombo and that of
Conseguina, the subterranean noise of which was heard in
Jamaica and on the highlands of Bogota in the year 1835,
like the fire of artillery. In Central America and the whole
southern part of the new continent, and generally from the
Chonos Archipelago, in Chili, to the most northern volcanoes
of Mount Edgecombe, on the small island near Sitka,* and
Mount Elias, on Prince William's Sound, for a length of
6400 geographical miles, the volcanic fissures have every
where broken out in the western part, or that nearest to the

* Mount Edgecombe, or the St. Lazarus mountain, upon the small
island (Croze's Island, near Lisiansky) which is situated to the west-
ward, near the northern half of the larger island Sitka or Baranow, in
Norfolk Sound, was seen by Cook, and is a hill partly composed of
basalt abounding in olivin, and partly of feldspathic trachyte.* Its
height is only, 2770 feet. Its last great eruption, which produced
much pumice-stone, was in the year 1796. (Lutke, Voyage autmr du
Monde, 1836, t. iii., p. 15.) Eight years afterward Captain Lisiansky
reached the summit, which contains a crater-lake. He found at that
time no signs of activity any where on the mountain.

256 coSxMos.

Pacific Ocean. "Where the line of the Central American vol-
canoes enters with the volcano of Conchagua into the state
of San Salvador, in the latitude of 13^° (to the north of the
Bay of Fonseca), the direction of the volcanoes changes at
once with that of the west coast. The series of the former
then strikes E.S.E. — -W.N.W. ; indeed, where the burning
mountains are again so closely approximated that five, still
more or less active, are counted in the short distance of 120
miles, the direction is nearly E. — W. This deviation cor-
responds with a great dilatation of the continent toward the
east in the peninsula of Honduras, where the coast tends also
suddenly, exactly east and west, from Cape Gracias a Dios
to the Gulf of Amatique for 300 miles, after it had been pre-
viously rurfhing from north to south for the same distance.
In the group of elevated volcanoes of Guatemala (lat. 14° 10^)
the series again acquires its old direction, N. 45° W., which
it continues as far as the Mexican boundary toward Chiapa
and the isthmus of Pluasacualco. Northwest of the volcano
of Soconusco to that of Tuxtla, not even an extinct trachytic
cone has been discovered ; in this quarter granite abounding
in quartz and mica-schist predominate.

The volcanoes of Central America do not crown the ad-
jacent mountain chains, but l-ise along the foot of the latter,
usually completely separated from each other. The greatest
elevations lie at the two extremities of the series. Toward
the south, in Costa Eica, both seas are visible from the sum-
mit of the Irasu (the volcano of Cartago), to which, besides
its elevation (11,081 feet), its central position contributes.
To the southeast of Cartago there stand mountains of ten or
eleven thousand feet : the Chiriqui (11,262 feet) and the Pico
Blanco (11,740 feet). We know nothing of the nature of
their rock, but they are probably unopened trachytic cones.
Farther toward the southeast, the elevations diminish in Ver-
agua to six and five thousand feet. This appears also to be
the average height of the volcanoes of Nicaragua and San Sal-
vador; but toward the northwestern extremity of the whole
series, not far from the new city of Guatemala, two volcanoes
again rise above 13,000 feet. The maxima consequently fall
into the third group of my attempted hypsometric classifica-
tion of volcanoes, coinciding with ^tna and the Peak of Ten-
erifife, while the greater number of the heights lying between
the two extremities scarcely exceed Vesuvius by 2000 feet.
The volcanoes of Mexico, New Granada, and Quito belong to
the fifth group, and usually attain an elevation of more #han
17,000 feet.

TRUE VOLCANOES. 257

Although the continent of Central America increases con-
siderably in breadth from the isthmus of Panama, through
Veragua, Costa Rica, and Nicaragua, to the latitude of 11^^,
the great area of the Lake of Nicaragua and the small eleva-
tion of its surface (scarcely 128 feet* above the two seas) gives
rise to such a degradation of the land exactly in this district,
that by it an overflow of air from the Caribbean Sea into the
Great South Sea is often caused, bringing danger to the voy-
ager in the so-called Pacific Ocean. The northeast storms
thus excited have received the name of Papagayos, and some-
times rage without intermission for four or five days. They
have the remarkable peculiarity that during their continu-
ance the sky usually remains quite cloudless. The name is
borrowed from the part of the west coast of Nicaragua be-
tween Brito or Cabo Desolado and Punta S. Elena (from 11°
22'' to 10° 50^), which is called Golfo del Papagayo, and in-
cludes the small bays of Salinas and S. Elena, to the south
of the Puerto de San Juan del Sur. On my voyage from
Guayaquil to Acapulco, I was able to observe the Papagayos,
in all their violence and peculiarity, for more than two whole
days (9 th — U-th March, 1803), allhough rather more to the
south, in less than U° 13^ of latitude. The waves rose higher
than I have ever seen them ; and the constant visibility of
the disk of the sun in the bright, blue arch of heaven enabled
me to measure the height of the waves by altitudes of the
sun taken upon the ridge of the wave and in the trough, by
a method which had not been tried at that time. All Span-
ish, English,! and American voyagers ascribe the above-de-
scribed storms of the Southern Ocean to the northeast trade-
wind of the Atlantic.

In a new workf which I have undertaken with much as-

* Even under the Spanish government in 1781, the Spanish engi-
neer, Don Jose Galisteo, had found for the surface of the Laguna of
Nicaragua an elevation only six feet greater than that ^ven by Baily
in his different levelings in 1838 (Humboldt, Relation Historique, t. iii.,
p. 321).

t See Sir Edward Belcher, Voyage round the World, vol. i., p. 185.
According to my chronometric longitude, I was in the Papagayo storm,
19° 11' to the west of the meridian of Guayaquil, and consequently
99° 9' west, and 880 miles west of the shore of Costa Eica.

X My earliest work upon seventeen linear volcanoes of Guatemala
and Nicaragua is contained in the Geographyiji Journal of Berghaus
{Hertha, bd. vi., 1826, p. 131-161). Beside^he old Chronista Fu-
entes (lib. ix., cap. 9), I could then only make use of the important
work of Domingo Juarros, Compendio de la Historia de la Civdad de
Guatemala, and of the three maps by Galisteo (drawn in 1781, at the

258 COSMOS.

siduity — partly from materials already published, and partly
from miinuscript notes — upon the linear volcanoes of Cen-

conimaud of the Mexican viceroy, Matias de Galvez), by Jose Rossi y
Rubi (Alcalde Mayor de Guatemala, 1800), and by Joaquin Ysasi and
Antonio de la Cerda (Alcalde de Granada), which I possessed princi-
pally in manuscript. In the French translation of 'his work upon the
Canary Islands, Leopold von Buch has given a masterly extension of
my first sketch (Descr. Physique des Isles Canaries, 1836, p. 500-514) ;
but the uncertainty of geographical synonyms and the confusion of
names caused thereby gave rise to many doubts, which have been for
the most part removed by the fine maps of Baily and Saunders ; by
Molina's Bosquejo de la Repuhlica de Costa Rica ; and by the great and
Very meritorious work of Squier (Nicaragua, its People and Monuments,
with Tables of the Comparative Heights of the Mountains in Central Amer-
ica, 1852, vol. i., p. 418, and vol. ii., p. 102). The important work
which is promised us by Dr. Oerstedt, under the title of Schilderung
der Naturverhaltnisse von Nicaragua und Costa Rica, besides the ad-
mirable botanical and geological discoveries which constitute the pri-
mary object of the undertaking, will also throw light upon the geog-
nostic nature of Central America. Dr. Oersted passed through that
region in various directions from 1846 to 1848, and brought back a
collection of rocks to Copenhagen. I am indebted to his friendly
communications for interesting corrections of my fragmentary work.
!From a careful comparison ofHhe materials with which I am acquaint-
ed, including those collected by Hesse, the Prussian c^sul-general in
Central America, which are of great value, I bring together the vol-
canoes of Central America in the following manner, proceeding from
south to north :

Above the central plateau of Cartago (4648 feet), in the republic
of Costa Rica (lat. 10° 9), rise the three volcanoes of Turrialva, Irasu,
and Reventado, of which the first two are still ignited.

Volcan de Turrialva* (height about 11,000 feet) is, according to
Oersted, only separated from the Irasu by a deep, narrow ravine.
Its summit, from w hich columns of smoke rise, has not yet been
ascended.
The volcano Irasu,* also called the volcano of Cartago (11,100 feet),
to the northeast of the volcano Reventado, is the principal vent
of volcanic activity in Costa Rica, but still remarkably accessible,
and toward the south divided into terraces in such a manner that
one may^on horseback, almost reach the elevated summit, from
which the two oceans, the sea of the Antilles and the Pacific, may
be seen at once. The cone of ashes and rapilli, which is about a
thousand feet in height, rises out of a wall of circumvallation (a
crater of elevation). In the flatter, northeastern part of the sum-
mit lies the true crater, of 7500 feet in circumference, which has
never emitted lava streams. Its eruptions of scoriae have often
(1723, 1726, 1821, 1847) been accompanied by destructive earth-
quakes, the effectarf which has been felt from Nicaragua or Rivas
to Panama (OeiHl). During a very recent ascent of the Irasu,
in the beginning of May, 1855, by Dr. Carl Hoftmann, the crater
of the summit and its eruptive orifices have been more accurate-
ly investigated. The altitude of the volcano is stated, from a

TRUE VOLCANOES. 259

tral America, twenty-nine volcanoes are numbered, whose
former or present varied activity may be stated with cer-

trigonometrical measurement by Galindo, at 12,000 Spanish feet,
or, taking the vara cast.—O-i^ of a toise, at 11,000 feet. (J3on-
plandia, Jahrgang 1856, No. 3.)

£1 Reventado (about 9500 feet), with a deep crater, of which the
southern margin has fallen in, and which was formerly filled with
water.

The volcano Barha (more than 8419 feet),*to the north of San Jose,
the capital of Costa Rica ; with a crater which contains several
small lakes.

Between the volcanoes Barba and Orosi there follows a series of
volcanoes which intersects the principal chain, running S.E. — N.W.
in Costa Eica and Nicaragua, almost in the opposite direction, east and
west. Upon such a fissure stand, farther to the eastwaM, Miravalles
and Tenorio (each of these volcanoes is about 4689 feet) ; in the cen-
tre, to the southeast of Orosi, the volcano Eincon, also called Rincon de
la Vieja* (Squier, vol. ii., p. 102), which exhibits small eruptions of
ashes every spring at the commencement of the rainy season ; and
farthest to the westward, near the little town of Alajuela, the volcano
Votos* (7513 feet), which abounds in sulphur. Dr. Oersted compares
this phenomenon of the direction of volcanic activity upon a trans-
verse fissure with the east and west direction, which I found in the
Mexican volcanoes froni sea to sea.

Orosi,'*' still active, in the most southern part of the state of Nica-
ragua (5222 feet) ; probably the Volcan del Papagayo, on the chart of
the Deposit o Jlidrograjico.

The two volcanoes Mandeira and Ometepec* (4157 and 5222 feet),
upon a small island in the western part of the Laguna de Nicaragua,
named by the Aztec inhabitants of the district after these two mount-
ains (ome tepetl signifies two mountains ; see Buschmann, Aztekische
Ortsnamen, p. 178 and 171). The insular volcano Ometepec, errone-
ously named Ometep by Juarros (^Hist. de Guatemala, t. i., p. 51), is
still in activity. It is figured by Squier (fbl. ii., p. 235).

The extinct crater of the island Zapatera, but little elevated above
the sea-level. The period of its ancient eruptions is quite unknown.

The volcano of Momohacho, on the western shore of the Laguna de
Nicaragua, somewhat to the south of the city of Granada. As this
city is situated between the volcanoes of Momobacho (the place is also
called Mombacho, Oviedo, Nicaragua, ed. Ternaux, p. 245) and Ma-
saya, the pilots indicate sometimes the one and sometimes the other
of these conical mountains by the indefinite name of the Volcano of
Granada.

The volcano Massaya (Masaya), which has already been treated of
in detail (p. 258-261), was" once a Stromboli, but has been extinct
since the great eruption of lava in 1670. According to the interesting
reports of Dr. Scherzer {Sitzungsberichte der Philos. Hist. Classe der
Akad. der Wiss. zu Wien, bd. xx., s. 58), dense clouds of vapor were
again emitted in April, 1853, from a newly-opened crater. The vol-
cano of Massaya is situated between the two lakes of Nicaragua and
Managua, to the west of the city of Granada. Massaya is not sy nony-

260 COSMOS.

tainty. The natives make the number more than one third
greater, taking into account a quantity of old eruptive ba-

mous with Nindiri ; but, as Dr. Oersted expresses himself, Massaya
and Nindiri* form a twin volcano, with two summits and two distinct
craters, both of which have furnished lava streams. The lava stream
of 1775 from the Nindiri reached the Lake of Managua. The equal
height of these two volcanoes, situated so close to each other, is stated
at only 2450 feet.

Volcan de Momotoiuho* (7034 feet), burning, and often giving forth
a thundering noise, but without smoking, in lat. 12° 28', at the north-
ern extremity of the Laguna de Managua, opposite to the small island
Momotombito, so rich in sculptures (see the representation of Momo-
tombo in Squier, vol. i., p. 233 and 302-312). The Laguna de Ma-
nagua lies 28 feet higher than the Laguna de Nicaragua, which is
more than double its size, and has no insular volcano.

From hence to the Bay of Fonseca or Conchagua, at a distance of
23 miles from the coast of the Pacific, a line of six volcanoes runs
from S.E. to N.W. ; closely approximated to each other, and bearing
the common name of Los Maribios (Squier, vol. i., p. 419; vol. ii.,
p. 123).

jm Nuevo,* erroneously called Volcan de las Pilas, because the erup-
tion of the 12th April, 1850, took place at the foot of this mountain ;
a great eruption of lava almost in the plain itself! (Squier, vol. ii.,
p. 105-110.)

Volcan de Telica,* visited, during its activity, by Oviedo as early as
the 16th century (about 1529), to the east of Chinendaga, near Leon
de Nicaragua, and consequently a little out of the direction previous-
ly stated. This important volcano, which emits much sulphurous va-
por from a crater 320 feet in depth, was ascended, a few years since,
by my scientific and talented friend Professor Julius Frobel. He
found the lava composed of glassy feldspar and augite (Squier, vol. ii.,
p. 115-117). At the summit, at an elevation of 3517 feet, there is a
crater in which the vapoi's deposit great masses of sulphur. At the
foot of the volcano is a mffe-spring (Salse ?).

The volcano JEl Viejo* the northernmost of the crowded line of six
volcanoes. It was ascended and measured in the year 1838 by Cap-
tain Sir Edward Belcher. The result of the measurement was 5559
feet, a more recent measurement, by Squier, gave G002 feet. This
volcano, which was very active in Dampier's time, is still burning.
The fiery eruptions»of scoriae* are frequently seen in the city of Leon.

The volcano Guanacaure, somewhat to the north, without the range
from El Nuevo to the Viejo, at a distance of only 14 miles from the
shore of the Bay of Fonseca.

The volcano Oonseguina* upon the cape which projects at the south-
ern extremity of the Bay of Fonseca (lat. 12° 50'), celebrated for the
fearful eruption, preceded by earthquakes, of the 23d January, 1 835.
The gi'eat darkness during the fall of ashes, similar to that which has
sometimes been caused by the volcano Pichincha, lasted for 43 hours.
At a distance of a few feet, fire-brands could not be perceived. Ees-
piration was obstructed, and a subterranean noise, like the discharge
of heavy artillery, was heard not only in Balize, on the peninsula of

TRUE VOLCANOES. 261

sins; which were probably only lateral eruptions on the de-
clivity of one and the same mountain. Among the isolated

Yucatan, but also upon the coast of Jamaica, and upon the plateau of
Bogota, in the latter case at an elevation of more than 8500 feet above
the sea, and at a distance of nearly five hundred and sixty geograph-
ical miles (Juan Galindo, in Silliman's American Journal, vol. xxviii.,
1835, p. 332-336 ; Acosta, Viojes d los Andes, 1849, p. 56 ; and Squier,
vol. ii., p. 110-113; figures, p. 163 and 165). Darwin {Journal of Re-
searches during the Voyage of the Beagle, 1845, p. 291) calls attention
to a remarkable coincidence of phenomena: After a long slumber,
Conseguina, in Central America, and Aconcagua and Corcovado (S.
lat. 32|° and 43^°), in Chili, broke out on the same day (accidentally?).

Volcano of Conchagua, or of Amalapa, at the north of the entrance
to the Bay of Fonseca, opposite to the volcano Conseguina, near the
beautiful Puerto de la Union, the harbor of the neighboring town of
San Miguel.

From the state of Costa Rica to the volcano of Conchagua, there-
fore, the close series of twenty volcanoes follows a direction from S.E.
to N.W. ; but on entering, near Conchagua, into the state of San Sal-
vador, which, in the short distance of 160 geographical miles, exhibits
five still more or less active volcanoes, the line, like the Pacific coast
itself, turns more E.S.E. — W.N.W., and indeed almost E. — W., while
on the eastern, Caribbean coast (toward the Cape Gracias a Dios) the
land suddenly bulges out in Honduras and los Mosquitos (see above,
p. 256). It is only, as there remarked, to the north of the high volca-
noes of Old Guatemala, toward the Laguna de Atitlan, that the former
general direction N. 45° W. again occurs, until at last, in Chiapa, and
on the isthmus of Tehuantepec, the abnormal direction E. — W. is again
manifested, but in non-volcanic chains. Besides Conchagua, the fol-
lowing four volcanoes belong t-o the state of^San Salvador :

The volcano of San 3Iiguel Bosotlan* (lat. 13° 35'), near the town of
the same name, the most beautiful and regular of trachytic cones
next to the insular volcano Ometepec, in the lake of Nicaragua (Squier,
vol. ii., p. 196). The volcanic forces are very active in Bosotlan, in
which a great eruption of lava occurred on the 20th of July, 1844.

Volcano of San Vicente,* to the west of the Rio de Lempa, between
the towns of Sacatecoluca and Sacatelepe. A great eruption of ashes
took place, according to Juarros, in 1643 ; and in January, 1835, a
long-continued eruption occurred with destructive earthquakes.

Volcano 'of /Saw Salvador (lat. 13° 47), near the city of the same
name. The last eruption was that of 1656. The whole surrounding
country is exposed to violent earthquakes; that of the 16th of April,
1854, which was preceded by no noises, overthrew nearly all the build-
ings in San Salvador.

Volcano of Izalco,* near the village of the same name, often pro-
ducing ammonia. The first eruption recorded in history occurred on
the 23d February, 1770; the last widely-luminous eruptions were in
April, 1798, 1805 to 1807, and 1825 (see above, p. 248, and Thompson,
Official Visit to Guatemala, 1829, p. 512).

Volcan dePacaya* (lat. 14° 23'), about 14 miles to the southeast of
the city of New Guatemala, on the small Alpine lake Amatitlan, a

262 COSMOS.

conical and bell-shaped mountains, which are there called
volcanoes, many may, indeed, consist of trachyte and dol-

very active and often flaming volcano ; an extended ridge with three
domes. The great eruptions of 1565, 1651, 1671, 1677, and 1775 are
known ; the last, which produced much lava, is described by Juarros
as an eye-witness.

Next follow the two volcanoes of Old Guatemala, with the singular
appellations jDe u4^?fa and DeFuego, near the coast, in latitude 14° 12'.

Volcan de Agua^ a trachytic cone n?ar Escuintla, higher than the
Peak of Teneriffe, surrounded by masses of obsidian (indications of
old eruptions ?). The volcano, which reaches into the region of per-
petual snow, has received its name from the circumstance that, in
September, 1541, a great inundation (caused by earthquake and the
melting of snow ?) was ascribed to it ; this destroyed the first-estab-
lished city of Guatemala, and led to the building of the second city,
situated to the north-northwest, and now called Antigua Guatemala.

Volcan de Fuego* near Acatenango, 23 miles in a west-northwest
direction from the so-called water- volcano. With regard to their rela-
tive position, see the rare map of the Alcalde Mayor, Don Jose Eossi,
y Rubi, engraved in Guatemala, and sent to me thence as a present:
JBosquejo del espacio que media entre los estremos de la Provincia de
Stichitepeques y la Capital de Guatemala, 1800. The Volcan de Fuego
is still active, but now much less so than formerly. The older great
eruptions were those of 1581, 1586, 1623, 1705, 1710, 1717, 1732, 1737,
and 1799, but it was not only these eruptions, but also the destructive
earthquakes which accompanied them, that moved the Spanish gov,-
ernment, in the second half of the last century, to quit the second seat
of the city (where the ruins of la Antigua Guatemala now stand), and
compel the inhabitants to settle farther to the north, in the new city
of Santiago de Guatemala. In this case, as at the removal of Rio-
bamba, and several other towns near the volcanoes of the chain of the
Andes, a dogmatic and vehement dispute was carried on in reference
to the difficult selection of a locality "of which it might be asserted,
according to previous experience, that it was but little exposed to the
action of neighboring volcanoes (lava streams, eruptions of scoriae, and
earthquakes !)." In 1852, during a great eruption, the Volcan de Fu-
ego poured forth a lava stream toward the shore of the Pacific. Cap-
tain Basil Hall measured, under sail, both the volcanoes of Old Gua-
temala, and found for the Volcan de Fuego 14,665 feet, and for the
Volcan de Agua 14,903 feet. The foundation of this measurement
has been tested by PoggendoriF. He found the mean elevation of the
two mountains to be less, and reduced it to about 13,109 feet.

Volcan d^ Quesaltenango* (lat. 15° 10'), burning since 1821, and
smoking, near the town of the same name ; the three conical mount-
ains which bound the Alpine lake Atitlan (in the mountain chain of
Solola) on the south, are also said to be ignited. The volcano of Ta-
jamulco, referred to hj Juarros, certainly can not be identical Avith the
volcano of Quesaltenango, as the latter is at a distance of 40. geograph-
ical miles to the N.W. of the village of Tajamulco, to the south of
Tejutla.

What are the two volcanoes of Sacatepeques and Sapotitlan, men-
tioned by Funel, or Brue's Volcan de Amilpasf

TRUE VOLCANOES. 263

erite, but, having always been unopened, have never exhib-
ited any igneous activity since the time of their upheaval.
Eighteen are to be regarded as still active; seven of these
have thrown up flames, scorige, and lava streams in the pres-
ent century (1825, 1835, 1848, and 1850)^ and two"^ at the
end of the last century (1775 and 1799). The deficiency of
lava streams in the mighty volcanoes of the Cordilleras of
Quito has recently given occasion to the repeated assertion
that this deficiency is equally general in the volcanoes of
Central America. Certainly, in the majority of cases, erup'
tions of scorise and ashes have been unaccompanied by any
effusion of lava — as, for example, at present in the volcano
of Izalco; but the descriptions which have been given by
eye-witnesses of the lava-producing eruptions of the four vol-
canoes Nindiri, El Nuevo, Conseguina, and San Miguel de
Bosotlan give an opposite testimony.!

I have purposely dwelt at length upon the details of the
position and close approximation of the linear volcanoes of
Central America, in the hope that some day a geognosist,
who has previously given a profound study to the active vol-
canoes of Europe and the extinct ones of Auvergne, the
Vivarais or the Eifel, and who also (this is of the*- greatest
importance) knows how to describe the mineralogical cgm-

The great volcano of Soconusco, situated on the borders of Chiapa,
28 geographical miles to the south of Cuidad Real, in lat. 16° 2'.

At the close of this long note I think I must again mention that the
barometric determinations of altitude here adduced are partly derived
from Espinache, and partly borrowed from the writings and maps of
Baily, Squier, and Molina.

* The following 1 8 volcanoes, constituting, therefore, nearly the half
of all those referred to by me as active in former or present times, are
to be regarded as at present more or less active : Irasu and Turrialva,
near Cartago, El Rincon de la Vieja, Votos(?) and Orosi ; the insular
volcano Omete})ec, Nindiri, Momotomba, El Nuevo, at the foot of the
trachytic mountain Las Pilas, Telica, El Viejo, Conseguina, San Mi-
guel Bosotlan, San Vicente, Izalco, Pacaya, Volcan de Fuego (de Gua-
temala), and Quesaltenango. The most recent eruptions are those
of El Nuevo, near Las Pilas, on the 18th April, 1850; San Miguel
Bosotlan, 1848; Conseguina and San Vicente, 1835; Izalco, 1825;
Volcan de Fuego, near New Guatemala, 1799 and 1852 ; and Pacaya,
1775.

t Compare Squier, Nicaragua, vol. ii., p. 103, with p. 106 and 111,
as also his previous small work On the Volcanoes of Central America,
1850, p. 7; Leopold de Buch, Ues Canaries, p. 506, where reference
is made to the lava stream which broke out of the volcano Nindiri in
1775, and which has been recently again seen by a very scientific ob-
server, Dr. Oersted.

264 COSMOS.

position of the different rocks in accordance with the present
state of our knowledge, may feel himself impelled to visit this
region, which is so near and so accessible. Even if the trav-
eler should devote himself exclusively to geognostic investi-
gations, there still remains much to be done here — especially
the oryctognostic determination of the trachytic, doleritic,
and melaphyric rocks ; the separation of the primitive mass
upheaved, and of the portion of the elevated mass which has
been covered over by subsequent eruptions ; the seeking out
and recognition of true, slender, uninterrupted lava streams,
which are only too frequently confounded with accumulations
of erupted scoriae. Conical mountains, which have never
been opened, rising in a dome or bell-like form, such as Chim-
borazo, are, therefore, to be clearly separated from volcanoes
which have been or still are, active, throwing out scorias and
lava streams, like Vesuvius and ^tna, or scoriae and ashes
alone, like Pichincha or Cotopaxi. I know nothing that
promises to impart a more brilliant impetus to our knowl-
edge of volcanic activity, which is still very deficient in multi-
plicity of observations in large and connected continental dis-
tricts. As the material results of such a labor, collections
of rocks^would be brought home from many isolated true vol-
canoes and unopened trachytic cones, together with the non-
volcanic masses which have been broken through by both ;
the subsequent chemical analyses, and the chemico-ggological
inferences deduced from the analyses, would open a field
equally wide and fertile. Central America and Java have
the unmistakable superiority over Mexico, Quito, and Chili,
that in a greater space they exhibit the most variously-formed
and most closely-approximated stages of volcanic activity.

At the point where the characteristic series of the volca-
noes of Central America terminates on the borders of Chiapa
with the volcano of Soconusco (lat. 16° 2^), there commences
a perfectly different system of volcanoes — the Mexican. The
isthmus of Huasacualco and Tehuantepec, so important for
the trade with the coast of the Pacific, like the state of Oaxa-
ca, situated to the northwest, is entirely without volcanoes,
and perhaps even destitute of unopened trachytic cones. It
is only at a distance of 160 geographical miles from the vol-
cano of Soconusco that the small volcano of Tuxtla rises, near
the coast of Alvarado (lat. 18° 28^. Situated on the eastern
slope of the Sierra de San Martin, it had a great eruption of
flames and ashes on the 2d of March, 1793. An exact astro-
nomical determination of the position of the colossal snowy

TRUE VOLCANOES. 265

mountains and volcanoes in the interior of Mexico (the old
Anahuac) led me, after my return to Europe, while inserting
the maxima of elevations in my chart of New Spain, to the
exceedingly remarkable result that there is in this place, from
sea to sea, a parallel of the volcanoes and greatest elevations
which oscillates by only a few minutes to and from the paral-
lel of 19°. The only volcanoes, and, at the same time, the
only mountains, covered with perpetual snow in the country,
and consequently elevations varying from 12,000 to 3000
feet — the volcanoes of Orizaba, Popocatepetl, Toluca, and
Colima — lie between the latitudes of 18° 59^ and 19° 20^,
and thus indicate the direction of a fissure of volcanic activity
of 360 geographical miles in length.* In the same direction
(lat. 19° 9 ), between the volcanoes of Toluca and Colima, at
a distance of 116 and 128 geographical miles from them, the
new volcano of Jorullo (4265 feet) rose on the 14th Septem-
ber, 1759, in a broad plain, having an elevation of 2583 feet.
The local position of this phenomenon in relation to the sit-
uation of the other Mexican volcanoes, and the circumstance
that the fissure from east to west, which I here indicate, in-
tersects the direction of the great mountain chain striking from

* See all the bases of these Mexican local determinations, and their
comparison with the observations of Don Joaquin Ferrer, in my Recueil
d' Observations Astron., vol. ii., p. 521, 529, and 536-550; and Essai
Politique su?- la Nouvelle Espagne, t. i., p. 55-59, and 170, t. ii., p. 173.
I bad myself early raised doubts with regard to the astronomical de-
termination of the position of the volcano of Colima, near the coast
of the Pacific {Essai Polit.^ t. i., p. 68 ; t. ii., p. 180). According to
angles of altitude taken by Captain Basil Hall while under sail, the
volcano is situated in lat. 19° 36', and consequently half a degree far-
ther north than I concluded to be its position from itineraries; cer-
tainly without absolute determinations for Selagua and Petatlan, upon
which I depended. The latitude, 19° 25', which I have given in the
text, is, like the determination of altitude (12,005 feet), from Captain
Beechey {Voyage, pt. ii., p. 587). The most recent map by Laurie
{The Mexican and Central States of America, 1853) gives 19° 20' for
the latitude. The latitude of Jorullo may also be wrong by 2 — 3
minutes, as I was then occupied entirely with geological and topo-
graphical investigations, and neither the sun nor stars were visible for
determinations of latitude. (See Basil Hall, Journal written on the
Coast of Chili, Peru, and Mexico, 1824, vol. ii., p. 379; Beechey, Voy-
age, pt. ii., p. 587 ; and Humboldt, Essai Polit., t. i., p. 68 ; t. ii., p.
180). In the true and exceedingly artistic views of the volcano of
Colima, drawn by Moritz Rugendas, which are preserved in the Ber-
lin Museum, we distinguish two adjacent mountains — the true volcano,
which constantly emits smoke, and is covered with but little snow, and
the more elevated Nevada, which rises far into the region of perpetual
snow.

Vol. v.— M

266

COSMOS.

south-soutlieast to north-northwest almost at right angles, are
geological phenomena no less important than the distance of
the eruption of JoruUo from the seas, the evidence of its up-
heaval which I have represented graphically in detail, the
innumerable fuming hornitos which surround the volcano,
and the fragments of granite, which I found immersed in
the lava poured forth from the principal volcano of JoruUo,
in a district which is destitute of granite for a long distance.
The following table contains the special local determina-
tions and elevations of the series of volcanoes of Anahuac,
upon a fissure which, running from sea to sea, intersects the
fissure of elevation of the great range of mountains :

Sequence from East to West.

Latitude.

Elevation

above the Sea,

in Feet.

Volcano of Orizaba

19° 2' 17"
19 10 3

18 59 47

19 11 33
19 9 0
19 20 0

17,879
15,705
17,726
15,168
4,265
12,005

Nevado Iztaccihuatl

Volcano of Toluca

Volcano of Colima

The prolongation of the parallel of volcanic activity in the
tropical zone of Mexico leads, at a distance of 506 miles west-
ward, from the shores of the Pacific to the insular group Ke-
villagigedo, in the vicinity of which Collnet saw pumice-stone
floating, and perhaps still farther on, at a distance of 3360 ge-
ographical miles, to the great volcano Mauna Roa (19o 28^),
without causing any upheaval of islands in the intervening
space !

The group of linear volcanoes of Quito and New Granada
includes a volcanic zone which extends from 2° S. lat. to
nearly 5° N. lat. The extreme boundaries of the area in
which the reaction of the interior of the earth upon its surface
is now manifested are the uninterruptedly active Sangay, and
the Paramo and Volcan de Euiz, the most recent conflagra-
tion of which was in the year 1829, and which was seen smok-
ing by Carl Degenhardt from the Mina de Santana, in the
province of Mariquita, in 1831, and from Marmato in 1833.
The most remarkable" traces of great eruptive phenomena next
to the Ruiz are exhibited from north to south, by the trun-
cated cone of the volcano of Tolima (18,129 feet), celebrated
by the recollection of the destructive eruption of the 12th
March, 1595 ; the volcanoes of Purace (17,006 feet) and So-
tara, near Popayan ; that of Pasto (13,450 feet), near the city
of the same name; of the Monte de Azufre (12,821 feet).

TRUE VOLCANOES. 267

near Tuquerres ; of Cumbal (15,618 feet) and of Chiles, in
the province de los Pastos ; then follow the historically cel-
ebrated volcanoes of the true highland of Quito, to the south
of the equator, of which four — namely, Pichincha, Cotopaxi,
Tungurahua, and Sangay — certainly can not be regarded as
extinct volcanoes. Although, to the north of the mountain
group of the Robles, near Popayan, as we shall shortly more
fully show in the tripartition of the vast chain of the Andes,
it is only the central Cordillera, and not the western one,
nearer to the sea-coast, that exhibits a volcanic activity ; on
the other hand, to the south of this group, where the Andes
form only two parallel chains, so frequently mentioned by
Bouguer and La Condamine in their writings, volcanoes are
so equally distributed, that the four volcanoes of the Pastos,
as well as Cotocachi, Pichincha, Iliniza, Carguairazo, and
Yana-Urcu, at the foot of Chimborazo, have broken out upon
the western chain, nearest to the sea ; and upon the eastern
Cordillera, Imbabura, Cayambe, Antisana, Cotopaxi, Tung-
urahua (opposite to Chimborazo toward the east, but still
nearly approximated to the middle of the narrow elevated
plateau), the Altar de los Collanes (Capac-Urcu), and San-
gay. If we include the northernmost group of the linear
volcanoes of South America in one view, the opinion so often
expressed in Quito, and to a certain extent founded on his-
torical documents, of the migration of the volcanic activity
and increase of intensity from north to south, acquires, at all
events, a certain amount of probability. It is true that in
the south, and indeed close to the colossal Sangay, which
acts like Stromboli, we find the ruins of the "Prince of
Mountains," Capac-Urcu, which is said to have exceeded
Chimborazo in height, but which fell in and became extinct
in the latter part of the 15th century (fourteen years before
the capture of Quito by the son of the Inca Tupac Yupangui),
and has. never again resumed its former activity.

The space of the chain of the Andes which is not occupied
by groups of volcanoes is far greater than is usually supposed.
In the northern part of South America, from the Volcan de
Ruiz and the conical mountain Tolima, the two most northern
volcanoes of the series of New Granada and Quito, over the
isthmus of Panama as far as the vicinity of Costa Pica, where
the series of volcanoes of Central America commences, there
is a country which is frequently and violently convulsed by
earthquakes, and in which flaming salses, but no true volcan-
ic eruptions, are known. The length of this tract amounts

268 COSMOS.

ii> 628 geographical miles. Nearly double this length (occu-
pying a space of 968 geographical miles) is a tract of country
free from volcanoes, from the Sangay, the southern termina-
tion of the group of New Granada and Quito, to the Chacani,
near Arequipa, the commencement of the series of volcanoes
of Peru and Bolivia — so complicated and various in the
same mountain chain must have been the coincidence of the
conditions upon which depends the formation of permanently
open fissures, and the unimpeded communication of the molt-
en interior of the earth with the atmosphere. Between the
groups of trachytic and doleritic rocks, through which the
volcanic forces become active, lie rather shorter spaces, in
which prevail granite, syenite, mica-schists, clay-slates, quartz-
ose porphyries, silicious con gloiji crates, and limestones, of
which (according to Leopold von Buch's investigation of the
organic remains brought home by Degenhardt and myself) a
considerable portion belong to the chalk formation. The
gradually increased frequency of labradoritic rocks, rich in
pyroxene and oligoclase, announces to the observant traveler
(as I have already elsewhere shown) the transition of a zone
hitherto closed and non-volcanic, and often very rich in sil-
ver in porphyries, destitute of quartz and full of glassy feld-
spar, into the volcanic regions, which still freely communi-
cate with the interior of the earth.

The more accurate knowledge which we have recently at-
tained of the position and boundaries of the five groups of
volcanoes (the groups of Anahuac or tropical Mexico, of
Central America, of New Granada and Quito, of Peru and
Bolivia, and of Chili) shows that, in the part of the Cordil-
leras which extends from 19i° north to 46° south latitude
(and, consequently, taking into account the curves caused by
alterations in the axial direction, for a distance of nearly
5000 geographical miles), not much* more than half (calcu-

* The following is the result of the determination of the length and
latitude of the five groups of linear volcanoes in the chain of the Andes,
as also the statement of the distance of the groups from each other : a
statement illustrating the relative proportions of the volcanic and non-
volcanic areas :

I. Group of the Mexican Volcanoes : The fissure upon which the vol-
canoes have broken out is directed from east to west, from the
Orizaba to the Colima, for a distance of 392 geographical miles,
between latitudes 19° and 19° 20'. The volcano of Tuxda lies
isolated 128 miles to the east of Orizaba, near the coast of the
Gulf of Mexico, and in a parallel (18° 28') which is half a degree
farther south.

TRUE VOLCANOES. 269

lation gives 2540 against 2428 geographical miles) is occu-
pied by volcanoes. If we examine the distribution of the
space free from volcanoes between the five volcanic groups,
we find the maximum distance of two groups from one an-
il. Distance of the Mexican group from the next group, that of Cen-
tral America (from the volcano of Orizaba to the volcano of So-
conusco, in the direction E.S.E. — W.N.W.), 300 miles.

III. Group of the Volcanoes of Central America : Its length from S.E.
to N.W., from the volcano of Soconusco to Turrialva, in Costa
Eica, more than 680 miles.

IV. Distance of the group of Central America from the series of
volcanoes of New Granada and Quito, 628 miles.

V. Group of the Volcanoes of New Granada and Quito : Its length from
the eruption in the Paramo de Ruiz to the north of the Volcan de
Tolima, to the volcano of Sangay, 472 miles. The portion of the
chain of the Andes between the volcano of Purace, near Popayan,
and the southern part of the volcanic mountain group of Pasto is
directed N.N.E. — S.S. W. Far to the eastward from the volcanoes
of Popayan, at the sources of the Rio Fragua, there is a very iso-
lated volcano, which I have inserted upon my general map of the
mountain group of the South American Cordilleras, from the
statements of missionaries from Timana, which were communi-
cated to me : distance from the sea-shore, 152 miles.

VI. Distance of the volcanic group of New Granada and Quito from
the group of Peru and Bolivia, 960 miles, the greatest length des-
titute of volcanoes.

Vil. Group of the Series of Volcanoes of Peru and Bolivia, from the
Volcan de Chacani and Arequipa to the volcano of Atacama (161:°
— 21J°), 420 miles.

VIII. Distance of the Group of Peru and Bolivia from the volcanic
group of Chili, 540 geographical miles. From the portion of the
desert of Atacama, on the border of which the volcano of San
Pedro rises, to far beyond Copiapo, even to the volcano of Co-
quimbo (30° 5'), in the long Cordillera to the west of the two prov-
inces Catamarca and Rioja, there is no volcanic cone.

IX. Group of Chili, from the volcano of Coquimbo to the volcano
San Clemente, 968 miles.

These estimates of the length of the Cordilleras, with the curvature
which results from the change in the direction of the axis, from .the
parallel of the Mexican volcanoes in ]9i° N. lat., to the volcano of
San Clemente in Chili (46° 8' S. lat.), give, for a distance of 4968
miles, a space of 2540 miles which is covered by five linear groups of
volcanoes (Mexico, Central America, New Granada with Quito, Peru
with Bolivia, and Chili); and a space probably quite free from volca-
noes of 2428 miles. The two spaces are nearly equal. I have ffiven
very definite numerical relations, as obtained by the careful criticism
of my own maps and those of others, in order to give rise to a greater
desire to improve them. The longest portion of the Cordilleras free
from volcanoes is that between the groups of New Granada with Quito,
and Peru with Bolivia. It is accidentally equal to that occupied by
the volcanoes of Chili.

270

COSMOS.

other between the volcanic series of Quito and Peru. This
is fully 960 miles, while the most closely approximated groups
are the first and second, those of Mexico and Central Amer-
ica. The four interspaces between the five groups are sever-
ally 300, 628, 960, and 540 miles. The great distance of
the southernmost volcano of Quito from the most northern
of Peru is, at the first glance, the more remarkable, because,
according to old custom, we usually term the measurement
of degrees upon the highland of Quito the Peruvian measure-
ment. Only a small southern portion of the Peruvian chain
of the Andes is volcanic. The number of volcanoes, accord-
ing to the lists which I have prepared after a careful criti-
cism of the newest materials, is as follows :

Names of the five Groups of Linear Vol-
canoes of the New Continent, from
l'J° 25' North, to 46° 8' South
Latitude.

Number of Vol-
canoes included
in each Group.

Number of Vol-
canoes which an
to be regarded
as still ipnited.

6

29
18
U
24

4
18
10

3
13

Group of Central Americaf

Group of New Granada and QuitoJ....
Group of Peru and Bolivia§

* The group of volcanoes of Mexico includes the volcanoes of Ori-
zaba,* Popocatepetl,* Toluca (or Cerro de San Miguel de Tutucuitla-
pilco), Jorullo,* Colima,* and Tuxtla.* Here, as in similar lists, the
still active volcanoes are indicated bj asterisks.

t The series of volcanoes of Central America is enumerated in the
notes on pages 257 and 268.

X The group of New Granada and Quito includes the Paramo y
Volcan de Euiz,* the volcanoes of Tolima, Purace,* and Sotara, near
Popayan ; the Volcan del Rio Fragua, an affluent of the Caqueta ; the
volcanoes of Pasto, El Azufral,* Cumbal,* Tuquerres,* Chiles, Imba-
buru, Cotocachi, Rucu-Pichincha, Antisana(?), Cotopaxi,* Tungura-
hua,* Capac-Urcu, or Altar de los Collanes(?), and Sangay.*

§ The group of Southern Peru and Bolivia includes from north to
south the following 14 volcanoes :

Volcano of Chacani (also called Charcani, according to Curzon and
Meyen), belonging to the group of Arequipa, and visible from the
town ; it is situated on the right bank of the Rio Quilca, in lat.
■ 16° 11', according to Pentland, the most accurate geological ob-
server of this region, 32 miles to the south of the Nevado de Chu-
quibamba, which is estimated at more than 19,000 feet in height.
Manuscript records in my possession give the volcano of Chacani a
height of fully 19,601 feet. Curzon saw a large crater in the
southeastern part of the summit.
Volcano of Arequipa* lat. 16° 20', 12 miles to the northeast of the
town. With regard to its height (18,879 feet ?), see p. 240. Thad-
daus Hanke, the botanist of the expedition of Malaspina (1796),
Samuel Curzon from the United States of North America (1811),

TRUE VOLCANOES. 271

According to these data the total number of volcanoes in
the five American groups is 91, of which 5Q belong to the

and Dr. Weddel (184:7), have ascended the summit. In August,
1831, Meyen saw large columns of smoke rising ; a year previous-
ly the volcano had thrown out scorice, but never lava streams
(Meyen's Beise urn die Erde^ th. ii., s. 33).

Volcan de Ornato, lat. 16° 50'; it had a violent eruption in the year
1667.

Volcan de Uvillas or Uvinas^ to the south of Apo ; its last eruptions
were in the 16th century.

Volca'i de Pichu-Pichu, 16 miles to the east of the town of Arequipa
(laf . 16° 25'), not far from the Pass of Cangallo, 9673 feet above
the sea.

Volcan Viejo, lat. 16° 55', an enormous crater, with lava, streams and
much pumice-stone.

The six volcanoes just mentioned constitute the group of Arequipa.

Volcan de Tacora or C'kipicani, according to Pentland's fine map of
the lake of Titicaca, lat. 17° 45', height 19,738 feet.

Volcan de Sahama* 22,354 feet in height, lat. 18° 7'; a truncated
cone of the most regular form** see p. 241. The volcano of Sa-
hama is (according to Pentland) 927 feet higher than the Chim-
borazo, but 6650 feet lower than Mount Everest, in the Himalaya,
which is now regarded as the highest peak of Asia. According
to the last official report of Colonel Waugh, of the 1st March, ] 856,
the four highest mountains of the Himalayan chain are ; Mount
Everest (Gaurischanka), to the northeast of Katmandu, 29,000 feet;
the Kuntschinjinga, to the north of Darjiling, 28,154 feet; the
Dhaulaqiri (Dhavalagirir), 26,825 feet; and Tschumalari (Cham-
alari), 23,946 feet.

Volcano of Pomorape, 21,699 feet, lat. 18° 8', almost a twin mount-
ain with the following volcano.

Volcano of Parinacota, 22,029 feet, lat. 18° 12'.

The group of the four trachytic cones Sahama, Pomarape, Parina-
cota, and Gualatieri, lying between the parallels of 18° 7' and 18°
25', is, according to Pentland's trigonometric measurement, higher
than Chimborazo, or more than 21,422 feet.

Volcano of Gmlatien* 21,962 feet, lat. 18° 25', in the BoHviau
province Carangas ; very active, according to Pentland {Hertha^
bd. xiii., 1829, s. 21).

Not far from the Sahama gronp, 18° 7' to 18° 25', the series of vol-
canoes and the entire chain of the Andes, which lies to the westward
of it, suddenly change their strike, and pass from the direction S.E.
— N.W. into that from north to south, which becomes general as far
as the Straits of Magellan. I have treated of this important turning-
point, the notch in the sliore near Arica (18° 28'), which has an an-
alogue on the west coast of Africa, in the Gulf of Biafra, in the first
volume of Cosmos, p. 292.

. Volcano of Ishiga, lat. 19° 20', in the province of Tarapaca, to the
west of Carangas.

272 COSMOS.

continent of South America. I reckon as volcanoes, besides
those which are still burning and active, those volcanic form-

Volcan de San Pedro de Atacama, on the northeastern border of the
Desierto of the same name, in lat. 22° 16', according to the new
plan of the arid sandy desert (Besierto) of Atacama, by Dr. Phi-
lippi, 16 miles to the northeast of the small town of San Pedro,
not far from the gi*eat Nevado de Chorolque.

There is no volcano from 20|° to 30°, and, after an interruption of
more than 568 miles, the volcanic activity first reappears in the vol-
cano of Coquimbo ; for the existence of a volcano of Copiapo (lat. 27°
28) is denied by Meyen, while it is asserted by Philippi, who is well
acquainted with the country.

II Our geographical and geological knowledge of the group of vol-
canoes which we include in the common name of the linear volca-
noes of Chili, is indebted for the first incitement to its completion,
and even for the completion itself, to the acute investigations of Cap-
tain Fitzroy in the memorable expedition of the ships Adventure and
Beagle, and to the ingenious and more detailed labors of Charles
Darwin. The latter, with his peculiar generalizing view, has grasped
the connection of the phenomena of earthquakes and eruptions of
volcanoes under one point of view. The great natural phenomenon
which destroyed the town of Copitipo on the 22d of November, 1822,
was accompanied by the upheaval of a considerable tract of country
on the coast ; and during the exactly -similar phenomenon of the 20th
February, 1835, which did so much inj-rv to the city of Concepcion,
a submarine volcano broke out, with fiery eruptions, near the shore
of the island of Chiloe, near Bacalao Head, and raged for a day and
a half. All this, depending upon similar conditions, has also occurred
formerly, and strengthens the belief that the series of rocky islands
which lies opposite to the Fjords of the main land, to the south of
Valdivia, and of the Fuerte Maullin, and includes Chiloe, the Arch-
ipelago of Chonos and Huaytecas, the Peninsula de tres Montes, and
the Islas de la Campana, De la Madre de Dios, De Santa Lucia and
Los Lobos, from 39° 53' to the entrance of the Straits of Magellan, is
the crest of a submerged western Cordillera projecting above the sea.
It is true that no open trachytic cone, no volcano, belongs to these
frdctis ex cequoj-e terris ; but individual submarine eruptions, some-
times followed and sometimes preceded by mighty earthquakes, ap-
pear to indicate the existence of this western fissure (Darwin, On the
Connection of Volcanic Phenomena, the Formation of Mountain Chains,
and the Effect of the same Powers, by which Continents are elevated: in
the Trans. Geol. Society, 2d series, vol. v., pt. 3, 1840, p. 606-615, and
629-631 ; Humboldt, Essai Politique sur la Nouvelle Espagne, t. i., p.
190, and t. ii., p. 287).

The series of twenty-four volcanoes included in the group of Chili
is as follows, counting from north to south, from the parallel of Co-
quimbo to 46° S. lat. :

(a.) Between the parallels of Coquimbo and Valparaiso :

Volcan de Coquimbo (lat. 30° 5'). Meyen, th. i., s. 385.

Volcano of Limari.

Volcano of Chuapri.

TRUE VOLCANOES. 273

atibns whose old eruptions belong to historic periods, or of
which the structure and eruptive masses (craters of elevation

Volcano of Aconcagua,* W.N.W. of Mendoza, lat. 32° 39' ; alti-
tude 23,004 feet, according to Kellet (see p. 241, note) ; but, ac-
cording to the most recent trigonometric measurement of the
engineer Amado Pissis (1854), only 21,301 feet; consequently,
rather lower than the Saharaa, which Fentland now assumes to
be 22,350 feet (Gilliss, United States Naval Astron. Exped. to Chilis
vol. i., p. 13). The geodetic basis of measurement of Aconca-
gua at 6797 metres, which required eight triangles, has been de-
veloped by M. Pissis, in the Anales de la Universidad de Chile^
1852, p. 219.

The peak of Tupungato is stated by Gilliss to be 22,450 English, or
21,063 Paris, feet in height, and in lat. 33° 22' ; but in the map
of the province of Santiago, by Pissis (Gilliss, p. 45), it is esti-
mated at 22,016 EngHsh, or 20,655 Paris, feet. The latter num-
ber is retained (as 6710 metres) by Pissis in the Anales de Chile^
1850, p. 12.

(6.) Between the parallels of Valparaiso and Concepcion :

Volcano of Maypu* according to Gilliss (vol. i., p. 13), in lat. 34°
17' (but in his general map of Chili, BS° 47', certainly errone-
ously), and 17,662 feet in height. Ascended by Meyen. The
trachytic rock of the summit has broken through upper Jurassic
strata, in which Leopold von Buch detected Exogyra Cotdoni,
Trigonia costata, and Ammonites biplex, from elevations of 9600
feet (Description Physique des Iks Canaries, 1836, p. 471). No
lava streams, but eruptions of flame and scorije from the crater.

Volcano of Peteroa,* to the east of Talca, in lat. 34° 53' ; a volca-
no which is frequently in activity, and which, according to Moli-
na's description, had a great eruption on the 3d December, 1762.
It was visited in 1831 by the highly-gifted naturalist. Gay.

Volcan de Chilian, lat 36° 2' ; a region which has been described by
the missionary Havestadt, of Miinster. In its vicinity is situated
the Nevado Descabezado (35° 1), which was ascended by Do-
meyko, and which Molina declared (erroneously) to be the high-
est mountain of Chili, ^ts height has been estimated by Gilliss
at 13,100 feet {United States Naval Astr. Exped., 1855, vol. i.,
p. 16 and 371).

Volcano of Tucapel, to the west of the city of Concepcion ; also
called Silla Veluda: perhaps an unopened trachytic mountain,
which is in connection with the active volcano of Antuco.

(c.) Between the parallels of Concepcion and Valdivia:

Volcano of Antuco,* lat. 37° 7' ; geognostically described in detail
by Poppig; a basaltic crater of elevation, from the interior of
which a trachytic cone ascends, with lava streams, which break
out at the foot of the cone, and more rarely from the crater at
the summit (Poppig, Reise in Chile and Peru, bd. i., s. 364). One
of these streams was still flowing in the year 1828. The inde-
fatigable Domeyko found the volcano in full activity in 1845, and
its height only 8920 feet (Pentland, in Mary SomervUle's Phys-
ical Geography, vol. i., p. 186). Gilhss states the height at 9242

]\r 2

274 COSMOS.

and eruption, lavas, scoriae, pumice-stones, and obsidians)
characterize them, without reference to any tradition, as
volcanoes which have long been extinct. Unopened tra-
chytic cones and domes, or unopened long trachytic ridges,
such as Chimborazo and Iztaccihuatl, are excluded. This
is also the sense given to the word volcano by Leopold von
Buch, Charles Darwin, and Friedrich Naumaun, in their
geographical narratives. I give the name of still active
volcanoes to those which, when seen from their immediate
vicinity, still exhibit signs of greater or less degrees of their
activity, and some which have also presented great and well-
attested eruptions in recent times. The qualification "seen
from their immediate vicinity" is of great importance, as
the present existence of activity is denied to many volcanoes,

feet, and mentions new eruptions in the year 1853. According
to intelligence communicated to me by the distinguished Ameri-
can astronomer, Gilliss, a new volcano rose out of the depths in
the interior of the Cordillera, between Antuco and the Descabe-
zado, ou the 25th of November, 1847, forming a hill* of 320 feet.
- The sulphureous and fiery eruptions were seen for more than a
3'ear by Domeyko. Far to the eastward of the volcano of An-
tuco, in a parallel chain of the Andes, Poppig states that there
are two other active volcanoes — Punhamuidda* andUnalavquen*.

Volcano of Callaqui.

Volcan de VUlarica* lat. 39° 14'.
. Volcano of Chinal, lat. 39° 35'.

Volcan de Panguipulli* lat. 40|, according to Major Philippi.

(c?.) Between the parallels of Valdivia and the southernmost Cape of
the Island of Chiloe :

Volcano of Ranco,

Volcano of Osorno or Llanquthue, lat^l° 9', height 7443 feet.

Volcan de Calbnco* lat. 41° 12'. ^

Volcano of Guanahuca (Guanegue ?).

Volcano of Minchinmadom, lat. 42° 48', height 7993 feet.

Volcan del Corcovado* lat. 43° 12', height 7509 feet.

Volcano of Yanteles (Yntales), lat. 43° 29', height 8030 feet.

Upon the last four volcanoes, see Captain Fitzroy, Erped. of the
Beagle, vol. iii., p. 275, and GiUiss, vol. i., p. 13.

Volcano of San Clemente, opposite to the Peninsula de Tres Montes,
which consists, according to Darwin, of granite, lat. 46° 8'. On
the gi-eat map of South America, by La Cruz, a more southern
volcano, De los Gigantes, is given, opposite the Archipelago de la
Madre de Dios, in lat. 51° 4'. Its existence is very doubtful.

The latitudes in the foregoing table of volcanoes are for the most
part derived from the maps of Pissis, Allan Campbell, and Claud©
Gay, in the admirable work of Gilliss (1855).

TRUE VOLCANOES. S75 '

because, when observed from the plain, the thin vapors, which
ascend from the crater at a great height, remam invisible to
the eye. Thus it was even denied, at the time of my Amer-
ican travels, that Pichincha and the great volcano of Mexico
(Popocatepetl) were still active, although an enterprising
traveler, Sebastian Wisse,* counted 70 still burning orifices
(fumaroles) around the great active cone of eruption in the
crater of Pichincha ; and I was myself a witness,! at the
foot of the volcano in the Malpais del Llano de Tetimpa, in
which I had to measure a base-line, of an extremely distinct
eruption of ashes from Popocatepetl.

In the series of volcanoes of New Granada and Quito,
which in 18 volcanoes includes 10 that are still active, and
is about twice the length of the Pyrenees, we may indicate,
from north to south, as four smaller groups or subdivisions :
the Paramo de Ruiz and the neighboring volcano of Tolima
(latitude, according to Acosta, 4° 55^ N.) ; Purace and Sota-
ra, near Popayan (lat. 2i°) ; the Volcanes de Pasto, Tuquerres
and Cumbal (lat. 2° 20^ to 0° 50') ; and the series of volca-
noes from Pichincha, near Quito, to the unintermittently act-
ive Sangay (from the equator to 2° S. lat.). This last sub-
division of the active group is not particularly remarkable
among the volcanoes of the New World, either by its great
length or by the closeness of its arrangement. We now
know, also, that it does not include the highest summit ; for
the Aconcagua in Chili (lat. 32° 390 of 23,003 feet, accord-
ing to Kellet, 23,909 feet, according to Fitzroy and Pent-
land, besides the Nevados of Sahama (22,349 feet), Parincota
(22,030 feet), Gualateiri (21,962 feet), and Pomarape (21,699
feet), all from between 18° 7^ and 18° 25^ south latitude,
are regarded as higher than Chimborazo (21,422 feet). Nev-
ertheless, of all the volcanoes of the New Continent, the
volcanoes of Quito enjoy the most widely-spread renown, for
to these mountains of the chain of the Andes, to this high
land of Quito, attaches the memory of those assiduous astro-
nomical, geodetical, optical, and barometrical labors, directed
to important ends, which are associated with the illustrious
names of Bouguer and La Condamine. Wherever intellectu-
al tendencies prevail, wherever a rich harvest of ideas has
been excited, leading to the advancement of several sciences
at the same time, fame remains, as it were, locally attached

* Humboldt, Kleinere Schri/ten, bd. i., s. 90.

t 24th of January, 1804. See my Essai Politique sur h Nouvelk
Espagne, t. i., p. 1G6.

*276 COSMOS.

for a long time. Such fame has in like manner belonged to
Mount Blanc, m the Swiss Alps — not on account of its height,
which only exceeds that of Monte Rosa by about 557 feet;
not on account of the danger overcome in its ascent — but on
account of the value and multiplicity of the physical and geo-
logical views which ennoble Saussure's name, and the scene
of his untiring industry. Nature appears greatest where, be-
sides its impression on the senses, it is also reflected in the
depths of thought.

The series of volcanoes of Peru and Bolivia, still entirely
belonging to the equinoctial zone, and, according to Pentland,
only covered with perpetual snow at an elevation of 16,945
feet (Darwin, Journal, 1845, p. 244), attains the maximum
of its elevation (22,349 feet) at about the middle of its length
in the Sahama group, between 18° 7^ and 18° 25^ south lati-
tude. There, in the neighborhood of Arica, appears a sin-
gular, bay-like bend of the shore, which corresponds with a
sudden alteration in the axial direction of the chain of the
Andes, and of the series of volcanoes lying to the west of it.
Thence, toward the south, the coast-line, and also the vol-
canic fissure, no longer strike from southeast to northwest,
but in the direction of the meridian, a direction which is
maintained until near the western entrance into the Straits
of Magellan, for a distance of more than two thousand miles.
A glance at the map of the ramifications and groups of mount-
ains of the chain of the Andes, published by me in the year
1831, exhibits many other similar agreements between the
outline of t\\e New Continent and the near or distant Cor-
dilleras. Thus, between the promontories of Aguja and San
Lorenzo (5^° to 1° S. lat.), both the coast-line of the Pacific
and the Cordilleras are directed from south to north, after
being directed so long from southeast to northwest, between
the parallels of Arica and Caxamarca; and in the same way
the coast-line and the Cordilleras run from southwest to
northeast, from the mountain group of Imbaburu, near Quito,
to that of Los Robles,* near Popayan. With regard to the geo-

* The niicha-schist mountain group de Los Robles (lat. 2 ° 2') and of
the Paramo de las Papas (lat. 2° 20') contains the Alpine lakes, La-
guna de S. lago and L. del Buey, scarcely six miles apart ; from the
former springs the Cauca, and from the latter the Magdalena, which,
being soon separated by a central mountain chain, only unite with
each other in the parallel of 9° 27', in the plains of Mompox andTen-
erife. The above-mentioned mountain group, between Popayan, Al-
maguer, and Timana, is of great importance in connection with the
geological question whether the volcanic chain of the Andes of Chili,

TRUE VOLCANOES. ' 277

logical causal connection of the agreement, which is so often
manifested between the outlines of continents and the direc-

Peru, Bolivia, Quito, and New Granada be connected with the mount-
ain chain of the Isthmus of Panama, and in this way with that of
Veraffua and the series of volcanoes of Costa Rica and Central Amer-
ica in general. In my maps of 1816, 1827, and 1831, the mountain
systems of which have been made more generally known by Brue in
Joaquin Acosta's fine map of New Granada (1847) and in other maps,
I have shown how the chain of the Andes undergoes a triple division
under the northern parallel of 2° 10' ; the western Cordillera running
between the valley of the Rio Cauca and the Rio Atrato ; the middle
one between the Cauca and the Rio Magdalena ; and the eastern one
between the valley of the Magdalena and the Llanos (plains), which
are watei-ed by the affluents of the Maranon and Orinoco. I have
been able to indicate the special direction of these three Cordilleras
from a great number of points which fall in the series of astronomical
local determinations, of which I obtained 152 in South America alone
by culmination of stars.

To the east of the Rio Dagua, and to the west of Cazeres, Rolda-
nilla, Toro, and Anserma, near Cartago, the western Cordillera runs
S.S.W. — N.N.E., as far as the Salto de San Antonio, in the Rio Cauca
(lat. 5° 14'), which lies to the southwest of the Vega de Supia. Thence
as far as the Alto del Viento (Cordillera de Abibe, or Avidi, lat. 7° 12'),
9600 feet in height, the chain increases considerably in elevation and
bulk, and amalgamates, in the province of Antioquia, with the inter-
mediate or Central Cordillera. Farther to the north, toward the
sources of the Rios Lucio and Guacuba, the chain ceases, dividing into
ranges of hills. The Cordillera occidental, which is scarcely 32 miles
from the coast of the Pacific, near the mouth of the Dagua, in the
Bahia de San Buenaventura (lat. 3° 50'), is twice this distance in the
parallel of Quibdo, in the Choco (lat. 5° 48'). This observation is of
some importance, because we must not confound with the western
chain of the Andes the country with high hills, and the range of hills,
which in this province, so rich in gold dust, runs from south to north,
from Novita and Tado, along the right bank of the Rio San Juan and
the left bank of the great Rio Atrato. It is this inconsiderable series
of hills that is intersected in the Quebrada de la Raspadura by the
canal of Raspadura {Canal des Monches), which unites two rivers (the
Rio San Juan or Noanama and the Rio Quibdo, a tributary of the
Atrato), and by their means two oceans (Humboldt, Essai Politique, t. i.,
p. 235) ; it was this, also, which was seen in the instructive expedition
of Captain Kellet between the Bahia de Cupica (lat. 6° 42'), long
and fruitlessly extolled by me, and the sources of the Napipi, which
falls into the Atrato. (See Humboldt, Op. cit., t. i., p. 231 : and Rob-
ert Fitzroy, Considerations on the Great Isthmus of Central America in
the Journal of the Royal Geogr. Soc, vol. xx., 1851, p. 178, 180, and
186.)

The middle chain of the Andes (Cordillera Central), constantly the
highest, reaching within the limit of perpetual snow, and, in its entire
extent, directed nearly from south to north, like the western chain,
commences about 35 miles to the northeast of Popayan with the Par-
amos of Guanacos, Huila, Iraca, and Chinche. Farther on toward
the north between Buga and Chaparral, rise the elongated ridge of the

278 COSMOS,

tion of near raountain chains (South America, Alleghanys,
Norway, Apennines), it appears difficult to come to any de-
cision.

Neveda de Baraguan (lat. 4° 11'), La Montana de Quindio, the snow-
capped, truncated cone of Tolima, the Volcano and Paramo de^Ruiz,
and the Mesa de Herveo. These high and rugged mountain deserts, to
which the name of Paramos is applied in Spanish, are distinguished
by their temperature and a peculiar character of vegetation, and rise
in the part of the tropical region which I here describe, accoi-ding to
the mean of many of my measurements, from 10,000 to 11,700 feet
above the level of the sea. In the parallel of Mariquita, of the Herveo
and the Salto de San Antonio, in the valley of the Cauca, there com-
mences a union of the western and central chains, of which mention
has already been made. This amalgamation becomes most remarkable
between the above-mentioned Salto and the Angostura and Cascada
de Caramanta, near Supia, Here is situated the high land of the prov-
ince of Antioquia, so difficult of access, which extends, according to
Manuel Restrepo, from 5i° to 8° 31'; in this we may mention, as
points of elevation from south to north, Anna, Sonson, to the north of
the sources of the Rio Samana, Marinilla, Rio Negro (6844: feet), and
Medellin (4847 feet), the plateau of Santa Rosa (8466 feet), and Valle
de Osos. Farther on, between Cazeres and Zaragoza, toward the con-
fluence of the Cauca and Nechi, the true mountain chain disappears,
and the eastern slope of the Cerros de San Lucar, which I saw from
Badillas (lat. 8° 1') and Paturia (lat. 7° 36')> during my navigation
and survey of the Magdalena, is only perceptible from its contrast
with the broad river plain.

The eastern Cordillera possesses a geological interest, inasmuch as it
not only separates the whole northern mountain system of New Gran-
ada from the low land, from which the waters flow partly by the Ca-
guan and Caqueta to the Amazons, and partly by the Guaviare, Meta,
and Apure to the Orinoco, but also unites itself most distinctly with
the littoral chain of Caraccas. What is called in systems of veins a
ralcing takes place there — a union of mountain chains which have been
elevated upon two fissures of very different directions, and probably
even at very different times. The eastern Cordillera departs far more
than the two others, from a meridional direction, diverging toward
the northeast, so that at the snowy mountains of Merida (lat. 8° 10') it
already lies five degrees of longitude farther to the east than at its issue
from the mountain group de Los Robles, near the Ceja and Timana.
To the north of the Paramo de la Suma Paz, to the east of the Purifi-
cacion, on the western declivity of the Paramo of Chingaza, at an alti-
tude of only 8760 feet, rises, over an oak forest, the fine, but treeless and
stern plateau of Bogota (lat. 4° 36'). It occupies about 288 geograph-
ical square miles, and its position presents a remarka-ble similarity to
that of the basin of Cashmere, which, however, according to Victor
Jacquemont, is about 3410 feet lower at the Wuller Lake, and belongs
to the southwestern declivity of the Hymalayan chain. The plateau
of Bogota and the Paramo de Chingaza are followed in the eastern
Cordillera of the Andes, toward the northeast, by the Paramos of
Guachaneque, above Tunja ; of Zoraca, above Sogamoso ; of Chita
(16,000 feet?), near the sources of the Rio Casanare, a tributary of
the Meta; of the Almorzadera (12,854 feet), near Socorro; of Cacota

TRUE VOLCANOES. 279

Although, in the series of volcanoes of Bolivia and Chili,
the western branch of the chain of the Andes, which approach-
es nearest to the Pacific, at present exhibits the greater part
of the traces of still existing volcanic activity, yet a very ex-
perienced observer, Pentland, has discovered at the foot of
the eastern chain, more than 180 geographical miles from the
sea-coast, a perfectly preserved but extinct crater, with un-
mistakable lava streams. This is situated upon the summit
of a conical mountain, near San Pedro de Cacha, in the val-
ley of Yucay, at an elevation of nearly 12,000 feet (lat. 14° 8^,
long. 71° 20'), southeast from Cuzco, where the eastern snowy
chain of Apolobamba, Carabaya, and Yilcanoto extends from
southeast to northwest. This remarkable point* is marked
by the ruins of a famous temple of the Inca Viracocha. The
distance from the sea of this old lava-producing volcano is

(10,986 feet), near Pamplona ; of Laura and Porquera, near La Grita.
Here, between Pamplona, Salazar, and Rosario (between lat. 7° 8' and
7° 50'), is situated the small mountain group, from which a crest ex-
tends from south to north toward Ocana and Valle de Upar to the
west of the Laguna de Maracaibo, and unites with the most advanced
mountains of the Sierra Nevada de Santa Marta (19,000 feet?). The
more elevated and vaster crest continues in the original northeasterly
direction toward Merida, Truxillo, and Barquisemeto, to unite there,
to the east^vard of 4;he Laguna de Maracaibo, with the granitic littoral
chain of Venezuela, to the west of Puerto Cabello. From the Grita
and the Paramo de Porquera the eastern Cordillera rises again at once
to an extraordinary height. Between the parallels of 8° 5' and 9° 7',
follow the Sierra Nevada de Merida (Mucuchies), examined by Bous-
singault, and determined by Codazzi trigonometrically at 15,069 feet;
and the four Pai'amos, De Timotes, Niquitao, Bocono, and de Las
Posas, full of the most beautiful Alpine plants. (See Codazzi, i?es2^-
men de la Geografia de Venezuela, 1841, p. 12 and 495; and also my
Asie Centrale, t. iii., p. 258-262, with regard to the elevation of the
perpetual snow in this zone.) The western Cordillera is entirely want-
ing in volcanic activity, which is peculiar to the central Cordillera
as far as the Tolima and Paramo de Ruiz, which however are sep-
arated fi'om the volcano of Purace by nearly three degrees of latitude.
The eastern Cordillera has a smoking hill near its eastern declivity,
at the origin of the Rio Fragna, to the northeast of Mocoa and south-
east of Timana, at a greater distance from the shore of the Pacific
than any other still active volcano of the New World. An accurate
knowledge of the local relations of the volcanoes to the arrangement
of the mountain chains is of the highest importance for the completion
of the geology of volcanoes. All the older maps, with the single ex-
ception of that of the high land of Quito, can only lead to error.

* Pentland, in Mrs. Somerville's P/rysicaZ Geography (1851), vol. i.,
p. 185. The Peak of Vilcanoto (17,020 feet), situated in lat. 14° 28',
forming a portion of the vast mountain group of that name, closes the
northern extremity of the plateau, in which the lake of Titicaca, a
small inland sea of 88 miles in length, is situated.

280 COSMOS.

far greater than that of Sangay, which also belongs to an
eastern Cordillera, and greater than that of Orizaba and
Jorullo.

An interval of 540 miles destitute of volcanoes separates
the series of volcanoes of Peru and Bolivia from that of Chili.
This is the distance of the eruption in the desert of Atacaraa
from the volcano of Coquimbo. At 2° 34^ farther to the
south, as already remarked, the group of volcanoes of Chili
attains its greatest elevation in the volcano of Aconcagua
(23,003 feet), which, according to our present knowledge, is
also the maximum of all the summits of the New Continent.
The average height of the Sahama group is 22,008 feet ; con-
sequently 58G feet higher than Chimborazo. Then follow,
diminishing rapidly in elevation, Cotopaxi, Arequipa(?), and
Tohma, between 18,877 and 18,129 feet in height. I give,
in apparently very exact numbers, and without alteration,
the results of measurements which are unfortunately com-
pounded from barometrical and trigonometrical determina-
tions, because in this way the greatest inducement will be
given to the repetition of the measurements and correction
of the results. In the series of volcanoes of Chili, of which
I have cited twenty-four, it is unfortunately for the most part
only the southern and lower ones, from Antuco to Yantales,
between the parallels of 37° 20^ and 43° 40^, that have been
hypsometrically determinied. These have the inconsiderable
elevation of from six to eight thousand feet. Even in Tierra
del Fuego itself the summit of the Sarmiento, covered with
perpetual snow, only rises according to Fitzroy, to 6821 feet.
From the volcano of Coquimbo to that of San Clemente the
distance is 968 miles.

With regard to the activity of the volcanoes of Chili, we
have the important testimony of Charles Darwin,* who re-
fers very decidedly to Osorno, Corcovado, and Aconcagua as
being ignited ; the evidence of Meyen, Poppig, and Gay, who
ascended Maipu, Antuco, and Peteroa ; and that of Domeyko,
the astronomer Gilliss, and Major Philippi. The number of
active craters may be fixed at thirteen, only five fewer than
in the group of Central America.

From the five groups of serial volcanoes of the New Con-
tinent, which we have been able to describe from astro-
nomical local determinations, and for the most part also hyp-
sometrically as to position and elevation, let us now turn to

* See Darwin, Journal of Researches in Natural History and Geology
during the Voyage of the Beagle, 1845, p. 275, 291, and 310.

TRUE VOLCANOES. 28]

the Old Continent, in which, in complete opposition to thq
New World, the greater part of the approximated volcanoes
belong not to the main land but to the islands. Most of the
European volcanoes are situated in the Mediterranean Sea,
and, indeed (if we include the great and repeatedly active
crater between Thera, Therasia, and Aspronisi), in the Tyr-
rhenian and JEgaean parts ; in Asia the most mighty volca-
noes are situated to the south and east of the continent, on
the large and small Sunda Islands, the Moluccas, and the
Philippines, in Japan, and the Archipelagoes of the Kurile
and Aleutian Islands.

In no othf^aregion of the earth's surface do such frequent
and such fr^'^ traces of the active communication between
the interior and exterior of our planet show themselves as
upon the narrowspace of scarcely 12,800 geographical (16,928
English) square miles between the parallels of 10° south and
14° north latitude, and between the meridians of the south-
ern point of Malacca and the western point of the Papuan
peninsula of New Guinea. The area of this volcanic island-
world scarcely equals that of Switzerland, and is washed by
the seas of Sunda, Banda, Solo, and Mindoro. The single
island of Java contains a greater number of active volcanoes
than the entire southern half of America, although this is^
and is only 544 miles in length, that is, only one seventh of
the length of South America. A new but fong-expected
light has recently been diffused over the geognostic nature of
Java (after previous very imperfect but meritorious works by
Horsfield, Sir Thomas Stamford Raffles, and Reinwardt), by
a learned, bold, and untiringly-active naturalist, Franz Jung-
huhn. After a residence of more than twelve years, he has
given the entire natural history of the country in an instruct-
ive work — Java, its Form, vegetable Covering, and internal
Structure. More than 400 elevations are carefully determ-
ined barometrically; the volcanic cones and bell-shaped
mountains, forty-five in number, are represented in profile,
and all but three* of them were ascended by Junghuhn.
More than half (at least twenty-eight) were found to be still
burning and active ; their remarkable and various profiles are
described with extraordinary clearness, and even the attain-
able history of their eruptions is investigated. No less im-
portant than the volcanic phenomena of Java are its sedi-
mentary formations of the tertiary period, which were en-
tirely unknown to us before the appearance of the complete
* Junghuhn, Java, bd. i., s. 79.

282 COSMOS.

work just mentioned, although they cover three fifths of the
entire area of the island, especially in the southern parts.
In many districts of Java there occur, as the remains of
former widely-spread forests, fragments, from three to seven
feet in length, of silicified trunks of trees, which all belong to
the Dicotyledons. For a countiy in which at present an
abundance of palms and tree ferns grows, this is the more re-
markable, because in the Miocene tertiary rocks of the brown-
coal formation of Europe, where arborescent monocotyledons
no longer thrive, fossil palms are not unfrequently met with.*
By a diligent collection of the impressions of leaves and fos-
silized woods, Junghuhn has been enabled tof'^e us, as the
first example of the fossil flora of a purely tropical region,
the ancient flora of Java, ingeniously elaborated by Goppert
from his collection.

As regards the elevation to which they attain, the volca-
noes of Java are far inferior to those of the three groups of
Chili, Bolivia, and Peru, and even to those of the two groups
of Quito with New Granada, and of Tropical Mexico. The
maxima attained by these American groups are : For Chili,
Bohvia, and Quito, 21,000 to 23,000 feet, and for Mexico,
18,000 feet. This is nearly ten thousand feet (about the
Leight ofiEtna) more than the greatest elevation of the vol-
canoes of Sumatra and Java. On the latter island the highest
still burning colossus is the Gunung Semeru, the culminating
point of the entire Javanese series of volcanoes. Junghuhn
ascended this in September, 1844; the average of his baro-
metric measurements gave 12,233 feet above the surface of
the sea, and consequently 1748 feet more than the summit
of -3^tna. At night the centigrade thermometer fell below
6°. 2 (43°. 2 Fahr.). The old Sanscrit name of Gunung Se-
meru was Mahd-Meru (the Great Meru) ; a reminiscence of
the time when the Malays received Indian civilization — a
reminiscence of the Mountain of the World in the north,
which, according to the Mahabharata, is the dwelling-place
of Brahma, Vishnu, and the seven Devarschi.f It is re-

* Op. ciL, bd. iii., s. lo/i ind Goppert, Die Tertidrflora avfderlnsel
Java nach den Entdeckvngea von Fr. Junghuhn (1854), s. 17. The ab-
sence of monocotyledons is, however, peculiar to the silicified trunks
of trees lying scattered upon the surface, and especially in the rivulets
of the district of Bantam ; in the subterranean carbonaceous strata, on
the contrary, there are remains of palm-wood, belonging to two genera
(Flabellariaimd Amesoneuron). See Goppert, s. 31 and 35.

fUpon the signification of the word Mcru, and the conjectures
which Burnouf communicated to me regarding its connection with

true" volcanoes. 283

markable that, as the natives of the plateau of Quito had
guessed, before my measurement, that Chimborazo surpassed
^all the other snowy mountains in the country, the Javanese
also knew that the Holy Mountain, Maha-Meru, which is but
at a short distance from the Gunung-Ardjuno (11,031 feet),
exhibited the maximum of elevation upon the island, and yet,
in this case, in a country free from snow, the greater dis-
tance of the summit from the level of the lower limit of per-
petual snow could no more serve as a guide to the judgment
than the height of an occasional temporary fall of snow.*

The elevation of the Gunung Semeru, which exceeds
11,000 (11,726 English) feet, is most closely approached
by four other mountains, which were found hypsometrically
to be between ten and eleven thousand feet. These are:
GunungI Slamat, or mountain of Tegal (11,116 feet), Gu-
nung Ardjuno (11,031 feet), Gunung Sumbing (11,029 feet),
and Gunung Lawu (10,726 feet). Seven other volcanoes
of Java attain a height of nine or ten thousand feet ; a re-
sult which is of the more importance as no summit of the
island was formerly supposed to rise higher than six thou-
sand feet.$ Of the five groups of North and South Ameri-

mira (a Sanscrit word for sea), see my Asie Ccntrale, t. i., p. 114-116;
and Lassen's Indische Alterthuviskunde, bd. i., s. 847, The latter is
inclined to regard the names as not of Sanscrit origin.

* See page 229.

t Gunung is the Javanese word for mountain, in Malayan, gunong^
which, singularly enough, is not farther disseminated over the enor-
mous domain of the Malayan language ; see the comparative table of
words in my brother's work upon the Kawi language, vol. ii., s. 249,
No. 62. As it is the custom to place this word gunung before the
names of mountains in Java, it is usually indicated in the text by a
simple G.

J Leopold de Buch, Description Physique des lies Canaries, 1836, p.
419. Not only has Java (Junghuhn, th, i., s. 61, and th. ii., s. 547)
a colossal mountain, the Semeru of 12,233 feet, which consequently
exceeds the peak of Teneriffe a little in height, but an elevation of
12,256 feet is also attributed to the Peak of Indrapura, in Sumatra,
which is also still active, but does not appear to have been so accu-
rately measured (th. i., s. 78, and profile Map No. 1). The next to
this in Sumatra, are the dome of Telaman, which is only one of the
summits of Ophir (not 13,834, but only 9603 feet in height), and the
Merapi (according to Dr. Horner, 9571 feet), the most active of the
thirteen volcanoes of Sumatra, which, however (th. ii,, s. 294, and
Junghuhn's Battaldnder, 1847, th. i., s. 25), is not to be confounded,
from the similarity of the names, with two volcanoes of Java — the
celebrated Merapi near Jopjakerta (9208 feet), and the Merapi which
forms the eastern portion of the summit of the volcano Idjen (8595
feet). In the Merapi it is thought that the holy name Meru is again
to be detected, combined with the Malayan and Javanese word opi, fire.

284 COSMOS.

can volcanoes, that of Guatemala (Central America) is the
only one exceeded in mean elevation hy the Javanese group.
Although in the vicinity of Old Guatemala the Volcan del
Fuego attains a height of 13,109 feet (according to the cal-
culation and reduction of PoggendorfF), and therefore 874
feet more than Gunung Semeru, the remainder of the Cen-
tral American series of volcanoes only varies between five
and seven thousand feet, and not, as in Java, between seven
and ten thousand feet. The highest volcano of Asia is not,
however, to be sought in the Asiatic Islands (the Archipel-
ago of the Sunda Islands), but upon the continent ; for upon
the peninsula of Kamtschatka the volcano Kljutschewsk
rises to 15,763 feet, or nearly to the height of the Kucu-
Pichincha, in the Cordilleras of Quito.

The principal axis* of the closely-approximated series of
the Javanese volcanoes (more than 45 in number) has a di-
rection W.N.W.— E.S.E. (exactly W. 12° N.), and there-
fore principally parallel to the series of volcanoes of the
eastern part of Sumatra, but not to the longitudinal axis of
the island of Java. This general direction of the chain of
volcanoes by no means excludes the phenomenon to which
attention has very recently been directed in the great chain
of the Himalaya, that three or four individual high summits
are so arranged together that the small axis of these partial
series form an oblique angle with the primary axis of the
chain. This phenomenon of fissure, which has been ob-
served and partially describedf by Hodgson, Joseph Hooker,
and Strachey, is of great interest. The small axes of the
subsidiary fissures meet the great axis, sometimes almost at
a right angle, and even in volcanic chains the actual maxi-
ma of elevation are often situated at some distance from the
major axis. As in most linear volcanoes, no definite pro-
portion is observed in Java between the elevation and the
size of the crater at the summit. The two largest craters
are those of Gunung Tengger and Gunung Raon. The for-
mer of these is a mountain of the third class, only 8704 feet
in height. Its circular crater is, however, more than 21,315
feet, and therefore nearly four geographical miles in diame-
ter. The flat bottom of the crater is a sea of sand, the sur-

* Junghuhn, Java, bd. i., s. 80.

t See Joseph Hooker, Sketch-Map of Sikhim, 1850, and in his
Himalayan Journals, vol. i., 1854, Map of part of Bengal ; and also
Strachey, Map of West-Nan^ in his Physical Geography of Western
TibeU 1853.

TRUE VOLCANOES. " 285

face of w6ich lies 1865 feet below the highest point of the
surrounding wall, and in which scoriaceous lava masses pro-
ject here and there from the layer of pounded rapilli. Even
the enormous crater of Kirauea, in Owhyhee, which is filled
with glowing lava, does not, according to the accurate trig-
onometrical survey of Captain Wilkes, and the excellent
observations of Dana, attain the size of that of Gunung
Tengger. In the middle of the crater of the latter there
rise four small cones of eruption, actual circumvallated fun-
nel-shaped chasms, of which only one, Bromo (the mythical
name Brahma, a word which has the signification of fire in
the Kawi, although not in the Sanscrit), is now not active.
Bromo presents the remarkable phenomenon that from 1838
to 1842 a lake was formed in its funnel, of which Junghuhn
has proved that it owes its origin to the influx of atmos-
pheric waters, which have been heated and acidulated by
the simultaneous penetration of sulphurous vapors.* Next
to Gunung Tengger, Gunung Raon has the largest crater,
but the diameter of this is about one half less. The view
into the interior is awe-inspiring. It appears to extend to
a depth of more than 2398 feet ; and yet the remarkable
volcano, 10,178 feet in height, which Junghuhn has ascend-
ed and so carefully described,! is not even named on the
meritorious map of Raffles.

Like almost all linear volcanoes, the volcanoes of Java
exhibit the important phenomenon that a simultaneity of
great eruptions is observed much more rarely in nearly ap-
proximated cones than in those which are widely separated.
When, in the night of the 11th and 12th of August, 1772,
the volcano Gunung Pepandajan (7034 feet) burst forth, the
most destructive eruption that has taken place upon the
island within historical periods, two other volcanoes, the
Gunung Tjerimai and Gunung Slamat, became ignited on
the same night, although they lie in a straight line at a dis-
tance of 184 and 352 miles from Pepandajan.f Even if the

* Junghuhn, Java, bd. ii., fig. ix., s. 572, 596, and 601-604. From
1829 to 1848 the small crater of eruption of the Bromo had eight fiery
eruptions. The crater-lake, which had disappeared in 1842, had
been again formed in 1848 ; but, according to the obsen'ations of B.
van Herwerden, the presence of the water in the chasm of the cal-
dron hsfll no effect in preventing the eruption of red-hot, widely-scat-
tered scoriae.

t Junghuhn, bd. ii., s. 624-641.

JThe G. Pepandajan was ascended in 1819 by Reinwardt, and in
1837 by Junghuhn. The latter, who has accurately investigated the

286 COSMOS.

volcanoes of a series all stand over one focus, the net of fis-
sures through which they communicate is, nevertheless, cer-
tainly so constituted that the obstruction of old vapor chan-
nels, or the temporary opening of new ones, in the course of
ages, render simultaneous eruption at very distant points
quite conceivable. I may again advert to the sudden dis-
appearance of the column of smoke which ascended from
the volcano of Pasto, when, on the morning of the 4th of
February, 1797, the fearful earthquake of lliobamba con-
vulsed the plateau of Quito between Tunguragua and Goto-
paxi.*

To the volcanoes of the island of Java generally a charac-
ter of rihhed formation is ascribed, to which I have seen noth-
ing similar in the Canary Islands, in Mexico, or in the Cor-
dilleras of Quito. The most recent traveler, to whom we
are indebted for such admirable observations upon the struc-
ture of the volcanoes, the geography of plants, and the psy-
chrometric conditions of moisture, has described the phenome-
non to which I here allude with such decided clearness that
I must not omit to call attention to this regulai-ity of form,
in order to furnish an inducement to new investigations.
"Although," says Junghuhn, "the surface of a volcano
10,974 feet in height, the Gunung Sumbing, when seen from
some distance, appears as an uninterruptedly smooth and
sloping face of the conical mountain, still on a closer exam-
ination, we find that it consists entirely of separate longi-
tudinal ridges or ribs, which gradually subdivide and become
broader as they advance downward. They run from the
summit of the volcano, or more frequently from an elevation
several hundred feet below the summit, down to the foot of
the mountain, diverging like the ribs of an umbrella." These
rib-like longitudinal ridges have sometimes a tortuous course
for a short distance, but are all formed by approximated
clefts of three or four hundred feet in depth, all directed in
the same way, and becoming broader as they descend. They
are furrows of the surface " which occur on the lateral slopes
of all the volcanoes of the island of Java, but differ consider-
ably from each other upon the various conical mountains, in

vicinity of the mountain, consisting of detritus intermingled with nu-
merous angular, erupted blocks of lava, and compared it with the
earliest reports, regards the statement, which has been disselhinated
by so many valuable works, that a portion of the mountain and an
area of sevei'al square miles sank during the eruption of 1772, as
greatly exaggerated (Junghuhn, bd. ii., s. 98 and 100).

* Cosmos, vol. v., p. 183, and Voyage aux Regions Equinox, t. ii., p. 16,

TRUE VOLCANOES. 287

their average depth and the distance of their upper origin
from the margin of the crater or from an unopened summit.
The Gunung Sumbing (11,029 feet) is one of those volcanoes
which exhibit the finest and most regularly formed ribs, as
the mountain is bare of forest trees and clothed with grass."
According to the measurements given by Junghuhn,* the
number of ribs increases by division in proportion as the de-
clivity decreases. Above the zone of 9000 feet there are, on
Gunung Sumbing, only about ten such ribs ; at an elevation
of 8500 feet there are thirty-two ; at 5500 feet, seventy-two ;
and at 3000 feet, more than ninety-five. The angle of in-
clination, at the same time, diminishes from 37° to 25° and
10^°. The ribs are almost equally regular on the volcano
Gunung Tengger (8702 feet), while on the Gunung Ringgit
they have been disturbed and coveredf by the destructive
eruptions which followed the year 1586. "The production
of these peculiar longitudinal ribs and the mountain fissures
lying between them, of which drawings are given, is ascribed
to. erosion by streams."

It is certain that the mass of meteoric water in this tropic-
al region is three or four times greater than in the temperate
zone ; indeed, the showers are often like water-spouts, for al-
though, on the whole, the moisture diminishes with the eleva-
tion of the strata of air, the great mountain cones exert, on
the other hand, a peculiar attraction upon the clouds, ai»d, as
I have already remarked in other places, volcanic eruptions
are in their nature productive of storms. The clefts and
valleys {Barrancos) in the volcanoe^|X)f the Canary Islands,
and in the Cordilleras of South AmOTca, which have become
of importance to the traveler from the frequent descriptions
given by Leopold von BuchJ and myself, because they open
up to him the interior of the mountain, and sometimes even
conduct him up to the vicinity of the highest summits, and
to the circumvallation of a crater of elevation, exhibit analo-
gous phenomena ; but although these also at times carry off
the £MHpmulated meteoric waters, the original formation of
the harrancos% upon the slopes of the volcanoes is probably

* Junghuhn, bd. ii., s. 241-246.

t Op. cit. sup., s. 566, 590 and 607-609.

X Leopold von Buch, Phys. Beschr. der Canarischen Inseln, s. 206,
218, 248, and 289. ,

§ Barranco and Barranca, both of the same meaning, and sufficient-
ly in use in Spanish America, certainly indicate properly a water-fur-
row or water-cleft : la quiebra que hacen en la tierra las corrientes de
las aguas — " una torrente que hace barrancas ;" but they also indicate

288 COSMOS.

not to be ascribed to these. Fissures, caused by folding in
the trachytic mass, which has been elevated while soft and
only subsequently hardened, have probably preceded all ac-
tions of erosion and the impulse of water. But in those places
where deep harrancos appeared in the volcanic districts visit-
ed by me on the declivities of bell-shaped or conical mount-
ains {en lasfaldas de los Cerros barrancosos), no trace was to
be detected of the regularity or radiate ramification with
which we are made acquainted by Junghuhn's works in the
singular outlines of the volcanoes of Java.* The greatest
analogy with the form here referred to is presented by the
phenomenon to which Leopold von Buch, and the acute ob-
server of volcanoes, Poulet Scrope, have already directed at-
tention, namely, that great fissures almost always open at a
right or obtuse angle from the centre of the mountain, radi-
ating (although undivided) in accordance with the normal
direction of the declivities, but not transversely to them.

The belief in the complete absence of lava streams upon
the island of Java,t to which Leopold von Buch appeared to
incline in consequence of the observations of Reinwardt, has
been rendered more than doubtful by recent observations.

any chasm. But that the word barranca is connected with barro, clay^
soft, moist loam, and also road-scrapings, is doubtful.

* Lyell, Manual of Elementary Geology^ 1855, chap, xxix., p. 497.
The most remarkable analogy with the phenomenon of regular rib-
bing in Java is presented by the surface of the Mantle of the Somma
of Vesuvius, upon the seventy folds of which an acute and accurate
observer, the astronomer J^us Schmidt, has thrown much light {Die
Ervption des Vesuvs im JSm, 1855, s. 101-109). According to Leo-
pold von Buch, these valley furrows are not originally rain furrows
(fiumare), but consequences of cracking (folding, etoilement) during the
first upheaval of the volcano. The usually radial position of the later-
al eruptions in relation to the axis of the volcano also appears to be
connected therewith (s. 129).

t "Obsidian, and consequently pumice-stones, are as rare in Java
as trachyte itself. Another very curious fact is the absence of any
stream of lava in that volcanic island. M. Reinwardt, who has him-
self observed a great number of eruptions, says expressly I^Kk there
have never been instances of the most violent and destructivc'miption
having been accompanied by lavas." — Leopold de Buch, Descr. des
Iks Canaries, p. 419. Among the volcanic rocks of Java, for which
the Cabinet of Minerals at Berlin is indebted to Dr. Junghuhn, dioritic
trachytes are most distinctly recognizable at Burungagung, s. 255 of
the Leidner catalogue, at Tjinas, s. 232, and in the Gunung Parang,
situated in the district Batu-gangi. This is consequently the identical
formation of dioritic trachyte of the volcanoes of Orizaba and Toluca,
in Mexico ; of the island Panaria, in the Lipari Islands, and of ^gina,
in the ^gean Sea !

TRUE VOLCANOES. 289

Junghuhn, indeed, remarks "that the vast volcano Gunung
Merapi has not poured forth coherent, compact lava streams
within the historical period of its eruptions, but has only-
thrown out fragments of lava (rubbish), or incoherent blocks
of stone, although for nine months in the year 1837 fiery
streams were seen at night running down the cone of erup-
tion."* But the same observant traveler has distinctly de-
scribed, in great detail, three black, basaltic lava streams on
three volcanoes — Gunung Tengger, Gunung Idjen, and Sla-

* Junghuhn, bd. ii., s. 309 and 314. The fiery streaks wliich were
seen on the volcano G. Merapi were formed by closely-approximated
streams of scorise {trainees de Jragmens), by non-coherent masses, Avhich
roll down during the eruption toward the same side, and strike against
each other from their very different weights on the steep declivity. In
the eruption of the G. Lamongan on the 26th March, 1847, a moving
line of scorias of this kind divided into two branches several hundred
feet below its point of origin. " The fiery streak," we find it express-
ly stated (bd. ii., s. 767), "did not consist of true fused lava, but of
fragments of lava rolling closely after one another." The G. Lamongan
and the G. Semeru are the two volcanoes of the island of Java, which
are found to be most similar, by their activity in long periods, to the
Stromboli, which is only about 2980 feet high, as they, although so re-
markably different in height (the Lamongan being 5340 and the Semeru
12,235 feet high), exhibited eruptions of scoriae, the former after pauses
of 15 to 20 minutes (eruptions of July, 1838, and March, 1847), and the
second of 1| to 3 hours (eruptions of August, 1836, and September,
1844) (bd. ii., s. 554 and 765-769). At Stromboli itself, together with
numerous eruptions of scoriae, small but rare effusions of lava also
occur, which, when detained by obstacles, sometimes harden on the
declivities of the cone. I lay great stress upon the various forms of
continuity or division, under which completely or partially fused mat-
ters are throAvn or poured out, whether from the same or different
volcanoes. Analogous investigations, undertaken under various zones,
and in accordance with guiding ideas, are greatly to be desired, from
the poverty and gi'eat one-sidedness of the views, to which the four
active European volcanoes lead. The question raised by me in 1802
and by my friend Boussingault in 1831 — whether the Antisana in the
Cordiheras of Quito has furnished lav<i streams ? which we shall touch
upon hereafter, may perhaps find its solution in the division of the
fluid matter. The essential character of a lava stream is that of a
uniform, coherent fluid — a band-like stream, from the surface of which
scales separate during its cooling and hardening. These scales, be-
neath which the nearly homogeneous lava long continues to flow, up-
raise themselves in part, obliquely or perpendicularly, by the inequal-
ity of the internal movement and the evolution of hot gases ; and when,
in this way, several lava streams, flowing together, form a lava lake, as
in Iceland, a field of detritus or fragments is produced on their cool-
ing. The Spaniards, especially in Mexico, call such a district, which
is very disagreeable to pass over, a malpais. Such lava fields, which
are often found in the plain at the foot of a volcano, remind one of
the frozen surface of a lake, with short, upraised ice-blocks.

Vol. v.— N

290 COSMOS.

mat.* On the latter the lava stream, after giving rise to a
water-fall, is continued into the tertiary rocks.f From such
true effusions of lava, which form coherent masses, Jung-
Imhn very accurately distinguishes, in the eruption of Gu-
nung Lamongan,t on the 6th of July, 1838, what he calls a
sioJie stream, consisting of glowing and usually angular frag-
ments, erupted in a row. " The crash was heard of the
breaking stones, which rolled down, like fiery points, either
in a line or without any order." I purposely direct especial
attention to the very various modes in which fiery masses
appear on the slopes of a volcano, because in the dispute
upon the maximum angle of fall of lava streams glowing
streams of stones (masses of scoriae), following each other in
rows, are sometimes confounded with continuous lava streams,
•As the important problem of the rarity or complete defici-
ency of lava streams in Java — a problem w^hich touches on
the internal constitution of volcanoes, and which, I must addy
has not been treated with sufficient earnestness — has recently
been so often spoken of, the present appears a fitting place in
which to bring it under a more general point of view. Al-
though it is very probable that in a group ro series of volca-
noes all the members stand in a certain common relation to
the general focus, the molten interior of the earth, still each
individual presents peculiar physical and chemical processes
as regards strength and frequency of activity, degree and
form of fluidity, and material difference of products — pecu-
liarities which can not be explained by the comparison of the
form, and elevation above the present surface of the sea.
The gigantic mountain Sangay is as uninterruptedly active

* The name of G. Idjen, according to Buschmann, may be explained
by the Javanese word hidjen, singly, alone, separately — a derivative
from the substantive hidji or widji, grain, seed, which with sa expresses
the number one. With regard to the etymology of G. Tengger, see
the important work of my brother upon the connections between Java
and India {Kawi-Sprache, bd. i., s. 188), where there is a reference to
the historical importance of the Tengger Mountain, which is inhabit-
ed by a small tribe of people, who, opposed to the now general Mo-
hammedanism of the island, have retained their ancient Indo-Javanic
faith. Junghuhn, who has very industriously explained the names of
mountains from the Kawi language, says (th. ii., s. 554), that in the
Kawi Tengger signifies hill ; the word also receives the same significa-
tion in Geriche's Javanese Dictionary (Jai;aawsc/i-?(eG?erc?t<i7sc/t Woorden-
boeky Amst., 1847). Slamat, the name of the high volcano of Tegal,
is the well-known Arabic word selamat, which signifies happiness and
safety.

t Junghuhn, bd. ii., Slamat, s. 153 and 163; Idjen, s. 698; Teng-
ger, t^- 773. t Bd. ii., s. 760-762,

TRUE VOLCANOES. 291

as the lowly Stromboli ; of two neighboring volcanoes, one
throws out pumice-stone without obsidian, the other both at
once ; one furnishes only loose cinders, the other lava flow-
ing in narrow streams. These characteristic processes, more-
over, in many volcanoes appear not to have been always the
s'ctme at various Epochs of their activity. To neither of the
two continents is rarity or total absence of lava streams to
be peculiarly ascribed. Remarkable distinctions only occur
in those groups with regard to which we must confine our-
selves to definite historical periods near to our own times.
The non-detection of single lava streams depends simultane-
ously upon many conditions. Among these we may instance
the deposition of vast layers of tufa, rapilli, and pumice-stone;
the simultaneous and non-simultaneous confluence of several
streams, fotming a widely-extended lava-field covered with
detritus; the circumstance that in a wide plain the small
conical eruptive cones, the volcanic platform, as it were, from
which, as at Lancerote, the lava had flowed forth in streams,
have long since been destroyed. In the most ancient condi-
tions of our unequally-cooling planet, in the earliest foldings
of its surface, it appears to me very probable that a frequent
viscid outflow of trachytic and doleritic rocks, of masses of
pumice-stone or perlite, containing obsidian, took place from
a composite net-work of fissures, over which no platform has
ever been elevated or built up. The problem of such simple
effusions from fissures deserves the attention of geologists.

In the series of ISIexican volcanoes, the greatest and, since
my American travels, the mgst celebrated phenomenon, is the
elevation of the newly-produced Jorullo, and its effusion of
lava. This volcano, the topography of which, founded on
measurements, I was the first to make known,* by its posi-
tion between the two volcanoes of Toluca and Colima, and
by its eruption on the great fissure of volcanic -activity,!
which extends from the Atlantic Ocean to the Pacific, pre-
sents an important geognostic phenomenon, which has con-
sequently been all the more the subject of dispute. Follow-
ing the vast lava stream which the new volcano poured out,
I suc*eded in getting far into the interior of the crater, and
in establishing instrumiits there. The eruption in a broad
and long-peaceful plain in the former province of Michuacan,
in the night from the 28th to the 29th of September, 1759,
at a distance of more than 120 miles from any other volcano,

* Atlas Gtographique et Physique, accompanying the Relation His-
toriqtie, 1814, pi. 28 and 29. f Cosjnos,\oh v., p. 264-266.

292 COSMOS.

was preceded for fully two (?) months, namely, from the 29th
of June in the same year, by an uninterrupted subterranean
noise. This differed from the wonderful bramidos of Guan-
axuato, which I have elsewhere described,* by the circum-
stance that it was, as is usually the case, accompanied by
earthquakes, which were not felt in the mountain city'»in
January, 1784. The eruption of the new volcano, about
3 o'clock in the morning, was foretold the day before by
a phenomenon which, in other eruptions, does not indicate
their commencement, but their conclusion. At the point
where the great volcano now stands, there was formerly a
thick wood of the Guayava {Psidium pyriferum), so much
valued by the natives on account of its excellent fruit. La-
borers from the sugar-cane fields (canaverales) of the Haci-
enda de San Pedro Jorullo, belonging to the rich Don An-
dres Pimentel, who was then living in Mexico, had gone out
to collect the fruit of the guayava. When they returned to
the farm (hacienda) it was remarked with astonishment that
their large straw hats were covered with volcanic ashes.
Fissures had, consequently, already opened in what is now
called the Malpais, probably at the foot of the high basaltic
dome El Cuiche, which threw out these ashes (rapilli) before
any change appears to have occurred in the plain. From
a letter of Father Joaquin de Ansogorri, discovered in the
Episcopal archives of Valladolid, which was written three
weeks after the day of the first eruption, it appears evident
that Father Isidro Molina, sent from the neighboring Jesuits'
College of Patzcuaro " to give^ spiritual comfort to the in-
habitants of the Playas de Jorullo, who were extremely dis-
quieted by the subterranean noise and earthquakes," was the
first to perceive the increasing danger, and thus caused the
preservation of the small population.

In the first hours of the night the black ashes already lay
a foot deep ; every one fled toward the hill of Aguasarco,
a small Indian village, situated 2409 feet higher than the
old plain of Jorullo. From this height (so runs the tradi-
tion) a large tract of land was seen in a state of fearful fiery
eruption, and "in the midst of the flames (as those Wio wit-
nessed the ascent of the mountain exjoressed themselves) there
appeared like a black castle (castillo negro) a great shape-
less mass (bulto grande)." From the small population of
the district (the cultivation of indigo and cotton was then
but very little carried on) even the force of long-continued
* Cosmos^ vol. i., p. 209, and vol. v., p. 172.

TRUE VOLCANOES. 293

earthquakes cost no human lives, although, as I learn from
manuscript record,* houses were overturned by them near

* In my Essai Politique sur la Nouvelle Espagne, in the two editions
of 1811 and 1827 (in the latter, t. ii., p. 165-175), I have, as the na-
ture of that work required, only given a condensed abstract from my
journal, without being able to furnish a topographical plan of the vi-
cinity or a chart of the altitudes. From the importance which has
been assigned to this great phenomenon of the middle of the last cen-
tury, I have thought it necessary to complete this abstract here* I
am indebted for particular details relating to the new volcano of Jo-
rullo to an official document, written three weeks after the day of the
first eruption, but only discovered in the year 1830 by a very scientific
Mexican clergyman, Don Juan Jose Pastor Morales ; and also to oral
communications from my companion, the Biscayan Don Ramon Es-
pelde, who had been able to examine living eye-witnesses of the first
eruption. Morales discovered in the archives of the Bishop of Mich u-
acan a report addressed on the 19th of October, 1759, by Joaquin de
Ansogorri, priest in the Indian village la Guacana, to his bishop. In
his instructive work {Aufenthalt und Eeisen in Mexico, 1836) Burkart
has also given a short extract from it (bd. i., s. 230). At the time of
my journey, Don Ramon Espelde was living on the plain of Jorullo,
and has the merit of having first ascended the summit of the volcano.
Some years afterward he attached himself to the expedition made on
the 10th of March, 1789, by the Intendente Corregidor, Don Juan
Antonio de Riano. To the same expedition belonged a well-informed
German, Franz Fischer, who had entered the Spanish service as a
mining com^rissar3^ By means of the latter the name of the Jorullo
first became known in Germany, as he mentioned it in a letter in the
Schriften der Gesellschaft der Bergbaukunde, bd. ii., s. 441. But the
eruption of the new volcano had already been referred to in Italy — in
Clavigero's Storia autica del Messico (Cesena, 1780, t. i., p. 42), and in
the poetical work, Rustlcatio Mexicana, of Father Raphael Landivar
(ed. altera, Bologna, 1782, p. 17). In his valuable work Clavigero er-
roneously places the production of the volcano, which he writes Ju-
ruyo, in the year 1760, and enlarges the description of the eruption
by accounts of the shower of ashes, extending as far as Queratoro,
which had been communicated to him in 1766 by Don Juan Manuel
de Bustamente, governor of the province of Valladolid de Michuacan,
as an eye-witness of the phenomenon. The poet Landivar, an enthu-
siastic adherent, like Ovid, of x)ur theory of upheaval, makes the co-
lossus rise, in euphonious hexameters, to the full height of three mil-
liaria, and finds the thermal springs (after the fashion of the ancients)
cold by day and warm at night. But I saw the thermometer rise to
1261^° in the water of the Rio de Cuitimba about noon.

In 1789, and consequently in the same year that the report of the
Governor Riano and the Mining Commissary Franz Fischer appeared
in the Gazeta de Mexico (in the fifth part of his large and useful JJic-
cionario Geogrqfico-historico de las Indias Occidentales 6 America, in the
article Xtirullo, p. 374, 375), Antonio de Alcedo gave the interesting
information that when the earthquakes commenced (29th of June,
1759) in the Playas, the western volcano of Colima, which was in erup-
tion, suddenly became quiet, although it is at a distance of " 70 leguas"
(as Alcedo says, according to my map only 112 geographical miles !)

294 COSMOS.

the copper mines of Inguaran, in the small town of Patzcu-
aro, in Santiago de Ario, and many miles farther, but not

from the Playas. "It is thought," he adds, "that the materials in
the bowels of the earth have met with obstacles to their following
their old course; and, as they have found suitable cavities (to the
east," they have broken out at Jorullo — para reventar en Xurullo). —
Accurate topographical statements regarding the neighborhood of the
volcano occur also in Jnan Jose Martinez de Lejarza's geographical
sketch of the ancient Taraskian country : Andlisis Estadistico de laPro-
vincia de Michuacan en 1822 (Mexico, 1824), p. 125, 129, 130, and 131.
The testimony of the author, living at Valladolid, in the vicinity of
Jorullo, that, since my residence in Mexico, no trace of an increased
activity has shown itself in the mountain, was the earliest contradic-
tion of the report of a new eruption in the year 1819 (Lyell, Princi-
ples of Geology^ 1855, p. 430). As the position of Jorullo in latitude
is not without importance, I have noticed that Lejarza, who otherwise
always follows my astronomical determinations of position, and who
gives the longitude of Jorullo exactly like myself as 2° 25' west of the
meridian of Mexico (101° 29' west of Greenwich), differs from me in
the latitude. Is the latitude attributed by him to the Jorullo (18° 53'
30 "), which comes nearest to that of the volcano of Popocatepetl (18°
59' 47"), founded upon recent observations unknown to me? In my
Recueil d'Observ. Astronomiques, vol. ii., p. 521, I have said expressly,
" Latitude supposce, 19° 8', deduced from good astronomical observa-
tions at Valladolid, which gave 19° 52' 8", and from the itinerary di-
rection." I only recognized the importance of the latitude of Jorullo
when subsequently I was drawing up the great map of Mexico in the
capital city and inserting the E. — W. series of volcanoes.

As in these considerations upon the origin of Jorullo I have repeat-
edly mentioned the traditions which still prevail in the neighborhood,
I will conclude this long note by referring to a very popular tradition,
which I have already touched upon in another work {Essai Politique sur
la Nouvelle Esjmgne, t. ii., 1827, p. 172): "According to the belief of
the natives, these extraordinary changes which we have just described
Ai'e the work of the monks, the greatest, perhaps, that they have pro-
duced in either hemisphere. At the Playas de Jorullo, in the hut that
we occupied, our Indian host told us that in 1759 the Capuchins be-
longing to the mission preached at the station of San Pedro, but that,
not having been favorably received, they charged this beautiful and
fertile plain with the most horrible and complicated imprecations,
prophesying that first of all the house would be devoured by flames
which would issue from the earth, and that afterward the surrounding
air would become cooled to such a degree that the neighboring mount-
ains would remain eternally covered with snow and ice. The former
of these maledictions having had such fatal consequences, the lower
class of Indians already see in the gradual cooling of the volcano the
presage of a perpetual winter."

Next to that of the poet. Father Landivar, the first printed account
of the catastrophe was ])robably that already mentioned in the Gazeta
de Mexico of the 5th of May, 1789 (t. iii.. Num. 30, p. 293-297) ; it
bears the modest title, Superficial y nada facultativa Descripcion deles-
tado en que se hallaba el Volcdn de Jorullo, la mahana del dia 10 de Marzo
de 1789, and was occasioned by the expedition of Eiaiio, Franz Fischer,

TRUE VOLCANOES. 295

beyond San Pedro Clmrumucu. In the Hacienda de Jorullo,
during the general nocturnal flight, they forgot to remove a
deaf and dumb negro slave. A mulatto had the humanity to
return and save him wliile the house was still standing. It
is still related that he was found kneeling, with a consecrated
taper in his hand, before the picture of Nuestra Senora de
Guadalupfe.

According to the tradition, widely and concordantly spread
among the natives, the eruption, during the first days, con-
sisted of great masses of rock, scorise, sand, and ashes, but
always combined with an effusion of muddy water, ipl the
memorable report, already mentioned, of the 19th of October,
1759, the author of which was a man Avho, possessing an
accurate knowledge of the locality, describes what had only
just taken place, it is expressly said : Que espele el dicho
Volcan arena, ceniza y agua. All eye-witnesses relate (1
translate from the description which the Intendant, Colonel
Kiaiio, and the German Mining Commissary, Franz Fischer,
who had passed into the Spanish service, have given of the
condition of the volcano of Jorullo on the 10th of March,
1789), " that before the terrible mountain made its appear-
ance (antes de reventar y aparecerse este terrible cerro) the
earthquakes and subterranean noises became more frequent ;
but on the day of the eruption itself the flat soil was seen to
rise perpendicularly (se observo, que el plan de la tierra se
levantaba perpendicularmente), and the whole became more
or less inflated, so that bhsters (vexigones) appeared, of which
the largest is now the volcano (de los que el mayor es hoy el
cerro del volcan). These inflated blisters, of very various
sizes, and partly of a tolerably regular conical form, subse-
quently burst (estas ampollas, gruesas vegigas 6 conos dife-
rentemente regulares en sus figuras y tamanos, reventaron
despues), and threw boiling-hot earthy mud from their or-
ifices (tierras hervidas y calientes), as M^ell as scoriaceous
stony masses (piedras cocidas? y fundidas), which are still
found, at an immense distance, covered with black stony
masses."

These historical records, which we might, indeed, wish to
see more complete, agree perfectly with what I learn from
the mouths of the natives fourteen years after the ascent of
Antonio de Riaiio. To the questions, whether *' the castle

and Espelde. Subsequently (1791), in the naval astronomical expedi-*
tion of Malaspina, the botanists Mocino and Don Martin Sesse visit-
ed Jorullo from the Pacific coast.

296 COSMOS.

mountain" was seen to rise gradually for months or years, or
whether it appeared from the very first as an elevated peak,
no answer could be obtained. Riano's assertion that farther
eruptions had taken place in the first sixteen or seventeen
years, and therefore up to ITT'G, was declared to be untrue.
According to the tradition, the phenomena of small eruptions
of water and mud which were observed during the first days
simultaneously with the incandescent scoriop are ascribed to
the destruction of two brooks, which, springing on the western
declivity of the mountain of Santa Ines, and consequently to
the^st of the Cerro de Cuiche, abundantly irrigated the
cane-fields of the former Hacienda de San Pedro de Jorullo,
and flowed onward far to the west to the Hacienda de la Pre-
sentacion. Near their origin, the point is still shown where
they disappeared in a fissure with their formerly cold waters
during the elevation of the eastern border of the Malpais.
Running below the hornitos, they reappear, according to the
general opinion of the people of the country, heated, in two
thermal springs. As the elevated part of the Malpais is
there almost perpendicular, they form two small water-falls,
which I have seen and represented in my drawing. For
each of them the previous name, H\o de San Pedro and Rio
de Cuitimba, has been retained. At this point I found the
temperature of the steaming water to be 126°*8. During
their long course the waters are only heated, but not acid-
ulated. The test papers, which I usually carried about with
me, underwent no change ; but farther on, near the Hacienda
de,la Presentacion, toward the Sierra de las Canoas, there
flows a spring impregnated with sulphureted hydrogen gas,
which forms a basin of 20 feet in breadth.

In order to acquire a clear notion of the complicated out-
line and general form of the surface of the ground, in which
such remarkable upheavals have taken place, we must dis-
tinguish hypsometrically and morphologically: 1. The po-
sition of the volcanic system of Jorullo in relation to the av-
erage level of the Mexican plateau ; 2. The convexity of the
Malpais, which is covered by thousands of hornitos ; 3. The
fissure upon which six large volcanic mountain masses have
arisen.

On the western portion of the Central Cordillera of Mex-
ico, which strikes from S.S.E. to N.N.W., the plain of the
Playas de Jorullo, at an elevation of only 2557 feet above
the level of the Pacific, forms one of the horizontal mount-
ain terraces which every where in the Cordilleras interrupt

TRUE VOLCANOES. 297

the line of inclination of the declivity, and consequently more
or less impede the decrease of heat in the superposed strata
of the atmosphere. On descending from the central plateau
of Mexico (whose mean elevation is 7460 feet), to the corn-
fields of Valladolid de Michuacan, to the charming lake of
Fatzcuaro, with*the inhabited islet Janicho, and into the
meadows around Santiago de Ario, which Bonpland and I
found adorned with the dahlias which have since become so
well known, we have not descended more than nine hundred
or a thousand feet. But in passing from Ario, on the steep
declivity over Aguasarco, into the level of the old plain of
JoruUo, we diminish the absolute elevation in this short dis-
tane« by from 3850 to 4250 feet.^ The roundish, convex
part of the upheaved plain is about 12,790 feet in diameter,
so that its area is more than seven square miles. The true
volcano of Jorullo and the five other mountains which rose
simultaneously with it upon the same fissure are so situated
that only a small portion of the Malpais lies to the east of
them. Toward the west, therefore, the number of hornitos
is much larger, and when in early morning I issued from the
Indian huts of the Playas de Jorullo, or ascended a portion
of the Cerro del Mirador, I saw the black volcano projecting
very picturesquely above the innumerable white columns of
smoke of the "little ovens" {hornitos). Both the houses of
the Playas and the basaltic hill Mirador are situated upon
the level of the old non-volcanic, or, to speak more cauti-
ously, unupheaved soil. Its beautiful vegetation, in which
a multitude of salvias bloom beneath the shade of a new se-
cies of fan palm (Corypha pumos), and of a new alder {Alnus
JoruUensis), contrasts with the desert, naked aspect of the
Malpais. The comparison of the height of the barometerf
at the point where the upheaval commences in the Playas,

*My barometric measurements give for Mexico 1168 toises (7470
feet), Valladolid 1002 toises (6409 feet), Patzcuaro 1130 toises (7227
feet), Ario 994 toises (6358 feet), Aguasarco 780 toises (4089 feet), for
the old plain of the Playas de Jorullo 404 toises (2584 feet) (Hum-
boldt, Observ. Astron., vol. i., p. 327, j!:^ivellement Barometrique, No.
366-370).

t If the old plain of the Playas be 404 toises (2584 feet), I find for
the maximum of convexity of the Malpais above the sea-level 487
toises (3115 feet); for the ridge of the great lava stream 600 toises
(3838 feet) ; for the highest margin of the crater 667 toises (4266
feet) ; for the lowest point of the crater at which we could establish
the barometer 644 toises (4119 feet). Consequently the elevation of
the summit of Jorullo above the old plain appeared to be 263 toises,
or 1682 feet.

N2

298

COSMOS.

with that at the point immediately at the foot of the vol-
cano, gives 473 feet of relative perpendicular elevation. The
house that we inhabited stood only about 500 toises (3197
feet) from the border of the Malpais. At that place there
was a small perpendicular precipice of scarcely twelve feet
high, from which the heated water of the htook (Rio de San
Pedro) falls down. The portion of the inner structure of
the soil which I could examine at the precipice showed
black, horizontal, loamy strata, mixed with sand (rapilli).
At other points which I did not see, Burkart has observed
" on the perpendicular boundary of the upheaved soil, where
the ascent of this is difficult, a light gray and not very dense
(weathered) basalt, with numerous grains of olivin."* ' This
accurate and experienced observer has, however, t like myself,
on the spot conceived the idea of a vesicular upheaval of the
surface effected by elastic vapors, in opposition to the opinion
of celebrated geognosists,{ who ascribe the convexity, which
I ascertained by direct measurement, solely to the greater
effusion of lava at the foot of the volcano.

The many thousand small eruptive cones (properly rather
of a roundish or somewhat elongated, oven-like form), which
cover the upheavedsur face pretty uniformly, are on the
average four to nine feet in height. They have risen almost
exclusively on the western side of the great volcano, as in-
deed the eastern part, toward the Cerro de Cuiche, scarcely
constitutes -213th of the entire area of the vesicular elevation
of the Playas. Each of the numerous hornitos is composed
of «veathered basaltic spheres, with fragments separated like

* Burkart, Aufenthalt und Reisen in Mexico in den Jahren, 1825-1834,
bd. i. (1836), p.' 227. f Op. cit. svp., bd. i., p. 227 and 230.

t Ponlett Scrope, Considerations on Volcanoes, p. 267 ; Sir Charles
Lyell, Principles of Geology, 1853, p. 429; Manual of Geology, 1855,
p. 580; Daubeny o« Volcanoes, p. 337. See also "on the elevation
hypothesis," Dana, Geology, in the United States Exploring Expedition,
vol. X., p. 369. Constant Prevost, in the Cowptes rendus, t. xli. (1855),
p. 866-876, and 918-923 : sur les Eruptions et le drapeau de Vivfailli-
bilite." See also, with regard to Jorullo, Carl Pieschel's instructive
description of the volcanoes of Mexico, with illustrations by Dr. Gum-
precht, in the Zeitschrift fur Allg. Erdkunde of the Geographical Society
of Berlin (bd. vi., s. 490-517); and the newly-published picturesque
views in Pieschel's Atlas der Vulkane der Repuhlih Mexico, 1856, tab.
13, 14, and 15. The Eoyal Museum of Berlin, in the department of
engravings and drawings, possesses a splendid and numerous collec-
tion of representations of the Mexican volcanoes (more than forty
sheets), taken from nature by Moritz Rugendas." Of the most western
of all Mexican volcanoes, that of Colima, alone, this great master has
furnished fifteen colored views.

TRUE VOLCANOES. 299

concentric shells ; I was frequently able to count from 24 to
28 such shells. The balls are flattened into a somewhat
spheroidal form, and are usually 15 — 18 inches in diameter,
but vary from one to three feet. The black basaltic *mass is
penetrated by hot vapors and broken up into an earthy form,
although the nucleus is of greater density, while the shells,
when detached, exhibit yellow spots of oxyd of iron. Even
the soft, loamy mass which unites the balls is, singularly
enough, divided into curved lamellae, which wind through
all the interstices of the balls. At the first glance I asked
myself whether the whole, instead of weathered basaltic
spheroids, containing but little olivin, did not perhaps pre-
sent masses disturbed in the course of their formation. But
in opposition to this we have the analogy of the hills of globu-
lar basalt, mixed with layers of clay and marl, which are
found, often of very small dimensions, in the central chain of
Bohemia, sometimes isolated and sometimes crowning long
basaltic ridges at both extremities. Some of the hornitos
are so much broken up, or have such large internal cavities,
that mules, when compelled to place their fore-feet upon the
flatter ones, sink in deeply, while in similar experiments
which I made the hills constructed by the termites resisted.
In the basaltic mass of the hornitos I found no immersed
scoriae, or fragments of old rocks which had been penetrated,
as is the case in the lavas of the great Jorullo. The appel-
lation i^To/'/ios or Hornitos is especially Justified by the cir-
cumstance that in each of them (I speak of the period when
1 traveled over the Playas dc Jorullo and wrote my journal,
18th of September, 1803) the columns of smoke break out,
not from the summit, but laterally. In the year 1780 cigars
might still be lighted when they were fastened to a stick and
pushed in to a depth of two or three inches ; in some places
the air was at that time so much heated by the vicinity of
the hornitos, that it was necessary to turn away from one's
proposed course. Notwithstanding the refrigeration which,
according to the universal testimony of the Indians, the district
had undergone within 20 years, I found the temperature in
the fissures of the hornitos^to range between 199° and 203° ;
and at a distance of twenty feet from some hills the tempera-
ture of the air was still 108°-5 and 116°*2, at a point where
no vapors reached me, the true temperature of the atmos-
phere of the Playas being at the same time scarcely 77°.
The weak sulphuric vapors decolorized strips of test paper,
and rose visibly, for some hours nfter sunrise, to a height of

300 COSMOS.

fully 60 feet. The view of the columns of smoke was most
remarkable early in a cool morning. Toward midday, and
even after 11 o'clock, they had become very low, and were
visible only from their immediate vicinity. ' In the interior
of many of the hornitos we heard a rushing sound like the
fall of water. The small basaltic hornitos are, as already
remarked,, easily destructible. When Burkart visited the
Malpais, 24 years after me, he found that none of the hor-
nitos were- still smoking, their temperature being in most
cases the same as that of the surrounding air, while many of
them had lost all regularity of form by heavy rains and me-
teoric influences. Near the principal volcano Burkart found
small cones, which were composed of a brownish-red con-
glomerate of rounded or angular fragments of lava, and only
loosely colierent. In the midst of the upheaved area, cover-
ed with hornitos, there is still to be seen a remnant of the
old elevation on which the buildings of the farm of San
Pedro rested. The hill, which I have indicated in my plan,
forms a ridge directed east and west, and its preservation at
the foot of the great volcano is n^ost astonishing. Only a
part of it is covered with dense sand (burned rapilli). The
projecting basaltic rock, grown over with ancient trunks of
FHcus indica and Psidium, is certainly, like that of the Cerro
del Mirador and the high mountain masses which bound the
plain to the eastward, to be regarded as having existed be-
fore the catastrophe.

It remains for me to describe the vast fissure upon which
a series of six volcanoes has risen, in the general direction
from south-southwest to north-northeast. The partial direc-
tion of the first three less-elevated volcanoes situated most
southerly is S.AV. — N.E. ; that of the three following near
S. — N. The fissure has consequently been curved, and has
changed its strike throughout its total length of 10,871 feet.
The direction here indicated of the linear but not contiguous
mountains is certainly nearly at right angles with the line
upon which, according to my observation, the Mexican vol-
canoes follow each other from sea to sea. But this differ-
ence is the less surprising if we consider that a great geog-
nostic phenomenon (the relation" of the principal masses to
each other across a continent) is not to be confounded with
the local conditions and direction of a single group. The
long ridge of the great volcano of Pichincha, also, is not in
the same direction as the series of volcanoes of Quito ; and
in non-volcanic chains, for example in the Himalaya, the

TRUE VOLCANOES. 301

culminating points are often situated, as I have already point-
ed out, at a distance from the general line of elevation of the
chain. They are situated upon partial snowy ridges, which
even form nearly a right angle with this general line of up-
heaval.

Of the six volcanic hills which have risen upon the above-
mentioned fissure, the first three, the more southern ones, be-
tween which the road to the copper mines of Inguaran pass-
es, appear, in their present condition, to be of least import-
ance. They are no longer open, and are entirely covered
with grayish-white volcanic sand, which, however, does not
consist of pumice-stone, for I have seen nothing either of
pumice or obsidian in this region. At Jorullo also, as at
Vesuvius, according to the assertion of Leopold von Buch
and Monticelli, the last covering-fall of ashes appears to have
been the white one. The fourth more northern mountain is
the large, true volcano of Jorullo, the summit of which, not-
withstanding its small elevation (4265 feet above the sea
level, 1151 feet above the Malpais at the foot of the volcano,
and 1681 feet above the old soil of the Playas), I had some
difficulty in reaching, when I ascended it with Bonpland and
Carlos Montufar on the 19th of September, 1803. We
thought we should be most certain of getting into the crater,
which was still filled with hot sulphurous vapors, by ascend-
ing the steep ridge of the vast lava stream, which burst forth
from the very summit. The course passed over a crisp, sco-
ria<;eous, clear-sounding lava, swelled up in a coke-like, or
rather cauliflower-like form. Some parts of it have a metal-
lic lustre : others are basaltic and full of small granules of
olivin. When we had thus ascended to the upper surface
of tbe lava stream at a perpendicular elevation of 711 feet,
we turned to the white ash cone, on which, from its great
steepness, we could not but fear that during frequent and
rapid slips we might be seriously wounded by the rugged
lava. The upper margin of the crater, on the southwestern
part of which we placed the instruments, forms a ring of a
few feet in width. We carried the barometer from the mar-
gin into the oval crater of the truncated cone. At an open
fissure air streams forth of a temperature of 200°*6. We
now stood 149 feet in perpendicular height below the margin
of the crater ; and the deepest point of the chasm, the attain-
ment of which we were compelled to give up on account of
the dense sulphurous vapors, appeared to be only about twice
this depth. The geognostic discovery which had the most

302 COSMOS.

interest for us was the finding of several white fragments,
three or four inches in diameter, of a rock rich in feldspar
baked into the black basaltic lava. I regarded these at first*
as syenite, but from the exact examination by Gustav Rose,
of a fragment which I brought with me, they probably belong
rather to the granite formation, which Burkart has also seen
emerging from below the syenite of the Rio de las Balsas.
" The inclosure is a mixture of quartz and feldspar. The
blackish-green spots appear to be not hornblende, but mica
fused with some feldspar. The white fragment baked in is
split by volcanic heat, and in the crack white, tooth-like,
fused threads run from one margin to the other."

To the north of the great volcano and the scoriaceous lava
mountain which it has vomited forth in the direction of the
old basalt of the Cerro delMortero follow the two last of the
six often-mentioned eruptions. These hills also were original-
ly very active, for the people still call the extreme mountain
of ashes El Volcancito. A broad .fissure, open toward the west,
bears the traces of a destroyed crater. The great volcano,
like the Epomeo in Ischia, appears to have only once poured
out a mighty lava stream. That its lava-pouring activity

* " M. Bonpland and myself were particularly astonished at finding,
encased in the basaltic, lithoid, and scorified lavas of the volcano of
Jorullo, white or greenish-white angular fragments of syenite^ com-
posed of a little amphibole and a great quantity of lamellar feldspar.
Where these masses have been split by heat the feldspar has become
filamentous, so that the margins of the crack are united in some places
by fibres elongated from the mass. In the Cordilleras of South Amer-
ica, between Fopayan and Almaguer at the foot of the Cen^o Bronco-
so, I have foiind actual fragments of gneiss encased in a trachyte
abounding in pyroxene. These phenomena prove that the trachytic
formations have issued from beneath the granitic crust of the globe.
Analogous phenomena are presented by the trachytes of the Siebenge-
hirge on the banks of the Rhine, and by the inferior strata of Phono-
lite {Porphyrschiefer) of the Bilincr Stein in Bohemia." (Humboldt,
Essai Geognostique sur le Gisement des Roches, 1823, p. 133 and 339.)
Burkart also {Aitfenthalt und Reisen in Mexico, bd. i.yS. 230) detected
inclosed in the black lava, abounding in olivin, of Jorullo, "blocks of
a metamorphosed syenite. Hornblende is rarely to be recognized dis-
tinctly. The blocks of syenite may certainly furnish an incontroverti-
ble proof that the seat of the focus of the volcano of Jorullo is either
in or below the syenite, which shows itself in considerable extent, a
few miles {leguas) farther south, on the left bank of the Rio de las
Balsas, flowing into the Pacific Ocean." Dolomieu, and, in 1832, the
excellent geognosist, Friedrich Hoffmann, found in Lipari, near Cane-
to, fragments of granite, formed of pale red feldspar, black mica, and
a little pale gray quartz, inclosed in compact masses of obsidian (Pog-
gendorfTs Annahn der JViysik. hd. xxvi., s. 49).

TRUE VOLCANOES. 303

endured after the period of its first eruption is not^ proved
historically ; for the valuable letter, so happily discovered,
of Father Joaquin de Ansogorri, written scarcely three weeks
after the first eruption, treats almost exclusively of the means
of making "arrangements for the better pastoral care of the
country people who had fled from the catastrophe and be-
come dispersed ;" and for the following thirty years we have
no records. As the tradition speaks very generally of fires
which covered so great a surface, it is certainly to be sup-
posed that all the six hills upon the great fissure, and the
portion of the Malpais itself in which the hornitos have ap-
peared, were simultaneously in combustion. The tempera-
ture of the surrounding air, which I measured, allows us to
judge of the heat which prevailed there 43 years previously ;
they remind one of the former condition of our planet, in
which the temperature of its atmospheric envelope, and with
this the distribution of organic life, might be modified by the
thermic action of the interior by means of deep fissures (un-
der any latitude and for long periods of time).

Since I described the hornitos which surround the volcano
of Jorullo, many analogous platforms in various regions of
the world have been compared with these oven-like little hills.
To me the Mexican ones, from their interior conformation,
appear still to stand in a very contrasting and isolated con-
dition. If all upheavals which emit vapors are to be called
eruptive cones, the hornitos certainly deserve the appellation
oi Fumaroles. But the denomination eruptive cones would
lead to the erroneous notion that there is evidence that the
hornitos have thrown out scoriae, or even, like many eruptive
cones, poured forth lava. Very different, for example (to
advert to a great phenomenon), are the three chasms in Asia
Minor, upon the former boundaries of Mysia and Phrygia, in
the ancient burning country (Katakekaumene), *' where it is
dangerous to dwell (on account of the earthquakes)," which
Strabo calls (f)vsJaL, or wind-bags, and which the meritorious
traveler,' William Hamilton, has rediscovered.* Eruptive
cones, such as are exhibited by the island of Lancerote near

* Strabo, lib. xiii., p. 579 and 628; Hamilton, Researches in Asia
Minor, vol. ii., chap. 39. The most western of the three cones, now
called Kara Devlit, is raised 532 feet above the plain, and has emitted
a great lava stream in the direction of Koula. Hamilton counted more
than thirty small cones in the vicinity. The three chasms {(i69poi and
ipvaai of Strabo) are craters situated upon conical mountains composed
of scoriae and lavas.

304 COSMOS.

Tinguaton, or by Lower Italy, or (of hardly 20 feet in height)
by the declivity of the great Kamtschatkan volcano, Awat-
scha,* which was ascended in July, 1824, by my friend and
Siberian companion, Ernst Hofmann, consist of scorias and
ashes surrounding a small crater, which has thrown them
out, and bas been in return buried by them. In the horni-
tos nothing like a crater is to be seen, and they consist — and
this is an important character — merely of basaltic balls, with
shell-like separated fragments, without any admixture of
loose angular scoriag. At the foot of Vesuvius, during the
great eruption of 1794 (and also in earlier times), eight dif-
ferent small craters of eruption {hocche nuove) were formed,
arranged upon a longitudinal fissure ; they are the so-called
parasitic cones of eruption, which poured forth lava, and are
even by this circumstance entirely distinct from the hornitos
of JoruUo. "Your hornitos," wrote Leopold von Buch to
me, " are not cones accumulated by erupted matters ; they
have been upheaved directly from the interior of the earth."
The production of the volcano of Jorullo itself was compared
by this great geologist with that of the Monte Nuovo in the
Phlegraean fields. The same notion of the upheaval of six
volcanic mountains upon a, longitudinal fissure forced itself
as the most probable upon Colonel Riano and the mining
commissary Fischer in 1789 (see ante, p. 295), upon myself
at the first glance in 1803, and upon Burkart in 1827.
With both the new mountains, produced in 1538 and 1759,
the same questions repeat themselves. Upon that of South-
ern Italy the testimonies of Falconi, Pietro Giacomo di To-
ledo, Francesco del Nero, and Porzio are circumstantial,
near the time of the catastrophe, and prepared by educated
observers. The celebrated Porzio, who was the most learned
of these observers, says : " Magnus terras tractus, qui inter
radices mentis, quern Barbarum incola; appellant, et mare
juxta Avernum jacet, sese erigere videbatur et montis subito
nascentis figuram imitari. Iste terrse cumulus aperto veluti
ore magnos ignes evomuit, pumicesque, et lapides, cineres-
que."t

* Erman, Reise um die Erde^ bd. iii., s. 538; Cosmos, vol. v., p,
236. Postels ( Voyage autour du Monde par le Cap. Lutke, partie hist.,
t. iii.,, p. 76) and Leopold von Buch {Description Physique des lies Ca-
naries, p. 448) mention the similarity to the hornitos of Jorullo. In
a manuscript most kindly communicated to me, Erman describes a
great number of truncated cones of scoriae in the immense lava-field to
the east of the Baidar Mountains, on the peninsula of Kamtschatka.

t Porzio, Opera omnia, Med., Phil., et Mathem. in vnum coUecta,

TRUE VOLCANOES. 305

From the geognostic description here completed of the vol-
cano of JoruUo we will pass to the more eastern parts of
Central America (Anahuac). Unmistakable lava streams,
the principal mass of which is usually basaltic, have been
poured out by the peak of Orizaba, according to the most
recent interesting observations of Pieschel (March, 1854)*
and H. de Saussure. The rock of the peak of Orizaba, like
that of the volcano of Toluca,t which I ascended, is com-
posed of hornblende, oligoclase, and a little obsidian ; while
the fundamental mass of Popocatepetl is a Chimborazo rock,
composed of very small crystals of oligoclase and augite. At
the foot of the eastern slope of Popocatepetl, westward of the
town La Puebla de los Angeles, in the Llano daTetimpa, where
I measured the base for the determination of the elevation
of the two great nevados (Popocatepetl and Iztaccihuatl)
which bound the valley of Mexico, I found, at a height of
7000 feet above the sea, an extensive and mysterious kind
of lava-field. It is called the Malpais (rough rubbish-field)
of Atlachayacatl, a low trachytic dome, on the declivity of
which the River Atlaco rises and runs at an elevation of
from 60 to 85 feet above the adjacent plain, from east to
west, and consequently at right angles to the volcanoes.
From the Indian village of San Nicolas de los Ranchos to
San Buenaventura, I calculated the length of the Malpais at
more than 19,200 fe^MWid its breadth at 6400 feet. It con-
sists of black, parti^iPlupraised lava-blocks, of a fearfully
wild appearance, and only sparingly coated here and there
with lichens, contrasting with the yellowish-white coat of
pumice-stone which covers every thing for a long distance
round. The latter consists here of coarsely fibrous fragments
of two or three inches in diameter, in which hornblende crys-
tals sometimes lie. This coarser pumice-stone sand is differ-
ent from the very finely granular sand which, near the rock

1736: according to DufrenDy, Mdmoires pour servir a une Description
Geologique de la France, t. iv., p. 272. All the genetic questions are
discussed very completely and with praiseworthy impartiality in the
9th edition of Sir Charles Lyell's Principles of Geology, 1853, p. 369.
Even Bouguer {Figwe de la Terre, 1749, p. Ixvi.) was not disinclined
to the idea of the upheaval of the volcano of Pichincha. He says :
"It is not impossible that the rock, which is burned and black, may
have been elevated by the action of subterranean fire." S§e also
p. xci.

* Zeitschrljl filr Allgemeine Erdkunde, bd, iv., s. 398.

t For the more certain determination of the minerals of which the
Mexican volcanoes . are composed, old and recent collections made by
myself and Pieschel have been compared.

306 COSMOS.

El Frayle and at the limit of perpetual snow, on the volcano
Popocatepetl, renders the ascent so dangerous, because, when
it is set in motion on steep declivities, the sand-mass, rolling
down, threatens to overwhelm every thing. Whether this
lava-field of fragments (in Spanish Malpais, in Sicily Sciarra
viva, in Iceland Odaada-IIraun) is due to ancient lateral erup-
tions of Popocatepetl, situated one above the other, or to the
somewhat rounded cone of Tetlijolo (Cerro del Corazon de
Piedra), I can not determine. It is also geognostically re-
markable that, farther to the east, on the road toward the
small fortress Perote, the ancient Aztec Pinahuizapan, between
Ojo de Agua, Venta de Soto, and El Portachuelo, the vol-
canic formation of coarsely fibrous, white, friable perlite*
rises beside a limestone (Marmol de la Puebla) which is
probably tertiary. This perlite is very similar to that of the
conical hill of Zinapecuaro (between Mexico and Valladolid),
and contains, besides laminae of mica and lumps of immersed
obsidian, a glassy, bluish-gray, or sometimes red, jasper-like
streaking. The wide " perlite district" is here covered with
a finely granular sand of weathered perlite, which might be.
taken at the first glance for granitic sand, and which, not-
withstanding its allied origin, is still easily distinguishable
from the true grayish-white pumice-stone sand. The latter
is more proper to the immediate vicinity of Perote — the pla-
teau 7460 feet in height between th^wo volcanic chains of
Popocatepetl and Orizaba, which stUp north and south.

When, on the road from Mexico to Vera Cruz, we begin
to descend from the heights of the non-quartzose, trachytic
porphyry of the Vigas toward Canoas and Jalapa, we again
twice pass over fields of fragments and scoriaceous lava — the
first time between the station Parage de Garros and Canoas
or Tochtlacuaya, and the second between Canoas and the
station Casas de la Hoya. The first point is called Loma de
Tablas, on account of the numerous upraised basaltic blocks
of lava containing abundance of olivin ; the second simply JEJl
^laljjais. A small ridge of the same trachytic porphyry, full
of glassy feldspar, which forms the eastern limit of the Arenal
(the perlitic sand-fields), near La Cruz Blanca and Rio Frio
(on the western declivity of the heights of Las Vigas), sepa-
rates .the two branches of the lava-field which have just been

* The beautiful marble of La Puebla comes from the quarries of
Tecali, Totomehuacan, and Portachuelo, to the south of the high tra-
chytic mountain, El Pizarro. I have also seen limestone cropping out
near the terrace pyramid of Cholula, on the way to La Puebla.

TRUE VOLCANOES. 807

mentioned — the Loma de Tablas, and the much broader Mal-
pais. Those of the country people who are well acquainted
with the district assert that the band of scoriae is elongated
toward the south-southeast, and consequently toward the
Cofre de Perote. As I have myself ascended the Cofre and
made many measurements on it,* I have been but little in-

* The Cofre de Perote stands nearly isolated to tlie southeast of
the Fuerte or Castillo de Perote, near the eastern slope of the great
plateau of Mexico ; but its great mass belongs to an important range
of heights, which, forming the margin of the slope, extends in a north
and south direction, from Cruz Blanca and Rio Frio toward Las Vigas
(lat. 19° 37' 37") past the Cofre de Perote (lat. 19° 28' 57", long. 97°
T 20"), to the westward of Xicochimalco and Achilchotla, to the Peak
of Orizaba (lat. 19° 2' 17", long. 97° 13' 56"), parallel to the chain (Po-
pocatepetl— Iztaccihuatl) which separates the cauldron valley of the
Mexican lakes from the plain of La Puebla. (For the grounds of
these determinations, see my Recueil d'Observ. Astron., vol. ii., p. 529-
532 and 547, and also Analyse de t At las du Mexique, or Essai Politique
swr la Nouvelle Espagne, t. i., p. 55-60). As the Cofre has raised itself
abruptly in a field of pumice-stone many miles in width, it appeared
to me in my winter ascent (the thermometer fell at the summit, on
the 7th of February, 1804, to 28°-4) to be extremely interesting that
the covering of pumice-stone, the thickness and height of which I
measured barometrically at several points both in ascending and de-
scending, rose more than 780 feet. The lower limit of the pumice-
stone, in the plain between Perote and Rio Frio, is 1187 toises (7590
feet) above the level of the sea ; the upper limit, on the northern de-
clivity of the Cofre, 1309 toises (8370 feet) ; thence through the Pina-
huast, the Alto de los Caxones (1954 toises = 12,496 feet), where I
could determine the latitude by the sun's meridian altitude, up to the
summit itself, no trace of pumice-stone was to be seen. During the
upheaval of the mountain a portion of the coat of pumice-stone of the
great Arenal, which has probably been leveled in strata by water, was
carried up. I inserted a drawing of this zone of pumice-stone in my
journal (February, 1804) on the spot. It is the same important phe-
nomenon which was described by Leopold von Buch in the year 1834
on Vesuvius, where horizontal strata of pumice-tufa were raised by the
elevation of the volcano to a greater height indeed, 1900 or 2000 feet
toward the Hermitage del Salvatore {Poggendorff' s Annalen, bd.xxxvii.,
s. 175-179). The surface of the dioritic trachyte rock on the Cofre, at
the point where I found the highest pumice-stone, was not withdrawn
from observation by snow. The limit <5f perpetual snow lies in Mex-
ico, under the latitudes of 19° or 19^°, only at the average elevation
of 2310 toises (14,770 feet); and the summit of the Cofre, up to the
foot of the small, house-like cubical rock where I set up the instru-
ments, reaches 2098 toises, or 13,418 feet, above the sea level. Ac-
cording to angles of altitude the cubical rock is 21 toises, or 134 feet,
in height; consequently, the total altitude, which can not be reached
on account of the perpendicular wall of the rock, is 13,552 feet above
the sea. I found only single spots of sporadic snow, the lower limit of
which was 12,150 feet, about 700 or 800 feet below the upper limit of
forest trees, in beautiful pine-trees : Pinus occidentalism mixed with Cu-

308 COSMOS.

clined to conclude, from the prolongation of the lava stream,
which is certainly very probable (it is so* represented in my
Profiles, tab. 9 and 1 1 , and in the Nivellement Barometrique\
that it may have flowed from this mountain, the form of
which is so remarkable. The Cofre de Perote, which is
nearly 1400 feet higher than the Peak of TeneriiFe, but in-
considerable in comparison with the giants Popocatepetl and
Orizaba, forms, like Pichincha, a long rocky ridge, upon the
southern extremity of which stands the small cubical rock
(La Pena), the form of which gave origin to the ancient Az-
tec name of Nauhcampatepetl. In ascending the mountain
I saw no trace of the falling in of a crater, or of eruptive or-
ifices on its declivities ; no masses of scoriae, and no obsidi-
ans, perlites, or pumice-stones belonging to it. The black-
ish-gray rock is very uniformly composed of much hornblende
and a species of feldspar, which is not glassy feldspar (sani-
dine) but oligoclase ; this would show the entire rock, which
is not porous, to be a dioritic trachyte. I describe the im-
pressions which I experienced. If the terrible, black lava-
field — Malpais — (upon which I have here purposely dwelt
in order to counteract the too one-sided Consideration of ex-
ertions of volcanic force from the interior) did not flow from
the Cofre de Perote itself at a lateral opening, still the up-

pressus sahinoides and Arbutus MadroTio. The oak, Quercus xalapensis,
had accompanied us only to an absohite elevation of 10,340 feet.
(Humboldt, Nivellement baromet?: des Cordilleres, Nos. 414-429.) The
name of Nauhcampatepetl, which the mountain bears in the Mexican
language, is derived from its peculiar form, which also induced the
Spaniards to give it the name of Cofre. It signifies "quadrangular
mountain" for nauhcampa, formed from nahuii, the numeral four, signi-
fies as an SL&serh, from four sides, but as an adjective (although the Dic-
tionaries do not state this), undoubtedly, quadrangular or four-sided, as
this signification is attached to the compound nauhcampa ixquich. An
observer very well acquainted with the country, M. Pieschel, supposes
the existence of an old crater-opening on the eastern declivity of the
Cofre de Perote {Zeitschrift fur Allgem. Erdkunde, herausg. von Gum-
precht, bd. v., s. 125). I drew the view of the Cofre, given in my Vues
des Cordilleres, pi. xxxiv., in the vicinity of the castle of San Carlos de
Perote, at a distance of about eight miles. The ancient Aztec name
of Perote was Pinahuizapan, and signifies (according to Buschmann)
the beetle pinahuiztU (regarded as an evil omen, and employed super-
stitiously in fortune-telling : see Sahagun, Hisioria Gen. de las Cosas
de Nueva Espana, t. ii., 1829, p. 10-11) on the water ; the name of this
beetle is derived from pinahua, to be ashamed. From the same verb
is derived the above-mentioned local name Pinahuast (pinahuaztli) of
this district, as well as the name of a shrub (M.imosa.cesd?) pinaliuihuiz-
tli, translated herba verecunda by Hernandez, the leaves of which fall
down when touched.

TRUE VOLCANOES. 309

heaval of this isolated mountain, 13,553 feet in height, may-
have caused the formation of the Soma de Tablas. During
such an upheaval, longitudinal fissures and net-vi^orks of fis-
sures may be produced far and wide by folding of the soil,
and from these molten masses may have poured directly,
sometimes as dense masses, and sometimes as scoriaceous
lava, without any formation of true mountain platforms
(open cones or craters of elevation). Do we not seek in vain
in the great mountains of basalt and porphyritic slate for
central points (crater mountains), or lower, circumvallated,
circular chasms, to^'hich their common production might be
ascribed? The eyeful separation of that which is genet-
ically different in phenomena — the formation of conical
mountains with permanently open craters and lateral open-
ings ; of circumvallated craters of elevation and Maars ; df
upraised, closed, bell-shaped mountains or open cones, or
matters poured out from coalescent fissures — is a gain to sci-
ence. It is so because the multiplicity of opinions which is
necessarily called forth by an enlarged horizon of observa-
tion, and the strict critical comparison of that which exists
with that which is asserted to be the only mode of produc-
tion, are most powerful inducements to investigation. Even
upon European soil, however, on the island of Euboea, so rich
in hot springs, a vast lava stream has been poured out,*
within the historical period, from a chasm in the great plain
of Lelanton, at a distance from any mountain.

In the volcanic group of Central America, which follows
the Mexican group toward the south, and in which eighteen
conical and bell-shaped mountains may be regarded as still
active, four (Nindiri, El Nuevo, Conseguina, and San Miguel
de Bosotlan) have been recognized as producing lava.f The
mountains of the third volcanic group, that of Popayan and
Quito, have already for more than a century enjoyed the rep-
utation of furnishing no lava streams, but only incoherent,
glowing scoriaceous masses, thrown out of the single sum-
mital crater, and often rolling down in a linear arrangement.
This was even the opinion^ of La Condamine, when he left

* Strabo, lib. i., p. 58 ; lib. vi., p. 269, ed. Casaubon ; Cosmos, vol. i.,
p. 237, and vol. v., p. 215.

f See page 263.

I "I have never known," says La Condamine, "lava-like matter in
America, although M. Bouguer and myself have encamped for whole
weeks and months upon the volcanoes, and especially upon those of
Picbincha, Cotopaxi, and Chimborazo. Upon these mountains I have
only seen traces of calcination, without liquefaction. Nevertheless, the

310 COSMOS.

the highlands of Quito and Cuenga in the spring of 1743.
Fourteen years afterward^vhen he returned from an ascent
of Vesuvius (4th of June, 1755), in which he accompanied
the. sister of Frederick the Great, the Margravine of Bai-
reuth, he had the opportunity of expressing himself warmly,
in a meeting of the French Academy, upon the want of true
lava streams (laves coulees par torrens de matieres liquejiees)
from tlie volcanoes of Quito. The Journal cTun Voyage en
Italie^ which was read at the meeting of the 20th of April,
1757, only appeared in 1762 in the Alemoires of the Acade-
my of Paris, and is of some geognostic iraDortance in the his-
tory of the recognition of old extinct vmcanoes in France,
because in this journal, La Condamine, with his peculiar
acuteness, and without knowing of the certainly earlier ob-
servations of Guettard,* expresses himself very decidedly
upon the existence of ancient crater lakes and extinct volca-

kind of blackish crystal, commonly called Piedra de Gallinago in Peru
(obsidian), of which I have brought home several fragments, and of
which a polished lens of seven or eight inches in diameter may be seen
in the cabinet of the Jai-din du Hoi, is nothing but a glass formed by
volcanic action. The materials of the stream of fire which flows con-
tinually from that of Sangai, in the province of Macas, to the south-
east of Quito, are no doubt lava, but we have only seen this mountain
from a distance, and I was no longer at Quito at the time of the last
eruptions of the volcano of Cotopaxi, when vents opened upon its flanks,
from which ignited and liquid matters were seen to issue in streams,
which must have been of a similar nature to the lava of Vesuvius" (La
Condamine, Journal de Voyage en Italie, in the Mevioires de PA cad. des
Sciences, 1757, p. 357, Historie, p. 12). The two examples, especially
the first, are not happily chosen. The Sangay was first scientifically
examined, in December of the year 1849, by Sebastian Wisse ; what
La Condamine, at a distance of 108 miles, took for luminous lava
flowing down, and "an effusion of burning sulphur and bitumen," con-
sists of red-hot stones and scoriaceous masses, which sometimes, press-
ed closely together, slip down on the steep declivities of the cone of
ashes {Cosmos, see above, p. 251). On Cotopaxi, as on Tungurahua,
Chimborazo, and Pichincha, or on Purace, and Sotara near Popayan,
I have seen nothing that could be looked upon as narrow lava streams,
which had flowed from these colossal mountains. The incoherent,
glowing masses of 5 — 6 feet in diameter, often containing obsidian,
which Cotopaxi has scattered abroad during its eruptions, impelled by
floods of melting snow and ice, have reached far into the plain, where
they form rows partially diverging in a radiate form. La Condamine
also says very truly elsewhere {Jom-nal du Voyage a VEquateur, p. 160) :
"These fragments of rock, as large as the hut of an Indian, form se-
ries of rays,"which start from the volcano as from a common centre."
* Guettard's memoir on the extinct volcanoes was read at the Acad-
emy in 1 752, consequently three years before La Condamine's journey
into Italy; but only printed in 1756, consequently during the Italian
travels of the astronomer.

TRUE VOLCANOES. 311

noes in Middle and Northern Italy, and in the south of
France.

This remarkable contrast between the narrow and un-
doubted lava streams of Auvergne thus early recognized,
and the often too absolutely asserted absence of any effusion
of lava in the Cordilleras, occupied me seriously during the
whole period of my expedition. All my journals are full of
considerations upon this problem, the solution of which I
long sought in the absolute elevation of the summits and in
the vastness of the circumvallation, that is to say, the sink-
ing of trachytic conical mountains from mountain plains of
eight or nine thousand (8500 — 9600 English) feet in eleva-
tion, and of great breadth. We now know, however, that a
volcano of Quito, 17,000 feet in height, which throws out
scori£e (that of Macas), is uninterruptedly much more active
than the low volcanoes Izalco and Stromboli ; we know that
the eastern dome-shaped and conical mountains, Antisana
and Sangay, have free slopes toward the plains of the Napo
and Pastaza ; and the western ones, Pichincha, Iliniza, and
Chimborazo, toward the affluents of the Pacific Ocean. In
many, also, the upper part projects without circumvallation
eight or nine thousand feet above the elevated plateaux.
Moreover, all these elevations above the sea-level, which is
regarded, although not quite correctly, as the mean elevation
of the earth's surface, are certainly inconsiderable as com-
pared with the depth which we may assume to be that of
the seat of volcanic activity, and of the necessary tempera-
ture for the fusion of rock-masses.

The only phenomena resembling narrow lava eruptions
which I discovered in the Cordilleras of Quito are those
presented by the colossal mountain Antisana, the height of
which I determined to be 19,137 feet (5833 metres) by a
trigonometrical measurement. As the structure furnishes
the most important criterion here, I will avoid the system-
atic denomination lava^ which confines the idea of the mode
of production within too narrow limits, and make use, but
quite provisionally, of the names ^^ rock-debris'" {Felstrummern)
or " detritus dikes " (Schuttiuiillen, trainees de masses volcan-
iques). The mighty mountain of Antisana, at an elevation
of 13,458 feet, forms a nearly oval plain, more than 12,500
toises (79,950 feet) in long diameter, from which the portion
of the mountain covered with perpetual snow rises like an
island. The highest summit is rounded off and dome-shaped.
The dome is united by a short, jagged ridge, with a truncat-

312

COSMOS.

ed cone lying toward the north. In the plateau, partly des-
ert and sandy, partly covered with grass (the dwelling-place
of a very spirited race of cattle, which, owing to the slight
atmospheric pressure, easily expel blood from the mouth and
nostrils when excited to any great muscular exertion), is sit-
uated a small farm (hacienda), a single house in which we
passed four days in a temperature varying between 38°*6.
and 48° '2. The great plain, which is by no means cir-
cumvallated as in craters of elevation, bears the traces of an
ancient sea-bottom. The Laguna Mica, to the westward of
the Altos de la Moya, is to be regarded as the residue of the
old covering of water. At the margin of the limit of per-
petual snow the Rio Tinajillas bursts forth, subsequently,
under the name of Rio de Quixos, becoming a tributary of
the Maspa, the Napo, and the Amazon. Two narrow, wall-
like dikes or elevations, which I have indicated upon the
plan of Antisana, drawn by me, as coulees de laves, and which
are called by the natives Volcan de la Hacienda and Yana
Volcan {Yayia signifies black or brown in the Qquechhua
language), pass like bands from the foot of the volcano at
the lower margin of the perpetual snow-line, and extend, ap-
parently with a very moderate declivity, in a direction N.E.
— S.W., for more than 2000 toises (12,-792 feet) into the
plain. With very little breadth they have probably an ele-
vation of 192 to 213 feet above the soil of the Llanos de la
Hacienda, de Santa Lucia, and del Cuvillan. Their decliv-
ities are every where very rugged and steep, even at the ex-
tremities. In their present state they consist of conchoidal
and usually sharp-edged fragments of a black basaltic rock,
without olivin or hornblende, but containing a few small
white crystals of feldspar. The fundamental mass has fi-e-
quently a lustre like that of pitch-stone, and contains an ad-
mixture of obsidian, which was especially recognizable in
very large quantity, and more distinctly in the so-called
Cueva de Antisana, the elevation of which we found to be
15,942 feet. This is not a true cavern, but a shed formed
by blocks of rock which had fallen against and mutually
supported each other, and which preserved the mountain
cowherds and also ourselves during a fearful hail-storm.
The Cueva lies somewhat to the north of the Yolcan de la
Hacienda. In the two narrow dikes, which have the ap-
pearance of cooled lava streams, the tables and blocks ap-
pear in part inflated like cinders, or even spongy at the
edges, and in part weathered and mixed with earthy detritus.

TRUE VOLCANOES. 313

Analogous but more complicated phenomena are presented
by another also band-like mass of rocks. On the eastern
declivity of the Antisana, probably about 1280 feet perpen-
dicularly below the plain of the hacienda, in the direction
of Pinantura and Pintac, there lie two small round lakes,
of which the more northern is called Ansango, and the
southern Lecheyacu. The former has an insular rock, and
is surrounded by rolled pumice-stone, a very important point.
Each of these lakes marks the commencement of a valley ;
the two valleys unite, and their enlarged continuation bears
the name of Yolcan de Ansango, because from the margins of
the two lakes narrow lines of rock debris, exactly like the
two dikes of the plateau which we have described above, do
not, indeed, fill up the valley, but rise in its midst like dams
to a height of 213 and 266 feet. A glance at the local plan
which I published in the *' Geographical and Physical At-
las " of my American travels (pi. 26), will illustrate these
conditions. The blocks are again partly sharp-edged, and
partly scorified and even burned like coke at the edges. It is
a basaltic, black, fundamental mass, with sparingly scattered
glassy feldspar ; some fragments are blackish-brown, and of
a dull, pitch-stone-like lustre. Basaltic as the fundamental
mass appears, however, it is entirely destitute of the olivin
which occurs so abundantly on the Rio Pisque and near
Guallabamba, where I saw basaltic columns of 72 feet in
height and 3 feet thick, which contained both olivin and
hornblende scattered in them. In the dike of Ans'ango nu-
merous tablets, cleft by weathering, indicate porphyritic
slates. All the blocks have a yellowish-gray crust from
weathering. As the detritus ridge (called los derrumbami-
entos, la reventazon, by the natives, who speak Spanish) may
be traced from the Rio del Molina, not far from the farm
of Pintac, up to the small crater-lakes surrounded by pum-
ice-stone (chasms filled with water), the opinion has grown
up naturally, and, as it were, of itself, that the lakes are the
openings from which the blocks of stone came to the surface.
A few years before my visiting the district, the ridge of frag-
ments was in motion for weeks upon the inclined surface,
without any perceptible previous earthquake, and some
houses near Pintac were destroyed by the pressure and shock
of the blocks of stone. The detritus ridge of Ansango is
still without any trace of vegetation, which is found, al-
though very sparingly, upon the two more weathered and
certainly older eruptions of the plateau of Antisana.

Vol. v.— O

314 ' COSMOS.

How is this mode of manifestation of volcanic activity, the
action of which I am describing, to be denominated ?* Have
we here to do with lava streams '? or only with semi-scorified
and ignited masses, which are thrown out unconnected, but
in chains pressed closely upon each other (as on Cotapaxi in
very recent times) ? Have the dikes of Yana Volcan and
Ansango been, perhaps, merely solid fragmentary masses,
which burst forth without any fresh elevation of temperature
from the interio;* of a volcanic conical mountain, in which
they lay loosely accumulated, and therefore badly supported,
their movement being caused by the concussion of an earth-
quake, impelled by shocks or falls, and giving rise to small
local earthquakes'? Is no one of the three manifestations of
volcanic activity here indicated, different as they are, appli-
cable in this case 1 and have the linear accumulations of rock
detritus been upheaved upon fissures in the spots where they
now lie (at the foot and in the vicinity of a volcano) ? The
two dikes of fragments in this so slightly inclined plateau,
called Volcan de la Hacienda and Yana Yolcan, which I
once considered, although only conjecturally, as cooled lava
streams, now appear to me, as far as I can remember, to
present but little in support of the latter opinion. In the
Volcan de Ansango, where the line of fragments may be
traced without interruption, like a river-bed, to the pumice
margins of two small lakes, the fall, or difference of level be-
tween Pinantura 1482 toises (9476 feet), andLecheyacu 1900
toises (12,150 feet), in a distance of about 7700 toises (49,239
feet), by no means contradicts what we now believe we know
of the small average angles of inclination of lava streams.
From the difference of level of 418 toises C2674 feet), there
is an inclination of 3° 6^. A partial elevation of the soil in
the middle of the floor of the valley would not appear to be
any hinderance, because the back swell of fluid masses im-
pelled up valleys has been observed elsewhere ; for example,
in the eruption of Scaptar Jokul in Iceland, in 1783 (Nau-
mann, Geognosie, bd. i., s. 160).

The word lava indicates no peculiar mineral composition
of the rock ; and when Leopold von Buch says that every

* " There are few volcanoes in the chain of the Andes," says Leo-
pold von Buch, "which have presented streams of lava, and none
have ever been seen around the volcanoes of Quito. Antisana, upon
the eastern chain of the Andes, is the only volcano of Quito upon
which M. de Humboldt saw, near the summit, something analogous to
a stream of lava ; this stream was exactly like obsidian" {Vescr. dis
lies Canaries, 1836, p. 468 and 488).

TRUE VOLCANOES. 315

thing is lava that flows in the volcano and attains new posi-
tions by its fluidity, I add that that which has not. again be-
come fluid, but is contained in the interior of a volcanic cone,
may change its position. Even in the first description* of
my attempt to ascend the summit of Chimborazo (only pub-
lished in 1837, in Schumacher's Astronomische Jahrhuch), I
expressed this opinion in speaking of the remarkable "frag-
ments of augitic porphyry which I collected on the 23rd of
June, 1802, in loose pieces of from twelve to fourteen inches
diameter, upon the narrow ridge of rock leading to the sum-
mit at an elevation of 19,000 feet. They had small, shining
cells, and were porous and of a red color. The blackest of
them are sometimes light like pumice-stone, and as though
freshly altered by fire. They have not, however, flowed out
in streams like. lava, but \\iXNG, probably been expelled at fis-
sures on the declivity of the previously upheaved, bell-shaped
mountain." This genetic explanation might find abundant
support in the assumptions of Boussingault, who regards the
volcanic cones themselves " as an accumulation of angular
trachytic fragments, upheaved in a solid condition, and heap-
ed up without any order. As after the upheaval the broken
rocky masses occupy a greater space than before they were
shattered, great cavities remain among them, movement be-
ing produced by pressure and shock (the action of the volcan-
ic vapor force being abstracted)." I am far from doubting
the partial occurrence of such fragments and cavities, which
become filled with water in the Nevados, although the beau-
tiful, regular, and, for the most, perfectly perpendicular tra-
chytic columns of the Pico de los Ladrillos, and Tablahuma
on Pichindha, and, above all, over the small basin Yana-
Cocha on Chimborazo, appear to me to have been formed on
the spot. My old and valued friend, Boussingault, whose
chemico-geognostic and meteorological opinions I am always
ready to adopt, regards what is called the Yolcan de Ansan-
go, and what now appears to me as an eruption of fragments
from two small lateral craters (on the western Antisana, be-
low Chussulongo), as upheavals of blocksj upon long fissures.

* Humboldt, Kkinere Schriften, bd. i., s. 161.

f "We differ entirely with regard to the pretended stream of An-
tisana toward Pinantura. I regard this stream (coulee) as a recent
upheaval analogous to those of Calpi (Yana Urcu), Pisque, and Jorul-
lo. The trachytic fragments have acquired a greater thickness toward
the middle of the stream. Their stratum is thicker toward Pinantura
than at points nearer Antisana. The fragmentary condition is an ef-
fect of local upheaval, and in the Cordillera of the Andes earthquakes

316 COSMOS.

As he has acutely investigated this region thirty years after
myself, he insists upon the analogy which appears to him to
be presented by the geognostic relations of the eruption of
Ansango to Antisana, and those of Yana Urcu (of which I
made a particular plan) to Chimborazo. I was the less in-
clined to beheve in a direct upheaval upon fissures through-
out the entire linear extent of the tract of fragments at An-
sango, because this, as I have already repeatedly mentioned,
leads, at its upper extremity, to the two chasms now filled
with water. Non-fragmentary, wall-like upheavals of great
length and uniform direction are, however, not unknown to
me, as I have seen and described them in our hemisphere,
in Chinese Mongolia, in granite banks with a floetz-like bed-
ding.*

Antisana had an eruptionf in the year 1580, and another
in the beginning of the last century, probably in 1728. Near
the summit of the north-northeast side, we observe a black
mass of rock, upon which even freshly-fallen snow does not
adhere. At this point a black column of smoke was seen
ascending for several days in the spring of 1801, at a time

may often be produced by heaping up" (letter from M. Boussingault,
dated August, 1834). See p. 256. In the description of his ascent of
Chimborazo (December, 1831), Boussingaulfsays, "The mass of the
mountain consists, in my opinion, of a heap of trachytic ruins piled
up on each other without any order. These trachytic fragments of a
volcano, which are often of enormous size, are upheaved in the solid
state ; their edges are sharp, and nothing indicates that they had been
in a fused or even a softened condition. Nowhere, on any of the equa-
torial volcanoes, do we observe any thing that would allow us to infer
a lava stream. Nothing has ever been thrown out from these craters
Qxcept masses of mud, elastic fluids, and ignited, more or less scorified
trachytic blocks, which have frequently been scattered to considerable
distances" CHumboldt, Kkinere Schrijien, bd. i., s. 200). With regard
to the first origin of the opinion of the upheaval of solid masses in the
form of heaped-up blocks, see Acosta in the Viajes a los Andes Ecua-
tonales par M. Boussingault^ 1849, p. 222, 223. The movement of the
heaped-up fragments, induced by earth-shocks and other causes, and
gradual filling up of the interstices, may, according to the assumption
of the celebrated traveler, produce a gradual sinking of volcanic mount-
ain peaks.

* Humboldt, Asie Centrale, t. ii., p. 296-301 (Gustav Rose, Mineral-
geognostische Reise nach dent Ural, deni Altai und dem Kasp. Mecre, bd.
i., s. 599). Narrow, much elongated granitic walls may have risen,
during the earliest foldings of the earth's crust, over fissures analogous
to the remarkable, still open ones, which are found at the foot of the
volcano of Pichincha ; as the Guaycos of the city of Quito, of 30 — 40
feet in width (see my Kleinere Schri/ten, bd. i., s. 24).

t La Condamine, Mesure des trois premiers Degree du Meridien dans
IHemisphere Austral, 1751, p. 56.

TRUE VOLCANOES. 317

when the summit was on all sides perfectly free from clouds.
6n the 16th of March, 1802, Bonpland, Carlos Montafar,
and myself reached a ridge of rock covered with pumice-
stone, and black, basaltic scoriae in the region of perpetual
snow, at an elevation of 2837 toises (18,142 feet), and con-
sequently, 2358 feet higher than Mont Blanc. The snow
was firm enough to bear us on many points near the ridge of
rock, which is so rare under the tropics (temperature of the
atmosphere 28°-8 — 34° -5). On the southern declivity, which
we did not ascend, at the Piedro de Azufre, where scales of
rocks sometimes separate of themselves by weathering, masses
of pure sulphur, of 10 — 12 feet in length and 2 feet in thick-
ness, are found ; sulphurous springs are wanting in the vi-
cinity.

Although in the eastern Cordillera the volcano of Anti-
sana, and especially its western declivity (from Ansango and
Pinantura, toward the village of Pedregal), is separated from
Cotopaxi by the extinct volcano of Passuchoa* with its wide-
ly distinguishable crater (La Peila), by the Nevado Sinchula-
hua and by the lower Ruminaui, there is still a certain re-
semblance between the rocks of the two giants. From Quin-
che onward the whole eastern chain of the Andes has pro-
duced obsidian, and yet El Quinche, Antisana, and Passuchoa
belong to the basin in which the city of Quito is situated ;
while Cotopaxi bounds another basin, that of Lactacunga,

* Passuchoa, separated by the farm El Tambillo from the Atacazo,
does not any more than tlie latter attain the region of perpetual snow.
The elevated margin of the crater, La Peila, has fallen in toward the
west, but projects toward the east like an amphitheatre. The tradi-
tion runs that at the end of the 16th century the Passuchoa, which
had previously been active, ceased its manifestations of activity on
the occasion of an eruption of Pichincha, which proves the communi-
cation between the vents of the opposite eastern and western Cordil-
leras. The true basin of Quito, closed*like a dam — on the north by
.a mountain group between Cotocachi and Imbaburo, and on the. south
by the Altos de Chisinche (between 0° 20' N. and 0° 41' S.), is for the
most part divided longitudinally by the mountain ranges of Ichimbio
and Poingasi. To the eastward lies the valley of Puembo and Chillo ;
to the westward the plain of Inaquito and Turubamba. In the eastern
Cordillera follow from north to south — Imbaburo, the Faldas de Gua-
mani, and Antisana, Sinchulahua, and the perpendicular black wall,
crowned with turret-like points, of Ruminaui (Stone-eye) ; in the
western Cordillera, Cotocachi, Casitagua, Pichincha, Atacazo, and Co-
razon, upon the slopes of which blooms the splendid Alpine plant, the
red Ranunculus Gusmani. This has appeared to me to be the place to
give, in brief terms, a moi'phological representation, drawn from my
own experience, of the form of a spot which is so important and classic-
al in respect to volcanic geology.

318 COSMOS.

Ham bate, and Eiobamba. The small knot of mountains oCAe
Altos of Chisinche separates the two basins like a dam ; aM,
what is remarkable enough, considering its sraallness, the
waters of the northern slope of Chisinche pass by the llios
de San Pedro, de Pito, and de Guailabamba into the Pacific,
while those of the southern declivity flow through the Rio
Alaques and the Rio de San Felipe into the Amazons and
Atlantic Ocean. The union of the Cordilleras by mountain
knots and dikes (sometimes low, like the Altos just mention-
ed ; sometimes equal to Mont Blanc in height, as on the road
over the Paso del Assuay) appears to be a more recent and
also a less important phenomenon than the upheaval of the
divided parallel mountain chain itself As Cotopaxi, the
greatest of the volcanoes of Quito, presents much analogy in
its trachytic rock with the Antisana, so also we again meet
with the rows of blocks (lines of fragments) which have al-
ready occupied us so long, even in greater number upon the
slopes of Cotopaxi.

It was especially our business, when traveling, to trace
these rows to their origin, or rather to the point where they
are concealed beneath the perpetual covering of snow. We
ascended upon the southwestern declivity of the volcano from
Mulalo (Mulahalo), along the Rio Alaques, which is formed
of the Rio de los 13anos, and the Rio Barrancas, up to Pan-
sache (12,006 feet), where we inhabited the spacious Casa
del Paramo in the grassy plain (El Pajonal). Although up
to this time much snow had fallen at night, we nevertheless
got to the eastward of the celebrated Cabeza del Inga, first
into the Quebrada and Reventazon de las Minas, and after-
ward still farther to the east, over the Alto do Suniguaicu,
to the chasm of the Lion Mountain (Puma-Urcu), where the
barometer only showed an elevation of 2263 toises, or 14,471
feet. Another line of figments, which, however we only
saw from a distance, has moved from the eastern part of the'
snow-clad,^sh-cone toward the Rio Negro (an affluent of the
Amazon) and Valle vicioso. It is uncertain whether these
blocks were all thrown out of the crater at the summit to a
great height in the air, as glowing, scoriaceous masses fused
only at the edges (some angular, some rounded, of six or
eight feet in diameter, rarely conchoidal like those of Anti-
sana), falling on the declivity of Cotopaxi, and hastened in
their movement by the rush of the melted snow-water; or
whether, without passing through the air they were forced out
through lateral fissures of the volcano, as the word reventa-

TRUE VOLCANOES. 319

:son would indicate. Soon returning from Suniguaicu and
the Quebrada del Mestizo, we examined the long and broad
ridge which, striking from N.W. to S.E., unites Cotopaxi
with the Nevado de Quelendana. Here the blocks arranged
in rows are wanting, and the whole appears to be a dam-
like upheaval, upon the ridge of which are situated the small
conical mountain El Morro, and, nearer to the horse-shoe
shaped Quelendana, several marshes and two small lakes
(Lagunas de Yauricocha and de Verdecocha). The rock of
El Morro and of the entire linear volcanic upheaval was
greenish-gray porphyritic slate, separated into layers of eight
inches thick, which dipped very regularly toward the east at
60°. Nowhere was there any trace of true lava streams.*

* It is particularly remarkable diat the vast volcano of Cotopaxi,
which manifests an enormous activity, although, indeed, usually only
after long periods, and acts destructively upon the neighborhood, es-
pecially by the inundations which it produces, exhibits no visible va-
pors between its periodical eruptions, when seen either in the plateau
of Lactacunga or from the Paramo de Pansache. From several com-
parisons with other colossal volcanoes, such a phenomenon is certainly
not to be explained from its height of 19,180 feet, and the great tenuity
of the strata of air and vapor corresponding with this elevation. No
other Nevado of the equatorial Cordilleras shows itself so often free
from clouds and in such great beauty as the truncated cone of Coto-
paxi, that is to say, the portion which rises above the limit of perpet-
ual snow. The uninterrupted regularity of this ash-cone is much
greater than that of the ash-cone of the Peak of TenerifFe, on which a
narrow projecting rib of obsidian runs down like a wall. Only the up-
per part of the Tungurahua is said formerly to have been distinguished
in an almost equal degree by the regularity of its form ; but the ter-
rible earthquake of the 4th of February, 1797, called the Catastrophe of
Hiobamba, has deformed the mountain cone of Tungurahua by fissures
and the falling in of parts and the descent of loosened wooded frag-
ments, as also by the accumulation of debris. At Cotopaxi, as even
Bouguer observed, the snow is mixed in particular spots with crumbs
of pumice-stone, when it forms a nearly solid mass. A slight ine-
quality in the mantle of snow is visible toward the northwest, where
two fissure-like valleys run down. Black rocky ridges ascending to
the summit are seen nowhere from afar, although in the eruptions of
the 24th of June and 9th of December, 1742, a lateral opening showed
itself half way up the snow-covered ash-cone. "There opened," says
Bouguer {Figure de la Terre, p. Ixviii. ; see also La Condamine, Jovr-
nal du Voyage a VEqnateur, p. 159), "a new mouth toward the middle
of the part constantly covered with snow, while the flame always is-
sued at the top of the truncated cone." Quite at the top, close to the
summit, some horizontal black streaks, parallel to each other, but in-
terrupted, are detected. When examined with the telescope under
various illuminations they appeared to me to be rocky ridges. The
whole of this upper part is steeper, and almost close to the truncation
of the cone forms a Avall-like ring of unequal height, which, however,
is not visible at a great distance with the naked eye. My de'scriptioH

320 COSMOS.

In the island of Lipari, which abounds in pumice-stone, a
lava stream of pumice-stone and obsidian runs down to the

of this nearly perpendicular uppermost circumA^allation has already at-
tracted the particular attention of two distinguished geologists — Darwin
(^Volcanic Islands, 1844, p. 83), and Dana {Geology of the U. S. Explor-
ing Expedition, 1849, p. 356). The volcanoes of the Galapagos Islands,
Diana's Peak in St. Helena, TenerifFe, and Cotopaxi, present analo-
gous formations. The highest point which I determined by angles of
altitude in the trigonometrical measurement of Cotopaxi, was situated
in a black convexity. It is, perhaps, the inner wall of the higher and
more distant margin of the crater ; or is the freedom from snow of the
protruding rock caused at once by steepness and the heat of the crater?
In the autumn of the year 1 800 the whole upper part of the ash-cone
was seen to be luminous, although no eruption, or even emission of
visible vapors, followed. On the other hand, in the violent eruption
of Cotopaxi, on the 4th of January, 1803, when during my residence
on the Pacific coast the thundering noise of the volcano shook the
windows in the harbor of Guayaquil (at a distance of 148 geographical
miles), the ash-cone had entirely lost its snow, and presented a most
threatening appearance. Was such a heating ever observed before ?
Even very recently, as we learn from that admirable and courageous
female traveler, Ida Pfeiffer (^Meine zweite Weltreise, bd. iii., s. 170),
the Cotopaxi had, in the beginning of April, 1854, a violent eruption
of thick columns of smoke, " through which the fire wound itself like
flashing flames." May this luminous phenomenon have been a conse-
quence of the volcanic lightning excited Ly vaporization? The erup-
tions have been frequent since 1851.

The great regularity of the snow-covered truncated cone itself ren-
ders it the more remarkable that to the southwest of the summit there
is a small, grotesquely-notched, rocky mass with three or four points at
the lower limit of the region of perpetual snow, where the conical
form commences. The snow remains upon it only in small patches,
probably on account of its steepness. A glance at my representation
{Atlas Piitoresque du Voyage, pi. 10) shows its relation to the ash-cone
most distinctly. I approached nearest to this blackish-gray, probably
basaltic rocky mass, in the Quebrada and Reventazon de Minas. Al-
though this widely visible hill, of very strange appearance, has been
generally known for centuries in the whole province as the Cabeza del
Inga, two vpry different hypotheses, nevertheless, prevail with regard
to its origin among the colored aborigines (Indios): according to the
one, it IS merely asserted that the rock is the fallen summit of the vol-
cano, which formerly ended in a point, without any statement of the
date at which the occurrence took place ; according to the second hy-
pothesis, this is placed in the year (1533) in which the Inca Atahuallpa
was strangled in Caxamarca, and thus connected with the terrible fiery
eruption of Cotopaxi, described by Herrera, which took place in the
same year, and also with the obscure prophecy of Atahuallpa's father,
Huayna Capac, regarding the approaching fall of the Peruvian empire.
Is that which is common to both hypotheses — namely, the opinion that
this fragment of rock formerly constituted the apex of the cone — the
traditional echo, or obscure remembrance of an actual occurrence?
The aborigines, it may be said, in their uncultivated state, would
probably notice facts and preserve them in remembrance, but would

TRUE VOLCANOES. 321

north of Caneto, from the well-preserved, extinct crater of
the Monte di Campo Bianco toward the sea, in which the
fibres of the former substance run, singularly enough, parallel
to the direction of the stream.* The extended pumice quar-
ries, four miles and a half from Lactacunga, present, accord-
ing to my investigation of the local conditions, an analogy
with this occurrence on Lipari. These quarries, in which
the pumice-stone, divided into horizontal beds, has exactly
the appearance of a rock in position, excited even the aston-
ishment of Bouguer in 1737. f " On volcanic mountains,"

be unable to rise to geognostic combinations. I doubt the correctness
of this objection. The idea that a truncated cone, "in losing its
apex," may have thrown it oif unbrol(pp, as large blocks were thrown
out during subsequent eruptions, may present itself even to very un-
cultivated minds. The terraced pyramid of Cholula, a work of the
Tolteks, is truncated. The natives could not suppose that the pyra-
mid Avas not originally completed. They therefore invented the fable
that an aerolite, falling from heaven, destroyed the apex ; nay, por-
tions of the aerolite were shown to the Spanish conquerors. More-
over, how can we place the first eruption of the volcano of Cotopaxi
at a period when the ash-cone (the result of a series of eruptions) was
already in existence ? It seems probable to me that the Cabeza del
Inga was produced at the spot which it now occupies ; that it was up-
heaved there, like the Yana Urcu at the foot of Chimborazo, and like
the Moro on Cotopaxi itself, to the south of Suniguaica, and to the
northwest qf the small lake Yurak-cocha (in the Qquechhua language,
the White Lake).

With regard to the name of the Cotopaxi, I have stated in the first
volume of my Kleinere Schriften (s. 463) that only the first part of it
could be explained from the Qquechhua language, being the word
ccotto, heap or mass, but that pacsi was unknown. La Condamine
(p. 53) explains the whole name of the mountain, saying, *' in the lan-
guage of the Incas the name signifies shining mass.''' Buschmann,
however, remarks that in this case 7>acst is replaced by the word />acsa,
which is certainly quite different from it, and which signifies lustre^
brilliancy^ especially the mild lustre of the moon ; to express * ' shining
mass,^' moreover, in accordance with the spirit of the Qquechhua lan-
guage, the position of the two words would have to be reversed —
pacsaccotto.

* Fried. Hoffmann, in Poggendbrff's Annalen, bd. xxvi., 1832, s.48.

t Bouguer, Figure de la Terre, p. Ixviii. How often, since the earth-
quake of the 19th July, 1698, has the little town of Lactacunga been
destroyed and rebuilt with blocks of pumice-stone from the subterra-
nean quarries of Zumbalica ! According to historical documents com-
municated to me during my sojourn in the country, from copies of the
old ones which^ave been destroyed, and from more recent original
documents partially preserved in the archives of the town, the destruc-
tions occurred in the years 1703 and 1736, on the 9th of December,
1742, 30th of November, 1744, 22d of February, 1757, 10th of Febru-
ary, 1766, and 4th of April, 1768 — therefore seven times in 65 years!
In the year 1802 I found four fifths of the town still in'ruins in conse-

02

"322 COSMOS.

he says, " we only find simple fragments of pumice-stone of
a certain size ; but at seven leagues to the south of Cotopaxi,
in a point which corresponds with our tenth triangle, pum-
ice-stone forms entire rocks, ranged in parallel banks of five
to six feet in thickness in a space of more than a square
league. Its depth is not known. Imagine what a heat it
must have required to fuse this enormous mass, and in the
very spot where it now occurs ; for it is easily seen that it
has not been deranged, and that, it has cooled in tWb place
where it was liquefied. The inhabitants of the neighborhood
have profited by this immense quarry, for the small town of
Lactacunga, with some very pretty buildings, has been entire-
ly constructed of pumice-g|one since the earthquake which
overturned it in 1698."

The pumice quarries are situated near the Indian village
of San Felipe, in the hills of Guapulo and Zumbalica, which
are elevated 512 feet above the plateau and 9990 feet above
the sea level. The uppermost layers of pumice-stone are,
therefore, five or six hundred feet below tlie level of Mulalo,
the once beautiful villa of the Marquis of Maenza (at the foot
of Cotopaxi), also constructed of blocks of pumice-stone, but
now completely destroyed by frequent earthquakes. The sub-
terranean quarries are at unequal distances from the two act-
ive volcanoes, Tungurahua and Cotopaxi : 32 miles from the
former, and about half that distance from the latter. They
are reached by a gallery. The workmen assert that from the
horizontal solid layers, of which a few are surrounded by loamy
pumice fragments, quadrangular blocks of 20 feet, divided by
no transverse fissures, might be procured. The pumice-stone,
which is partly white and partly bluish-gray, consists of very
fine and long fibres, with a silky lustre. The parallel fibres
have sometimes a knotted appearance, and then exhibit a sin-
gular structure. ' The knots are formed by roundish particles
of finely porous pumice-stone, from 1 — 1^ line in breadth,
around which long fibres curve so as to inclgse them. Brown-
ish-black mica in small six-sided tables, white crystals of oli-
goclase, and black hornblende are sparingly scattered in it ;
on the other hand, the glassy feldspar, which elsewhere (Ca-
maldoli, near Naples) occurs in pumice-stone, is entirely want-
ing. The pumice-stone of Cotopaxi is very different from that
of the quarries of Zumbalica:* its fibres are short, not paral-

quence of the great earthquake of Riobamba on the 4th of February,
1797.

* This difference has also been recognized by the acute Abich
iUeher Natur und Zvsammenham; vnlkanischer Bildurnjen, 1841, s. 83).

TRUE VOLCANOES. 323

lei, but curved in a confused manner. Magnesia mica, how-
ever, is not peculiar to pumice-stone, for it is also found in the
fundamental mass of the trachyte* of Cotopaxi. At the more
southern volcano,' Tungurahua, pumice-stone appears to be
entirely wanting. There is no trace of obsidian in the vi-
cinity of the quarries of Zumbalica, but I have found black
obsidian with a conchoidal fracture in very large masses, im-
mersed in bluish-gray weathered perlite, among the blocks
thrown out from Cotopaxi, and lying near Mulalo. Of this
fragments are preserved in the Koyal Collection of Minerals
at Berhn. The pumice-stone quarries here described, at a
distance of sixteen %iiles from the foot of Cotopaxi, appear,
therefore, to judge from their mineralogical nature, to be quite
foreign to that mountain, and only to stand in the same rela-
tion to it which all the volcanoes of Pasto and Quito, occu-
pying many thousand square miles, present to the volcanic
focus of the equatorial Cordilleras. Have these pumice-stones
been the centre and interior of a proper crater of elevation,
the external wall of which has been destroyed in the numer-
^ ous convulsions which the surface of the earth has here un-
dergone? or have they been deposited here upon fissures in
apparent rest during the most ancient foldings of the earth's
crust ? For the assumption of aqueous sedimentary alluvia,
such as are often exhibited in volcanic tufaceous masses mix-
ed with remains of plants and shells, is attended with still
greater difficulties. Up

The same questions are suggested by the great mass of
pumice-stone, at a distance from all intumescent volcanic
platforms, which I found on the Rio Mayo, in the Cordillera

* The rock of Cotopaxi has essentially the same mineralogical com-
position as that of the nearest volcanoes, Antisana and Tungurahua.
It is a trachyte, composed of oligoclase and augite, and consequently
a Chimborazo rock : a proof of the identity of the same kind of vol-
canic mountain in masses in the opposite Cordilleras. In the speci-
mens collected by me in 1802, and by Boussingault in 1831, the funda-
mental mass is partly light or greenish gray, M^ith a pitch-stone-like
lustre and translucent at the edges; partly black, nearly resembling
basalt ; with large and small pores, which possess shining walls. The
inclosed oligoclase is distinctly limited ; sometimes in very brilliant
crystals, very distinctly striated on the cleavage planes ; sometimes in
small fragments, and difficult of detection. The intermixed augites
are brownish and blackish green, and of very variable siae. Dark
laminae of mica and black metallic grains of magnetic iron are rarely
and probably quite accidentally sprinkled through the mass. In the
pores of a mass containing much oligoclase there was some native sul-
phur, probably deposited by the all-penetrating sulphurous vapors.

324 COSMOS.

of Pasto, between Mamendoj and the Cerro del Pulpito, 36
miles from the active volcano of Pasto. Leopold von Buch
has also called attention to a similar perfectly isolated erup-
tion of pumice-stone described by Meyen; which, consisting
of bowlders, forms a hill of 320 feet in height, near the vil-
lage of ToUo, to the east of Valparaiso, in Chili. The vol-
cano Maypo, which upheaves Jurassic strata in its rise, is two
full days' journey from this eruption of pumice-stone.* The
Prussian embassador in Washington, Friedrich von Gerolt,
to whom we are indebted for the first colored geognostic map
of Mexico, also mentions "a subterranean quarry of pumice-
stone at Bauten," near Huichapa, 32 n|^les to the southeast
of Queretaro, at a distance from all volcanoes. f The geo-
logical explorer of the Caucasus, Abich, is inclined to believe,
from his own observation, that the vast eruption of pumice-
stone near the village Tschegem, in the little Kabarda, on the
northern declivity of the central chain of the Elburuz, is, as
an effect of fissure, much older than the elevation of the very
distant conical mountain just mentioned.

If, therefore, the volcanic activity of the earth, by radia-,
tion of heat into space during the diminution of its original
temperature, and in the contraction of the superior cooling
strata, produces fissures and wrinkles {fractures et rides), and
therefore simultaneous sinking of the upper and upheaval of
the lower pafts,| we must naturally regard, as the measure

* *^Re volcano of Maypo (S. lat. 34° 15'), which has never ejected
pumice-stone, is at a distance of two days' journey from the ridge of
Tollo, which is 320 feet in height, and entirely composed of pumice-
stone, inclosing vitreous feldspar, brown crystals of mica, and small
fragments of obsidian. It is, therefore, an (independent) isolated erup-
tion, quite at the foot of the Andes and close to the plain." Leop. de
Buch, Desc. Phys. des Iks Canaries, 1836, p. 470.

t Federico de Gerolt, Cartas Geognosticas de los Prindpales Distritos
Mineraks de Mexico, 1827, p. 5.

t On the solidification and formation of the cnists of the earth, see
Cosmos, vol. i., p. 172, 173. The experiments of Bischof, Charles De-
ville, and Delesse have thrown a new light upon the folding of the body
of the earth. See also the older, ingenious considerations of Babbage,
on the occasion of his thermic explanation of the problem presented
by the temple of Serapis to the north of Puzzuoli, in the Quarterly
Journal of the Geological Society of London, \o\. iii., 1847, p. 186 ;
Charles Deville, Sur la Diminution de Densite dans les Roches en pas^
sant de Vetat cristallin a Vetat vitreux, in the Comptss rendus de VAcad.
des Sciences, t. xx., ]845, p. 1453 ; Delesse, Sur les Effets de la Fusion,
t. XXV., 1847, p. 455 ; Louis Frapolli, Sur la Caracth'e Gcologique, in the
Bull, de la Soc. Geol de France, 2me. serie, t. iv., 1847, p. 627 ; and,
above all, Elie de Beaumont, in his important work, Notice sur les Sys-

TRUE VOLCANOES. 325

and evidence of this activity in the various regions of the
earth, the number of recognizable volcanic platforms (open,
conical, and dome-shaped mountains) upheaved upon fissures.
This enumeration has been repeatedly and often very imper-
fectly attempted : eruptive hills and solfataras, belonging to
one and the same system, have been referred to as distinct
volcanoes. The magnitude of the space in the interior of
continents which has liitherto remained closed to all scien-
tific investigation, has not been so great an obstacle to the
solidity of this work as is commonly supposed, as islands and
regions near the coast are generally the principal seat of
volcanoes. In a numerical investigation, which can not be
brought to a full conclusion in the present state of our knowl-
edge, much is already gained when we attain to a result which
is to be regarded as a lower limit, and when we can determ-
ine with great probability upon how many points the fluid
interior of our earth has remained in active communication
with the atmosphere within the historical period. Such an
activity usually manifests itself simultaneously in eruptions
from volcanic platforms (conical mounfains), in the increas-
ing heat and inflammability of thermal springs and naphtha
wells, and in the increased extent of circles of commotion,
phenomena which all stand in intimate connection and in
mutual dependence.* Here again, also, Leopold von Buch
has the great merit of having (in the supplements to the Phys-
ical Description of the Canary Islands) for the first time under-
taken to bring the volcanic system of the whole earth, after

thnes de Montagnes, 1852, t. iii. The following three sections deserve
the particular attention of geologists : Considerations sur les Soulcve-
vients dus a une diminution lente et progressive du volume de la Terre, p.
1330; Sur FEcrasement Transversal nomme refoulement />ar Saussure,
covime une des causes de l elevation des Chaines de Montagnes, p. 1317,
1333, and 1346; Sur la Contraction que les Roches fondues dprouvent en
cristalllsant, tendant des le commencement au refroidissement du Globe a
rendre sa masse interne plus petite que la capacitc de son enveloppe extcri-
eure, p. 1235.

* "The hot springs of Saragyn at the height of fully 5600 feet are
remarkable for the part played by the carbonic acid gas which trav-
erses them at the period of earthquakes. At this epoch the gas, like
the carbonated hydrogen of the peninsula of Apscheron, increases in
volume, and becomes heated, before and during the earthquakes in the
plain of Ardebil. In the peninsula of Apscheron the temperature rises
36°, until spontaneous inflammation occurs at the moment when and
the spot where an igneous eruption takes })lace, which is always prog-
nosticated by earthquakes in the provinces of Chemakhi and A| sche-
ron." Abich, in the Melanges Physiques et Chijniques, t. ii., 1855, p.
364-365 (see Cosmos, vol. v., p. 169).

326

COSMOS.

the fundamental distinction of Central and Linear Volcanoes,
under one cosmical point of view. My own more recent, and,
probably for this reason, more complete enumeration, under-
taken in accordance with principles which I have already in-
dicated (p. 233 and 257), and therefore excluding unopened
bell-shaped mountains and mere eruptive cones, gives, as the
probable lower numerical limit (nombre limite inferieur), a result
which differs considerably from all previous ones. It is an
attempt to indicate the volcanoes which have been active
within the historical period.

The question has been repeatedly raised whether in thosa
parts of the earth's surface in which the greatest number ot
volcanoes are crowded together, and the reaction of the inte-
rior of the earth upon the hard (solid) crust manifests the
most activity, the fused part may not lie nearer to the sur-
face ? Whatever be the course adopted to determine the av-
erage thickness of the solid crust of the earth in its maximum :
whether it be the purely mathematical one which is present-
ed by theoretical astronomy,* or tlie simpler course, found-
ed upon the law of the increase of heat with depth and the
temperature of fusion of rocks,! still the solution of this prob-

* "W". Hopkins, Researches on Physical Geology in the Phil. Transact,
for 1839, pt. ii., p. 311, for 1840, pt. i., p. 193, and for 1842, pt. i., p. 43 ;
also with regard to tlie necessary relations of stability of the external
surface ; Theory of Volcanoes in the British Association Report for 1847,
p. 45-49.

t Cosmos, vol. v., p. 38-40; Naumann, Geognosie, bd. i., p. 66-76;
Bischof, Wdrmelehre, s. 383 ; Lyell, Principles of Geology, 1853, p. 536
-547 and 562. In the very interesting and instructive work. Souvenirs
d'un Naturaliste, by A. de Quatrefages, 1854, t. ii., p. 469, the upper
limit of the fused liquid strata is brought up to the small depth of 20
kilometres "as most of the silicates fuse at 1231°." "This low esti-
mate," as Gustav Rose observes, "is founded in an error. The tem-
perature of 2372"^, which is given by Mitscherlich as the melting point
of granite ( Cosmos, vol. i., p. 25), is certainly the minimum that we can
admit. I have repeatedly had granite placed in the hottest parts of a
porcelain furnace, and it was always but imperfectly fused. The mica
alone fuses with the feldspar to form a vesicular glass ; the quartz be-
comes opaque, but does not fuse. This is the case with all rocks which
contain quartz ; and this means may even be made use of for the de-
tection of quartz in rocks, in which its quantity is so small that it can
not be discovered with the naked eye ; for example, in the syenite of
Plauen, and in the diorite which we brought in 1829 ft-om Alapajewsk,
in the Ural. All rocks which contain no quartz, or any other miner-
als so rich in silica as granite, such as basalt, for example, fuse more
readily than granite to form a perfect glass in the porcelain furnace ;
but not over the spirit lamp with a double current, which is neverthe-
less certainly capable of j)roducing a temperature of 1231°." In Bis-

TRUE VOLCANOES. 827

lem presents a great number of values which are at present
undetermined. Among, these we have to mention the influ-
ence of an enormous pressure upon fusibility ; the different
conduction of heat by heterogeneous rocks ; the remarkable
enfeebling of conductibility with a great increase of tempera-
ture, treated of by Edward Forbes ; the unequal depth of the
oceanic basin ; and the local accidents in the connection and
nature of the fissures which lead down to the fluid interior !
If the greater vicinity of the upper limit of the fluid interior
in particular regions of the earth may explain the frequency
of volcanoes and the greater multiplicity of communication
between the depths and the atmosphere, this vicinity again
may depend either upon the relative average differences of
elevation of the sea-bottom and the continents, or upon the
unequal perpendicular depth at which the surface of the molt-
en fluid mass occurs, in various geographical longitudes and
latitudes. But where does such a surface commence? Are
there not intermediate degrees between perfect solidity and
perfect mobility of the parts? — states of transition which
have frequently been referred to in the discussions relative to
the plasticity of some Plutonic and volcanic rocks which have
been elevated to the surface, and also with regard to the move-
ment of glaciers. Such intermediate states abstract them-
selves from mathematical considerations, just as much as the
condition of the so-called fluid interior under an enormous
pressure. If it be not even very probable that the tempera-
ture every where continues to increase with the depth in ar-
ithmetical progression, local intermediate disturbances may
also occur, for example, by subterranean basins (cavities in
the hard mass), which are from time to time partially filled
from below with fluid lava and vapors resting upon it.*
Even the immortal author of the Protogoca allows these cav.
ities to play a part in the theory of the diminishing central
heat : " Fostremo credibile est contrahentem se refrigeratione
crustam hullas reliquisse, ingentes pro rei magnitudine id est
sub vastis fornicibus cavitates."^ The more improbable it is

cliof's remarkable experiments on the fusion of a globule of basalt,
even this mineral appeared, from some hypothetical assumptions, to
require a temperature 264° higher than the melting point of copper.
{Wdrmekkre des Innern unsers Erdkorpers^ s. 473.)

* Cosmos, vol. v., p. 162. See also with regard to the unequal dis-
tribution of the icy soil, and the depth at which it commences, inde-
pendently of geographical latitude, the remarkable observations of
Captain Franklin, Erman, Kupffer, and especially of Middendorff {loc.
cit. sup., s. 42, 47 and 167).

t Leibnitz in the Protogoca ; § 4.

328 COSMOS.

that the thickness of the crust already solidified is the same
in all regions, the more important is the consideration of the
number and geographical position of the volcanoes which
have been open in historical periods. Such an examination
of the geography of volcanoes can only be perfected by fre-
quently-renewed attempts.

I. Europe.
jEtna,

Volcano in the Liparis,
Stromholi,
Ischia,
Vesuvius,
Santorin,
Lemnos,

All belong to the great basin of the Mediterranean, but to
its European and not to its African shores ; and all these
seven volcanoes are still, or have been, active in known his-
torical periods ; the burning mountain Mosychlos in Lemnos,
which Homer names the favorite seat of Hephaestos, was
only destroyed and sunk beneath the waves of the sea by
earthquakes, together with the island of Chryse, after the
time of the great Macedonian {Cosmos, vol. i., p. 246 ; Ukert,
Geogr. der Griechen und Romer, th. ii., abth. 1, s. 198). The
great upheaval of the three Kaimenes in the middle of the
Gulf of Santorin (partly inclosed by Thera, Therasia, and
Aspronisi), which has been repeated several times within
about 1900 years (from 186 B.C. to 1712 of our epoch), had
in their production and disappearance a remarkable similar-
ity with the relatively unimportant phenomenon of the tem-
porary formation of the islands which were called Graham,
Julia, and Ferdinandea, between Sciacca and Pantellaria.
Upon the peninsula of Methana, which has already been fre-
quently mentioned (Cosmos, vol. i., p. 240; vol. v., p. 218),
there are distinct traces of volcanic eruptions in the reddish-
brown trachyte which rises from the limestone near Kaime-
nochari and Kaimeno (Curtius, Pelop,, bd. ii., s. 439).

Of pre-historic volcanoes with fresh traces of the emission
of lava from craters there are, counting from north to south,
those of the Eifel {Mosenberg, Geroldstein), farthest to the
north ; the great crater of elevation in which Schemnitz is
situated; Auvergne (Chahie des Puys or of the Monts Domes

TRUE VOLCANOES. 329

le Cone du Cantal, les Monts-Dore) ; Vivarais, in which the an-
cient lavas have broken out from gneiss {Coupe d'y Asac, and
the cone of MontpezcCt) ; Velay : eruptions of scoriae from
which no lava issue ; the Euganean hills ; the Alban mount-
ains, llocca Monfina and Vultur, near Teano and Melfi ; the
extinct volcanoes about Olot and Castell Follit, in Catalo-
nia ;* the island group, Las Columhretes, near the coast of
Valencia (the sickle-shaped larger island Columbraria of the
Romans, upon which Montcolibre, latitude 39° 54^ accord-
ing to Captain Smyth, is full of obsidian and cellular tra-
chyte) ; the Greek island Nisyros, one of the Carpathian
Sporades, of a perfectly round form, in the middle of which,
at an elevation of 2270 feet according to Ross, there is a
deep, walled cauldron, with a strongly detonating solfatara,
from which at one time radiating lava streams poured them-
selves into the sea, where they now form small promontories,
and furnished volcanic millstones in Strabo's time (Ross, Rei-
sen aufden griechischen Insehi, bd. ii., s. 69, and 72-78). For
the British islands we have here still to mention, on account
of the antiquity of the formations, the remarkable effects of
submarine volcanoes upon the strata of the lower silurian
formation (Llandeilo strata), cellular volcanic fragments be-
ing baked into these strata, while, according to Sir Roderick
Murchison's important observation, even the eruptive trap-
masses penetrate into lower silurian strata in the Corndon
mountains (Shropshire and Montgomeryshire) ;"!■ the dike-phe-
nomena of the isle of Arran ; and the other points in which
the interference of volcanic activity is visible, although no
traces of true platforms are to be discovered.

II. Islands of the Atlantic Ocean.

The volcano Esk, upon the island of Jan Mayen, ascended
by the meritorious Scoresby, and named after his ship ; height
scarcely 1600 feet. An open, not ignited summit-crater ; ba-
salt, rich in pyroxene and trass.

Southwest of the Esk, near the North Cape of Egg IsJand,

* With regard to Vivarais and Velay, see the very recent and ac-
curate researches of Girard, in his Geologischen Wanderungen, bd. i.
(1856), s, 161, 173, and 214. The ancient volcanoes of Olot were dis-
covered by the American geologist Maclure in 1808, visited by Lyell
in 1830, and well described and figured by the latter in his Manual of
Geology, 1855, p. 535-542.

t Sir Roderick Murchison, SUuria, p. 20, and 55-58 (Lyell, Manualj
p. 563).

330 COSMOS.

another volcano, which in April, 1818, presented high erup-
tions of ashes every four months.

The Beerenberg, 6874 feet in height, in the broad, north-
eastern part of Jan May en (lat. 71° 4^), is not known to be
a volcano.*

Volcanoes of Iceland : Oerafa, Hecla, Rauda-Kamba . . .

Volcano of the island of Pico,t in the Azores : a great
eruption of lava from the 1st May to the 5th June, 1800.

The Peak of Teneriffe.

Volcano of Fogo^ one of the Cape de Verd Islands.

Pre-historic Volcanic Activity. — This on Iceland is less defin-
itely attached to certain centres. If we divide the volca-
noes of the island, with Sartorius von Waltershausen, into
two classes, of which those of the one have only had a sin-
gle eruption, while those of the other repeatedly emit lava
streams at the same principal fissure, we must refer to the
former, Rauda-Kamba, Scaptar, EUidavatan, to the south-
east of Reykjavik . . . . ; to the second, which exhibits a per-
manent individuality, the two highest volcanoes of Iceland
Oerafa (more than 6390 feet) and Snaefiall, Hecla, etc. Snae-
fiall has not been in activity within the memory of man, while
Oerafa is known by the fearful eruptions of 1362 and 1727
(Sart. von Waltershausen, Skizze von Island, s. 108 and 112).
In Madeira,§ the two highest mountains, the conical Pico
Ruivo, 6060 feet in height, and the Pico de Torres, which is
but little known, covered on their steep declivities with sco-
riaceous lavas, can not be regarded as the central point of the
former volcanic activity on the whole island, as in many
parts of the latter, especially toward the coasts, eruptive ori-
fices, and even a large crater, that of the Lagoa, near Ma-
chico, are met with. The lavas, thickened by confluence,
can not be traced far as separate streams. Remains of an-
cient dicotyledonous and fern-like vegetation, carefully inves-
tigated by Charles Bunbury, are found buried in upheaved

* Scoresby's Account of the Arctic Regions, vol. i., p. 155-169, tab.
V. and vi.

t Leop. von Buch., Descr. des lies Canaries, p. 357-369, and Land-
grebe, Naturgeschichte der Vulkane, 1855, bd. i., s. 121-136; and with
regard to the circumvallations of the craters of elevation {Caldeiras)
upon the islands of St. Michael, Fayal, and Terceira (from the maps
of Captain Vidal) (see page 216). The eruptions of Fayal (1672) and
Saint George (1580 and 1808) appear to be dependent upon the prin-
cipal volcano, the Pico. % See pages 236 and 249.

§ Results of the observations upon Madeira, by Sir Charles Lyell
and Hartung, in the Manual of Geology, 1855, p. 5L5-525,

TRUE VOLCANOES. 331

strata of volcanic tufa and loam, sometimes covered by more
recent basalt. Fernando de Noronha, lat. 3° 50^ S. and 2°
27'' to the east of Pernambuco ; a group of very small isl-
ands ; phonolitic rocks containing hornblende — no crater,
but vein-fissures filled with trachytic and basaltic amygda-
loid, penetrating white tufa layers.* The island of Ascen-
sion, highest summit 2868 feet; basaltic lavas with more
glassy feldspar than olivin sprinkled through them, and well-
bounded streams traceable up Jto the eruptive cone of tra-
chyte. The latter rock of light colors, often broken up like
tufa, predominates in the interior and southeast of the island.
The masses of scoriae thrown out from Green Mountain in-
close immersed angular fragments! containing syenite and
granite, which remind one of the lavas of Jorullo. To the
westward of Green Mountain there is a large open crater.
Volcanic bombs, partly hollow, of as much as ten inches in
diameter, lie scattered about in innumerable quantities, to-
gether with large masses of obsidian. St. Helena : the whole
island volcanic, the beds of lava in the interior rather felds-
pathic ; basaltic toward the coast, penetrated by innumera-
ble, dikes as at Flagstaff Hill. Between Diana Peak and
Nestlodge, in the central series of mountains, are the curved
and crescentic shaped fragments of a wider, destroyed crater
full of scoriae and cellular lava (" the mere wreckj of one
great crater is left"). The beds of lava are not limited, and
consequently can not be traced as true streams of small
breadth. Tristan da Cunha (lat. 37° 3^ S., long. 11° 26'
W.), discovered as early as 1506 by the Portuguese ; a small
circular island of six miles in diameter, in the centre of which
a conical mountain is situated, described by Captain Denham
as about 8300 feet in height, and composed of volcanic rock
(Dr. Petermann's Geogr. MittheiL, 1855, No. iii., s. 84). To
the southeast, but in 53° S. lat., lies the equally volcanic
Thompson's Island ; and between the two, in the same direc-
tion, Gough Island, also called Diego Alvarez. Deception

* Darwin, P'olcanic Islands, 1844, p. 23, and Lieutenant Lee, Cruise
of the United States Brig Dolphin, 1854, p. 80. .

t See the admirable description of Ascension in Darwin's Volcanic
Islands, p. 40 and 41.

X Darvvin, p. 84 and 92, with regard to " the great hollow space, or
valley southward of the central curved ridge, across which the half of
the crater must once have extended. It is interesting to trace the
steps by which the structure of a volcanic district becomes obscured
and finally obliterated." (See also.Seale, Geognosy of the Island of
St. Helena, p. 28.) '

332 COSMOS.

Island, a slender, narrowly-opened ring (S. lat. 62° 55^), and
Bridgeman's Island, belonging to the South Shetlands group ;
both volcanic, with layers of ice, pumice-stone, black ashes,
and obsidian ; perpetual eruption of hot vapois (Kendal,
Journal of the Geographical Society, vol. i., 1831, p. 62). In
February, 1842, Deception Island was seen to produce flames
simultaneously at thirteen points in the ring (Dana, in Unittd
States Exploring Expedition, vol. x., p. 548). It is remark-
able that, as so many islan(J^ in the Atlantic Ocean are vol-
canic, neither the entire flat islet of St. Paul* (Penedo de S.
Pedro), one degree to the north of the equator ; nor the Falk-
lands (with thin quartzose clay-slate). South Georgia or Sand-
wich land appear to offer any volcanic rock. On the other
hand, a region of the Atlantic Ocean, about 0° 20^ to the
south of the equator, longitude 22° W., is regarded as the
seat of a submarine volcano.f In this vicinity Krusenstern
saw black columns of smoke rise out of the sea (19th of May,
1806); and in 1836 volcanic ashes, collected at the same
point (southeast from the above-mentioned rock of St. Paul)
on two occasions, .were exhibited to the Asiatic Society of
Calcutta. According to very accurate investigations by Daus-
sy, singular shocks and agitation of the sea, ascribed to the
commotion of the sea-bottom by earthquakes, have been ob-
served in this volcanic region, as it is called in the new and
beautiful American chart of Lieutenant Samuel Lee {Track
of the Surveying Brig Dolphin, 1854), five times between 1747
and Krusenstern's circumnavigation of the globe, and seven
times from 1806 to 1836. But during the recent expedition
of the brig Dolphin (January, 1852), as previously (1838),
during Wilkes's exploring expedition, nothing remarkable
was observed, although the brig was ordered, " on account of
Krusenstern's volcano," to make investigations with the lead
between the equator and 7° S. lat., and about 18° to 27° long.

III. Africa.

It is stated by Captain Allan that the volcano Mongo-ma
Leba, in the CaiAeroon Mountains (4° 12^ N. lat.), westward
of the mouth of the river of the same name, in the Bight of

* St. Paul's Eocks. (See Danv-in, p. 31-33 and 125.)

t Daussy on the probable existence of a submarine volcano in the

Atlantic, in the Comptes rendus de VAcad. des Sciences, t. vi., 1858, p.

512 ; Darwin, Volcanic Islands, p. 92; Lee, Cruise of the United States

Brig Dolphin, p. 2-55, and 61. •

TRUE VOLCANOES. 333

Biafra, and eastward of the Delta of the Kowara, or Niger,
emitted an eruption of lava in the year 1838. The four
high volcanic islands of Annabon, St. Thomas, Isla do Prin-
cipe, and San Fernando Po, which run on a fissure in a di-
rect linear series from S.S.W. to N.N.E., point to the Came-
roons, which, according to the measurements of Captain Owen
and Lieutenant Boteler, rises to the great altitude of nearly
13,000 feet.*

A volcano (?) a little to the west of the snowy mountain
Kignea, in Eastern Africa, about 1° 20^ S. lat., was discov-
ered by the missionary Krapf in 1849, near the source of the
River Dana, about 320 geographical miles northwest of the
coast of -Mombas. In a parallel nearly two degrees more
southerly than the Kignea is situated another snowy mount-
ain, the Kihmandjaro, w^^ich was discovered by the mis-
sionary Rebmann in 1847, perhaps scarcely 200 geographical
miles from the ^ame coast. A little to the westward lies a
third snowy mountain, the Doengo Engai, seen by Captain
Short. The kn^^wledge of the existence of these mountains
is the result of laOorious and hazardous researches.

Evidences of pre-historical volcanic action in the great con-
tinent, the interior of which between the seventh degree north
and^'the twelfth degree south latitude (the parallels of Ada-
maua and the Lubalo Mountain, which acts as a water-shed)
still remains so unexplored, are furnished, according to Riip-
pell, by the country surrounding the Lake Tzana, in the king-
dom of Gondar, as well as by the basaltic lavas, trachytes,
and obsidian strata of Shoa, according to Rochet d'Hericourt,
whose mincralogical specimens, quite analogous to those of "
Cantal and Mont Dore, may have been examined by Dufre-
noy (Coniptes rendus, t. xxii., p. 806-810). Though the con-
ical mountain Koldghi, in Kordofan, is not now seen either in
a burning or smoking state, yet it appears that the existence
of a black, porous, and vitrified rock has been ascertained
there.!

In Adamaua, south of the great Benue River, rise the iso-
lated mountain masses of Bagele and Alantika, which from
their conical and dome-like forms appeared to Dr. Earth, on
his journey from Kuka to lola, to resemble trachyte mount-

* Gumprecht, Die Vulkanische Tlidtigkeit avf dem Festlande von Af-
rika^ in Arahien und aufden Inselndes Rothen Meeres, 1849, s. 18.

t Cosmos, vol. i., p. 245, note J. For the whole of the phenomena
hitherto known in Africa, see Landgrebe, Naturgeschkhte der Vulkane,
bd. i., s. 195-219.

334 cos^K)s.

ains. According to Petermann's notices from the note-books
of Overweg (of whose researches natural science was so ear-
ly deprived), that traveler found in the district of Gudsheba,
westward of the Lake of Tshad, separate basaltic cones, rich
in olivin and columnar in form, which were sometimes inter-
sected by layers of the red, clayey sandstone, and sometimes
by those of quartzose granite.

The small number of now ignited volcanoes in the undi-
vided continents, whose coast-lands are sufficiently known, is
a very remarkable phenomenon. Can it be that in the un-
known regions of Central Africa, especially south of the equa-
tor, large basins of water exist, analogous to Lake Uniames
(formerly called by Dr. Cooley, N'yassi), on whose scores rise
volcanoes, like the Demavend, near the Caspian Sea? Much
as the natives are accustomed to move about over the coun-
try, none of them have hitherto brought us the least notice
of any such thing !

IV. Asia.

a. The Western and Central part.

The volcano of Demavend,* in a state of ignition, but, ac-
cording to the accounts of Olivier, Morier, and Taylor Thom-
son (1837), smoking only moderately, and not uninterrupt-
edly.

The volcano of Medina (eruption of lava in 1276).

The volcano of Djebel el Tir (Tair or Tehr), an insular
mountain 895 feet high, between Loheia and Massaua, in the
Eed Sea.

* The height of Demavend above the sea was given by Ainsworth at
14,695, but, after correcting a barometrical result probably attributable
to an error of the pen {Asie Centrale, t. iii., p. 327), it amounts, accord-
ing to Ottman's tables, to fully 18,633 feet. A somewhat greater ele-
vation, 20,085 feet, is given by the angles of altitude worked by my
friend Captain Lemm, of the Russian navy, in the year 1839, and
which are certainly very correct, but the distance is not trigonomet-
rically laid down, and rests on the presumption that the volcano of
Demavend is 66 versts distant from Teheran (one equatorial degree
being equal to 104-^ versts). Hence it would appear that the Persian
volcano of Demavend, covered with perpetual snow, situated so near
the southern shore of the Caspian Sea, but distant 600 geographical
miles from the Colchian coast of the Black Sea, is higher than the great
Ararat by about 2989 feet, and the Caucasian Elburuz by probably 1600
feet. On the Demavend, see Ritter, Erdkunde vonAsien, bd. vi., abth.
i., s. 551-571 ; and on the connection of the name Albordj, taken from
the mythic and therefore vague geography of the Zend nation, with the
modern name Elburz (Koh Alburz of Kazwini) and Elburuz, see 3id.,
s< 43-49, 424, 552, and 555.

* TRUE VOLCANOES. 335

The volcano of Peshan, northward of Kutsche, in the great
mountain chain of the Thian-schan or Celestial Mountains, in
Central A'sia ; eruptions of lava within the true historical
period, from the year 89 up to the beginning of the 7th cen-
tury of our era.

The volcano of Ho-cheu, called also sometimes in the very
circumstantial Chinese geographies the volcano of Turfan ;
120 geographical miles from the great Solfatara of Urumtsi,
near the eastern extremity of the Thian-schan, in the direc-
tion of the beautiful fruit country of Hami.

The volcano of Demavend, which rises to a height of up-
ward of 19,000 feet, lies nearly 36 geographical miles from
the southern shore of the Caspian Sea, in Mazenderan, and
almost at the sjime distance from Resht and Asterabad, on
the chain of the Hindu-kho, which slopes suddenly down to
the west in the direction of Herat and Meshid. I have else-
where (Asie Centrale, t. i., p. 124-129 ; t. iii., p. 433-435)
mentioned the probability that the Hindu-kho of Chitral and
Kafiristan is a westerly continuation of the mighty Kuen-lun,
which bounds Thibet toward the north and intersects the Bo-
lor Mountains in the Tsungling. The Demavend belongs to
the Persian or Caspian Elburz, a system of mountains which
must not be confounded with the Caucasian ridge of the
same name (now called Elburuz), and which lies 7^° farther
north and 10° farther west. The word Elburz is a corrup-
tion of Alborj, or Mountain of the World, which is connected
with the ancient cosmogony of the Zends.

While the volcano of Demavend, according to the gener-
ality of geognostic views on the direction of the mountain
chains of Central Asia, bounds the great Kuen-lun chain
near its western extremity, another igneous appearance at
its eastern extremity, the existence of which I was the first
to announce (Asie Centrale, t. ii., p. 427 and 483), deserves
particular notice. In the course of the important researches
which I recommended to my respected friend and colleague
in the Institute, Stanislas Julien, with the view of deriving
information from the rich geographical sources of old Cliinese
literature on the subject of the Bolor, the Kuen-lun, and the
Sea of Stars, that intelligent investigator discovered, in the
great Dictionary published in the beginning of the 18th cen-
tury by the Emperor Yong-ching, a description of the " eter-
nal flame" which issues from an opening in the hill called
Shin-khien, on the eastern slope of the Kuen-lun. This lu-
minous phenomenon, however deeply seated it may be, can

336 COSMOS.

not well be termed a. volcano. It appears to me rather to
present an analogy with the Chimasra in Lycia, near DeUk-
tash and Yanartash, which was so early known to the Greeks.
This is a stream of fire, an issue of gas constantly kindled by
volcanic action in the interior of the earth (see page 243^
note t)-

Arabian writers inform us, though for the most part with-
out quoting any precise year, that lava eruptions have taken
place during the Middle Ages on the southwestern shore of
Arabia, in the insular chain of the Zobayr, in the Straits of
Bab-el-Mandeb and Aden (Wellsted, Travels in Arabia, vol.
ii., p. 466-468), in Hadhramaut, in the Strait of Ormuz, and
at different points in the western portion of the Persian Gulf.
These eruptions have always occurred on a ^oil which had
already been in pre-historical times the seat of volcanic ac-
tion. The date of the eruption of a volcano at Medina it-
self, 12^° northward of the Straits of Bab-el-Mandeb, was
found by Burckhardt in Samhudy's Chronicle of the famous
city of that name in the Hedjaz. It took place on the 2d
November, 1276. According to Seetzen, however, Abulma-
hasen states that an igneous eruption had occurred there in
1254, which is twenty-two years earlier (see Cosmos, vol. i.,
p. 246). The volcanic island of Djebeltair, in which Vincent
recognized the " burned-out island" of the Periplus Maris Ery-
thrcei, is still active, and emits smoke, according to Botta and
the accounts collected by Ehrenberg and Russegger {Eeisen in
Europa, Asien, und Africa, bd. ii., th. 1, 1843, s. 54). For in-
formation respecting the entire district of the Straits of Bab-
el-Mandeb, with the basaltic island of Perim — the crater-like
circumvallation, within which lies the town of Aden — the
island of Seerah with streams of obsidian, covered with pum-
ice— the island groups of the Zobayr and the Farsan (the
volcanic nature of the latter was discovered by Ehrenberg in
1825), I refer my readers to the interesting researches of Eit-
ter, in his Erdkunde von Asien, bd. viii., abth. 1, s. 664-707,
889-891, and 1021-1034.

The volcanic mountain chain of the Thian-schan {Asie Cen-
trale, t.i., p. 201-203 ; t. ii., p. 7-51), a range which intersects
Central Asia between Altai and Kuen-lun from east to west,
formed at one period the particular object of my investiga-
tions, so that I have been enabled to add to the few notices
obtained by Abel-Remusat from the Japanese Encyclopaedia,
some fragments of greater importance discovered by Klaproth,
Neumann, and Stanislas Julien (Asie Centrale, t. ii., p. 39-50

TRUE VOLCANOES. 337

and 335-364). The length of the Thian-schan is eight times
greater than that of the Pyrenees, if we include the Asferah,
which is on the other side of the intersected meridian chain
of the Kusyurt-Bolor, stretching westward as far as the me-
ridian of Samarcand, and in which Ibn Haukal and Ibn-al-
Vardi describe streams of fire, and notice luminous ("?) fissures
emitting sal ammoniac (see the account of Mount Botom, iit
supra). In the history of the dynasty of Thang it is expressly
stated that on one of the slopes of the Pe-shan, which contin-
ually emits fire and smoke, the rocks burn, melt, and flow to
the distance of several li, like a " stream of melted fat. The
soft mass hardens as it cools." It is impossible to describe
more characteristically the appearance of a stream of lava.
Moreover, in the forty -ninth book of the great geography of
the Chinese empire, which was printed at Pekin from 1789
to 1804 at the expense of the state, the burning mountains
of the Thian-schan are described as '^ still active." Their
position is very central, being nearly equidistant (1520 geo-
graphical miles) from the nearest shore of the Frozen Ocean
and from the mouth of the Indus and Ganges, 1020 miles
from the Se%of Aral, I72»and 208 miles from the salt-lakes
of Issikal and Balkasch. Information respecting the flames
issuing from the mountain of Turfan (Hotscheu) has also been
furnished by the pilgrims of Mecca, who were officially exam-
ined at Bombay in the year 1835 {Journal of the Asiatic Soc.
of Bengal, vol. iv., 1835, p. 65'?^664). When may we hope
to see the volcanoes of Peschan and Turfan, Barkul and Hami
explored by some scientific traveler, by way of Gouldja on the
Hi, which may be easily reached?

The better knowledge now possessed of the position of the
volcanic mountain chain of the Thian-schan has very natu-
rally given rise to the question whether the fabulous terri-
tory of Gog and Magog, where "eternal fire" is said to burn
at the bottom of the River El Macher, is not in some way
connected with the eruptions of the Peschan or the volcano
of Turfan. This Oriental myth, which had its origin West-
ward of the Caspian Sea, in the Pylis Albanice, near Der-
bend, has traveled, like all other myths, far toward the East.
Edrisi gives an account of the journeyings of one Salam el
Terdjeman, the dragoman of one of the Abbasside califs, in
the first half of the 9th century, from Bagdad to the Land
of Darkness. He proceeded through the steppe of Baschkir
to the snowy mountain of Cocaia, which is surrounded by
the great wall of Magog (Madjoudj). Amedee Jaubert, to

Vol. v.— P

338

COSMOS.

whom we are indebted for important supplements to the
Nubian geographers, has shown that the fires which burn on
the slope of the Cocaia have nothing volcanic in their nature.
(Asie -Centrale, t ii., p. 99.) Edrisi places the Lake of Te-
hama farther to the south. I think I have said enough to
show the probability of the Tehama being identical with the
great Lake of Balkasch, into which the Hi flows, and which
is only 180 miles farther south. A century and a half later
than Edrisi, Marco Polo placed the wall of Magog among
the mountains of In-schan, to the east of the elevated plain
of Gobi, in the direction of the River Hoang-ho and the Chi-
nese Wall, respecting which, singularly enough, the famous
Venetian traveler is as silent as he is on the subject of the
use of tea. The In-shan, the limit of the territory of Pres-
ter John, may be regarded as the eastern prolongation of the
Thian-schan (Asie Centrale, t. ii., p. 92-104).

The two conical volcanic mountains, the Petschan "and
Hotshen of Turfan, which formerly emitted lava, and which
are separated from each other at a distance of about 420
geographical miles by the gigantic block of mountains called
the Bogdo-Oola, crowned with eternal snow and ice, have
long been erroneously considered an isolated volcanic group.
I think I have shown that the Volcanic action north and
south of the long chain of the Thian-schan here, as well as
in the Caucasus, stands in close geognostic connection with
the limits of the circle of terrestrial commotion, the hot-
springs, the solfataras, the sal ammoniacal fissures, and beds
of rock salt.

According to the view I have already frequently express-
ed, and in which the writer most profoundly acquainted with
the Caucasian mountain system (Abich) now coincides, the
Caucasus itself is only a continuation of the ridge of the vol-
canic Thian-schan and Asferah, on the other side of the great
Aralo-Caspian depression.* This is, therefore, the place, in
connection with the phenomena of the Thian-schan, to cite
as beionging to pre-historical periods the four extinct volca-
noes of Elburuz, 18,494 feet in height; Ararat, 17,112 feet;
Kasbegk, 16,532 feet; and Savalan, 15,760 feet high.f In

* Asie Centrale, t. ii., p. 9, and 54-58. See also p. 199, note *, of
the present vokime.

t Elburuz, Kasbegk, and Ararat, according to communications from
Struve, Asie Centrale, t. ii., p. 57. The height of the extinct volcano
of Savalan, westward of Ardebil, as given in the text, is founded on
a measurement of Chanykow. See Abich, in the Melanges Phys. et
Chim., t. ii., p. 361. To save tedious repetition in the citation of the

TRUE VOLCANOES. 339

point of height, these mountains sland between Cotopaxi and
Mont Blanc. The great Ararat (Agri-dagh), ascended for
the first time on the 27th of September, 1829, by Friedrich
von Parrot, several times during 1844 and 1845 by Abich,
and lastly, in 1850, by Colonel Chodzko, is dome-shaped,
like Chimborazo, with two' extremely small elevations on the
border of the summit, but without any crater at the apex.
The most extensive and probably the latest pre-historical
lava eruptions of Ararat have all issued below the limit of
perpetual snow. The nature of these eruptions is two-fold ;
they are sometimes trachytic with glassy feldspar, inter-
spersed with pyrites which readily weather, and sometimes
doleritic, composed of labradorite and augite, like the lavas
of ^tna. The doleritic lavas of Ararat are considered by
Abich to be more recent than the trachytic. The points of
emission of the lava streams, which are all beneath the limit
of perpetual snow, are frequently indicated (as, for example,
in the extensive grassy plain of Kip-ghioU, on the northwest-
em slope) by eruptive cones and by small craters encircled
by scorise. Although the deep valley of St. James, which
extends to the very summit of Ararat, and gives a peculiar
character to its form, even when seen at a distance, exhibits
much resemblance to the Val del Bove on ^tna, and dis-
plays the internal structure of the dome, yet there ik this
striking difference between them, that in the valley of St.
James massive trachytic rock alone is found, and no streams
of lava, beds of scorise or rapilli.* The Great and Little
Ararat, the first of which is shown by the geodetic labors of
Wasili Fedorow, to be 3^ 4^^ more northerly, and 6^ 42^^
more westerly than the other, rise on the southern edge of
the great plain through which the Araxes flows in a large
bend. They both stand on an elliptic volcanic plateau, whose
major axis runs southeast and northwest. The Kasbegk
and the Tshegem have likewise no summit crater, although
the former has thrown out vast eruptions toward the north,
in the direction of Wladikaukas. The greatest of all these
extinct volcanoes, the trachytic cone of the Elburuz, which
has risen out of the talc and dioritic schistous mountains,

sources on which I have drawn, I would here explain that every thing
in the geological section of Cosmos relating to the important Caucasian
isthmus is borrowed from manuscript essays o^the years 1852 and 1 855,
communicated to me by Abich in the kindest and friendliest manner
for my unrestricted use.

* Abich, Notice Eplicative cfune Vue de V Ararat, in the Bulletin de
la Soc. de Geographic de France, 4eme serie, t. i., p. sIg.

340 COSMOS.

rich in granite, of the valley of the River Backsan, has a
crater lake. Similar crater lakes occur in the rugged high-
lands of Kely, from which streams of lava flow out between
eruption-cones. Moreover, the basalts are here, as w^ell as
in the Cordilleras of Quito, widely separated from the tra-
chyte system ; they commence from twenty-four to thirty-two
miles south of the chain of the Elburuz, and of the Tsehe-
gem, on the upper Phasis or lihion valley.

p. The Northeastern Portion {the Peninsula of Kamtschatka).

The peninsula of Kamtschatka, from CapeLopatka, which,
according to Krusenstern, is in lat. 51° 3^, as far north as to
Cape Ukinsk, belongs, in common with the island of Java,
Chili, and Central America, to those regions in which the
greatest number of volcanoes, and it may be added, of still
active volcanoes, are compressed :vvithin a very small area.
Fourteen of these are reckoned in Kamtschatka within a
range of 420 geographical miles. In Central America I find
in a space of 680 miles, from the volcano of Coconusco to
Turrialva, in Costa Rica, twenty-nine volcanoes, eighteen of
which are still burning ; in Peru and Bolivia, over a space
of 420 miles, from the volcano Chacani to that of San Pedro
de AlRcama, fourteen volcanoes, of which only three are at
present active ; and in Chili, over a space of 960 miles, from
the volcano of Coquimbo to that of San Clemente, twenty-
four volcanoes. Of the latter, thirteen are known to have
been active within the periods of time embraced in historical
records.

Our acquaintance with the Kamtschatkan volcanoes, in
respect to their form, the astronomical determination of their
position, and their height, has been vastly extended in recent
times by Krusenstern, Horner, Hoffman, Lenz, Liitke, Pos-
tels, Captain Beechey, and, above all, by Adolph Erman.
The Peninsula is intersected lengthwise by two parallel
mountain chains, in the most easterly of which the volcanoes
are accumulated. The loftiest of these attain a height of
from 11,190 to 15,773 feet. They lie in the following order
from south to north.

The Opalinskian volcano (the Pic Koscheleff of Admiral
Krusenstern), lat. 51* 21'. According to Captain Chwos-
tow, this mountain rises to the height of the Peak of Tene-
riffe, and was extremely active at the close of the 18th cen-
tury.

TRUE VOLCANOES. 341

The Hodutka Sopka (51° 35^). Between this and the one
just noticed there lies an unnamed volcanic cone (51° 32^),
which, however, according to Postels, seems, like the Hodut-
ka, to be extinct.

4?oworotnaja Sopka (52° 22^) ; according to Captain Bee-
chey, 7930 feet high (Erman's Reise, t. iii., p. 253 ; Leop. von
Buch, lies Can., p. 447).

Assatschinskaja Sopka (52° 2^) ; great discharges of ashes,
particularly in the year 1828.

The Wiljutschinsker volcano (52° 52'') ; according to Cap-
tain Beechey, 7373 feet ; according to Admiral Liitke, 6744
feet high. Distant only 20 geographical miles from the har-
bor of Petropolowski, on the north side of the Bay of Torinsk,

Awatschinskaja, or Gorelaja Sopka (53° 17''); according
to Erraan, 8910 feet high ; first ascended during the expedi-
tion of La Perouse, in 1787, by Mongez and Bernizet; after-
ward by my dear friend and Siberian fellow-traveler, Ernst
Hofraann (in July, 1824, during the circumnavigation of the
globe by Kotzebue ; by Postels and Lenz during the expedi-
tion of Admiral Liitke in 1828, and by Erman in September,
1829. The latter made the important geognostic observation
that the upheaving trachyte had pierced through slate and
graywacke (a silurian rock). The still smoking volcano had
a terrific eruption in Octolber, 1837, there having previously
been a slight one in April, 1828 (Postels, in Liitke, Voyage, t.,
bd., s. 67-84 ; Erman, Reise, Hist. Bericht, bd. iii., s. 494, and
534-540).

In the immediate neighborhood of the. Awatscha-volcano
(see page 236) lies the Koriatskaja or Strjeloschnaja Sopka
(lat. 53° 19'), 11,210 feet high, according to Lutke, t. iii., p.
84. This mountain is rich in obsidian, which the Kamtschat-
kans so late as the last century made into arrow-heads, as the
Mexicans and the ancient Greeks used to do.

Jupanowa Sopka, lat., according to Erman's calculation
{Reise, bd. iii., s. 469), 53° 32'. The summit is pretty flat,
and the traveler just mentioned expressly states " that this
Sopka, on account of the smoke it emits, and its perceptible
subterranean rumbling, is always compared to the mighty
Schivvelutsch, and reckoned among the undoubted igneous
mountains." Its height, as measured by Liitke from the sea,
is 9055 feet.

Kronotskaja Sopka, 10,609 feet, at the lake of the same
name, lat. 54° 8'; a smoking crater on the summit of the very
sharp-pointed conical mountain (Liitke, Voyage, t. iii., p. 85).

342 COSMOS.

The volcano Schiwelutsch, 20 miles southeast of Jelowka,
respecting which we possess an admirable work by Erman
{Reise, bd. iii., s. 261-317 ; and Phys. Bioh., bd. i., s. 400-403),
previous to whose journey the mountain was almost unknown.
Northern peak, lat. 56^ 40^, height 10,544 feet ; southern pAk,
lat. 56° 39^ height 8793 feet. When Erman ascended the
Schiwelutsch in September, 1829, he found it smoking vehe-
mently. Great eruptions took place in 1739, and between
1790 and 1810; the latter consisting, not of flowing, melted
lava, but of ejections of loose volcanic stones. C. von Ditt-
mar relates that the northern peak fell in during the night
from the 17th to the 18th of February, 1854. At that time
an eruption, which still continues, took place, accompanied
by genuine streams of lava.

Tolbatschinskaja Sopka; smoking violently, but in earli-
er times frequently changing the openings througluwhich it
ejected its ashes. According to Erman, lat. 55° 51^, and
height 8313 feet.

Uschinskaja Sopka ; closely connected with the Kliuts-
chewsker volcano ; lat. 56° 0^, height 1 1,723 feet (Buch, Can.,
p. 452 ; Landgrebe, Volkane, vol. i., p. 375). •

Kliutschewskaja Sopka (56° 4^), the highest and most act-
ive of all the volcanoes of the peninsula of Kamtschatka ;
thoroughly examined by Erman, both geologically and hyp-
sometrically. According to KraschenikofF's report, the Kli-
utschewsk had great igneous eruptions from 1727 to 1731, as
also in 1767 and 1795. On the 11th of September, 1829,
Erman performed the hazardous feat of ascending the volca-
no, and was an eye-witness of the ejection of red-hot stones,
ashes, and vapor from the summit, while at a great distance
below it an immense stream of lava flowed from a fissure on
the western declivity. Here, also, the lava is rich in obsidian.
According to Erman {Beoh., vol. i., p. 400-403 and 419) the
geographical latitude of the volcano is 56° V, and its height
in September, 1829, was, on a very accurate calculation,
15,763 feet. In August, 1828, on the other hand. Admiral
Liitke, on taking angles of altitude at sea, at a distance of
160 knots (40 nautical miles), found the summit of Kliuts-
chewsk 16,498 feet high (Voyage, t. iii., p. 86; Landgrebe,
Vulkane, bd. i., s. 375-386). This measurement, and a com-
parison of the admirable outline drawings. of Baron von Kit-
tlitz, who accompanied Liitke's expedition on board the Se-
niawin, with what Erman himself observed in September,
1829, led the latter to the conclusion that, in this short pe-

TRUE VOLCANOES. 343

riod of thirteen months, great changes had taken place in
the form and height of the summit. " I am of opinion," says
Erman {Reise, vol. iii., p. 359), "that we can scarcely be wrong
in assuming the height of the summit in August, 1828, to have
been 266 feet more than in September, 1829, during my stay
in the neighborhood of Kliutschi, and that therefore its height
at the former of these periods must have been 16,029 feet."
In the case of Vesuvius, I found, by my own calculations
(founde^on Saussure's barometrical measurement in 1778),
of the K»cca del Palo, the highest northern margin of the
crater, that up to the^ year 1805 — that is to say, in the course
of thirty-two years — this northern margin of the crater had
sunk 35^ feet ; while from 1773 to 1822, or forty-nine years,
it had risen (apparently) 102 feet {Views of Nature, 1850, p.
376-378), In the year 1822 Monticelli and Covelli calcu-
lated the Rocca del Palo at 3990 feet, and I at 4022 feet; I
then gave 3996 as the most probable result for that period.
In the spring of 1855, thirty-three years later, the delicate
barometrical measurements of the Olmutz astronomer, Julius
Schmidt, again brought out 3990 feet {Neue Bestimm. Am.
Vesuv., 1856, s. i., 16 and 33). It would be curious to know
how much should here be attributed to imperfection of meas-
urement and barometrical formula. Investigations of this
kind might to be multiplied on a larger scale and with greater
certainty if, instead of often-repeated completed trigonomet-
rical operations or, in the case of accessible summits, the
more practicable though less satisfactory barometrical meas-
urement, operators would confine themselves to determining,
even to fractions of seconds, at comparative periods of twen-
ty-five or fifty years, the simple angle of altitude of the mar-
gin of the summit, from the same point of observation, and
one which could with certainty be found again. On account
of the influence of terrestrial refraction, I would recommend
that, in each of the normal epochs, the mean result of three
days' observations at different hours should be taken. In
order to obtain not only the general result of the increase or
diminution of the angle, but also the absolute amount of the
change in feet, the distance would required to be determined
previously only once for all. What a rich source of knowl-
edge, relative to the twenty volcanic Colossi of the Cordille-
ras of Quito, would not the angles of altitude, determined for
more than a century by the labors of Bouguer and La Con-
daraine, have provided had those travelers accurately desig-
nated as fixed and permanent points the stations whence they

344 COSMOS.

measured the angles of altitude of the summits. According
to C. von Dittmar, the Kliutschewsk was entirely quiescent
since the eruption of 1841, until the lava burst forth again
in 1853. The falling in, however, of the summit of the
Schiwelutsch interrupted the new action {Bulletin de la Classe
Physico-Mathem. de VAcad. des Sc. de St. Petershourg, t. xiv.,
1856, p. 246).

Four more volcanoes, mentioned in part by Admiral Lutke,
and in part by Postels — namely, the Apalsk, still ^moking,
to the southeast of the village of Bolscheretski, the^lchischa-
pinskaja Sopka (lat. 55° 11^), the cone of Krestowsk (lat.
56° 4^), near the Kliutschewsk group, and the Uscnkowsk —
I have not cited in the foregoing series, from want of more
exact specification. The central mountain range of Kamts-
chatka, especially in the plain of Baidaren, lat. 57° 20^, east-
ward of Sedanka, presents (as if it had been " the field of an
ancient crater of about four wersts, that is to say, the same
number of kilometres, in diameter") the remarkable geolog-
ical phenomenon of effusions of lava and scorias from a blis-
tery and often brick-colored volcanic rock, which in its turn
has penetrated through fissures in the earth at the greatest
possible distance from any frame-woik of raised cones (Er-
man, Reise, bd. iii., 221, 228, and 273 ; Buch, lies Canaries,
p. 454). The analogy is here very striking with what I
have already circumstantially explained regarding the Mal-
pays, the problematical fields of debris in the elevated plain
of Mexico, (see p. 297).

V. Islands of Eastern Asia.

From Torres Strait, which in the 10th degree of south-
ern latitude separates New Guinea and Australia, and from
the smoking volcano of Flores to the most northern of the
Aleutian Isles (lat. 55°), there is a multitude of islands, for
the most part volcanic, which, considered in a general geo-
logical point of view, it would be somewhat difficult, on ac-
count of their genetic connection, to divide into separate
groups, and which increase considerably in circumference to-
ward the south. Beginning at the north, we first observe
that the curved series* of the Aleutians, issuing from the

* See Dana's remarks on the curvatures of ranges of islands, whose
convexity in the South Sea is almost always directed toward the south
or southeast, in th^ United States Exploring Exjiedition by Wilkes, vol.
X. {Geology, by James Dana), 1849, p. 419.

TRUE VOLCANOES. 345

American peninsula of Alaska, connect the old and the new
continents together by means of the island Attn, near Cop-
per Island and Behring's Island, while to the south they close
in the waters of Behring's Sea. From Cape Lopatka at the
southern extremity of the peninsula of Kamtschatka, we find
succeeding each other, in the direction from north to south,
first the Archipelago of the Kuriles, bounding on the east
the Saghalien or Ochotsk Sea, rendered famous by La Pe-
rouse ; next Jesso, probably in former times connected with
the island of Krafto* (Saghalin or Tschoka) ; and, lastly, the
tri-insular empire of Japan, across the narrow Strait of Sau-
gar (Niphon, Sitkok, and Kiu-Siu, according to ^iebold's ad-
mirable map, between 41° 32^ and 30° 18^). From the vol-
cano of Kliutschewsk, the northernmost on the east coast of
the peninsula of Kamtschatka, to the most southern Japan-
ese volcano island of Tanega-Sima, in the Van Diemen's
Channel, explored by Krusenstern, the direction of the igne-

* The island of Saghahn, Tschoka, or Tarakai, is called by the Jap-
anese mariners Krafto (written Karafuto). It lies opposite the mouth
of the Amoor (the Black River, Saghalian Ula), and is inhabited by
the Ainos, a race mild in disposition, dark in color, and sometimes
rather hairy. Admiral Krusenstern was of opinion, as were also pre-
viously the companions of La Perouse (1787) and Broughton (1797),
that Saghalin was connected with the Asiatic continent by a narrow
sandy isthmus (lat. 52° 5') ; but, from the important Japanese notices
communicated by Franz von Siebold, it appears that, according to a
chart drawn up in the year 1808, by Mamia Rinso, the chief of an
imperial Japanese commission, Krafto is not a peninsula, but an isl-
and surrounded on all sides by the sea (Ritter, Erdkunde, von Asien,
vol. ii., p. 488). The conclusion of Mamia Rinso has been very re-
cently completely verified, as mentioned by Siebold, when the Russian
fleet lay at anchor in the year 1855 in the Bale de Castries (lat. 51°
29'), near Alexandrowsk, and consequently to the south of the con-
jectured isthmus, and yet was able to retire into the mouth of the
Amoor (lat. 52° 24'). In the narrow channel in which the isthmus
was formerly supposed to be, there were in some places only five fath-
oms water. The island is beginning to acquire some political impor-
tance on account of the proximity of the great stream of Amoor or
Saghalin. Its name, pronounced Karafto or Krafto, is a contraction
of Kara-fu-to, whicli signifies, according to Siebold, " the island bor-
dering on Kara." In the Japano-Chinese language Kara denotes the
most northerly part of China (Tartary), and fu, according to the learn-
ed writer just mentioned, signifies, " lying close by." Tschoka is a
corruption of Tsyokai, and Tarakai originates from a mistake in the
name of a single village called Taraika. According to Klaproth (Asia
Polyglotta. p. 301), Taraikai, or Tarakai, is the native Aino name of
the whole island. Compare Leopold Schrenk's and Captain Bernard
Wittingham's remarks, in Petermann's Geogr. Mittheilungen, 1856, s.
176 and 184. See also Perry, Expedition to Japan, vol. i., p. 468.

P2

346 COSMOS.

ous action, as indicated in the numerous rents of the earth's
crust, is precisely from northeast to southwest. The range
is carried on by the island of Jakuno-Sima, on which a
conical mountain rises to the height of 5838 feet (1780 me-
tres), and which separates the two straits of Van Diemen
and Colnet — by the Linschote Archipelago of Siebold — by
Captain Basil Hall's sulphur island, Lung-Huang-Schan, and
by the small group of the Loo-choo and Majico-sima, which
latter approaches within a distance of 92 geographical miles
the eastern margin of the great island of the Chinese coasts,
Formosa or Tay-wan.

Here at Formosa (N. lat. 25°-26°) is the important point
where, instead of the lines of elevation from N.E. to S.W. '
those in the direction from north to south commence, and
continue nearly as fiir as the parallel 5° or 6° of southern lati-
tude. They are recognizable in Formosa and in the Philip-
pines (Luzon and Mindanao) over a space of fully twenty de-
grees of latitude, intersecting the coasts, sometimes on one
side and sometimes on both, in the direction of the meridian.
They are likewise visible on the east coast of the great isl-
and of Borneo, which is connected by the So-lo Archipelago
with Mindanao, and by the long, narrow island of Palawan
with Mindoro. So also in the western portions of the Cel-
ebes, with their varied outline, and Gilolo, and, lastly (which
is especially remarkable), in the longitudinal fissures on which,
at a distance of 1400 geographical miles eastward of the group
of the Philippines and in the same latitude, the range of vol-
canic and coral islands of Marian or the Ladrones have been
upheaved. Their general direction* is north, and 10° east.

Having pointed out in the parallel of the carboniferous
island of Formosa the turning point at which the direction
of the Kuriles from N.E. to S.W. is changed to that from
north to south, I must now observe that a new system of fis-
sures commences to the south of Celebes and the south coasts
of Borneo, which, as we have already seen, is cut from east
to west. The greater and lesser Sunda islands, from Timor-
lant to West-Bali, follow chiefly for the space of 18° of longi-
tude, the mean parallel of 8° south latitude. At the western

* Dana, Geology of the Pacific Ocean, p. 16. Corresponding with the
meridian lines of the southeast Asiatic island world, the shores of Co-
chin-China from the Gulf of Tonquin, those of Malacca from the Gulf
of Siam, and even those of New Holland south of the 25th degree of
latitude are for the most part cut oflF, as it were, in the direction from
north to south.

TRUE VOLCANOES. 347

extremity of Java the mean axis runs somewhat more to-
ward the north, nearly E.S.E. and W.N.W., while from the
Strait of Sunda to the southernmost of the Nicobar Isles the
direction is from S.E. to N.W. The whole volcanic fissure
of elevation (E. to W., and S.E. to N.W.) has consequently
an extent of about 2700 geographical miles, or eleven times
the length of the Pyrenees. Of this space, if we disregard
the slight deviation toward the north in Java, 1620 miles
belong to the east and west direction, and 1080 to the south-
east and northwest.

Thus do general geological considerations on form and
range lead uninterruptedly, in the island world on the east
coast of Asia (over the immense space of 68° of latitude),
from the Aleutian Isles and Behring's Sea to the Moluccas
and the Great and Little Sunda Isles. The greatest variety
in the configuration of the land is met with in the parallel
zone of 5°- north and 10° south latitude. It is very remark-
able how generally the line of eruption in the larger portions
is repeated in a neighboring smaller portion. Thus a long
range of islands lies near the south coast of Sumatra and
parallel to it. We find the same appearances in the smaller
phenomena of the mineral veins as in the greater ones of the
mountain ranges of whole continents. Accompanying debris,
running by the side of the principal vein, and secondary
chains (chaines accompagnantes) lie frequently at considerable
distances from each other. They indicate similar causes and
similar tendencies of the formative action in the folding in of
the crust of the earth. The conflict of powers in the con-
temporaneous openings of fissures in opposite directions ap-
pears sometimes to occasion strange formations in juxtapo-
sition, as may be seen in the Molucca Islands, Celebes, and
Kilolo.

After developing the internal geological connection of the
East and South Asiatic insular system, in order not to devi-
ate from the long-adopted, though someirftet arbitrary, geo-
graphical divisions and nomenclature, w^Race the southern
limit of the Eastern Asiatic insular range (the turning-point)
at Formosa, where the line of direction runs oflT from the
N.E.— S.W. to the N.— S., in the 24th degree of north lati-
tude. The enumeration proceeds again from north to south,
beginning with the eastern, and more American, Aleutian
Islands.

The Aleutian Isles, which abound in volcanoes, include,
in the direction from east to west, the Eox Islands, among

348 COSMOS.

which are the largest of all, Unimak, Unalaschka, and Ura-
nak — the Andrejanowsk Isles, of which the most famous
are Atcha, with three smoking. volcanoes, and the great vol-
cano of Tanaga, already delineated by Sauer — the Eat Isl-
ands, and the somewhat distant islands of Blynia, among
which, as has been already observed, Attn forms the connect-
ing link to the Commander group (Copper and Behring's Isles),
near Asia. There seems no ground for the often-repeated
conjecture that the range of continental volcanoes in the di-
rection of N.N.E. and S.S.W., on the peninsula of Kamts-
chatka, first commences where the volcanic fissure of up-
heaval in the Aleutian Islands intersects the peninsula be-
neath the ocean, the Aleutian fissure thus forming, as it were,
a channel of conduction. According to Admiral Liitke's
chart of the Kamtschatkan Sea (Behring's Sea), the island of
Attn, the western extremity of the Aleutian range, lies in
lat. 52° 46^, and the non-volcanic Copper and Behring's Isl-
ands in lat. 54° 30^ to 55° 20^, while the volcanic range of
KamtBchatka commences under the parallel of 56° 40^ with
the great volcano of Schivvelutsch, to the west of Cape Stol-
bowoy. Besides, the direction of the fissures of eruption is
very diflferent, indeed, almost opposite. The highest of the
Aleutian volcanoes, on Unimak, is 8076 feet according to
Lutke. Near the northern extremity of Umnak, in the month
of May, 1796, there arose from the sea, under very remark-
able circumstances, which have been admirably described
in Otto von Kotzebue's ^' Entdeckungsreise''' (bd. ii., s. 106),
the island of Agaschagokh (or St. Johannes Theologus),
which continued burning for nearly eight years. According
to a report published by Krusenstern, this island was, in the
year 1819, nearly sixteen geographical miles in circumfer-
ence, and was nearly 2240 feet high. On the island of Una-
laschka the proportions of the trachyte, containing much
hornblende, of the volcano of Matuschkin (5474 feet) to the
black porphyry (l|yind the neighboring granite, as given by
Chamisso, would^reserve to be investigated by some scientific
observer acquainted with the conditions of modern geology,
and able to examine carefully the mineralogical character of
the different kinds of rocks. Of the two contiguous islands
of the Pribytow group, which lie isolated in the Kamtschat'
kan Sea, that of St. Paul is entirely volcanic, abounding in
lava and pumice, while St. Geoi*ge's Island, on the contrary,
contains only granite and gneiss.

According to th^most exact enumeration we yet possess,

TRUE VOLCANOES. ^ 349

the range of the Aleutian Isles, stretching over 960 geo-
graphical miles, seems to contain above thirty-four volcanoes,
the greater part of them active in modern historical times.
Thus we see here (in 54° and 60° latitude, and 160°-196°
west longitude) a strip of the whole floor of the ocean be-
tween two great continents in a constant state of formative
and destructive activity. How many islands in the course
of centuries, as in the group of the Azores, may there not be
near becoming visible above the surface of the ocean, and
how many more which, after having long appeared, have
sunk either wholly or partially unobserved ! For the min-
gling of races, and the migration of nations, the range of the
Aleutian Islands furnishes a channel from thirteen to four-
teen degrees more southerly than that of Behring's Straits,
by which the Tchutches seem to have crossed from America
to Asia, and even to the other side of the Eiver Anadir.

The range of the Kurile Islands, from the extreme point
of Kamtschatka to Cape Broughton (the northernmost prom-
ontory of Jesso), in a longitudinal space of 720 geographical
miifes, exhibits from eight to ten volcanoes, still for the most
part in a state qf ignition. The northernmost of these, on
the island of Alaid, known for its great eruptions in the years
1770 and 1793, is well worthy of being accurately measured,
its height being calculated at from 12,000 to 15,000 feet. The
much less lofty Pic Sarytshew (4193 feet according to Horner)
on Mataua, and the southernmost Japanese Kuriles, Urup,
Jetorop, and Kunasiri, have also been very active volcanoes.

We now come in the order of succession of the volcanic
range to Jesso, and the three larger Japanese Islahds, re-
specting which the celebrated traveler, Herr von Siebold, has
kindly communicated to me a large and important work for
assistance in my Cosmos. This will serve to correct what-
ever was defective in the notices which I borrowed from the
great Japanese Encyclopedia in my Fragmevs de Geologie et
de Climatologie Asiatiques (tom. i., p. 217-234), and in Asie
Centrale (tom. ii., p. 540-552).

The large island of Jesso, which is very quadrangular in
its northern portion (lat. 41^° to 45^°) separated by the
Strait of Saugar, or Tsugar, from Niphon, and by that of La
Peroyse from the island of Krafto (Kara-fu-to), bounds by
its northeast cape the Archipelago of the Kuriles ; but not
far from the northwest Cape Romanzow, on Jesso, which
stretches a degree and a half more northward in the Strait
of La Perouse, lies, in latitude 45° 11^, the volcanic Pic de

350 ^ COSMOS.

Langle (5350 feet), on the little island of Eisiri. Jesso itself
seems also to be intersected by a range of volcanoes, from
Broughton's Southern Volcano Bay nearly all the way to the
North Cape, a circumstance the more remarkable, as, on the
narrow island of Krafto, which is almost a continuation of
Jesso, the naturalists of La Perouse's expedition found, in the
Baie de Castries, fields of red porous lava and scoriae. On
Jesso itself Siebold counted seventeen conical mountains, the
greater number of which appear to be extinct volcanoes.
The Kiaka, called by the Japanese Usaga-Take, or Mortar
Mountain, on account of a deeply-hollowed crater, and the
Kajo-hori are both said to be still in a state of ignition.
(Commodore Perry noticed two volcanoes from Volcano Bay,
near the harbor of Endermo, lat. 42° 17^.) The lofty Manye
(Krusenstern's conical mountain Pallas) lies in the middle of
the island of Jesso, nearly in lat. 44°, somewhat to the E.N.E.
of Bay Strogonow.

" The historical books of Japan mention only six active
volcanoes before and since our era — namely, two on the isl-
and of Niphon, and four on the island of Kiu-siu. Thefbl-
canoes of Kiu-siu, the nearest to the peninsud© of Corea, reck-
oning them in their geographical position from south to north,
are, (1) the volcano of Mitake, on the islet of Sayura-sima,
in the Bay of Kagosima (province of Satsuma), which lies
open to the south, lat. 31° 33^ long. 130° 41^ ; (2) the vol-
cano Kirisima (lat. 31° 45^), in the district of Naka, prov-
ince of Finga ; (3) the volcano Aso jama, in the district Aso
(lat. 32° 45^), province of Figo; (4) the volcano of Vunzen,
on the peninsula of Simabara (lat. 32° 44^), in the district of
Takaku. The height of this volcano amounts, according to
a barometrical measurement, only to 1253 metres, or 4110
English feet, so that it is scarcely a hundred feet higher than
Vesuvius (Rocca del Palo). The most violent eruption of
the volcano of Vunzen on record is that of February, 1793.
Vunzen and Aso jama both lie east-southeast of Nangasaki."

" The volcanoes of the great island of Niphon, again reck-
oning from south to north, are, (1) the volcano of Fusi jama,
scarcely 16 geographical mileS distant from the southern
coast, in the district of Fusi, province of Suruga (lat. 35° 18^,
long. 138° 35''). Its height, measured in the same way as
the volcano of Vunzen, or Kiu-siu, by some young Japanese
instructed by Siebold, amounts to 3793 metres, or 12,441
feet ; it is, therefore, fully 320 feet higher than the Peak of
Teneriffe, with v/hich it has been already compared by Kamp-

TRUE VOLCANOES. 351

fer (Wilhelm Heine, Meise nach Japan, 1856, bd. ii., s. 4).
The upheaval of this conical mountain is recorded in the fifth
year of the reign of Mikado VI. (286 years before our era)
in these (geognostically remarkable) words : ' In tlie country
of Omi a considerable quantity of land sinks, an inland lake
is formed, and the volcano Fusi makes its appearance.' The
most violent historically recorded eruptions within the Chris-
tian era are those of 799, 800, 863, 937, 1032, 1083, and
1707 : since the latter period the mountain has been tranquil.
(2) The volcano of Asama jama, the most central of the act-
ive volcanoes in the intei:ior of the country, distant 80 geo-
graphical miles from the south-southeast, 52 miles from the
north-northwest coast, in the district of Saku (province of
Sinano), lat. 36° 22^, long. 138° 38'; thus lying between
the meridians of the two capitals, Mijako and Jeddo. The
Asama jama had an eruption as early as the year 864, con-
temporaneously with the Fusi jama ; that of the month of
July, 1783, was particularly violent and destructive. Since
that time the Asama jama has maintained a constant state
of activity.

" Besides these volcanoes two other small islands with
smoking craters have been observed by European mariners,
namely, (3) the small island of Ivogasima, or Ivosima {sima
signifies island, and ivo sulphur ; ga is merely an affix mark-
ing the nominative), Krusenstern's lie du Volcan, south of
Kiu-siu, in Van Diemen's Strait, 30° 43' N. lat., and 130°
18' E. long., distant only fifty-four miles from the above-
mentioned volcano of Mitake ; the height of the volcano is
2364 feet (715 met.). This island is mentioned byLinscho-
ten, so early as 1596, in these words : ' The island has a vol-
cano, which is a sulphur, or fiery mountain.' It occurs also
on the oldest Dutch sea-charts under the name of Vulcanus
(FrJ von Siebold, Atlas von Jap. Reiche, tab. xi.). Krusen-
stern saw it smoking in 1804, as did Captain Blake in 1838,
and Gue'rin and De la Roche Poncie in 1846. The height
of the cone, according to the latter navigator, is 2345 feet
(715 met.). The rocky islet mentioned as a volcano by
Landgrebe in the Naturgeschichte der Vulkane (bd. i., s. 355),
and which, according to Kampfer, is near Firato (Firando),
is undoubtedly Ivo-sima, for the group to which Ivo-sima
belongs is called Kiusiu ku sima, i. <?., the nine islands of Kiu-
siu, and not the ninety-nine islands. A group of this de-
scription occurs near Firato, northward of Nagasaki, and no-
where else in Japan. (4) The island of Ohosima (Barne-

352 COSMOS.

velde's Island ; Krusenstern's He de Vries), which is consid-
ered part of the province of Idsu, on Niphon, and lies in
front of the Bay of Vodavara, in 34° 42^ N. lat., and 136°
. 26' E. long. Broughton saw smoke issuing from the crater
in 1797, a violent eruption of the volcano having taken place
a short time previous. From this island a range of smaller
volcanic isles stretches out in a southerly direction as far as
Fatsi-syo (33° 6' N. lat.) and continues as far as the Bonin
Islands (26° 30" N. lat., and 142° 5' E. long.), which, accord-
ing to A. Postels (Liitke, Voyage autour du Monde dans les
annees 1826-29, t. lii., p. 117), are likewise volcanic, and
are subject to very violent earthquakes."

"These, then, are the eight volcanoes historically known
to be active in Japan Proper, in and near the islands of Kiu-
siu and Niphon. But in addition to these volcanoes a range
of conical mountains must also be cited, some of which,
marked by very distinct and often deeply indented craters,
appear to be volcanoes long since extinct. One of these is
the conical mountain of Kaimon, Krusenstern's Pic Horner,
in the southernmost corner of the island of Kiu-siu, on the
coast of Van Diemen's Strait, in the province of Satsum (lat.
31° 9^), scarcely six geographical miles S.S.W. from the act-
ive volcano of Mitake. Another is the Kofusi, or Little Fusi,
on Sikok; and another is on the islet of Kutsunasima, in the
province of. Ijo (lat. 33° 45"), on the eastern coast of the
great straits of Suvo Nada or Van der Capellen, which sep-
arate the three great portions of the Japanese empire, Kiu-
siu, Sikon, and Niphon. On the latter, or principal island,
nine such conical mountains, probably trachytic, are reck-
oned, the most remarkable of which are, the Siri jama (or
White Mountain), in the province of Kaga, lat. 36° 5", and
the Tsyo Kai-san, in the province of Deva (lat. 39° 10"),
both of which are considered loftier than the southerly vol-
cano of Fusi jama, which is upward of 12,360 feet high. Be-
tween, these two, in the province of Jetsigo, lies the Jaki
jama (or Flame Mountain, lat. 36° 53"). The two northern-
most conical mountains in the Saugar Strait, in sight of the
great island of Jesso, are, (1) the Ivaki jama, called byKru-
senstern, whose illustrations of the geography of Japan have
gained him immortal honor, the Pic Tilesius (lat. 40° 42^) ;
and (2) the Jake jama (the Burning Mountain, lat. 41° 20"),
in Nambu, at the northeastern extremity of Niphon, with
igneous eruptions from the remotest times."

In the continental portion of the neighboring peninsula of

TRUE VOLCANOES. 353

Corea, or Korai (which, in the parallels of 34° and 34^°, is
almost united with Kiusiu by the islands Tsu sima and Iki),
notwithstanding its great similarity in form to the peninsula
of Kamtschatka, no volcanoes have hitherto been discovered.
The volcanic action seems to be confined to the adjoining
islands. Thus, in the year 1007, the island volcano of Tsin-
raura, called by the Chinese Tanlo, rose from the sea. A
learned Chinese, named Tien-kong-chi, was sent to describe
the phenomenon and to execute a picture of it.* But it is
especially on the island of Se-he-sure (the Quelpaerts of the
Dutch) that the mountains exhibit every where a volcanic
conical form. The central mountain rises, according to
Broughton and La Perouse, to the height of 6395 feet. How
many volcanic effects may there not yet remain to be discov-
ered in the Western Archipelago, where the Kiag of the Co-
reans styles himself the Sovereign of 10,000 islands!

From the Pic Horner (Kaimon ga take), on the west side
of the southern extremity of the Kiusiu, in the Japanese tri-
insular empire, there stretches out, in a curve which lies open
toward the west, a small range of volcanic islands, comprising
first, between the Van Diemen and Colnet Straits, the Ja-
kuno sima and the Tanega sima ; second, south of the Strait
of Colnet, in the Linschoten groupf of Siebold (the Archipel
Cecile of Captain Guerin), which extends as far as the paral-
lel of 29°, the island of Suvase sima, the volcano island of
Captain Belcher (lat. 29° 39^ and long. 129° 41^, rising,
according to De la Koche Poncie, to a height of 2800 feet
(855 met.) ; third, Basil Hall's sulphur island, the Tori-sima,
or Bird Island of the Japanese, the Lung-hoang-shan of Pere
Gaubil, in lat. 27° 51^ ^d long. 128° 14^, as fixed by Cap-
tain De la Roche Poncie in 1848. As this island is also
called Iwosima, care must be taken not to confound it with
its more northerly namesake in Van Diemen's Straits. It
has been admirably described by Captain Basil Hall. Be-
tween the parallel of 26° and 27° of latitude comes in suc-
cession the Lieu-thieu, or Loo-choo Islands, as the natives
call them, of which Klaproth published a separate map in
1824, and more^ to the southwest the small Archipelago of
Majicosima, which approaches the great island of Formosa,
and is considered by me to be the closing point of the eastern

* Compare the translations of Stanislaus Julien from the Japanese
Encyclopedia, in mj Asie Centr-ale, t. ii., p. 551.

t Compare Kaart van den Zuid-en Zuidwest-Kust van Japan door F.
von Siebold, 1851.

354 COSMOS.

Asiatic islands. Close to the east coast of Formosa (lat. 24°)
a great volcanic eruption in the sea was observed by Lieu-
tenant Boyle in 1853 (Commodore Perry, Expedition to Japan,
vol. i.j p. 500). Among the Bonin Islands (Buna-sima of the
Japanese, lat. 26^° to 27|°, and long. 142° 15^, that called
Peel's Island has several craters abounding in. sulphur and
scoriae, which do not appear to have been long extinct (Per-
ry, i., p. 200 and 209).

VI. Islands of Southern Asia.

We comprehend under this division Formosa (Tayvan), the
Philippines, the Sunday Islands, and the Moluccas. Klap-
roth first made us acquainted with the volcanoes of Formosa
by information extracted from Chinese sources, which are
always so c<j|mous in their descriptions of nature.* They are
four in number, and of these the Chy-kang (Red Mountain),
whose crater contains a hot-water lake, has experienced great
igneous eruptions. The small Baschi Islands and the Babu-
yans, which so late as 1831, according to Meyen's testimony,
experienced a violent eruption of fire, connect Formosa with
the Philippines of which the smallest and most broken islands
abound most in volcanoes. Leopold von Buch enumerates
nineteen lofty isolated conical mountains upon them, which
in the country are called volcanes, though probably some of
them are closed trachytic domes. Dana is of opinion that
in southern Luzon there are only two active volcanoes — that
of Taal, which rises in the Laguna de Bongbong, with an en-
circling escarpment which incloses another lagoon (see page
232) ; and in the southern portion of the peninsula of Cama-
rines the volcano of Albay, or Maygn, which the natives call
Isaroe. The latter, which is 3197 feet high, experienced
great eruptions in the years 1800 and 1814. In the northern
portion of Luzon granite and mica-slate, and even sediment-
ary formations, together with coal, are diffused, f

* Compare my Fragmens de Geologic et de CUmatologie Asiatiques^
t. i., p. 82, which appeared immediately after my return from my Si-
berian expedition, and the Asie Centrale, in which the opinion ex-
pressed by Klaproth, and which I formerly adopted, respecting the
probability of the connection of the snowy mountains of the Himalaya
with the Chinese province of Yiman and with Nanling, northwest-
ward of Canton, has been confuted by me. The mountains of For-
mosa, upward of 11,000 feet high, as well as Ta-yu-ling, which bounds
Fukian to the Avestward, belong to the system of meridian fissures of
Upper Assam, in the country of the Burmese, and in the group of , the
Philippines.

f Dana's Geology, in theExplor. Exped., yo\. x., p. 540-545; Ernest

TRUE VOLCANOES. 355

The far-stretching group of the Soolo (Solo) Islands, which
Are fully one hundred in number, and which connect Minda-
nao and Borneo, is partly volcanic, and partly intersected by
coral-reefs. Isolated unopened trachytic cone-shaped peaks
are indeed often called Vulcanes by the Spaniards.

If we carefully examine all that lies to the south of the
fifth degree of north latitude (to the south of the Philippines)
between the meridians of the* Nicoba'rs and the northwest
of New Guinea, thus taking in the Sunda Islands, great and
small, and the Moluccas, we shall find as the result, given in
the great work of Dr. Junghuhn, that *'in a circle of islands
which surround the almost continental Borneo there are one
hundred and nine lofty fire-emitting mountains, and ten mud
volcanoes." This is not merely an approximate calculation,
but an actual enumeration.

Borneo, the Giava Maggiore of Marco Polo,* has hitherto
furnished us with no certain proofs of the existence of any
active volcano upon it ; but, indeed, it is only a few narrow
strips of the shore that we are acquainted with (on the
northwest side, as far as the small coast-island of Labuan,
and as far as Cape Balambangan ; on the west coast at the
mouth of the Pontianak j and on the southeastern point in
the district of Banjermas-Sing, on account of the gold, dia-
mond, and platinum washings). It is not even believed that
the highest mountain of the whole island, and perhaps even
of the whole South Asiatic island world, the double-peaked

Hofmann, Geogn. Beoh.avfder Reise von Otto v. Kotzebue, p. 70; Leop.
de Buch, Description Physique des Iks Canaries, p. 435-439. See the
large and admirable chart of the Islas Filipinas, by the Pilot Don An-
tonio Morati (Madrid, 1852), in two plates.

* Marco Polo distinguishes (Part iii., cap. 5 and 8) Giava Minore
(Sumatra), where he remained for five months, and where he de-
scribes the elephants, which were i^to be found in Java itself (Hum-
boldt, Examen Grit, de IHist. de la mogr., t. ii., p. 218), from what he
had before described as Giava (Maggiore), la quale, secondo dicono i mari-
nai, che bene lo sanno, e Visola piii grande che sia al mondo — which, as the
sailors say, who know it well, is the largest island in the world. This
assertion is even to this day true. From the outlines of the chart of
Borneo and Celebes, by James Brooke and Captain Rodney Mundy,
I find the area of Borneo 51,680 square geographical miles, nearly
equal to that of the island of New Guinea, but only one tenth of the
continent of New Holland. Marco Polo's account of the great quantity
of gold and treasure which the "Mercanti di Zaiton e del Mangi" ex-
ported from thence shows that by Giava Maggiore he meant Borneo
(as also did Martin Behaim on the Niirnberg globe of 1492, and Johann
Euysch in the RonH^ edition of Ptolemy, dated 1508, which is so im-
portant for. the history of the discovery of America.)

356

COSMOS.

Kina Bailu at the northern extremity, distant only thirty-
two geographical miles from the Pirate coasts, is a volcano.
Captain Belcher makes it 13,695 feet high, which is nearly
4000 feet higher than the Gunung Pasaraan (Ophir) of Su-
matra.* On the other hand Rajah Brooke mentions a much
lower mountain in the province of Sarawak, whose name,
Gunung Api (Fire Mountain in the Malay tongue), as well as
the scoriae which lie around it, lead to the conclusion that it
was once volcanically active. Large deposits of gold sand
between quartz veins, the abundance of tin washed down on
both shores of the rivers, and the feldspathic porphyryf of
the Carambo Mountains, indicate a great extenSion of what
are called primitive and transition rocks. According to the
only certain information which we possess from a geologist
(Dr. Ludwig Horner, son of the meritorious Zurich astrono-
mer and circumnavigator of the globe), there are found in
the southeastern portion of Borneo, united in several profita-
bly worked washings, precisely as in the Siberian Ural, gold,
diamonds, platinum, osmium, and iridium (but not yet palla-
dium). Formations of serpentine, euphotide, and syenite, ly-
ing in great proximity, belong to a range of rocks 3411 feet
high, that of the Ratuhs Mountains. {

The still active volcanoes on the remaining three great
Sunda Islands are reckoned by Junghuhn as follows : On
Sumatra from six to seven, on Java from twenty to twenty-
three, on Celebes eleven, and on Flores six. Of the volca-
noes.of the island of Java we have already (see above page
281) treated in detail. In Sumatra, which has not hitherto
been completely investigated, out of nineteen conical mount-
ains of volcanic appearance there are six still active.^ Those
ascertained to be so are the following : The Gunung Indra-
pura, about 12,256 feet in height, according to angles of al-
titude measured from the sea^ and probably of equal height
with the more accurately measured Semeru or Maha-Meru,

* Captain Mundy's chart (coast of Borneo Proper, 3847,) gives, it
is true, 14,000 English feet. See a doubt of this datum in Junghuhn's
Java, bd. ii., s. 580. The colossal Kina Bailu is not a conical mount-
ain. In shape it much more resembles the basaltic mountains which
occur under all latitudes, and which form a long ridge with two term-
inal summits.

t Brooke's Borneo and Celrbes, vol. ii., p. 382, 384, and 386.

X Horner, in the V^erhandelingen van het Bataviaasch Genootschap van
Kunsten en Wetenschappen, Deel xvii. (1839), s. 284; Asie Centrale, t.
ii., p. 534-537. H

§ Junghuhn, Java, bd. ii., s. 809 (Battaldnder^d. i., S..39).

TRUE VOLCANOES. 357

on Java ; the Gunung Pasaman, called also Ophir (9602
» feet), with a nearly extinguished crater, ascended by Dr. L.
Horner ; the sulphureous Gunung Salasi, with eruptions of
ashes in 1833 and 1845 ; the Gunung Merapi (9751), also
ascended by Dr. L. Horner, accompanied by Dr. Korthal, in
the year 1834, the most active of all the volcanoes of Suma-
tra, and not to be confounded with the two similarly-named
mountains of Java ;* the Gunung Ipu, a smoking truncated
cone ; and the Gunung Dempo, in the inland country of Ben-
kula, reckoned at 9940 feet high.

Four islets forming trachitic cones, of which the Pic Re-
cata and Panahitam (Prince's Island) are the highest, rise
above the sea in the Strait of Sunda, and connect the vol-
canic range of Sumatra with the crowded field of Java ; and
in like manner, the eastern extremity of Java, with its vol-
cano of Idjen, forms, through the medium of the active vol-
canoes of Gunung Batur and Gunung Agung, on the neigh-
boring island of Bali, a connection with the long chain of
the smaller Sunda Islands. Here, again, the range is con-
tinued eastward from Bali, by the smoking volcano of Rind-
jani, on the island of Lombok, 12,363 feet high, according
to the trigonometrical measurement of M. Melville de Carn-
bee ; by the Temboro (5862 feet), on the Sumbava, or Sam-
bava, whose eruption of ashes and pumice in April, 1815,
obscured the surrounding atmosphere, and was one of the
greatest which history has recorded ;t and by six conical
mountains still partially smoking, on Flores

The large and many-armed island of Celebes contains six
Volcanoes, which are not yet all extinct ; they lie all together,
on the narrow northeastern peninsula of Menado. Beside it
spout out streams of hot melted sulphur, into the orifice of
one of which, near the road from Sonder to Lamovang, a
great traveler and intrepid observer. Count Carlo Vidua, iml
Piedmontese friend, sank and met his death fpom the buraf
he received. As the small island of Banda, in the Moluccas,
consists of the volcano of Gunung Api, which was active
from 1586 to 1824, and is about 1812 feet high, in the same
way the larger island of Ternate is likewise formed by a sin-
gle conical mountain, 5756 feet high, the Gunung Gama
Lama, whose violent eruptions from 1838 to 1849, after
more than a century and a half of entire quiescence, are de-
scribed at ten different periods. During the eruption of the
3d of February, 1840, according to Junghuhn, a stream of

♦ See page 283, note t- ^ Java, bd. ii., s. 818-828.

358 COSMOS.

lava poured out of a fissure near the fort of Toluko, and
flowed down to the shore,* " partly issuing in the form of a
connected and thoroughly molten stream, and partly consist-
ing of glowing fragments which rolled down and were forced
along the plain by the weight of the succeeding masses." If
to the more important volcanic cones here individually men-
tioned we add the numerous small island volcanoes which
can not be here noticed, the total number of the igneous
mountains situated to the southward of the parallel of Cape
Serangami, on Mindanao, one of the Philippines, and between
the meridians of the northwest Cape of New Guinea on the
east, and of the Nicobar and Andaman groups on the west,
amounts, as has been already stated, to the large number of
109.t This calculation is made in the belief that "on Java
forty-five volcanoes, for the most part cone-shaped, and pro-
vided with craters, may be counted." Of these, however, only
21, and only 42 to 45, of the whole number of 109, are recog-
nized as now active, or as having been so at any period with-
in the range of history. The mighty Pic of Timor formerly
served like Stromboli as a light-house to mariners. On the
small island of Pulu Batu (called also P. Komba), a little to
the north of Floris, a volcano was seen in 1850 to pour a
stream of glowing lava down to the sea-shore. The same
thing was observed in 1812, and again in the spring of 1856,
in respect to the Pic on the greater Sangir Island, between
Magindaiiao and Celebes. Junghuhn doubts whether the
famous conical mountain of Vavani or Ateti, on Amboina,
ejected any thing more than hot mud in 1674, and con-
siders the island at present as only a solfatara. The great
group of the South Asiatic islands is connected by the divi-
sion of the Western Sunda Islands with the Nicobar and
Andaman, Isles of the Indian Ocean, and by the division of
,e Moluccas and Philippines with the Papuas, the Pellew
'ands and €)arolinas of the South Sea. We shall first,
however, proceed with the less numerous and more dispersed
groups of the Indian Ocean.

VII. Vhe Indian Ocean.

This comprehends the space between the west coast of the
peninsula of Malacca, or of the Birman country to the east
coast of Africa, thus inclosing in its northern division the
Bay of Bengal and the Arabian and Red Seas. We pursue
the chain of volcanic activity in the Indian Ocean in the
direction from northeast to southwest.

* Junghuhn's Java^ vol. ii., p. 840-842. t Ibid.^ p. 853.

TRUE VOLCANOES. 359

Barren Island, in the Bay of Bengal, a little to the east
of the great Andaman Island (lat. 12° 15'), is correctly con-
sidered an active cone of eruption, issuing out of a crater of
upheaval. The sea forces its way through a narrow open-
ing and fills an internal basin. The appearance presented
by this island, which was discovered by Horsburgh in 1791,
is exceedingly instructive for the theory of the formation of
volcanhs structures. We see here in a complete and perma-
nent form what nature exhibits in only a cursory way at
8antorin, and at other points of the earth's surface.* The
eruptions in November, 1803, were, like those of Sangay, in
the Cordilleras of Quito, very distinctly periodical, recurring
at intervals often minutes (Leop. von Buch, in the AbhandL
derBerl Akademie, 1818-1819, s. 62).

The island of Narcondam, to the north of Barren Island,
has likewise exhibited volcanic action at a former period, as
has also the cone mountain of the island of Cheduba, which
lies more to the north, near the shore of Arracan (10° 52^).
(Silliman's American Journal, vol. xxxviii., p. 385.)

The most active volcano, judging from the frequency of
the lava eruptions, not only in the Indian Ocean, but in al-
most the whole of the south hemisphere between the merid-
ians of the west coast of New Holland and the east coast of
America, is that on the island of Bourbon, in the grOup of
the Mascareignes. The greater part of the island, particu-
larly the western portion and the interior, is basaltic. Ke-
cent veins of basalt, with little admixture of olivin, run
through the older rock, which abounds in olivin ; beds of
lignite are also inclosed in the basalt. The culminating
points of the Mountain Island are the Gt'os Morne and the
Trois Salazes, the height of which La Caille overestimated at
10,658. The volcanic action is now limited to the southern-,
most portion, the "Grand pays brule." The summit of the
volcano of Bourbon, which Hubert describes as emitting,
nearly every year, two streams of lava, which frequently ex-
tend to the sea, is, according to Berth's measurement, 8000
feet high."t It exhibits several cones of eruption which have
received distinct names, and which alternately send forth
eruptions. The eruptions from the summit are infrequent.

* Leop. von Buch, in the AbhandL der Akad. der Wiss. zu Berlin,
1818 and 1819, s. 62 ; Lyell, Princ. of Geology (1853), p. 447, where
a fine representation of the volcano is given.

t Bory de St. Vincent, Voyage aux Quatre Isles d'Afrique, t. ii., p.
429.

o60 COSMOS.

The lavas contain glassy feldspar, and are therefore rather
trachytic than basaltic. The shower of ashes frequently con-
tains olivin in long, fine threads, a phenomenon which like-
wise occurs at the volcano of Owhyhee. A violent eruption
of these glassy threads, covering the whole island of Bour-
bon, occurred in the year 1821.

All that we know of the great neighboring terra incog-
nita of Madagascar is the extensive dispersion of puITiice at
Tintingue, opposite the French island of St. Marie, and the
occurrence of basalt, to the south of the Bay of Diego Sua-
rez, near the northernmost Cap d'Ambre, surrounded by
granite and gneiss. The southern central ridge of the Am-
bohistmene Mountains is calculated (though with little cer-
tainty) at about 11,000 feet. Westward of Madagascar, in
the northern outlet of the Mozambique Channel, the largest
of the Comoro Islands has a burning volcano (Darwin, Coral
Reefs, p. 122).

the small volcanic island of St. Paul (38° 38^, south of
Amsterdam, is considered volcanic, not only on account of
its form, which strongly reminds us of that of Santorin, Bar-
ren Island and Deception Island, in the group of the New
Shetland Isles ; but likewise on account of the repeatedly-
observed eruptions of fire and vapor in modern times. The
very characteristic drawing given by Valentyn in his work
on the Banda Islands, relative to the expedition of Willem
de Vlaming (November, 1696), corresponds exactly, as do
also the statements of the latitudes, with the representations
in the atlas of Macartney's expedition and Captain Black-
wood's survey (1842). The crater-shaped, circular bay, near-
ly an English mile across, is every where surrounded by pre-
cipitous rocks which fall perpendicularly in the interior, with
the exception of a narrow opening, through which the sea
enters at flood-tide ; while those which form the margin of
the crater fall away externally, with a gentle slope.*

The island of Amsterdam, which lies 50'' of latitude far-
ther toward the north (37° 48^), consists, according to Val-
entyn's representation, ^f a single, well-wooded, somewhat
rounded mountain, from the highest ridge of which rises a
small cubical rock, almost the same as at the Cofre de Pe-
rote, on the higher plains of Mexico. During the expedition
of D'Entrecasteaux (March, 1792), the island was seen for
two whole days entirely enveloped in flames and smoke.

* Valentyn, Beschryvirig van Oud en Nieiiw Oost Indim, Deel ill.,
(1726), p, 70 ; Het Eyland St. Paulo. (Compare Lyell, Princ, p. 446.)

TRUE VOLCANOES. 361

The smell of the smoke seemed to indicate the combustion
of wood and earth ; columns of vapor were indeed thought to
rise here and there from the ground near the shore, but the
naturalists who accompanied the expedition were decidedly
of opinion that the mysterious phenomenon could by no
means be ascribed to an eruption* of the high mountain, like

* We were unable," says D'Entrecasteaux, " to form any conjectil^e
as to the cause of the burning on the island of Amstei'dam. The isl-
and was in flames throughout its whole extent, and we recognized
distinctly the smell of burned wood and earth. We had felt nothing
to lead us to suppose that the fire was the eifect of a volcano" (t. ii.,
p. 45). A few pages before, he says, " We remarked, however, as we
sailed along the coast, from which the flames were rather distant, lit-
tle puffs of smoke, which seemed to come from the earth like jets ; yet
we could not distinguish the least trace of fire around them, though
we were very close to the land." These jets of smoke, which appeared
at intervals, were considered by the naturalists of the expedition as cer-
tain proofs of subterranean fire. Are we to conclude from this that
there were actual combustions of earth — conflagrations of lignite, the
beds of which, covered with basalt and tufa, occur in such abundance
on volcanic islands (as Bourbon, Kerguelen-land, and Iceland)? The
Surtarbi-and, on the latter island, derives its name from the Scandi-
navian myth of the fire-giant Surtr causing the conflagration of the
world. The combustion of earth, however, causes no flame, in gfen-
eral. As in modern times the names of the island of Amsterdam'^nd
St. Pauf are unfortunately often confounded on charts, I would here
observe, in order to prevent mistakes in ascribing to one observations
which apply to the other, they being very different in formation,
though lying almost under one and the same meridian, that originally
(as early as the end of the 17th century) the south island was called
St. Paul and the northern one Amsterdam. Vlaming, their discov-
erer, assigned to the first the latitude of 38° 40', and to the second
that of 37° 48' south of the equator. This corresponds in a remark-
able manner with the calculation made by D'Entrecasteaux a century
later, on the occasion of the expedition in search of La Perouse ( Voy-
age, t. i., p. 43-45), namely, for Amsterdam, according to Beautemps-
Beaupre, 37° 47' 46" (long. 77° 71')ffor St. Paul 38° 38'. This near
coincidence must be considered accidental, as the points of observa-
tion were cwtainly not exactly the same. On the other hand Captain
Blackwood, in his Admiralty chart of 1842, gives 38° 44', and longi-
tude 77° 37' for St. Paul. On the charts given in the original editions
of the voyages of the immortal circumnavigator Cook — those, for in-
stance of the first and second expedition ( Voyage to the Sotith Pole and
Round the World, London, 1777, p. 1), as well as of the third and last
voyage ( Voyage to the Pacific Ocean, published by the Admiralty, Lon-
don, 1784, in 2d edition, 1785), and even of all the three expeditions
(^A General Chart, exhibiting the Discoveries of Captain Cook in his Third
and Two Preceding Voyfiges, by Lieutenant Henry Roberts) — the isl-
and of St. Paul is very correctly laid down as the most southernly of
the two ; but in the text of the voyage of D'Entrecasteaux (t. i., p. 44)
it is mentioned, by way of censure (whether with justice or not I am
unable to say, although I have sought after the editions in the libraries
Vol. v.— Q

362 COSMOS.

that of a volcano. More certain evidences of former genuine
volcanic action on the island of Amsterdam may be found in
the beds of pumice-stone {uitgehranden jmimsteen), mention of
which is made so early as by Valentyn, according to Vla-
ming's Ship Journal of 1696.

To the southeast of the Cape of Good Hope lie Marion's,
or Prince Edward's Island (47° 2^), and Possession Island
(lat. 46° 28^, and long. 51° 560, forming part of the Grozet
group. Both of them exhibit traces of former volcanic ac-
tion— small conical bills,* with eruption openings surround-
ed by columnar basalt.

More eastward, and almost in the same latitude, we come

of Paris, Berlin, and Gottingen), " that on the special chart of Cook's
last expedition the island of Amsterdam is set down as more to the
south than St. Paul." A similar reversal of the appellations, quite
opposed to the intention of the discoverer, Willem de Vlaming, was
frequent in the first third of the present century — as, for example, on
the older and excellent maps of the world by Arrowsmith and Purdy
(1833) — but there was more than a special chart of Cook's third voy-
age operating to cause it. There was, 1st, the arbitrary entry on the
maps of Cox and Mortimer ; 2d, the circumstance that, in the atlas
of Lord Macartney's voyage to China, though the beautiful volcanic
island represented smoking is very correctly named St. Paul, under
lat. 158° 42', yet it is absurdly added, " commonly called Amsterdam,"
and, what is still worse, in the narrative of the voyage itself, StauiUtin
and Dr. Gillan uniformly called this " island still in a state of inflam-
mation" Amsterdam, and they even add (p. 226, after having given the
correct latitude in p. 219) "that St. Paul is lying to the northward of
Amsterdam ; and, 3d, there is the same confusion of names by Bar-
row {Voyage to Cochin CJiina in the Years 1792 and 1793, p. 140-157),
who also gives the name of Amsterdam to the southern island, emit-
ting smoke and flames, assigning to it at the "same time the latitude
38° 42'. Malte-Brun {Precis de la Gcographie Universelle, t. v., 1817,
p. 146) very properly blames Barrow, but he errs in also blaming M.
de Rossel and Beautemps-Beaupre. Both of the latter writers give as
the latitude of the island of Am^h*dam, which is the only one they rep-
resent, 37° 47', and that of the island of St. Paul, because it lies 50'
more to the south, 38° 38' {Voy. de D^Entrecasteaux, 1808, t. i., p.
40-46) ; and to show that the design represents the true island of Am-
sterdam, discovered by Willem de Vlaming, Beautemps-Beaupre adds
in his atlas a copy of the thickly-wooded island of Amsterdam from
Valentyn. I may here observe that, the celebrated navigator, Abel
Tasman, having in 1642, along with Middelburg, called the island of
Tonga-Tabu (lat. 21^°), in the Tonga group, by the name of Amster-
dam (Burney, Chronolog. Hist, of the Voyages and Discoveries in the
South Sea or Pacific Ocean, part iii., p. 81 and 437), he has also been
sometimes erroneously cited as the discoverer of Amsterdam and St.
Paul, in the Indian Ocean. See Leidenfrost, Histor. Handivortenbuchy
bd. v., s. 310.

* Sir James Ross, Voyage in the Southern and Antarctic Regions, vol.
i., p. 46, and 50-56.

TRUE VOLCANOES. 363

to Kerguelen's Island (Cook's Island of Desolation), for the
first geological account of which we are indebted to the suc-
cessful and important expedition of Sir James Ross. In the
harbor called by Cook Christmas harbor (lat. 48° 41^, long.
69° 2^), basaltic lavas, several feet thick, are found inclosing
the fossil trunks of trees ; there also is seen the singular and
picturesque Arched Rock, a natural passage through a narrow
projecting wall of basalt. In the neighborhood are conical
mountains, the highest of which rise to 2664 feet, with ex-
tinct craters — masses of green-stone and porphyry, traversed
by beds of basalt — and amygdalaid with drusy masses of
quartz, at Cumberland Bay. The most remarkable of all are
the numerous beds of coal, covered with trap-rock (dolerite, as
at Meissner in Hessian ?), of a thickness of from a few inches
to four feet at the outcrop.*

If we ta^e a general survey of the Indian Ocean, we shall
find the northwesterly extremity of the Sunda range in Su-
matra, which is curved, carried on through the Nicobars and
the Great and Little Andamans ; while the volcanoes of Bar-
ren Island, Narcondam, and Cheduba, almost parallel to the
coasts of Malacca and Tenasserim, run into the eastern por-
tion of the Bay of Bengal. Along the shores of Orissa and
Coromandel, the eastern portion of the bay is destitute of isl-
ands, the great island of Ceylon bearing, like that of Mada-
gascar, more of the character of a continent. Opposite the
western shore of the Indian peninsula (the elevated plain of
Neilgherry and the coasts of Canara and Malabar) a range of
three Archipelagoes, lying in a direction from north to south,
and extending from 14° north to 8° south latitude (the Lac-
cadives, the Maldives, and the Chagos), is connected by the
shallows of Sahia de Malha and Cargados Carajos with the
volcanic group of the Mascareignes and Madagascar. The
whole of this chain, so far as can be seen, is the work of cor-
al polyfes — true Atolls, or lagoon-reefs ; in accordance with
Darwin's ingenious conjecture that at this part a large extent
of the floor of the ocean forms, not an area of upheaval, but
an area of subsidence.

VIII. The South Sea, or Pacific

If we compare that portion of the earth's surface now cov-
ered with water with the aggregate area of the terra firma

* Sir James Boss, Voyage in the Southern and Antarctic Regions, vol.
i., p. 68-82.

364 COSMOS.

(nearly* in the proportion of 2-7 to 1), we can not but be
astonished, in a geological point of view, at the small number
of volcanoes which still continue active in the oceanic region.
The South Sea, the superficies of which is nearly one sixth
greater than that of the whole terra firma of our planet —
which in the equinoctial region, from the Archipelago of
Galapagos to the Pellew Islands, is nearly two fiftlis of the
whole circumference of the earth in breadth — exhibits fewer
smoking volcanoes, fewer openings through which the inte-
rior of the planet still continues in active communion with
its atmospheric envelope than does the single island of Java.
Mr. James Dana, the talented geologist of the great American
exploring expedition (1838-1842), under the command of
Charles Wilkes, basing his views on his own personal investi-
gations, aided by a careful comparison of all previous reliable
observations, and especially by a comprehensive examination
of the different opinions on the forms, the distribution, and
the axial direction of the island groups, on the character of
the difFerQnt kinds of rocks, and the periods of the subsidence
and upheaval of extensive tracts of the floor of the ocean, has
the indisputable merit of having shed a new light over the
island world of the South Sea. . In availing myself of his
work, as well as of the admirable writings of Charles Dar-
win, the geologist of Captain Fitzroy's expedition (1832-
1836), without always particularizing them, I trust that the
high respect in which 1 have for so many years held those
gentlemen will secure me from the chance of having my mo-
tives misinterpreted.

It is my intention to avoid altogether the divisional terms
of Polynesia, Micronesia, Melanesia, and Malaisia,! which are

* The result of Prof. Rigaud's levelings at Oxford, according to Hal-
ley's old piethod. See my Asie Centrale, t. i., p. 189.

> D'Urville, Voy. de la Corvette V Astrolabe, 1826-1829, Atlas, pi. i.
— 1st. Polynesia is considered to contain tl;e eastern portion of the
South Sea (the Sandwich Islands, Tahiti, and the Tonga Arihipelago ;
.and also New Zealand); 2. Micronesia and Melanesia form the west-
ern portion of the South Sea ; the former extends from Kauai, the
westernmost island of the Sandwich group, to near Japan and the
Philippines, and reaches south to the equator, comprehending the Ma-
rians (Ladrones), the Carolinas and the Pellew Islands ; 3d. Melane-
sia, so called from its dark-haired inhabitants, bordering on the Malai-
sia to the northwest, embraces the small Archipelago of Viti, or Fee-
jee, the New Hebrides and Solomon's Islands ; likewise the larger isl-
ands of New Caledonia, New Britain, New Ireland, and New Guinea.
The terms Oceania and Polynesia, often so contradictory in a geograph-
ical point of view, are taken from Malte-Brun (1813) and from Lesson
(1828).

TRUE VOLCANOES. 366

not only extremely arbitrary, but founded on totally different
principles drawn from the number and size, or the complex-
ion and descent of the inhabitants, and to commence the enu-
meration of the still active volcanoes of the South Sea with
those which lie to the north of the equator. I shall after-
ward proceed in the direction from east to west, to the isl-
ands situated between the equator and the parallel of 30°
south latitude. The numerous basaltic and trachytic islands,
with their countless craters, formerly at different times erup-
tive, must on no account be said to be indiscriminately scat-
tered.* It is admitted, with respect to the greater number
of them, that their upheaval has taken place on widely ex-
tended fissures and submarine mountain chains, which run in
directions governed by fixed laws of region and grouping, and
which, just as we see in the continental mountain chains of
Central Asia, and of the Caucasus, belong to different sys-
tems; but the circumstances which govern the area over
which at any one particular time the openings are simultane-
ously active, probably depend, from the extremely limited
number of such openings, on entirely local disturbances, to
which the conducting fissures are subjected. The attempt to
draw lines through three now simultaneous volcanoes, whose
respective distances amount to between 2400 and 3000 geo-
graphical miles asunder, without any intervening cases of
eruption (I refer to throe volcanoes now in a state of ignition

* " The epithet scattered, as applied to the islands jof the ocean (in
the arrangement of the (];roups), conve_ys a very incorrect idea of their
positions. There is a system in their arrangement as regular as in the
mountain heights of a continent, and ranges of elevation are indicated
as grand and extensive as any continent presents." Geology, by J.
Dana, United States Exploring Expedition, under command of Charles
Wilkes, vol. x. (1849), p. 12. Dana calculates that there are in the
whole of the South Sea, exclusive of the small rock islands, about 350
basaltic or trachytic and 290 coral islands. He divides them into
twenty-five groups, of which nineteen in the centre have the direction
of their axis N. 50°— 60<^ W., and the remaining N. 20°— 30° E. It
is particularly remarkable that these numerous islands, with a few ex-
ceptions, such as the Sandwich Islands and New Zealand, all lie be-
tween 23° 28' of north and south latitude, and that there is such an
immense space devoid of islands eastward from the Sandwich and the
Nukahiva groups as far as the American shores of Mexico and Peru.
Dana likewise draws attention to a circumstance which forms a con-
trast to the insignificant number of the now active volcanoes, namely,
that if, as is probable, the Coral Islands, when lying between entirely
basaltic islands, have likewise a basaltic foundation, the number of
submarine and subaerial volcanic openings may be estimated at more
than a thousand (p. 17 and 24). ,;

366 COSMOS.

— Mouna Loa, with Kilauea on its eastern declivity; the
cone mountain of Tanna, in the New Hebrides ; and Assump-
tion Island in the North Ladrones), would afford us no in-
formation in regard to the general formation of volcanoes in
the basin of the South Sea. The case is quite different if
we limit ourselves to single groups of islands, and look back
to remote, perhaps pre-historic, epochs when the numerous
linearly-arranged, though now extinct, craters of the Ladrones
(Marian Islands), the New Hebrides, and the Solomon's Isl-
ands were active, but which certainly did not become gradu-
ally extinguished in a direction either from southeast to north-
west or from north to south. Though I here name only vol-
canic island chains of the high seas, yet the Aleutes and oth-
er true coastian islands are analogous to them. General con-
clusions as to the direction of a cooling process are deceptive,
as the state of the conducting medium must operate tempo-
rarily upon it, according as it is open or interrupted.

Mouna Loa, ascertained by the exact measurement* of the
American exploring expedition under Captain "Wilkes to be
13,758 feet in height, and consequently 1600 feet higher than
the Peak of Teneriffe, is the largest volcano of the South Sea
Islands, and the only one that still remains really active in
the whole volcanic Archipelago of the Hawaii or Sandwich
Islands. The summit craters, the largest of which is nearly
13,000 feet in diameter, exhibit ifi their ordinary state a solid
bottom, composed of hardened lava and scoriae, out of which
rise small cones of eruption, exhaling vapor. The summit
openings are, on the whole, not very active, though in June,
1832, and in January, 1843, they emitted eruptions of sever-
al weeks' duration, and even streams of lava of from 20 to 28
geographical miles in length, extending to the foot of Mouna
Kea. The fall (inclination) of the perfectly connected flow-
ing stream! was chiefly 6°, frequently 10°, 15°, and even 25°.
The conformation of the Mouna Loa is very remarkable, from
the circumstance of its having no cone of ashes, like the Peak
of Teneriffe, Cotopaxi, and so many other volcanoes ; it is
likewise almost entirely deficient in. pumice, { though the

* See Cosmos, vol. v., p. 238, note J.

t Dana, Geology of the United States Explor. Exped., p. 208 and 210.

X Dana, p. 193 and 201. The absence of cinder-cones is likewise
very remarkable in those volcanoes of the Eifel which emit streams of
lava. Reliable information, however, received by the missionary Dib-
ble from the mouths of eye-witnesses, proves that an eruption of ashes
may notwithstanding occur from the summit crater of Mouna Loa, for
he was told that, during the war carried on by Kamehameha against

TRUE VOLCANOES. 367

blackish-gray, and more trachytic than basaltic, lavas of the
summit abound in feldspar. The extraordinary fluidity of
the lavas of Mouna Loa, whether issuing from the summit
crater (Mokua-weo-weo) or from the sea of lava (on the east-
ern declivity of the volcano, at a height of only 3969 feet
above the sea), is testified by the glass threads, sometimes
smooth and sometimes crisped or curled, which are dispersed
by the wind all over the island. This hair glass, which is
likewise thrown out by the volcano of Bourbon, is called
Fele's hair by the Hawaiians, after the tutelary goddess of
the country.

Dana has ably demonstrated that Mouna Loa is not the
central volcano of the Sandwich Islands, and that Kilauea is
not a solfatara.* The basin of Kilauea is 16,000 feet (about
2§ geographical miles) across its long diameter, and 7460 feet
across its shorter one. The steaming, bubbling, and foaming
mass which forms the true lava pool does not, however, under
ordinary circumstances, fill the whole of this cavity, but mere-
ly a space whose long diameter measures 14,000 feet and its
breadth 5000 feet. The descent to the edge of the crater is
graduated. This great phenomenon produces a wonderful
impression of silence and solemn repose. The approach of
an eruption is not here indicated by earthquakes or subterra-
nean noises, but merely by a sudden rising and falling of the
surface of the lava, sometimes to the extent of from 300 or
400 feet up to the complete filling of the whole basin. If,
disregarding the immense difference in size, we were to com-
pare the gigantic basin of Kilauea with the small side craters
(first described by Spallanzani) on the declivity of Stromboli,
at four fifths of the height of the mountain, the summit of

the insurgents in the year 1789, an eruption of hot ashes, accompanied
by an earthquake, enveloped the surrounding country in the darkness
of night (p. 183). On the volcanic glass threads (the hair of the god-
dess Fele, who, before she went to settle at Hawaii, inhabited the now
extinct volcano of Hale-a-Kala — or the House of the Sun — on the isl-
and of Maui) see p. 179 and 199-200.

* Daaa, p. 205. "The term Solfatara is wholly misapplied. A sol-
fatara is an area with streaming fissures and escaping sulphur vapors,
and without proper lava ejections ; while Kilauea is a vast crater with
extensive lava ejections and no sulphur, except that of the sulphur
banks, beyond what necessarily accompanies, as at Vesuvius, violent
volcanic action." The structural frame of Kilauea, the mass of the
great lava basin, consists also, not of beds of ashes or fragmentary
rocks, but of horizontal layers of lava, arranged like limestone. Dana,
p. 193. (Compare Strzelecki, Phys. Descr, of New South Wales, 1845,
p. 105-111.)

368 COSMOS.

which has no opening — that is to say, with basins of boiling
lava of from 30 to 200 feet in diameter only — we must not
forget that the fiery gulfs on the slope of Stromboli throw
out ashes to a great height, and even pour out lava. Though
the great lava lake of Kilauea (the lower and secondary cra-
ter of the active volcano of Mouna Loa) sometimes threatens
to overflow its margin, yet it never actually runs over so as
to produce true streams of lava. These occur by currents
from below, through subterranean channels, and the forma-
tion of new eruptive openings at a distance of from 16 to 20
geographical miles, consequently at points very much lower
than the basin. After these eruptions, occasioned by the
pressure of the immense mass of lava in the basin of Kilauea,
the fluid surface sinks in the basin.*

Of the two other high mountains of Hawaii, Mouna Kea
and Mouna Hualalai, the former is, according to Captain
Wilkes, 190 feet higher than Mouna Loa. It is a conical
mountain on whose summit there no longer exists any term-
inal crater, but only long extinct mounds of scoriae. Mouna
Hualalai* is fully 10,000 feet high, and is still burning. In
the year 1801 an eruption took place, during which the lava
reached the sea on the western side. It is to the three colos-
sal mountains of Loa, Kea, and Hualalai, which rose from
the bottom of the sea, that the island of Hawaii owes its
origin. In the accounts given of the numerous ascents of
Mouna Loa, among which that of the expedition of Captain
Wilkes was based on investigations of twenty-eight days' du-
ration, mention is made of falls of snow with a degree of cold
from 23 to 17^ Fahr. above zero, and of single patches of
snow, which could be distinguished with the aid of the teles-

* This remarkable sinking of the surface of the lava is confirmed
by the relations of numerous voyagers, from Ellis, Stewart, and Doug-
las to the meritorious Count Strzelecki, Wilkes's expedition and the
remarkably observant missionary Coan. During the great eruption of
June, 1840, the connection of the rise of the lava in the Kilauea with
the sudden inflammation of the crater of Arare, situated so far below
it, was most decidedly shown. The disappearance of the lava poured
forth from Arare, its renewed subterranean course, and final reappear-
ance in greater quantity, do not quite admit of an absolute conclusion
as to identity, because numerous lava-yielding longitudinal fissures
opened simultaneously below the line of the floor of the Kilauea basin.
It is likewise veiy worthy of observation, as bearing on the internal
constitution of this singular volcano of Hawaii, that in June, 1832,
both craters, that of the summit and that of Kilauea, poured out and
occasioned streams of lava, so that they were simultaneously actire.
(Compare Dana, p. 18-4, 188, 193, and 196.)

TRUE VOLCANOES. 369

cope at the summit of the volcano, but nothing is ever said
of perpetual sno\v.(*) I have already observed, in a former
part of this work, that the Mouna Loa (13,758 feet) and the
ilouna Kea (13,950 feet) are respectively more than 1000 and
821 feet lower than the lowest limit of perpetual snow, as
found by me in the continental mountains of Mexico under
19^° latitude. On a small island the line of perpetual snow
should lie somewhat lower, on account of the less elevated
temperature of the lower strata of air in the hottest season
of the tropical zone, and on account of the greater quantity
of water held in solution in the upper atmosphere.

The volcanoes of Tafoa* and Amargura* in the Tonga
group are both active, and the latter had a considerable erup-
tion of lava on the 9th of July, 1847. t It is extremely re-
markable, and is in entire accordance with the stories of the
coral animals avoiding the shores oT volcanoes, either at the
time or shortly before, in a state of ignition, that the Tonga
islands of Tafoa and the cone of Kao, which abound in coral
reefs, are entirely destitute of those creatures.^

Next follow the volcanoes of Tanna* and Ambrym,* the
latter westward of Mallicollo, in the Archipelago of the New
Hebrides. The volcano of Tanna, first described by Keinhold
Forster, was found in a full state of eruption on Cook's dis-
covery of the island in 1774. It has since remained con-
stantly active. Its height being only 458 feet, it is one of
the lowest fire-emitting cones, along with the volcano of
Mendana, hereafter to be noticed, and the Japanese volcano
of Kosima. There is a great quantity of pumice on Mal-
licollo.

Matthew's Rock,* a very small smoking rock island, about
1183 feet high, the eruption of which was observed by D'Ur-
ville in January, 1828. It lies eastward of the southern
point of New Caledonia.

The volcano of Tinakoro,* in the group of Vanikoro or
Santa Cruz.

In the same Archipelago of Santa Cruz, fully 80 geograph-
ical miles N.N.W. of Tinakoro, the volcano* seen by Men-
dana so early as 1595 rises out of the sea to a height of about
213 feet (lat. 10° 23^ S.). Its eruptions have sometimes

(*) Wilkes, p. 114, 140, and 157 ; Dana, p. 221. From the perpetu-
al transmutation of the r and /, Mauna Loa, is often written Roa, and
Kilauea, Kirauea. t Dana, p. 25 and 138. ,

X Dana, Geology of the United States Exploring Exped.^ p. 138. (See
Darwin, Structure of Coral Reefs, p. 60.)

Q2

370 COSMOS.

been periodical, occurring every ten minutes, and at other
times, as on the occasion of the expedition of D'Entrecas-
teaux, the crater itself and the column of vapor were undis-
tinguishable from each other. '^

In the Solomon's group the volcano of the island of Se^
sarga* is in a state of ignition. On the coast of Guadalca^
nar, in this neighborhood, and therefore also af the southeast
end of the long range of islands toward the Vanikoro or
Santa Cruz group, volcanic eruptive action has likewise been
observed.

In the Ladrones, or Marian Islands, at the north end of
the range, which seems to have been upheaved from a me-
ridian fissure, Guguan,* Pagon,* and the Volcan grande of
Asuncion, are said to be stillin a state of activity.

The direction of the coasts of the small continent of New
Holland, and particularly the dev'iation from that direction
seen in the east coast in 25° south latitude (between Cape
Hervey and Moreton Bay), seem to be reflected in the zone
of the neighboring eastern islands. The great southern isl-
and of New Zealand, and the Kermadec and Tonga groups,
stretch from the southwest to the northeast; while, on the
other hand, the northern portion of the north island of New
Zealand (from the Bay of Plenty to Cape Oton), New Cale-
donia and New Guinea, the New Hebri(^s, the Solomon's
Isles, New Ireland, and New Britain, run in a direction from
S.E. to N.W., chiefly N. 48° W. Leopold von Buch(*) first
drew attention to this relation between continental masses
and neighboring islands in the Greek Archipelago and the
Australian Coral Sea. The islands of the latter sea, too,
are not deficient, as both Forster (Cook's companion) and La
Billardiere formerly observed, in granite and mica-slate, the
quartzose rocks formerly called primeval. Dana has like-
wise collected them on the northern island of New Zealand,
to the west of Tipuna, in the Bay" of Islands.!

New Holland exhibits only on its southern extremity (Aus-
tralia Felix), at the foot and to the south of the Grampian
Mountains, fresh traces of former igneous action, for we learn
from Dana that a number of volcanic cones and deposits of

(*) Leop. von Buch, Description Phys, deslles Canaries, 1836,'p. 893
and 403-405.

t See Dana, 3id., 438-446, and on the fresh traces of ancient vol-
canic action in New Holland, p. 453 and 457 ; also on the. many basaltic
columns in New South Wales and Van Diemen's Land, p. 495-510;
and E. de Strzelecki, Phys. Descr. ofNexo South Wales, p. 112.

TRUE VOLCANOES. 371

lava are found to the northwest of Port Philip, as also in
the direction of the Murray River (Dana, p. 453).

On New Britain* there are at least three cones on the
west coast, which have been observed within the historical
era, by Tasman, Dampier, Cartaret, and La Billardiere, in a
state of ignition and throwing out lava.

There are two active volcanoes in New Guinea,* on the
northeastern coast, opposite New Britain and the Admiralty
Islands, which abound in obsidian.

In New Zealand, of which the geology of the north island
at least has been illustrated by the important work of Ernst
DiefFenbach, and the admirable investigations* of Dana, ba-
saltic and trachytic rocks at various points break through
the generally diffused Plutonic and sedimentary rocks. This
example is the case in a very limited area near the Bay of
Islands (lat. 35° 2''), where the ash-cones, crowded with dis-
tinct craters, Turoto and Poerua rise; and again, more to
the southeast (between 37^° and 39i° lat.), where the vol-
canic floor runs quite across the centre of the north island, a
distance of more than 160 geographical miles from northeast
to southwest, from the Bay of Plenty, on the east, to Cape
Egmont, on the west. This zone of volcanic action liere
traverses (as we have already seen it to do on a much larger
scale in the Mexican Continent), in a diagonal fissure from
northeast to southwest, the interior chain of mountains which
runs lengthwise in a north and south direction, and which
seems to give its form to the whole island. On the ridge of
this chain stand, as it were, at the points of intersection, the
lofty cone of Tongariro^^ (6198), whose crater is found on
the top of the ash-cone, Bidwill, and, somewhat more to the
south, Ruapahu (9006 feet). The northeast end of the zone
is formed in the Bay of Plenty (lat. 38^) by a constantly
smoking solfatara, the island volcano of Puhia-i-wakati(*)*
(White Island). Next follow to the southwest, on the shore
itself, the extinct volcano of Putawaki (Mount Edgecombe),
8838 feet high, probably the highest snowy mountain on New
Zealand ; and in the interior, between Mount Edgecombe and
the still burning Tongariro,* which has poured forth some
streams of lava, a lengthened chain of lakes, partly consist-
ing of boiling water. The lake of Taupo, which is surround-

(*) Ernst Dieffenbach, Travels in New Zealand, 1843, vol. i., p. 337,
355, and 401. DiefFenbach calls White Island "a smoking solfatara,
but still in volcanic activity" (p. 358 and 407), and on the chart, " in
continual ignition."

372 COSMOS.

ed by beautiful glistening leucite and sanidine sand, as well
as by mounds of pumice, is nearly 24 geographical miles long,
and lies in the centre of the north island of New Zealand, at
an elevation, according to Dieffenbach, of 1337 feet above
the surface of the sea. The ground for two English square
miles round is entirely covered with solfataras, vapor holes,
and thermal springs, the latter of which form, as at the Gey-
ser, in Iceland, a variety of silicious precipitates. (*) West-
^ward of Tongariro,* the chief seat of volcanic action, whose
crater still ejects vapors and pumice-stone ashes, and at a dis-
tance of only sixteen miles from the western shore, rises the
volcano of Taranaki (Mount Egmont), 8838 feet high, which
was first ascended and measured by Dr. Ernst DieiFenbach in
November, 1840. The summit of the cone, which in its out-
line more resembles Tolima than Cotopaxi, terminates in a
plain, out of which rises a steep ash-cone. No traces of pres-
ent activity, such as are seen on the volcano of the White
Island* and on Tongariro,* are visible, nor any connected
stream of lava. The substance composed of very thin scales,
and having a ringing sound, which is seen projecting with
sharp points like fish-bones, from among the scoriae, in the
same manner as on one side of the Peak of TenerifFe, resem-
bles porphyritic schist, or clink-stone.

A narrow, long-extended, uninterrupted accumulation of
island groups, erupted from northwestern fissures, such as
New Caledonia and New Guinea, the New Hebrides and
Solomon's Island, Pitcairn, Tahiti, and the Paumotu Islands,
traverses the great Ocean in the Southern hemisphere in a
direction from west to east, for a length of 5400 geograph-
ical m^es, between the parallels of latitude of 12° and 27°,
from the meridian of the east coast of Australia as far as
Easter Island, and the rock of Sala y Gomez. The western
portions of this crowd of islands (New Britain,* the New
Hebrides,* Vanikoro* in the Archipelago of Santa Cruz, and
the Tonga group*) exhibit at the present time, in the middle
of the nineteenth century, inflammation and igneous action.
New Caledonia, though surrounded by basaltic and other
volcanic islands, has nevertheless nothing but. Plutonic rock,t
as is the case with Santa MariaJ in the Azores, according to

(*) Dana, p. 445-448 ; Dieffenbach, vol. i„ p, 331, 339-341 and 397.
On Mount Egmont, see vol i., p. 131-1.^)7.

t Darwin, Volcaiiic Islands, p. 125; Dana, p. 140.

t L. de Buch, Bescr. des I. Can., p. 365. On the three islands here
named, however, phonolite and basaltic rock are also found along with

TRUE VOLCANOES. 373

Leopold voK Buch, and with Flores and Graciosa, according
to Count Bedemar. It is this absence of volcanic action in
New Caledonia, where sedimentary formations with seams
of coal have lately been discovered, that the great develop-
ment of living coral reefs on its shores is ascribed. The
Archipelago of the Viti, or Feejee Islands, is at once basaltic
and trachytic, though distinguished only by hot springs in
the Savu Bay on Vanua Lebu.* The Samoa group (Navi-
gator's Islands), northeast of the Feejee Islands, and nearly
north of the still active Tonga Archipelago, is likewise ba-
saltic, and is moreover characterized by a countless number
of eruption craters linearly arranged, which are surrounded
by tufa-beds with pieces of coral baked into them. The Peak
of Tafua, on the island of Upolu, one of the Samoa group,
presents a remarkable degree of geognostic interest. It must
not, however, be confounded with the still enkindled Peak of
Tafua, south of Amargura, in the Tonga Archipelago. The
Peak of Tafua (2138 feet), which Dana firstf ascended and
measured, has a large crater entirely filled with a thick for-
est, and crowned by a regularly rounded ash-cone. There is
here no trace of any stream of lava ; yet on the conical mount-
ain of Apia (2576 feet), which is likewise on Upolu, as well
as on the Peak of Fao (3197 feet), we meet with fields of
scoriaceous lava (Malpais of the Spaniards), the surface of
which is, as it were, crimped, and often twisted like a rope.
The lava-fields of Apia contain narrow subterranean cavities.
Tahiti, in the centre of the Society Islands, far more tra-
chytic than basaltic, exhibits, strictly speaking, only the ruins
of its former volcanic frame-work, and it is difficult to trace
the original form of the volcano in those enormous masses,
looking like ramparts and chevaux-de-frise, with perpendicu-
lar precipices of several thousand feet in depth. Of its two
highest summits, Aorai and Orohena, the former was first
ascended and investigated by that profound geologist Dana. J
The trachytic mountain, Orohena, is said to equal -^tna in
height. Thus, next to the active group of the Sandwich Isl-
ands, Tahiti contains the highest rock of eruption in the
whole range of the ocean between the continents of America

Plutonic and sedimentary strata. But these rocks may have made
their appearance above the surface of the sea on the first volcanic up-
heaval of the island from the bed of the ocean. No traces are said to
have been found of fiery eruptions or of extinct volcanoes.

* Dana, p. 343-350. t Dana, p. 312, 318, 320, and 323.

X Leop. von Buch, p. 383 ; Darwin, Vole. IsL, p. 25 ; Darwin, Coral
Reefs, p. 138; Dana, p. 286-305 and 364.

374 COSMOS.

and Asia. There is a feldspathic rock on the small islands
of Borabora and Maurua, near Tahiti, designated by late
travelers with the name of syenite, and by Ellis in his Poly*
nesian researches described as a granitic aggregate of feldspar
and quartz, which, on account of the breaking out of porous,
scoriaceous basalt in the immediate neighborhood, merits a
much more complete mineralogical investigation. Extinct
craters and lava streams are not now to be met with on the
Society Islands. The question occurs : Are the craters on
the mountain tops destroyed; or did the high and ancient
structures, now riven and transformed, continue closed at the
top like a dome, while the veins of basalt and trachyte poured
immediately forth from fissures in the earth, as has probably
been the case at many other points of the sea's bottom ? Ex-
tremes of great viscidity or great fluidity in the matter poured
out, as well as the varying width or narrowness of the fis-
sures through which the effusion takes place, modify the shapes
of the self-forming volcanic mountain strata, and, where fric-
tion produces what is called ashes and fragmentary subdivis-
ion, give rise to small and for the most part transitory cones
of ejection, which are not to be confounded with the great
terminal cinder-cones of the permanent structural frames.

Close by the Society Islands, in an easterly direction, are
the Low Islands, or Paumotu. These are merely coral isl-
ands, with the remarkable exception of the small basaltic
group of Gambler's and Pitcairn's Islands.* Volcanic rock,
similar to the latter, is also found in the same parallel (be-
tween 25° and 27° south latitude), 1260 geographical miles
farther to the east, in the Easter Island (Waihu), and proba-
bly also 240 miles farther east, in the rocks Sala y Gomez.
On Waihu, where the loftiest conical peaks are scarcely a
thousand feet high. Captain Beechey remarked a range of
craters, none of which appeared, however, to be burning.

In the extreme east, toward the New Continent, the range
of the South Sea Island terminates with one of the most act-
ive of all island groups, the Archipelago of Galapagos, com-
posed of five great islands. Scarcely any where else, on a
small space of barely 120 or 140 geographical miles in diam-
eter, has such a countless number of conical mountains and
extinct craters (the traces of former communication between
the interior of the earth and the atmosphere) remained visi-
ble. Darwin calculates the number of the craters at nearly
two thousand. When that talented observer visited the Gala-
* Dana, p. 137.

TRUE VOLCANOES. 375

pagos in the expedition of the Beagle, under Captain Fitzroy,
two of the craters were simultaneously in a state of igneous
eruption. On all the islands, streams of a very fluid lava
may be seen which have forked off into different channels,
and have often run into the sea. Almost all are rich in
augite and olivin ; some, which are more of a trachytic
character, are said to contain albite* in large crystals. It
would be well, in the perfection to which mineralogical sci-
ence is now brought, to institute investigations for the pur-
pose of discovering whether oligoclase is not contained in
these porphyritic trachytes, as at Teneriffe, Popocatepetl, and
Chimborazo, or else labradorite, as at ^tna and Stromboli.
Pumice is entirely wanting on the Galapagos, as at Vesuvius,
where, although it may be present, it is not produced, nor
is hornblende any where mentioned to have been found in
them ; consequently the trachyte formation of Toluca, Ori-
zaba, and some of the volcanoes of Java, from which Dr.
Junghuhn has sent me some well-selected solid pieces of lava
for examination by Gustav Rose, does not prevail here. On
the largest and most westerly island of the Galapagos group,
Albemarle, the cone mountains are ranged in a line, and con-
sequently on fissures. Their greatest height, however, reaches
only to 4636 feet. The Western Bay, in which the Peak of
Narborough, so violently inflamed in 1825, rises in the form
of an island, is described by Leopold von Buchf as a crater
of upheaval, and compared to Santorino. Many margins of
craters on the Galapagos are formed of beds of tufa, which
slope off in every direction. It is a very remarkable circum-
stance, seeming to indicate the simultaneous operation of
some great and wide-spread catastrophe, that the margins of
all the craters are disrupted or entirely destroyed toward the
soiSth. A part of what in the older descriptions is called
tufa, consists of palagonite beds, exactly similar to those of
Iceland and Italy, as Bunsen has ascertained by an exact

* Danvin, Vole. Isl, p. 104, 110-112, and 114. When Darwin says
so decidedly that there is no trachyte on the Galapagos, it is because
he limits the term trachyte to the common feldspar, i. e., to orthoclase,
or orthoclase and sanidine (glassy feldspar). The enigmatical frag-
ments imbaked in the lava of the small and entirely basaltic crater of
James Island contain no quartz, although they appear to rest on a
Plutonic rock (see above, p. 367 et seq.). Several of the volcanic cone
mountains on the Galapagos Islands, have at the orifice a narrow cyl-
indrical, annular addition, exactly like what I saw on Cotopaxi; "in
some parts the ridge is surmounted by a wall or parapet perpendicular
on both sides." Darwin, Vole. IsL, p. 83.

t L. von Buch, p. 376.

376 COSMOS.

analysis of the tufas of Chatham Island.(*) This island, the
most easterly of the whole group, and whose situation is fixed
by careful astronomical observations by Captain Beechey, is,
according to my determination of the longitude of the city of
Quito (78° AV 8^^), and according to Acosta's Mapa de la
Nueva Granada of 1849, 536 geographical miles distant from
the Punta de S. Francisco.

IX. Mexico.

The six Mexican volcanoes, Tuxtla,* Orizaba, Popocate-
petl,* Toluca, Jorullo,* and Colima,* four of which have been
in a state of igneous activity within the historical era, were
enumerated in a former place,! and described in their geog-
nostically remarkable relative position. According to recent
investigations by Gustav Eose, the formation of Chimborazo
is repeated in the rock of Popocatepetl, or great volcano of
Mexico. This rock also consists of oligoclase and augite.
Even in the almost black beds of trachyte, resembling pitch-
stone, the oliglocase is recognizable in very small acute-an-
gled crystals. To this same Chimborazo and TenerifFe forma-
tion belongs the volcano of Colima, which lies far to the west,
near the shore of the South Sea. I have not myself seen this
volcano, but we are indebted to Herr PieschelJ (since the

(*) Bunsen, in Leonhard's Jahrh.filr Mineralogic, 1851, s. 856 ; also in
Poga;end., ylnna/ew der Pftysik, bd. Ixxxiii., s. 223.

t^See above, p. 279-281.

t See Pieschel, Ueher die Vulkane von Mexico, in the Zeitschrift fur
allgem. Erdkunde, bd. vi., 1856, s. 86 and 489-532. The assertion
there made (p. 86), "that never mortal has ascended the steep summit
of the Pico del Fraile," that is to say, the highest peak of the volcano
of Toluca, has been confuted by my barometrical measurement made
upon that very summit (which is, by-the-way, scarcely 10 feet in wid^h)
on the 29th of September, 1803, and published first in 1807, and again
recently by Dr. Gumprecht in the same volume of the journal alaove
referred to (p. 489). The doubt raised on this point was the more
singular, as it was from this very summit of the Pico del Fraile, whose
tower-like sides are certainly not very easy to climb, and at a height
scarcely 600 feet less tlian that of Mont Blanc, that I struck off the
masses of trachyte which are hollowed out by the lightning, and which
are glazed on the inside like vitreous tubes. An essay was inserted
so early as 1819 by Gilbert, in volume Ix. of his Annates der Physik,
(s. 261), on the specimens placed by me in the Berlin Museum, as well
as in several Parisian collections (see also Annates de Chimie et de Phy-
sique, t. xix., 1822, p. 298). In some places the lightning has bored
such regular cylindrical tubes (as much as three inches in length), that
they can be looked through from end to end, and in those cases the
rock surrounding the openings is likewise vitrified. I have also brought
with me pieces of trachyte in my collections, in which the whole sur-

TRUE VOLCANOES. 377

spring of 1855) for a very instructive view of the different
kinds of rocks collected by him, as well as for his interesting
geological notices on the volcanoes of the whole Mexican
highlands, all of which he has personally visited. The vol-
cano of Toluca, whose highest summit (the Pico del Fraile),
though narrow and difficult to climb, I ascended on the 29th
of September, 1803, and found barometrically to be 15,166
feet high, has a totally different mineralogical composition
from the still active Popocatepetl and the igneous mountain
of Colima ; this must not, however, be confounded with an-
other still higher summit, called the Snow mountain. The
volcano of Toluca consists, like the Peak of Orizaba, the Puy
de Chaumont in the Auvergne and -5Cgina, of a combination
of oligoclase and hornblende. From this brief sketch it will
be seen, and it is well deserving of notice, that in the long
range of volcanoes which extend from ocean to ocean there
are not two immediately succeeding each other which are of
similar mineralogical composition.

X. The Northwestern Districts of America (northward
of the parallel of Rio Gila.)

In the section which treats of the volcanic action on the
eastern Asiatic Islands,* particular notice has been drawn
to the bow-like curve in the direction of the fissure of up-
heaval from which the Aleutian Islands have risen, and which
manifests an immediate connection between the Asiatic and
American continents — between the two volcanic peninsulas
Kamtschatka and Aliaska. At this point is the outlet, or
rather the northern boundary, of a mighty gulf of the Pacific
Ocean, which, from the 150 degrees of longitude embraced by
it under the equator, narrows itself down between the term-
inal points .of these two peninsulas to 37° of longitude. On
the American continent, near the sea-shore, a number of more

face is vitrified without any tube-like perforation, as is the case at the
little Ararat and at Mont Blanc. Herr Pieschel first ascended the
double-peaked volcano of Colima, in October, 1852, and reached the
crater, from which he then saw nothinf]^ but sulphureted-hydrogen va-
por rising in a cloud ; but Sonneschmid, who vainly attempted to as-
cend Colima, in February, 1796, gives an account of an immense ejec-
tion of ashes in the year 1770. In the month of March, 1795, on the
other hand, red-hot scoria; were visibly thrown out in a column of fire
at night. " To the northwest of th<5 volcano of Colima a volcanic
branch fissure runs along the shore of the South Sea. Extinct craters
and ancient lava streams are recognized in what are called the Volca-
noes of Ahuacatlan (on the road from Guadalaxara to San Bias) and
Tepic." (Pieschel, Ibid., p. 529.) * See above, p. 344-349.

378 COSMOS.

or less active volcanoes has become known to mariners within
the last seventy or eighty years, but this group lay hitherto,
as it were, isolated, and unconnected with the volcanic range
of the Mexican tropical region, or with the volcanoes which
were believed to exist on the peninsula of California. If we
include the range of extinct trachytic cones as intermediate
links, we may be said to have obtained insight into their im-
portant geological connection over a gap of more than 28^
of latitude, between Durango and the new Washington terri-
tory, northward of West Oregon. The study of the physical
condition of the earth owes this important step in advance to
the scientifically well-prepared expeditions which the govern-
ment of the United States has fitted out for tlie discovery of
the best road from the plains of the Mississippi to the shores
of the South Sea. All the departments of natural history
have derived advantage from those undertakings. Great
tjacts of country have been found, in the now explored terra
incognita of this intermediate space, from very near the Kocky
Mountains on their eastern slope, to a great distance beyond
their western descent, covered with evidences of extinct or
still active volcanoes (as, for instance, in the Cascade Mount-
ains). Thus, setting out from New Zealand, and ascending
first a long way to the northwest through New Guinea, the
Sunda Islands, the Philippines, and Eastern Asia, to the
Aleutians; and then descending toward the south through
the northwestern, the Mexican, the Central American, and
South American territories to the terminating point of Chili,
we find the entire circuit of the basin of the Pacific Ocean,
throughout an extent of 26,400 geographical miles, sur'
rounded by a range of recognizable memorials of volcanic
action. Without entering into the details of exact geograph-
ical bearings and of the perfected nomenclature, a cosmical
view such as this could never have been obtained.

Of the circuit of the great oceanic* basin here indicated
(or, as there is but one united mass of water over the whole
earth, we ought rather to say the circumference of the larg-
est of those portions of it which penetrate between conti-
nents) it remains for us now to describe the tract of country
which extends from Rio Gila to Norton's and Kotzebue's

* The tei*m "Grand Ocean," used to designate the basin of the
South Sea by that learned geographer, my friend Contre-amiral de
Fleurieu, the editor of the Introduction Historique au Voyage de Mar-
chand, confounds the whole with a part, and consequently leads to
misapprehension.

TRUE VOLCANOES. 379

Sounds. Analogies drawn in Europe from the Pyrenees or
the Alpine chain, and in South America from the Cordilleras
of the Andes, from South Chili to the fifth degree of north
latitude in New Granada, supported by fanciful delineations
in maps, have propagated the erroneous opinion that the
Mexican mountains, or at least their highest ridge, can be
traced along like a wall, under the name of the Sierra Madre,
from southeast to northwest. But though the mountainous
part of Mexico is a mighty swelling of the land running con-
nectedly in the direction above stated between two seas to
the height of from 5000 to 7000 feet, yet on the top of this,
in the same way as in the Caucasus and in Central Asia,
still loftier ranges of mountains, running in partial and very
various directions, rise to about 15,000 and 17,800 feet.
The arrangement 'of these partial groups, erupted from fis-
sures not parallel to each other, is in its bearings for the most
part independent of the ideal axis which may be drawn
through the entire swell of the undulating flattened ridge.
These remarkable features in the formation of the soil give
rise to a deception which is strengthened by the pictorial
effect of the beautiful country. The colossal mountains cov-
ered with perpetual snow, seem, as it were, to rise out of a
plain. The spectator confounds the ridge of the soft swell-
ing land, the elevated plain, with the plain of the low lands ;
and it is only from the change of climate, the lowering of the
temperature, under the same degree of latitude, that he is re-
minded of the height to which he has ascended. The fissure
of upheaval, frequently before mentioned, of the volcano of
Anahuac (running in a direction from east to west between
19° and 19^° lat.) intersects* the general axis of the swell-
ing land almost at right angles.

The conformation here described of a considerable portion
of the surface of the earth, which only began to be estab-
lished by careful measurements since the year 1853, must
not be confounded with those swellings of the soil which are
met with inclosed between two mountain chains, which bound
them, as it were, like walls — as in Bolivia, at the Lake of
Titicaca; and in Central Asia, between the Himalaya and
Kuen-liin. The former of these, the South American eleva-
tion, which at the same time forms the bottom of a valley,

* On the axes of the greatest elevations and of the volcanoes in the
tropical zone of Mexico, see above, p. 264 and 300. Compare also
Essai Pol sur laNouv. Esp., t. i., p. 257-268, t. ii., p. 173; Views of
Nature, p. .37.

380 COSMOS.

is on an average, according to Pentland, 12,847 feet above
the level of the sea ; the latter, or Tliibetian, according to
Captain Henry Strachey, Joseph Hooker, and Thomas Thom-
son, is upward of 14,996. Tlie wish expressed by me half a
century since, in my circumstantial '■^Analyse de C Atlas Gco-
graphiqiie et Physique de Uoyaume de laNouvelle Espange (§ xi v.),
that my profile of the elevated plain between Mexico and Gu-
anaxuato might be continued by measurements over Durango
and Chihuahua as far as Santa Fe del Nuevo Mexico, is now
completely realized. The length of way, reckoning only one
fourth for the inflections, amounts to more than 1200 geo-
graphical miles, and the characteristic feature of this so long
unobserved configuration of the earth (the soft undulation of
the swelling, and its breadth in a transverse section, amount-
ing sometimes to 240 or 280 geographical miles) is manifest-
ed by the fact that the distance (from Mexico to Santa Fe),
comprising a difference of parallels of fully 16° 20' about the
same as that from Stockholm to Florence, is traveled over in
four-wheeled carriages, on the ridge of the table-land, with-
out the advantage of artificially prepared roads. The possi-
bility of such a medium of intercourse was known to the
Spaniards so early as the end of the 16th century, when the
viceroy, the Conde de Monterey,* planned the first settlements
from Zacatecas.

In confirmation of what has- been stated in a general way
■ respecting the relative heights between the capital of Mexico
and Santa Fe del Nuevo Slexico, I here insert the chief ele-
ments of the barometrical levelings, which have been com-
pleted from 1803 to 1847. I take them in the direction from
north to south, so that the most northerly, placed at the top
of the list, may correspond more readily with the bearings of
our charts if

* By Juan de Onate, 1594. Memoir of a Tour to Northern Mexico
in 1846 and 1847, by Dr. Wislizenus. On the influence of the con-
figuration of the soil (the wonderful extent of the table-land) on the
internal commerce and the intercourse of the tropical zone with the
north, when once civic order, legal freedom, and industry increase in
these parts, see Essai Pol., t. iv., p. 38, and Dana, p. 612.

t In this survey of the elevations of the soil between Mexico and
Santa Fe del Neuvo Mexico, as well as in the similar but more imper-
fect table which I have given in the Views of Nature, p, 208, the letters
Ws, Bt, and Ht, attached to the numerals, denote the names of the
observer. Thus, Ws stands for Dr. Wislizenus, editor of the very in-
structive and scientific il/e?HOzVo/'a Tour to Northern Mexico, connected
with Colonel Doniphan's Expedition in 1846 and 1847 (Washington,
1848) ; Bt the Chief Counselor of Mines, Burkart ; and Ht for my-

TRUE VOLCANOES. 881

Santa Fe del Nuevo Mexico (lat. 35° 410, height 7047
feet, Ws.

self. At the time when I was occupied, from March, 1803, to Febru-
ary, 1804r, with the astronomical determinations of places in the trop-
ical part of New Spain, and ventured, from the materials I could dis-
cover and examine, to design a map of that country, of which my re-
spected friend Thomas Jefferson, then President of the United States,
during my residence in Washington, caused a copy to be made, there
existed as yet in the interior of the country, on the road to Santa Fe,
no determinations of latitude north of Durango (lat. 24° 25'). Ac-
cording to the two manuscript journals of the engineers Rivera, Lafo-:
ra, and Mascaro, of the years 1724 and 1765, discovered by me in the
archives of Mexico, and which contained dii'ections of the compass and
computed partial distances, a careful calculation showed for the im-
portant station of Santa Fe, according to Don Pedro d^ Rivera, lat.
36° 12', and long. 105° 52' 30". (See my Atlas Gcogr. et Phys. du Mex-
iqm, tab. 6, andEssai Pol., t. i., p. 75-82.) I took the precaution, in the
analysis of my map, to note this result as a very uncertain one, seeing
that in the valuations of the distances, as well as in the directions of
the compass, uncorrected for the magnetic variation, and unaided by
objects in treeless plains, destitute of human habitations, over an ex-
tent of more than 1200 geographical miles, all the errors can not be
compensated (t. i., p. 127-131). It happens that the result here given,
as compared with the most recent astronomical observations, turns
out to be much more erroneous in the latitude than in the longitude —
being in the former about thirty-one, and in the latter scarcely twen-
ty-three minutes. I was likewise fortunate enough to determine, near-
ly correctly, the geographical position of the Lake Timpanogos, now
generally called the Great Salt Lake, while the name of Timpanogos
is now only applied to the river which falls into the little Utah Lake, a
fresh-water lake. In the language of the Utah Indians a river is called
oq-toahbe, and by contraction ogo alone ; timpan means rock, so that
Timpan-ogo signifies rock-river (Fremont, Explor. Exped., 1845, p.
273). Buschmann explains the word thnpa as derived from the Mexi-
can tetl, stone, while in pa he finds a substantive teripination of
the native North-Mexican languages; to ogo he attributes the general
signification of water : see his work. Die Spuren der Aztekischen Sprache
im nordlichen Mexico, s. 354-356 and 351. Compare Expedition to the
Valley of the Qreat Salt Lake of Utah, by Captain Howard Stansbury,
1852, p. 300, and Humboldt, Views of Nature, p. 206. My map gives
to the Montagnes de Sel gemme, somewhat to the east of the Laguna de
Timpanogos, lat, 40° 7', long. 111° 48' 30"; consequently nay first
conjecture differs 39 minutes in latitude, and 17 in longitude. The
most recent determinations of the position of Santa Fe, the capital of
New Mexico, with which I am acauainted, are, 1st, by Lieutenant
Emory (1846), from numerous astraHtiical obsei-vations, lat. 35° 44'
6"; and, 2d, by Gregg and Dr. WislfiCTius (1848), perhaps in another
locality, 35° 41' 6". The longitude, according to Emory, is 1^ 4' 18",
in time from Greenwich, and therefore 106° 5.' in the equatorial cir-
cle ; according to Wislizenus, 108° 22' from Paris {New Mexico and
California,' hy Emory, Document No. 41, p. 36; Wisl., p. 29), Most
maps err in making the latitudes of places in the neighborhood of
Santa Fe too far to the north. The height of the city of Santa Fe

382 COSMOS.

Albuquerque* (lat. 35° 8^, height 4849 feet, Ws.

Paso del Norte,t on the Rio Grande del Norte (lat. 29° 48"),
height 3790 feet, Ws.

Chihuahua (lat. 28° 32^, 4638 feet, Ws.

Cosiquiriachi, 6273 feet, Ws.
. Mapimi, in the Bolson de Mapimi (lat. 25° 540, 4*^82 feet,
Ws.

Parras (lat. 25° 32^, 4986 feet, Ws.

SaltiUo (lat. 25° lO'), 5240 feet, AVs.

Durango (lat. 24° 25^, 6849 feet, according to Oteiza.

FresniUo (lat. 23° 10"), 7244 feet, Bt.

Zacatecas (lat. 22° 50^, 9012 feet, Bt.

San Luis Potosi (lat. 22° 8^, 6090 feet, Bt.

Aguas Calientes (lat. 21° 53^, 6261 feet, Bt.

Lagos (lat. 21° 200, 6376 feet, Bt.

Villa de Leon (lat. 21° 70, 6134 feet, Bt.

Silao, 5911 feet, Bt.

Guanaxuato (lat. 21° 0' 15'0, 6836 feet, Ht.

Salamanca (lat. 20° 400, 5762 feet, Ht.

Celaya (lat. 20° 380, 6017 feet, Ht.

Queretaro (lat. 20° 36^ 39^0, 6363 feet, Ht.

San Juan del Rio, in the state of Queretaro (lat. 20° 300>
6490 feet, Ht.

Tula (lat. 19° 570, 6733 feet, Ht.

Pachuca, 8140 feet, Ht.

Moran, near Real del Monte, 8511 feet, Ht.

Huehuetoca, at the northern extremity of the great plain
of Mexico (lat. 19° 480, 7533 feet, Ht.

Mexico (lat. 19° 25" 45^0, 7469 feet, Ht.

Toluca (lat. 19° 160, ^^25 feet, Ht.

Venta de Chalco, at the southeastern extremity of the great
plain of Puebla, 7712 feet, Ht.

San Francisco Ocotlan, at the western extremity of the
great plain of Puebla, 7680 feet, Ht.

Chplula, at the foot of the ancient graduated Pyramid,
(lat. 19° 20, 6906 feet, Ht.

above the level of the sea, accoj^g to Emory, is 6844 ; according to
Wislizenus, fully 7046 feet (iVB measurement 6950) ; it therefore
resembles that of the Spliigen raa Gotthard passes in the Swiss Alps.

* The latitude of Albuquerque is taken from the beautiful special
map, entitled Map of the Territory of New Mexico, by Kern, 1851.
Its height, according to Emory (p. 166), is 4749 feet; according to
Wislizenus (p. 122), 4858.

t For the latitude of the Paso del Norte compare Wi«liz., p. 125,
Met. Tables 8-12, Aug., 1846.

TRUE VOLCANOES. 383

La Puebla de los Angeles (lat. 19° 0^ 15'0, 7201 feet, Ht.

(The village of Las Vigas marks the eastern extremity of
the elevated plain of Anahuac, lat. 19° 37^; the height of
the village is 7814 feet, Pit.)

Thus, though previous to the commencement of the 19tli
century, not a single altitude had been barometrically taken
in the whole of New Spain, the hypsometrical and in most
cases also astronomical observations for thirty-two places in
the direction from north to south, in a zone of nearly 16^°
of latitude, between the town of Santa Fe and the capital of
Mexico have been acomplished. We thus see that the surface
of the wide elevated plain of Mexico assumes an undulating
form, varying in the centre from 5850 to 7500 feet in height.
The lowest portion of the road from Parras to Albuquerque
is even 1066 feet higher than the highest point of Vesuvius.

The great though gentle* swelling of the soil, whose high-
est portion we have just surveyed, and which from south to
north, from the tropical part to the parallels of 42° and 44°.
so increase in extent from east to west that the Great Basin,
westward of the great Salt Lake of the Mormons, has a di-
ameter of upward of 340 geographical miles, with a mean
elevation of nearly 5800 feet, differs very considerably from
the rampart-like mountain chains by which it is surmounted.
Our knowledge of this configuration is one of the chief points
of Fremont's great hypsometrical investigations in the years
1842 and 1844. This swelling of the soil belongs to a dif-
ferent epoch from that late upheaval which we call mountain
chains and systems of varied direction. At the point where,
about 32° lat., the mountain mass of Chihuahua, according
to the present settlement of the boundaries, enters the western
territory of the United States (in the provinces taken from
Mexico), it begins to bear the not very definite title of the
Sierra Madre. A decided bifurcation,! however, occurs in

* Compare Eremont, Report of the Exploring Exped. in 1 842, p. 60 ;
Dana, Geology of the United States Expl. Exped., p. 611-G13; and for
South America, Alcide D'Orbigny, Voy. dans I'Avicrique Mcrid.^ Atlas,
pi. viii., De Geologic spcciale, fig. 1.

t For this bifurcation and the correct denomination of the east and
west chains see the large special map of the Territory of New Mexico,
by Parke and Kern, 1851 ; Edwin Johnson's Map of Railroads, 1854 ;
John Bartlett's Map of the Boundary Commission, ,1854; Explorations
and Surveys from the Mississippi to the Pacific in 1853 and 1 854, vol. i.,
p. 15 ; and, above all, the admirable and comprehensive work of Jules
Marcou, Geologist of the Southern Pacific R. E. Survey, under the
command of Lieutenant "Whipple, entitled i?esMme explicatif d'une Carte
Geologique des Etats Unis et d'un Prqfil Geologique allant de la Vallee du

384 COSMOS.

the neighborhood of Albuquerque, and at this bifurcation the
western chain still maintains the general title of the Sierra
Madre, while the eastern branch has received from lat. 36°
10^ forward (a lit-tle to the north of Santa Fe), from Amer-
ican and English travelers, the equally ill-chosen, but now

Mississippi mix cotes de V Or/can Pacijique^ p. 113-116; also in t\ie, Bul-
letin de la Socicte Geologique de la France, 2e Serie, t. xii., p. 813. In
the elongated valley closed by the Sierra Madre, or Rocky Mountains,
lat. 35° 38|-°, the separate groups of which the western chain of the
Sierra Madre and the eastern chain of the Rocky Mountains (Sierra
de Sandia) consist, bear different names. To the first chain belong,
reckoning from south to north, the Sierra de las Grullas, the S. de
los Mimbres (Wislizenus, p. 22 and 54), Mount Taylor (lat. 35° 15'),
the S. de Jemez, and the S. de San Juan ; in the eastern chain the
Moro Peaks, or Sierra de la Sangre de Cristo, are distinguished from
the Spanish Peaks (lat. 37° 32') and th& noithwesterly tending White
Mountains, which close the elongated valley of Taos and Santa Ee.
Professor Julius Frobel, whose examination of the volcanoes of Cen-
tral America I have already noticed {Cosmos, above, p. 260), has with
much ability elucidated the indefinite geographical appellation of Si-
erra Madre on the older maps ; but he has at the same time, in a treat-
ise entitled Remarks contributing to the Physical Geography of the North
American Continent (9th Annual Report of the Smithsonian Institution,
1855, p. 272-281), given expression to a conjecture which, after having
examined all the materials within my reach, I am unable to assent to,
namely, that the Rocky Mountains are not to be regarded as a con-
tinuation of the Mexican mountain range in the tropical* zone of Ana-
huac. Uninterrupted mountain chains, like those of the Apennines,
the Swiss Jura, the Pyrenees, and a great part of the German Alps,
certainly do not exist from the 19th to the 44th degrees of latitude,
from Popocatepetl, in Anahuac, as far as to the north of Fremont's
Peak, in the Rocky Mountains, in the direction from S.S.E. to N.N.W. ;
but the immense swelling of the surface of the land, which goes on in-
creasing in breadth toward the north and northwest, is continuous from
tropical Mexico to Oregon, and on this swelling (or elevated plain),
which is itself the great geognostic phenomenon, separate groups of
mountains, running in often varying directions, rise over fissures which
have been formed more recently and at diffbrent periods. These snper-
imposed groups of mountains, which, however, in the Rocky Mountains
are for an extent of 8 degrees of latitude connected together almost
like a rampai-t, and rendered visible to a great distance by conical
mountains, chiefly trachytic, from 10,000 to 12,000 feet high, produce
an impression on the mind of the traveler which is only the more pro-
found from the circumstance that the elevated plateau which stretches
far and wide aj^ound him assumes in his eyes the appearance of a plain
of the level country. Though in reference to the Cordilleras of South
America, a considerable part of which is known to me by personal in-
spection, we speak of double and triple ranges (in fact, the Spanish
expression Las Cordilleras de los Ailftcs refers to such a disposition and
partition of the chain), we must not forget that even here the direc-
tion of the separate ranges of mountain groups, whether in long ridges
or in consecutive domes, are by no means parallel, either to one an-
other or to the direction of the entire swell of the land.

TRUE VOLCANOES. 385

universally accepted title of the Rocky Mountains. The two
chains form a lengthened valley, in which Albuquerque,
Santa Fe', and Taos lie, and through which the Eio Grande
del Norte flows. In lat. 38^° this valley is closed by a chain
running east and west for the space of 88 geographical miles,
while the Rocky Mountains extend undivided in a meridional
direction as far as lat. 41°. In this intermediate space rise,-
somewhat to the east, the Spanish Peaks — Pike's Peak (5800
feet), which has been beautifully delineated by Fremont,
James's Peak (11,434 feet), and the three Park Mountains,
all of which inclose three deep valleys, the lateral walls of
which rise up, along with the eastern Long's Peak, or Big Horn,
to a height of 9060 and 11,191 feet.* On the eastern bound-
ary, between Middle and North Park, the mountain chain all
at once changes its direction, and runs from lat. 40^° to 44°
for a distance of about 2C0 geographical miles from south-
east to northwest. In this intermediate space lie the south
Pass (7490 feet), and the famous Wind River Mountains, so
singularly sharp pointed, together with Fremont's Peak (lat.
43° 8^), which reaches the height of 13,567 feet. In the par-
allel of 44°, in the neighborhood of the Three Tetons, where
the northwesterly direction ceases, the meridian direction of
the Rocky Mountains begins again, and continues about as
far as I^ewis and Clarke's Pass, which lies in lat. 47° 2^, and

* Fremont, Explor. Exped., p. 281-288. Pike's Peak, lat. 38° 50',
delineated at p. 114 ; Long's Peak, 40° 15' ; ascent of Fremont's Peak
(13,570 feet) p. 70. The Wind River Mountains take their name from
the source of a tributary to the Big Horn River, whose waters unite
with those of the Yellow Stone River, which falls into the Upper Mis-
souri (lat. 47° 58', long. 103° 6' 30"). See the delineations of the
Alpine range, rich in mica-slate and granite, p. G6 and 70. I have in
all cases retained the English names given by the North American
geographers, as- their translation into a pure German nomenclature
has often proved a rich source of confusion. To help the comparison
of the direction and length of the meridian chain of the Ural, Avhich,
according to the careful investigations of my friend and traveling com-
panion, Colonel Ernst Hofmann, takes a curve at the northern extrem-
ity toward the east, and which, from the Truchmenian Mountain Airuk-
Tagh (481°) to the Sablja Mountains (65°), is fully 1020 geographical
miles in length, with those of the Rocky Mountains, I would here re-
mind the reader that the latter chain runs between the parallels of
Pike's Peak and Lewis and Clarke's Pass, from 105° 9' 30" into 112°
9' 30" of longitude. The chain Af the Ural, which, within the same
space of 17 degrees of latitude, deviates little from the meridian of
59°0'30", likewise changes its direction under the parallel of 65°, and
attains under lat. 67^° the meridian of Q&° 5' 30". Compare Ernst
Hofmann, Der ncirdlirJie Ural und das Kiistengebirge Pac-Choi, 1856, s.
191 and 297-305, with Humboldt, Asie Centrak (1843), t. i., p. 447.
Vol. v.— R

386 COSMOS.

long. 112° 9^ SO'". Even at this point the chain of the
Rocky Mountains maintains a considerable height (5977 feet) ;
but, from the many deep river-beds in the direction of Flat-
head River (Clarke's Fork), it soon decreases to a more regu-
lar level. Clarke's Fork and Lewis or Snake River unite in
forming the great Columbia River, which will one day prove
an important channel for commerce. (Explorations for a Hail-
. road from the Mississippi River to the Facific Ocean, made in
1853-1854, vol. i., p. 107.)

As in Bolivia, the eastern chain of the Andes farthest re-
moved from the sea, that of Sorata (21,287 feet) and Illimani
(21,148 feet), furnish no volcano now in a state of ignition,
so also, in the western parts of the United States, the vol-
canic action on the coast chain of California and Oregon is
at present very limited. The long chain of the Rocky Mount-
ains, at a distance from the shores of the South Sea vary-
ing from 480 to 800 geographical miles, without any trace
of still existing volcanic action, nevertheless shows, like the
eastern chain of Bolivia, in the vale of Yucay,* on both of
its slopes volcanic rock, extinct craters, and even lavas in-
closing obsidian, and beds of scoriae. In the chain of the
Rocky Mountains which we have here geographically de-
scribed, in accordance with- the admirable observations of
Fremont, Emory, Abbot, Wislizenus, Dana, and Jules Mar-
cou, the latter, a distinguished geologist, reckons three groups
of old volcanic rock on the two slopes. For the earliest no-
tices of the vulcanicity of this district we are also indebted to
the investigations made by Fremont since the years 1842 and
1843 {Report of the Exploring Expedition to the Rocky Mount-
ains in 1842, and to Oregon and North California in 1843-44,
p. 164, 184, 187, and 193).

On the eastern slope of the Rocky Mountains, on the south-
western road from Bent's Fort, on the Arkansas River, to
Santa Fe del Nuevo Mexico, lie two extinct volcanoes, the
Raton Mountains! with Fisher's Peak, and the hill of EI
Cerrito, between Galisteo and Pera Blanca. The lavas of
the former cover the whole district between the Upper Ar-
kansas and the Canadian River. The Perperino and the
volcanic scoriae, which are first met with even in the prairies,

* See above, p. 279.

t According to the road-map of 1855, attached to the general report
of the Secretary of State, Jefferson Davis, the Raton Pass rises to au
elevation of as much as 7180 feet above the level of the sea. Compare
also Marcou, Resume explicatif d'vne Carte Geol, 1855, p. 113.

TRUE VOLCANOES. 387

on approaching the Rocky Mountains from the east, belong
perhaps to old eruptions of the Cerrito, or of the stupendous
Spanish Peaks (37° 320- This easterly volcanic district of
the isolated Raton Mountains forms an area of 80 geograph-
ical miles in diameter; its centre lies nearly in latitude
36° 50^.

On the western slope most unmistakable evidences of an-
cient volcanic action are discernible over a wider space, which
has been traversed by the important expedition of Lieutenant
Whipple throughout its whole breadth from east to west.
This variously-shaped district, though interrupted for fully
120 geographical miles to the north of the Sierra de Mogo-
yon, is comprised (always om the authority of Marcou's geo-
logical chart) betwji^latitude 33° 48^ and 35° 40^ so that
instances of eruptSBroccur farther south than those of the
Raton Mountains. Its centre falls nearly in the parallel of
Albuquerque. The area here designated divides into two
sections,' that of the crest of the Rocky Mountains nearer
Mount Taylor, which terminates at the Sierra de Zuiii,* and
the western section, .called the Sierra de San Francisco. The
conical mountain of Mount Taylor, 12,256 feet high, is sur-
rounded by radiating lava streams, which, like Malpays still
destitute of all vegetation, covered over with scoriae and pum-
ice-stone, wind along to a distance of several miles, precisely
as in the district around Hecla. About ^2 geographical miles
to the west of the present Pueblo de Zuni rises the lofty vol-
canic mountain of San Francisco itself. It has a peak which
has been calculated more than 16,000 feet high, and stretches
away southward from the Rio Colorado Chiquito, where, far-
ther to the west, the Bill William Mountain, the Aztec Pass
(6279 feet), and the Aquarius Mountains (8526 feet) follow.
The volcanic rock does not terminate at the confluence of
the Bill William Fork with the great Colorado, near the vil-
lage of the Mohave Indians (lat. 34°, long. 114°); for, on

* We must be careful to distinguish, to the west of the mountain
ridge of Zuni, where the Paso de Zuni attains an elevation of as much
as 7943 feet, between Zuni viejo, the old dilapidated town delineated
by Mollhausen on Whipple's expedition, and the still inhabited Pueblo
de Zuiii. Forty geogi'aphical miles north of the latter, near Fort De-
fiance, there still exists a very small and isolated volcanic district. Be-
tween the village of Zuni and the descent to the Kio Colorado Chiquito
(Little Colorado) lies exposed the petrified forest which Mollhausen
admirably delineated in 1853, and described in a treatise which he
sent to the Geographical Society of Berlin. According to Marcou
(Resume explic. d'tme Carte G^oL, p. 59), fossil trees and ferns are min-
gled with the silicified coniferse.

388 COSMOS.

the other side of the Rio Colorado, at the Soda Lake, sev-
eral extinct but still open craters of eruption may be recog-
nized.*

Thus we find here, in the present New Mexico, in the vol-
canic group commencing at the Sierra de San Francisco, and
ending a little to the westward of the Rio Colorado Grande,
or del Occidente (into which the Gila falls), over a distance
of 180 geographical miles, the old volcanic district of the
Auvergne and the Vivarais repeated, and a new and wide
field opened up for geological investigation.

Likewise on the western slope, but 540 geographical miles
more to the north, lies the third ancient volcanic group of
the Rocky Mountains, that of Fremont's Peak, and the two
triple mountains, whose names, the Trois Tetons and the
Three Buttes,| correspond well wifl(|jtheir conical forms.
The former lie more to the west than the latter, and conse-
quently farther from the mountain chain. They exhibit
wide-spread, black banks of lava, very much rent, and with
a scorified surface.J

Parallel with the chain of the Rocky Mountains, some-
times single and sometimes double, run several ranges in which
their northern portion, from lat. 46° 12'', are still the seat of
volcanic action. First, from San Diego to Monterey (32^°
to o6|°), there is the coast range, specially so called, a con-
tinuation of the ridge of land on the peninsula of Old, or
Lower, California ; then, for the most part 80 geographical
miles distant from the shore of the South Sea, the Sierra
Nevada (de Alta California), from 36° to 40f ° ; then again,
commencing from the lofty Shasty Mountains, in the parallel
of Trinidad Bay (lat. 41° 10^), the Cascade range, which con-
tains the highest still-ignited peak, and which, at a distance
of 104 miles from the coast, extends from south to north far
beyond the parallel of the Fuca Strait. Similar in their
course to this latter chain (lat. 43°-46°), but 280 miles dis-

* All on the authority of the profiles of Marcou and the above-cited
road-map of 1 855.

t The French appellations, introduced by the Canadian fur-hunters,
are generally used in the country and on English maps. According to
the most recent calculations, the relative positions of the extinct vol-
canoes are as follows: Fremont's Peak, lat. 43° 5', long. 110° 9' 30";
Trois Tetons, lat. 43° 38', long. 110° 49' 30"; Three Buttes, lat. 43° 20',
long. 112° 41' 30 "; Fort Hall, lat. 43° 0', long. 111° 24' 30".

t Lieutenant Mullan, on Volcanic Formation, in the Reports of Ex-
plor. Surveys, vol. i. (1855), p. 330 and 348; see also Lambert's and
Tinkham's Reports on the Three Buttes, Ibid.y p. 167 and 226-230,
and Jules Marcou, p. 115.

TRUE VOLCANOES. 389

taut from the shore, are the Blue Mountains,* which rise in
their centre to a height of from 7000 to 8000 feet. In the
central portion of Old California, a little farther to the north,
near the eastern coast or bay in the neighborhood of the
former Mission of San Ignacio,.in about 28° north latitude,
stands the extinct volcano known as the " Volcanes de las
Virgenes," which I have given on my chart of Mexico. This
volcano had its last eruption in 1746 ; but we possess no re-
liable information either regarding it or any of the surround-
ing districts. (See Venegas, Noticia de la California, 1757,
t. i., p. 27 ; and Duflot de Moras, Exploration de V Oregon et
de la Californie, 1844, t. i., p. 218 and 239.)

Ancient volcanic rock has already been found in the coast
range near the harbor of San Francisco, in the Monte del
Diablo, which Dr. Trask investigated (3673 feet), and in the
auriferous elongated valley of the Rio del Sacramento, in a
trachytic crater now fallen in, called the Sacramento Butt,
which Dana has delineated. Farther to the north, the Shasty,
or Tshashtl Mountains, contain basaltic lavas, obsidian, of
which the natives make arrow-heads, and the talc-like ser-
pentine which makes its appearance on many points of the
earth's surface, and appears to be closely allied to the vol-
canic formations. But the true seat of the still-existing igne-
ous action is the Cascade Mountain range, in which, covered
with eternal snow, several of the peaks rise to the height of
16,000 feet. I shall here give a list of these, proceeding from
south to north. The now ignited and more or less active
volcanoes will be (on the plan heretofore adopted ; see above,
p. 68, note *) distinguished by a star. The high conical
mountains not so distinguished are probably partly extinct
volcanoes, and partly unopened trachytic domes.

Mount Pitt, or M'Laughlin (lat. 42° 30^, a little to the
west of Lake Tlamat ; height 9548 feet.

Mount Jefferson, or Vancouver (lat. 44° 35^), a conical
mountain.

Mount Hood (lat. 45° 10^), decidedly an extinct volca-
no, covered with cellular lava. According to Dana, this
mountain, as well as Mount St. Helen's, which lies more
northerly in the volcanic range, is between 15,000 and

* Dana,p. 616-620; Blue Mountains, p. 649-651; Sacramento i^utt,
p. 630-643; Shasty Mountains, p. 614; Cascade range. On the Monte
Diablo range, perforated by volcanic rock, see also John Trail, on
the Geology of the Coast Mountains and "the Sierra Nevada, 1854, p.
13-18.

V •

'S90 COSMOS.

16,000 feet high, though somewhat lower(*) than the latter.
Mount Hood was ascended in August, 1853, by Lake, Tra-
vaillot, and Heller.

Mount Swalahos, or Saddle Hill, S.S.E. of Astoria,! wrth
a fallen in, extinct crater.

Mount St. Helen's,* north of the Columbia River (lat.
46° 12^); according to Dana, not less than 15,000 feet
high. J Still burning, and always smoking from the sum-
mit crater. A volcano of very beautiful, regular, conical
form, and covered with perpetual §now. There was a
great eruption on the 23d of November, 1842 ; which, ac-
cording to Fremont, covered every thing to a great distance
round with ashes and pumice.

Mount Adams (lat. 46° 18^), almost exactly east of the
volcano of St. Helen's, more than 112 geographical miles
distant from the coast, if it be true that the last-named
and still active mountain is only 76 of those miles inland.

Mount Regnier,* also written Mount Rainier (lat. 46°
48^), E.S.E. of Fort Nisqually, on Puget's Sound, which is
connected with the Fuca Strait. A burning volcano ; ac-
cording to Edwin Johnson's road-map of 1854, 12,330 feet
high. It experienced severe eruptions in 1841 and 1843.

Mount Olympus (lat. 47° 50^), only 24 geographical miles
south of the Strait of San Juan de Fuca, long so famous in
the history of the South Sea discoveries.

Mount Baker,* a large and still active volcano, situated
in the territory of Washington (lat. 48° 48^), of great (un-
measured ?) height (not yet determined), and regular conic-
al form.

Mount Brown (16,000 feef?) and, a little more to the
east, Mount Hooker (16,750 feet?), are cited by Johnson

(*) Dana (p. 61 5 and 640) estimated the volcano of St. Helen's at 16,000
feet, and Mount Hood, of course, under that height, while according to
others Mount Hood is said to attain the great height of 18,316 feet,
which is 2521 feet higher than the summit of Mont Blanc, and 4730
feet higher than Fremont's Peak, in the Rocky Mountains. Accord-
ing to this estimate (Langrebe, Nahirgeschichte der Vnlkane, bd. i., s.
497), Mount Hood would be only 571 feet lower than the volcano Co-
topaxi ; on the other harid, Mount Hood, according to Dana, exceeds
the highest summit of the Rocky Mountains by 2586 feet at the utmost.
I am always desirous of drawing attention to variantes lectiones such as
these.

t Dana, Geology of the United States Exploring Expedition^ p. 640 and
643-*45.

X Variously estimated previously at 10,178 feet by Wilkes, and 13,535
feet by Simpson.

TRUE VOLCANOES. 391

as lofty, old volcanic trachytic mountains, under lat. 52^°,
and long. 117° 40^ and 119° 40^ They are, therefore, re-
markable as being more than 300 geographical miles dis-
• tant from the coast-
Mount Edgecombe,* on the small Lazarus Island, near
Sitka (lat. 57° 3'). Its violent igneous eruption in 1796
has already been mentioned by me see above, p. 255).
Captain Lisiansky, who ascended it in the first years of
the present century, found the volcano then unignited. Its
height(*) reaches, according to Ernst Hofmann, 3039 feet;
according to Lisiansky, 2801 feet. Near it are l^t springs
which issue from granite, as on the road from theValles de
Aragua to Portocabello.

Mount Fairweather, or Cerro de Buen Tiempo ; accord-
ing to Malaspina, 4489 metres, or 14,710 feet highf (lat.
58° 35^). Covered with pumice-stone and probably ignited
up to a short time back, like Mount Elias.

The volcano of Cook's Inlet (lat. C0° 8^) ; according to
Admiral Wrangel, 12,065 feet high, and considered by that
intelligent mariner, as well as by Vancouver, to be an act-
ive volcano.J

Mount Elias (lat. 60° 17^ long. 136° 10^ 30"') ; accord-
ing to Malaspina's manuscripts, which I found in the Ar-
chives of Mexico, 5441 metres, or 17,854 feet ; according
to Captain Denham's chart, from 1853 to 1856, the height
is only 14,970 feet.

What M'C!lure, in his account of the Northwest Passage,
<ialls the volcano of Franklin's Bay (lat. 69° 57^, long. 127°),
eastward of the moiith of the Mackenzie River, seems to be
a kind of earth-Jire, or salses, throwing out hot, sulphurous
vapors- An eye-witness, the missionary Miertsching, inter-
preter to the expedition on board the ship Investigator, found
from thirty to forty columns of smoke rising from fissures in
the earth, or from small conical mounds of clays of various
colors. The sulphurous odor was so strong that it was scarce-
ly possible to approach the columns of smoke within a dis-
tance of twelve paces. No rock or other solid masses could

(*) Karsten's Archiv.fiXr Minerahgie, bd. i., 1829, s. 243-
t Humboldt, Essai Polit. sur la Nouv. Esp.^ t. i., p. 266, torn, ii.,
p. 310.

X According to a manuscript which I was permitted to examine in
the year 1803, in the Archives of Mexico, the whole coast of Nutka,
as far as what was afterward called " Cook's Inlet," was visited during
the expedition of Juan Perez, and Estevan Jose Martinez, in the year
1774.

392 COSMOS.

be discovered in the immediate vicinity. Lights were seen
from the ship at night, no ejections of mud, but great heat
of the bed of the sea, and small pools of water containing
sulphuric acid were observed. The district merits a careftil
investigation, and the phenomenon stands quite unconnected
there, like the volcanic action of the Cerro de Buen Tiempo,
or of Mount Elias in the Californian Cascade range (M'Clure,
Discovery of the Northwest Passage, pi 99 ; Papers relative to
the Arctic Expedition, 1854, p. 34; Miertsching's Reise-Tage-
huch ; Gnadau, 1855, s. 46).

I hav%hitherto treated the volcanic vital activities of our
planet in their intimate connections as if forming an ascend-
ing scale of the great and mysterious phenomenon of a reac-
tion of its fused interior upon its surface, clothed with ani-
mal and vegetable organisms. I have considered next in
order to the almost purely dynamic effects of the earthquake
(the wave of concussion) the thermal springs and salses, that
is to say, phenomena produced, with or without spontaneous
ignition, by the permanent elevation of temperature commu-
nicated to the water-springs and streams of gas, as well as
by diversity of chemical mixture. The highest, and in its
expressions the most complicated grade of the scale is pre-
sented by the volcanoes, which call into action the great and
varied processes of crystalline rock-formation by the dry
method, and which consequently do not simply reduce and
destroy, but appear in the character of creative powers, and
form the materials for new combinations. A considerable
portion of very recent, if not of the most recent, mountain
strata is the work of volcanic action, whether effected, as in
the present day, by the pouring forth of molten masses at
many points of the earth at peculiar conical or dome-shaped
elevated stages, or, as in the early years of our planet's exist-
ence, by the immediate issuing forth of basaltic and trachytio
rock by the side of the sedimentary strata, from a net-work of
open fissures, without the intervention of any such structures.

In the preceding pages I have most carefully endeavored
to determine the locality of the points at which a communi-
cation has long continued open between the fluid interior of
the earth and the atmosphere. It now remains to sum up
the number of these points, to separate out of the rich abund-
ance of the volcanoes which have been active in very re-
mote historical periods those which are still ignited at the
present day, and to consider these according to their division
into continental and insular volcanoes. If all those which,

•true volcanoes. 893

in this enumeration, I think I may venture to consider the
lowest limit of the number, were simultaneously in action,
th'eir influence on the condition of the atmosphere, and its
climatic, and especially its electric relations, would certainly
be extremely perceptible ; but as the eruptions do not take
place simultaneously, but at different times, their effect is di-
minished, and is confined within very narrow and chiefly
mere local limits. In great eruptions there occur around the
crater, as a consequence of the exhalation, volcanic storms,
which, being accompanied by lightning and torrents of rain,
often occasion great ravages ; but these atmospheric phenom-
ena have no generally extended results. For that the re-
markable obscurity (known by the name of the dry fog) which
for the space of several months, from May to August of the
year 1783, overspread a very considerable part of Europe
and Asia, as well as the North of Africa — while the sky was
seen pure and untroubled at the top of the lofty mountains
of Switzerland — could have been occasioned by the unusual
activity of the Icelandic volcanicity, and tlie earthquakes of
Calabria, as is even now sometimes maintained, seems to me
very improbable, on account of the magnitude of the effect
produced.* Yet a certain apparent influence of earthquakes,
in cases where they occupy much space, in changing the com-
mencement of the rainy season, as in the highland of Quito
and Riobamba (in February, 1797), or in the southeastern
countries of Europe and Asia Minor (in the autumn of 1856),
might, indeed, be viewed as the isolated influence of a volcanic
eruption.

In the following table the first figures denote the number
of the volcanoes cited in the preceding pages, while the sec-
ond figures, inclosed in parentheses, denote the number of
those which in recent times have given evidence of their ig-
neous activity. v

Number of Volcanoes on the Earth.

I. Europe (above, p. 328, 329) 7 (4)

II. Islands of the Atlantic Ocean (p. 329-332) 14 (8)

III. Africa (p. 332-334) 3 (1)

IV. Asia— Continental 25 (15)

(1) Western and Central (p. 334-340) 11 (6)

(2) The Peninsula of Kamtschatka (p. 340-344). 14 (9>
V. Eastern Asiatic Islands (p. 344-354) 69 (54)

[* A similar fog overspread the Tyrol and Switzerland in 1755, just
before the great earthquake which destroyed Lisbon. It appeared to
be composed of earthy particles reduced to an extreme degree of fine-
ness.— Tr.]

R2

394 COSMOS. H

VI. South Asiatic Islands (p. 281-391, 354-358) 120 (56)

VII. Indian Ocean (p. 358-363, and note * at p. 361, 362) 9 (5)
VIII. South Sea (p. 363-376; 364, note f; 365, note *;

366, note* 40 (26)

IX. America — Continental 115 (53)

(1) Soutli America 56 (26)

(a) Chili (p. 270, note |i at p. 272-274) 24 (13)

(6) Peru and Bolivia (p. 270-275, note § at

p. 270-272) U (3)

(c) Quito and Kew Granada (p. 270, note J). 18 (10)

(2) Central America (p. 245, 255-264, 270, 309,

note t at p. 257, notes * and t at p. 263).... 29 (18)

(3) Mexico, south of the Rio Gila (p. 264, 266,

270, 291-309, note at 293-5, notes at p. 297,
298, 302, 303; 376-401, note t at p. 376, and
notes on p. 377-82) 6 (4)

(4) Northwestern America, north of the Gila (p.

383-392) 24 (5)

TheAntilles* 5 (3)

Total 407 (225)

* In the Antilles the volcanic activity is confined to what are called
the " Little Antilles," three or four still active volcanoes having broken
out on a somewhat curvilinear fissure running from south to north,
nearly parallel to the volcanic fissure of Central America. In the
course of the considerations induced by the simultaneousness of the
earthquakes in the valleys of the rivers Ohio,*Mississippi, and Arkan-
sas, with those of the Orinoco, and of the shore of Venezuela, I have
already described the little sea of the Antilles, in its connection with
the Gulf of Mexico and the great plain of Louisiana, between the Al-
leghanies and the Rocky Mountains, on geognostic views, as a single
ancient basin (Voyage aux Regions Equinoxiales, t. ii., p. 5 and 19 ; see
also above, p. 10). This basin is intersected in its centre, between 18°
and 22° lat., by a Plutonic mountain range from Cape Catoche, of the
peninsula of Yucatan, to Tortola and Virgeh gorda. Cuba, Hayti,
and Porto Rico form a range running from west to east, parallel with
the granite and gneiss chain of Caraccas. On the other hand, the
Little Antilles, which are for the most part volcanic, unite together
the Plutonic chain just alluded to (that of the Great Antilles) and that
of the shore of Venezuela, closing the southern portion of the basin
on the east. The still active volcanoes of the Little Antilles lie be-
tween the parallels of 13° to 16J°, in the following order, reckoning
from south to north :

The volcano of the island of St. Vincent, stated sometimes at 3197
and sometimes at 5052 feet high. Since the eruption of 1718 all re-
mained quiet, until an immense ejection of lava took place on the 27th
of April, 1812. The first commotions commenced as early as May,
•1811, near the crater, three months after the island of Sabrina, in the
Azores, had risen from the sea. They began faintly in the mountain
valley of Caraccas, 3496 feet above the surface of the sea, in Decem-
ber of the same year. The complete destruction of the great city took
place on the 26th of March, 1812. As the earthquake which destroyed
Cumana, on the 14th of December, 1796, was with justice ascribed
to the eruption of the volcano of Guadaloupe (the end of September,

TRUE VOLCANOES. 395

The result of this laborious work, on which I have long

1796), in Hke manner the destruction of Caraccas appeal's to have been
the effect of the reaction of a southerly volcano of the Antilles — that
of St. Vincent. The frightful subterranean noise, like the thundering
of cannon, produced by a violent eruption of the latter volcano on the
30th of April, 1812, was heard on the distant grass-plains (Llanos) of
Calabozo, and on the shores of the Rio Apure, 192 geographical miles
farther to the West than its junction with the Orinoco (Humboldt,
Voyage, t. ii., p. 14). The volcano of St. Vincent had thrown out no
lava since 1718, but on the 30th of April a stream of lava flowed from
the summit crater and in four hours reached the sea-shore. It was a
very striking circumstance, and one which has been confirmed to me
by very intelligent coasting mariners, that the noise was very much
stronger on the open sea, far from the island, than near the shore.

The volcano of the island of St. Lucia, commonly called only a sol-
fatara, is scarcely 1200 to 1800 feet high. In the crater are several
small basins periodically filled with boiling water. In the year 1766
an ejection of scoriae and cinders is said Xo have been o)3served, which
is certainly an unusual phenomenon in a solfatara ; for, although the
cai'eful investigations of James Forbes and Poulett Scrope leave no
room to doubt that an eruption took place from the Solfatara of Poz-
zuoli in the year 1198, yet one might be inclined to consider that
event as a collateral effect produced by the great neighboring volcano,
Vesuvius (see Forbes, in the Edinb. Journal of Sckyice^ vol. i., p. 128,
and Poulett Scrope, in the Transact, of the Geol. Soc, 2d Ser., vol. ii.,
p. 346). Lancerote, Hawaii, and the Sunda Islands furnish us with
analogous examples of eruptions at exceedingly great distances from
the summit craters, the peculiar seat of action. It is true the sol-
fatara of Pozzuoli was not disturbed on the occasion of great erup-
tions of Vesuvius in the years 1794,. 1822, 1850, and 1855 (Julius
Schmidt, Ueher die Eruption des Vesuvs im Mai, 1855, p. 156), though
Strabo (lib. v., p. 245), long before the eruption of Vesuvius, speaks
of fire, somewhat vaguely, it is true, in the scorched plains of Dica-
archia, near Cumoea and Phlegra. Dicaarchia in Hannibal's time re-
ceived the name of Puteoli from the Romans, who colonized it.
**Some are of opinion," continues Strabo, ^'on account of the bad
smell of the water, that the whole of that district, as far as Baiaj and
Cumoea, is so called because it is full of sulphur, fire, and warm wa-
ter. Some think that on this account Cumoea (Cumanus ager) is

called also Phlegra ;" and then again Strabo mentions discharges

of fire and water (" irpoxoaQ tov Trvpbg kuI tov t'^aroe")-

The recent volcanic action of the island of Martinique, in the Mon-
tague Pelee (according to Dupuget, 4706 feet high), the Vauclin and
the Pitons du Carbet, is still more doubtful. The great eruption of
vapor on the 22d of January, 1792, described by Chisholm, and the
shower of ashes of the 5th of August, 1851, deserve to be more thor-
oughly inquired into.

The Soufriere de la Guadeloupe, according to the older measure-
ments of Amic and Le Boucher, 5435 and 5109 feet high, but, accord-
ing to the latest and very correct calculations of Charles Sainte-Claire
Deville, only 4867 feet high, exhibited itself on the 28th of Septem-
lier, 1797, 78 days before the great earthquake and the destruction of
the town of Cumana, as a volcano ejecting pumice (Rapport fait au

396 COSMOS.

been occupied, having in all cases consulted the original

General Victor Hugues par Amic et Hapel sur le Volcan de la Basse
Terre, dans la nuit du 7 au 8 Vendemiaire, an G, pag. 46 ; Humboldt,
Voyage, X. i., p- 316).- The lower part of the mountain is dioritic rock;
the volcanic cone, the summit of which is open, is trachyte, containing
labradorite. Lava does not appear even to have flowed in streams
from the mountain called, on account of its usual condition, the Sou-
friere, either from the summit crater or from the lateral fissures, but
the ashes of the eruptions of Sept., 1797, Dec, 1836, and Feb., 1837,
examined by the excellent and much lamented Dufrenoy, with his pe-
culiar accuracy, were found to be finely pulverized fragments of lava,
in which feldspathic minerals (labradorite, rhyakolite, and sanidine)
were recognizable, together with pyroxene. (See Lherminier, Daver,
Elie de Beaumont, and Dufrenoy, in the Comptes rendus de I Acad, des
Sc, t. iv., 1837, p. 294; 651 and 743-749). Small fragments of quartz
fiave also been recognized by Deville in the trachytes of the soufriere,
together with the crystals of labradorite (Comptes rendus, t, xxxii., p.
675), while Gustav Rose even found hexagonal dodecahedra of quartz
in the trachytes of the volcano of Arequipa (Meyen, Reise um die Erde^
bd. ii., s. 23).

The phenomena here described, of the temporary ejection of very
various mineral productions from the fissure openings of a soufriere,
remind us very forcibly that what we are accustomed to denominate a
solfatara, soufriere, or fumarole denotes, properly speaking, only cer-
tain conditions of volcanic action. Volcanoes which have once emit-
ted lava, or, when that failed, have ejected loose scoriae of considera-
ble volume ; or, finally, the same scoriaj pulverized by trituration, pass,
on a diminution of their activity, into a state in which they yield
only sulphur, sublimates of sulphurous acid, and aqueous vapor. If
as such we were to call them semi-volcanoes, it would readily convey
the idea that they are a peculiar class of volcanoes, Bunsen, to whom,
along with Boussingault, Senarmont, Charles Deville, and Danbree,
science is indebted for such important advances for their ingenious
and happy application of chemistry to geology, and especially to the
volcanic processes, shows " how, when in sulphur sublimations, which
almost always accompany volcanic eruptions, the masses of sulphur in
the form of vapor come in contact with the glowing pyroxene rocks,
the sulphurous acid is generated by the partial decomposition of the
oxyd of iron contained in those rocks. If the volcanic action then
sinks to a lower temperature, the chemical action of that zone then
enters into a new phase. The sulphurous combinations of iron, and
perhaps of metals of the earths and alkalies there produced, com-
mence their operation on the aqueous vapor, and the result of the al-
ternate action is the generation of sulphureted hydrogen and the prod-
ucts of its decomposition, disengaged hydrogen and sulphur vapor."
The sulphur fumaroles outlive the great volcanic eruptions for centu-
ries. The muriatic acid fumaroles belong to a different and later pe-
riod. They seldom assume the character of permanent phenomena.
The muriatic acid in the gases of craters is generated in this way : the
common salt which so often occurs as a product of sublimation in vol-
canoes,* particularly in Vesuvius, is decomposed in higher tempera-
tures, under the co-operation of aqueous vapor and silicates, and forms
muriatic acid and soda, the latter combining with the silicates j)resent.

TRUE VOLCANOES. 397

sources of information (the geological and geographical ac-

Muriatic acid fumaroles, which, in Italian volcanoes, are not unfre-
qnently on the most extensive scale, and are then generally accompa-
nied by immense sublimations of common salt, seem to be of a very
unimportant character in Iceland. The concluding stages in the chro-
nological series of all these phenomena consist in mere emanations of
carbonic acid. The hydrogen contained in the volcanic gases has
hitherto been almost entirely overlooked. It is present in the vapor
springs of the gi-eat solfataras of Krisuvik and Reykjalidh, in Iceland,
and is, indeed, at both those places combined vi^ith sulphureted hydro-
gen. When the latter come in contact with sulphuric acid, they are
both mutually decomposed by the separation of the sulphur, so that
they can never occur together. They are, however, not unfrequently
met with on one and the same field of fumaroles in close proximity to
each other. Unrecognizable as was the sulphureted hydrogen gas in
the Icelandic solfataras just mentioned, it failed, on the other hand,
entirely in the solfataric condition assumed by the crater of Hecla
shortly after the eruption of the year 1845 — that is to say, in the first
phase of the volcanic secondary action. Not the smallest trace of sul-
phureted hydrogen could be detected, either by the smell or by re-
agents, while the copious sublimation of sulphur, the smell of which
extended to a great distance, afforded indisputable evidence of the
presence of sulphurous acid. In fact, on the approach of a lighted
cigar to one of these fumaroles those thick clouds of smoke were pro-
duced which Melloni and Piria have noticed as a test of the smallest
trace of sulphureted hydrogen {Comptes rendus, t. xi., 1840, p. 352;
and Poggendorft''s Annalen, Erganzungsband, 1842, s. 511). As it
may, however, be easily seen by experiment that even sulphur itself,
when sublimated with aqueous vapor, produces the same phenomenon,
it remains doubtful whether any trace whatever of sulphureted hy-
drogen accompanied the emanations from the crater of Hecla in 1845,
and of Vesuvius in 1843 (compare Kobert Bunsen's admirable and
geologically important treatise on the processes of formation of the
volcanic rock of Iceland, in Poggend., AnnaL, bd. Ixxxiii., 1851, s.
241, 244, 246, 248, 254, and 256 ; serving as an extension and rectifi-
cation of the treatises of 1847 in Wohler's and Liebig's Annalen der
Chemie nnd Pharmacie, bd. Ixii., s. 19). That the emanations from
the solfatara of Pozzuoli are not sulphureted hydrogen, and that no
sulphur is deposited from them by contact with the atmosphere, as
Breislak has conjectured (^Essai Min6ralogiqne sur la Soufriere de Poz-
zuoli, 1792, p. 128-130), was remarked by Gay-Lussac when I visited
the Phlegraean Fields with him at the time of the great eruption of
lava in the year 1805. That acute observer, Archangelo Scacchi,
likewise decidedly denies the existence of sulphureted hydrogen {Me-
iiioiie Geologiche sulla Caivpania, 1849, p. 49-121), Piria's test seeming
to him only to prove the presence of aqueous vapor: "Son di avviso
che lo solfo emane mescolato a i vapori acquei senza essere in chimica
combinazione con altre sostanze" — " I am of opinion that the sulphur
emanates mixed with aqueous vapors without being in combination
with other substances." An actual analysis, however, long looked for
by m«, of the gases ejected by the solfatara of Pozzuoli, has been very
recently published by Charles Sainte-Claire Deville and Leblanc, and
has completely established the absence of sulphureted hydrogen

398 COSMOS.

counts of travels), is that, out af 407 volcanoes cited by me,
225 have exhibited proofs of activity in modern times. Pre-
vious statements of the number* of active volcanoes have
given sometimes about 30 and sometimes about 50 less, be-
cause they were prepared on diiFerent principles. In the di-
vision made by me, I have confined myself to those volcanoes
which still emit vapors, or which have had historically cer-
tain eruptions in the 19th or in the latter half of the 18th
century. There are doubtless instances of the intermission
of eruptions which extend over four centuries and more, but
these phenomena are of very rare occurrence. We are ac-
quainted with the lengthened series of the eruptions of Ve-
suvius in the years 79, 203, 512, 652, 983, 11^8, and 1500.
Previous to the great eruption of Epomeo on Ischia, in the
year 1302, we are acquainted only with those which occurre'd
in the 36th and 45th years before our era; that is to say, 55
years before the eruption of Vesuvius. .

Strabo, who died at the age of 90 under Tiberius (99 years
after the occupation of Vesuvius by Spartacus), and whom
no historical account of any former eruption had ever reached,
describes Vesuvius notwithstanding as an ancient and long
extinct volcano. "Above the places" (Herculaneum and

(^Comptes rendus de VAcad. d. Sc, t. xliii., 185G, p. 746). Sartorius
von Waltershausen, on the other hand, observed on cones of eruption
of ^tna, in 1811, a strong smell of sulphureted hydrogen, where in
other years sulphurous acid only was perceived. Nor did Charles De-
ville discover any sulphureted hydrogen at Girgenti, or in the Maca-
lube, but a small portion of it on the eastern declivity of ^tna, in the
spring of Santa Venerina. It is remarkable that throughout the im-
portant series of chemical analyses made by Boussingault on gas-ex-
haling volcanoes of the Andes (from Purace and Tolima to the ele-
vated plains of Las Pastos and Quito) both muriatic acid and sulphuret-
ed hydrogen (hydrogene sulfureux) are wanting.

* The following numbers are given in older works as those of the
volcanoes still in a state of activity: By Werner, 193; by Caesar von
Leonhard, 187; by Arago, 175 {Astronomie Popidaire, t. iii., p. 170);
variations which, as compared with my results, all show a difference
ranging from ^ to -J^ in a downward direction, occasioned partly by
diversity of principle in judging of the igneous state of a volcano, and
partly by a deficiency of materials for forming a correct judgment. It
is well known, as I have previously remarked, and as we learn from
historical experience, that volcanoes which have been held to be ex-
tinct have, after the lapse of very long periods, again become active,
and therefore the result which I have obtained must be considered as
rather too low than too high. Leopold von Buch, in the supplement
to his masterly description of the Canary Isles, and Landgrebe, in his
Geography of Volcanoes, have not attempted to give any general nu-
merical result.

TRUE VOLCANOES. 399

Pompeii), he says, " lies the Mount Vesuios, covered round
by the most beautiful farms, except on the summit. This is
indeed for the most part pretty smooth, but on the whole un-
fruitful, and having an ashy appearance. It exhibits fissured
hollows of red-colored rock, as if it were corroded by fire, so
that it might be supposed that this place had formerly burned
and had gulfs of fire, which, however, had died away when
the fuel became consumed." (Strabo, lib. v., page 247, Ca-
saub.) This description of the primitive form of Vesuvius
indicates neither a cone of cinders nor a crater-like hollow-
ing* of the ancient summit, such as, being walled in, could
have served Spartacusj and his gladiators for a defensive
strong-hold.

* This description is, therefore, totally at variance with the often-
repeated representation of Vesuvius, according to Strabo, given in
Poggendorfl"'s Annakn der Physik, bd. xxxvii., s. 190, tafel 1. It is a
very late Avriter, Dio Cassius, under Septimius Severus, who first speaks,
not (as is frequently supposed) of the production of several summits,
but of the changes of form which the summits have undergone in the
course of time. He records (quite in confirmation of Strabo) that the
mountain formerly had every where a fiat summit. His words are as
follows (lib. Ixvi., cap. 21, ed. Sturz, vol. iv., 1824, p. 240) : "For Vesu-
vius is situated by the sea near Naples, and hgis numerous sources of
fire. The whole mountain was formerly of uniform height, and the
fire arose from its centre, for at this part only is it in a state of com-
bustion. Outwardly, however, the whole of it is still, down to our
times, devoid of fire. But while the exterior is always without con-
flagration, and the centre is dried up (heated) and converted into cin-
ders, the peaks round about it have still their ancient height. But the
whole of the igneous part, being ^onsumed by length of time, has be-
come hollow by sinking in, so that the whole mountain (if we may com-
pare a small thing with a great) resembles an amphitheatre." ( Coinp.
Sturz, vol. vi., Annot. ii., p. 568.) This is a clear description of those
mountain masses which, since the year 79, have formed the margins
of the crater. The explanation of this passage, by referring it to the
Atrio del Cavallo, appears to me erroneous. According to the large
and excellent hypsometrical work of that distinguished Olmutz astron-
omer, Julius Schmidt, for the year 1855, the Punta Nasone of the.
Somma is 3771 feet, the Atrio del Cavallo,* at the foot of the Punta
Nasone, 26GI, and the Punta or Rocca del Palo (the highest edge of
the crater of Vesuvius ta the north, p. 112-116) 3992 feet high. My
barometrical measurements of 1822 ( Views of- Nature, p. 376-377) gave
for fhe same three points 3747 feet, 2577 feet, and 4022 feet, showing
a difference of 24, 84, and 30 feet respectively. The floor of the Atrio
del Cavallo has, according to Julius Schmidt {Eruption des Vesuvs im
Mai, 1855, p. 95), undergone great alterations of level since the erup-
tion of February, 1850.

t Velleius Paterculus, who died under Tiberius, mentions Vesuvius,
it is true, as the mountain which Spartacus'occupied with his gladia-
tors (ii., 30) ; while Plutarch, in his Biography of Crassus, cap. ii.,
speaks only of a rocky district having a single narrow entrance. The

400 COSMOS.

Diodorus Siculus, likewise (lib. iv., cap. 21, 5), who lived
under Caasar and Augustus, in his account of the progress of
Hercules and his-battles with the giants in the Phlegrasan
Fields, describes "what is now called Vesuvius as a Aocpog,
which, like ^tna in Sicily, once emitted a great deal of fire,
and (still) shows traces of its former ignition." He calls the
whole space between Cuma3 and Naples the Phlegrsean Fields,
as Poljbius does the still greater space between Capua and
Nola (lib. ii., cap. 17); while Strabo (lib. v., page 246) de-
scribes, with much local truth the neighborhood of Puteoli
(Dicaearchia), where the great solfatara lies, and calls it
'irlpalarov dyopd. *In later times the name of ra (pXeypala
■nedla is ordinarily confined to this district, as at this day
geologists place the mineralogical composition of the lavas
of the Phlegi'iean Fields in opposition to those from the
neighborhood of Vesuvius. The same opinion that in an-
cient times there was fire burning within Vesuvius, and that
that mountain had formerly had eruptions, is most distinctly
expressed in the architectural work of Vitruvius (lib. ii., cap.
6), in a passage which has hitherto not been sufficiently re-
garded: "Non minus etiam memoratur, antiquitus crevisse
ardores et abundavifese sub Vesuvio monte, et inde evomuisse
circa agros flammara. Ideoque nunc qui spongia siv e pumex
Pompejanus vocatur, excoctus ex alio genere lapidis, in banc
redactus esse videtur generis qualitatem. Id autem genus
spongiae, quod inde eximitur, non in omnibus locis nascitur,
nisi circum -ZEtnam, et collibus Mysias, qui a Graecis KaraKe-
Kavfj.EV0L nominantur." (It is also related that in ancient
times the fire increased and abounded beneath Mount Vesu-
vius, and vomited out flame from thence on the fields around.
So that now what is called spongia, or Pompeian pumex,
baked out of some other kind of stone, seems to have been
reduced to this kind of substance. But that kind of spongia
which is got out of there is not produced in all places, only
around ^tna and on the hills of Mysia, which are called by
the Greeks KaTaKEKavfitvoL.) Now it can no longer be
doubted, since the investigations of Bockh and Hirt, ^hat

servile war of Spartacus took place in the 681st year of Rome, or 152
years before the eruption of Vesuvius described by Pliny (24th of
August, 79 A.D,). The circumstance that Florus, a writer who lived in
the time of Trajan, and who, being acquainted with the eruption just
referred to, knew what was hidden in the interior of the mountain,
calls it " cavus," proves ifbthing, as others have already observed, for
its earlier configuration {Florus, lib. i., cap. 16, "Vesuvius mons, JEtneei
ignis imitator;" lib. iii., cap. 20, "fauces cavi mentis").

TRUE VOLCANOES. 401

Vitruvius lived in the tiine of Augustus,* and consequently
a full century before the eruption of Vesuvius at which the
elder Pliny met his death. The passage thus quoted, there-
fore, and the expression pumex Pompeianus (thus connecting
pumice-stone with Pompeii), present a special geological in-
terest in relation to the question raised as to whether, ac-
cording to the acute conjecture of Leopold von Buch,t
Pompeii was overwhelmed only by the pumiceous tufa-beds
thrown up on the first formation of Mount Somma ; these
beds, which are of submarine formation, covering in horizon-
tal layers the whole level between the Apennine range and
the west coast of Capua as far as Sorento, and from Nola to
the other side of Naples ; or whether Vesuvius itself, entire-
ly contrary to its present habit, ejected the pumice from its
interior.

Both Carmine Lippi,J who (1816) describes the tufa cov-
ering of Pompeii as an aqueous deposit, and his ingenious op-
ponent Archangelo Scacchi,§ in the letter addressed to the
Cavaliere Francesco Avellino (1843), have directed attention
to the remarkable phenomenon that a portion of the pumice
of Pompeii and Mount Somma contains small fragments of
chalk which have not lost their carbonic acid, a circumstance
which, on the supposition that they have been exposed to a
great pressure during their igneous formation, can excite but
little surprise. I have myself had the opportunity of seeing
specimens of 'this pumice-Btone in the interesting geological
collections of my learned friend and academical colleague,
Dr. Ewald. The similarity of the mineralogical constitution
at two opposite points naturally gives rise to the question —
whether that which covers Pompeii has been thrown down,
as Leopold von Buch supposes, during the eruption of the

* At all events, Vitruvius wrote earlier than the elder Pliny, as is
evident, not merely because he is three separate times cited by Pliny
in his list of authorities, so unjustly attacked by the English translator
Newton (lib. xvi., xxxv., and xxxvi.), but because in book xxxv., cap. 14,
s. 170-172, as has been distinctly proved by SiUig (vol. v., 1851, p. 277)
and Brunn {Diss, deauctorum indicihus Plinianis, Bonnae, 1856, p. 55-60),
a passage has actually been extracted from Vitruvius by Pliny himself.
See also Sillig's edition of Pliny, vol. v., p. 272. Hirt, in his Essay on
the Pantheon, places the date of Vitruvius's writings on architecture
between the years 16 and 14 of our era.

t PoggendorfTs Annakn, bd. xxxviU, s. 175-180.

X Carmine Lippi : Ftt iljuoca a I'acqua che sottero Pompei ed Ercola-
wof (1816), p. 10.

§ Scacchi, Osservazioni critiche sulla maniera come fa seppellita VArX"
tica Pompei, 1843, p. 8-10.

402 COSMOS.

year 79, from the declivities of Somma; or whether, as
Scacchi maintains, the newly-opened crater of Vesuvius has
ejected pumice simultaneously on Pompeii and on Somma?
What was known as pumex Pompejanus in the time of Vitru-
vius, under Augustus, carries us back to eruptions before the
time of Pliny ; and from the experience we have respecting
the variable nature of the formations in different ages and
different circumstances of volcanic activity, we should be as
little warranted in absolutely denying that, since its first ex-
istence, Vesuvius could have ejected pumice, as we should be
in absolutely taking it for granted that pumice — that is to
say, the fibrous or porous condition of a pyrogenous mineral
— could only be formed where obsidian or trachyte with
vitreous feldspar (sanidine) were present.

Although, from the examples w^hich have been cited of the
length of the periods at which the revival of a slumbering
volcano may take place, it is evident that much uncertainty
must still remain, yet it is of great importance to verify
the geographical distribution of burning volcanoes for a de-
terminate period. Of the 225 open craters through which,
in the middle of the 19th century, the molten interior of the
earth maintains a volcanic communication with the atmos-
phere, 70, that is to say, one third, are situated on the con-
tinents, and 155, or two thirds, on the islands of our globe.
Of the 70 continental volcanoes, 53, or three fourths, belong
to America, 15 to Asia, 1 to Europe, and one or two to that
portion of the continent of Africa hitherto known to us. In
the South-Asiatic Islands (the Sundas and Moluccas), as
well as in the Aleutian and Kurile Islands, the greatest num-
ber of the island volcanoes are situated in a very limited
space. The Aleutian Isles contain, perhaps, more volcanoes
active in late historical times than the whole continent of
South America. On the whole surface of the earth, the tract
containing the greatest number of volcanoes is that which
ranges between 73° west and 127° east longitude, and be-
tween 47° south and 66° north latitude, in a direction from
southeast to northwest.

If we suppose the great gulf of the sea known under the
name of the South Sea, or South Pacific Ocean, to be cos-
mically bounded by the parallel of Behring's Straits, and
that of New Zealand, which is also the parallel of South
Chili and North Patagonia, we shall find — and this result is
very remarkable — in the interior of the basin, as well as
around it (on its Asiatic and American continental bounda-

TRUE VOLCANOES. 403

ries), 198, or nearly seven eighths of the 225 still active vol-
canoes of the whole earth. The volcanoes nearest the poles
are, so far as our present geographical knowledge goes, in
the northern hemisphere the volcano Esk, on the small isl-
and of Jan Meyen, in lat. 71° V, and west long. 7° 30^30^^;
and in the southern hemisphere Mount Erebus, whose red
flames are visible even by day, and which Sir James Ross,*
on his great southern voyage of discovery in 1841, found to
be 12,400 feet high, or about 240 feet higher than the I^eak
of Teneriffe, in lat. 77° 33' and long. 166° 58' 30'' east.

The great number of volcanoes on the islands and on the
shores of continents must have early led to the investigation
by geologists of the causes of this phenomenon. I have al-
ready, in another place (CosmoSy vol. i., p. 243), mentioned
the confused theory of Trogus Pompeius under Augustus,
who supposed that the sea-water elicited the volcanic fire.
Chemical and mechanical reasons for this supposed effect of
the sea have been adduced to the latest times. The old hy-
pothesis of the sea-water penetrating into the volcanic focus
seemed to acquire a firmer foundation at the time of the dis-
covery of the metals of the earth by Davy^ but the great dis-
coverer himself soon abandoned* the theory to which even
Gay-Lussac inclined,! in spite of the rare occurrence, or total
absence of hydrogen gas. Mechanical, or rather dynamical
causes, whether sought for in the contraction of the upper
crust of the earth and the rising of continents, or in the lo-
cally diminished thickness of the inflexible portion of the
earth's crust, might, in my opinion, offer a greater appear-
ance of probabilty. It is not difficult to imagine that at tlie
margins of the upheaving continents which now form the
more or less precipitous littoral boundary visible over the
surface of the sea, fissures have been produced by the simul-
taneous sinking of the adjoining bottom of the sea, through
which the ^communication with the molten interior is pro-
moted. On the ridge of the elevations, far from that area of
depression in the oceanic basin, the same occasion for the
existence of such rents does not exist. Volcanoes follow the
present sea-shore in single, sometimes double, and sometimes
even triple parallel rows. These are connected by short

* Sir James Ross, Voyage to the Antarctic Regions, vol. i., p. 217,
220, and 364.

t Gay-Lussac, Reflexions swr ks Volcans in the Annales de Chimie et
de Physiqueyt. xxii., 1823, p. 429; see above, p. 163, note *; Arago,
(Euvres completes, t. iii., p. 47.

404

COSMOS.

chains of mountains, raised on transverse fiussres, and form-
ing mountain nodes. The range nearest to the shore is fre-
quently (but by no means always) the most active, while the
more distant, those more in the interior of the country, ap-
pear to be extinct or approaching extinction. It is some-
times thought that, in a particular direction in one and the
same range of volcanoes, an increase or diminution in the
frequency of the eruptions may be perceived, but the phenom-
ena of renewed activity after long intervals of rest render this
perception very uncertain.

As many incorrect statements of the distance of volcanic
activity from the sea are circulated, either through ignorance
of, or inattention to, the exact localities both of the volcanoes
and of the nearest points of the coast, I shall here give the
following distances in geographical miles (each being equal
to about 2030 yards, or 60 to a degree) : In the Cordilleras
of Quito, the volcano of Sangay, which discharges uninter-
ruptedly, is situated in the most easterly direction, but its
distance from the sea is still 112 miles. Some very intelli-
gent monks attached to the mission of the Indios Andaquies,
at the Alto Putumayo, have assured me that on the upper
Rio de la Fragua,* a tributary of the Caqueta, to the east-
ward of the Ceja, they had seen smoke issue from a conical
mountain of no great height, and whose distance from the
coast must have been 160 miles. The Mexican volcano of
Jorullo, which was elevated above the surface in September,
1759, is 84 miles from the nearest point of the sea-shore (see
above, p. 296-303); the volcano of Pococatepetl is 132
miles ; an extinct volcano in the eastern Cordilleras of Bo-
livia, near S. Pedro de Cacha, in the vale of Yucay (see
above, p. 279), is upward of 180 miles; the volcanoes of
the Siebengebirge, near Bonn, and of the Eifel (see above, p.
221-227), are from 132 to 152 miles; those of Auvergne,
Velay, and Vivarais,! distributing them into thuee separate

* The position of the Volcan de la Fragua, as reduced at Timana,
is N. lat. 1° 48', long. 75° 30' nearly. Compare the Carte Hypso-
metrique des Noeuds deMontagnes dans ks Cordilleres, in the large atlas
in my travels, 1831, pi. 5 ; see also pi. 22 and 24. This mountain ly-
ing isolated and so far to the east, ought to be visited by a geologist
capable of determining the longitude and latitude astronomically.

t In these three gi'oups, which, according to the old geographical
nomenclature, belong to Auvergne, the Vivarais, and the Velay, the
distances given in the text are those of the northernmost parts of each
group as taken from the Mediterranean Sea (between the Golfe d'Aigues
Mortes and Cette). In the first group, that of the Puy de Dome, a
crater erupted in the granite near Manzat, called Le Gour de Tazena,

TRUE VOLCANOES. 405

groups (the group of the Puy de Dome, near Clermont, with
the Mont Dore, the group of the Cantal, and the group of
the Puy and Mezenc), are severally 148, 116, and 84 miles
distant from the sea. The extinct volcanoes of Olot, south
of the Pyrenees, west of Gerona, with their distinct and
sometimes divided lava streams, are distant only 28 miles
from the Catalonian shores of the Mediterranean ; while, on
the other hand, the undoubted, and to all appearances very
lately extinct, volcanoes in the long chain of the Rocky
Mountains, in thew northwest of America, are situated at a
distance of from 600 to 680 miles from the shore of the
Pacific.

A very abnormal phenomenon in the geographical distri-
bution of volcanoes is the existence in historical times of act-
ive, and partially, perhaps, even of burning volcanoes in the
mountain chain of the Thian-shan (the Celestial Mountains),
between the two parallel chains of the Altai and the Kuen-
lun. The existence of these volcanoes was first made known
by Abel-Remusat and Klaproth, and I have been enabled, by
the aid of the able and laborious investigations of Stanislas
Julien, to treat of them fully in my work on Central Asia.*

is taken as the most northerly point (TJozet, in the Mem. de la Socicte
Geol. de France, t. i., 1844, p. 119). Farther south than the group of
the Cantal, and therefore nearest the sea-shore, lies the small volcanic
district of La Guiolle, near the Monts d'Aubrac, northwest of Chirac,
and distant scarcely 72 geographical miles from the sea. Compare
the Carte Gcologiqne de la France, 1841.

* Humboldt, Asie Centrale, t. ii., p. 7-61, 216, and 335-364; Cos^
7nos, vol. i., p. 245. The mountain lake of Issikul, on the northern
slope of the Thian-shan, which was lately visited for the first time by
Russian travelers, I found marked on the famous Catalonian map of
1374,* which is preserved as a treasure among the manuscripts of the
Paris library. Strahlenberg, in his work entitled Der nbrdliche und
bstliche Theil von Europa njtd Asien (Stockholm, 1730, s. 327), has the
merit of having first represented the Thian-shan as a peculiar and in-
dependent chain, without, however, being aware of its ^Icanic action.
He gives it the very indefinite name of Mousart, which — as the Bolor
was designated by the general title of Mustag, which particularizes
nothing, and merely indicates snow — has for a whole century occa-
sioned an erroneous representation, and an absurd and confused no-
menclature of the mountain ranges to the north of the Himalaya, con-
founding meridian and parallel chains with each other. Mousart is a
corruption of the Tartaric word Muztag, synonymous with our expres-
sion snowy chain, the Sierra Nevada of the Spaniards, the Himalaya in
the Institutes of Menu — signifying the habitation (alaya) of snow (hima),

[* This curious Spanish map was the result of the great commercial
relations which existed at that time between Majorca and Italy, Egypt
and India. See a more full notice of it in Asie Centrale, loc. cit. — Tr.]

406 COSMOS.

The relative distances of the volcano of Pe-shan (Mont
Blanc) with its lava streams, and the still burning igneous

and the Sineshan of the Chinese. Eleven hundred years before Strah-
lenberg wrote, under the dynasty of Sui, in the tinae of Dagobert, King
of the Franks, the Chinese possessed maps, constructed by order of
the government, of the countries lying between the Yellow River and
the Caspian Sea, on which the Kuen-liin and the Thian-shan were
marked. It was undoubtedly these two chains, but especially the first,
as I think I have shown in another place (Asie Cent?:, t. i., p. 118-129,
194-203, and t. ii., p. 413-425), which, when the march of the Mace-
donian army had brought the Greeks into clo*r acquaintance with
the interior of Asia, spread among their geographers the knoAvledge
of a belt of mountains extending from Asia Minor to the eastern sea,
from India and Scythia to Thinae, thus cutting the whole continent
into two halves (Strabo, lib. i., p. 68 ; lib. xi., p. 490). Dicaearchus,
and after him Eratosthenes, denominated this chain the elongated Tau-
rus; the Himalaya chain is included under this appellation. "That
which bounds India on the north," we are expressly told by Strabo
(lib. XV., p. 689), " from Ariane to the eastern sea, is the extremest por-
tions of the Taurus, which are separately called by the natives Paro-
pamisos, Emodon, Imaon, and other names, but which the Macedo-
nians call the Caucasus." In a previous part of the book, in describ-
ing Bactriana and Sogdiana (lib. xi., p. 519), he says, "the last por-
tion of the Taurus, which is called Imaon, touches the Indian (eastern)
Sea." The terms " on this side and on that side the Taurus" had ref-
erence to what was believed to be a single range, running east and
west ; that is to say, a parallel chain. Strabo was aware of this, for
he says, " the Greeks call the half of the region of Asia looking to the
north this side the Taurus, and the half toward the south that side'^
(lib. ii. p. 129). In the later times of Ptolemy, however, when com-
mei-ce in general, and particularly the silk trade, became animated,
the appellation of Imaus was transferred to a meridian chain, the Bo-
lor, as many passages of the 6th book show(^sie Centr., t. i., p. 146-
162). The line in which, parallel to the equator, the Taurus range
intersects the whole region, according to Hellenic ideas, was first called
by Dicaearchus, a pupil of the Stagirite, a Diaphragma (partition wall),
because, by means of j^jrpendicular lines drawn from it, the geograph-
ical width of other points could be measured. The diaphragma was
the parallel of Rhodes, extended on the west to the pillars of Hercules,
and on the east to the coast of Thinse {Agathemeros in Hudson's Geogr.
Gr. Min., vol. ii., p. 4). The divisional line of Dicaarchus, equally
interesting in a geological and an orographical point of view, passed
into the work of Eratosthenes, who mentions it in the 3d book of his
description of the earth, in illustration of his table of the inhabited
world. Strabo places so much importance on this direction and par-
tition line of Eratosthenes that he (lib. i., p. 65) thinks it possible
"that on its eastern extension, which at Thinse passes through the
Atlantic Sea, there might be the site of another inhabited world, or
even of several worlds ;" although he does not exactly predict that
they will be found to exist. The expression "Atlantic Sea" may^seem
remarkable as used instead of the " Eastern Sea," as the South Sea
(the Pacific) is usually called, but as our Indian Ocean, south of Ben-
gal, is called in Strabo the Atlantic South Sea, so were both seas to

TRUE VOLCANOES. 407

mountain (Hotschen) of Turfan, from the shores of the Polar
Sea and the Indian Ocean, are almost equally great, about
1480 and 1520 miles. On the other hand, the distance of
Pe-shan, whose eruptions of lava are separately recorded
from the year 89 of our era up to the 7th century in Chi-
nese works, from the great mountain lake of Issikul to the
descent of the Temurtutagh (a western portion of the Thian-
shan), is only 172 miles ; while from the more northerly
situated lake of Balkasch, 148 miles in length, it is 208 miles
distant.* The great Dsaisang lake, in the neighborhood of
which I was during my stay in the Chinese Dsungarei in
1829, is 360 miles distant from the volcanc^s of Thian-shan.
Inland waters are, therefore, not wanting, but they are cer-
tainly not in such propinquity as that which the Caspian Sea
bears to the still active volcano of Demavend, in the Persian
Mazenderan.

While, however, basins of water, whether oceanic or in-
land, may not be requisite for the maintenance of volcanic
activity — yet, if islands and coasts, as I am inclined to be-
lieve, abound more in volcanoes only because the elevation
of the latter, produced by internal elastic forces, is accom-
panied by a neighboring depression in the basin of the sea,t

the southeast of India considered to be connected, and were frequently
confounded together. Thus we read, hb. ii., p. 130, " India, the larg-
est and most favored country, which terminates at the Eastern Sea
and at the Atlantic South Sea;" and again, lib. xv., p. 689, "the
southern and eastern sides of India, which are much larger than the
other sides, run into the Atlantic Sea," in which passage, as well as
in the one above quoted regarding Thinse (lib. i., p. 65), the expres-
sion, *' Eastern Sea" is even avoided. Having been uninteiTuptedly
occupied since the year 1792 with the strike and inclination of the
mountain strata, and their relation to the bearings of the ranges of
mountains, I have thought it right to point attention to the fact that,
taken in the mean, the equatorial distance of the Kuen-liin, throughout
its whole extent, as well as in its western prolongation by the Hindu-
Kho, points toward the basin of the Mediterranean Sea and the Straits
of Gibraltar {Asie Centr., t. i., p. 118-127, and t. ii,, p. 115-118), and
that the sinking of the bed of the sea in a great basin which is vol-
canic, especially in the northern margin, may very possibly be con-
nected with this upheaval and folding in. My friend, Elie de Beau-
mont, so thoroughly acquainted with all that relates to geological bear-
ings, is opposed to these views on loxodromical principles (Notice sur
les Systemes de Montagnes, 1852, t. ii., p. 667).

* See above, p. 336.

t See Arago, Sur la cause de la de'pression d'une grandee parte de
I'Asie et sur le phenomSne que les pentes les plus rapides des chaines
de montagnes sont fgene'ralement) tournees vers la mer la plus voisine,
in his Astronomie Populaire, t. iii., p. 1266-1274.

408 COSMOS.

so that an drea of elevation borders on an area of depression,
and that at this bordering-line large and deeply penetrating
fissures and rents are produced — it may be supposed that in
the central Asiatic zone, between the parallels of 41° and
48°, the great Aralo-Caspian area of depression, as well as
the large number of lakes, whether disposed in ranges or
otherwise, between the Thian-shan and the Altai-Kurts-
chura, may have given rise to littoral phenomena. We
know from tradition that many small basins now ranged in
a row, like a string of beads {lacs a chapelet), once upon a
time formed a single large basin. Many large lakes are seen
to divide and form smaller ones from the disproportion be-
tween precipitation and evaporation. A very experienced
observer of the Kirghis Steppe, General Genz of Orenburg,
has conjectured that there formerly existed a water commu-
nication between the Sea of Aral, the Aksakal, the Sary-
Kupa, and the Tschagli. A great furrow is observed, run-
ning from southwest to northeast, which may be traced by
the way of Omsk, between Irtisch and Obi, through the
steppe of Barabinsk, which abounds in lakes, toward the
moory plains of the Saraoiedes, toward Beresow and the
shore of the Arctic Ocean. With this furrow is probably
connected the ancient and wide-spread tradition of a Bitter
Lake (called also the Dried Lake, Hanhai), which extended
eastward and southward from Hami, and in which a por-
tion of the Gobi, whose salt and reedy centre was found by
Dr. von Bunge's careful barometrical measurement to be only
2558 feet above the level of the sea, rose in the form of an
island.* It is a geological fact, which has not hitherto re-
ceived its due share of attention, that seals, exactly similar to
those which inhabit the Caspian Sea and the Baikal in shoals,
are found upward of 400 miles to the east of the Baikal, in
the small fresh-water lake of Oron, only a few miles in cir-
cumference. The lake is connected with the Witim, a tribu-
tary of the Lena, in which there are no seals.f The present
isolation of these animals and their distance from the mouth
of the Volga (fully 3600 geographical miles) form a remark-
able geological phenomenon, indicative of an ancient and ex-
tensive connection of waters. Can it be that the numerous

* Klaproth, Asia Polyghtta, p. 232, and Memoires relati/s d VAsie
(fj-om the Chinese Encyclopedia, published by command of the Em-
peror Kanghi, in 1711), t, ii., p. 342; Humboldt, Asie Centrale, t. ii.,
p. 125 and 135-143.

t Pallas, Zoographia Rosso-Asiatica, 1811, p. 115.

TRUE VOLCANOES. 409

depressions to which, throughout a large tract* of country,
this central part of Asia has been exposed, have called forth
exceptionally, on the convexity of the continental swelling,
conditions similar to- those produced on the littoral borders
of the fissures of elevation ? .^

From reHable accountsnrendered to the Emperor Kanghi,
we are acquainted with the existence of an extinct volcano
far to the east, in the northwestern Mantschurei, in the
neighborhood of Mergen (probably in lat. 48^° and long.
122° 20'' east). The eruption of scoriae and lava from the
mountain of Bo-shan or Ujun-Holdongi (the Nine Hills),
from 12 to 16 miles in a southwesterly direction from Mar-
gen, took place in January, 1721. The mounds of scoria3
thrown out on that occasion, according to the report of the
persons sent by the Emperor Kanghi to investigate the cir-
cumstances, were 24 geographical miles in circumference ; it
was likewise mentioned that a stream of lava, damming up
the water of the River Udelin, had formed a lake. In the
7th century of our era the Bo-shan is said to have had a
previous igneous eruption. Its distance from the sea is
about 420 geographical miles, similar to that of the Hima-
laya,* so that it is upward of three times more distant than

* It is not in the Himalaya range, near the sea (some portions of it,
between the colossal Kuncliinjinga and Shamalari, approach the sliore
of the Bay of Bengal within 428 and 37G geograpliical miles), that the
volcanic action has first burst forth, but in the third, or interior, parallel
chain, the Thian-shan, nearly four times as far removed from the
same shore, and that under very special circumstances, the subsidence
of ground in the neighborhood deranging strata and causing fissures.
We learn, from the study of the geographical works of the Chinese,
first instigated by me and afterward continued by my friend Stanislas
Julien, that the Kuen-liin, the northern boundary range of Thibet, the
Tsi-shi-shan of the Mongols, also possesses in the hill of Shin-Khieu
a cavern emitting uninterrupted flames {Asie Centrale, t. ii., p. 427-467
and 483). The phenomenon seems to be quite analogous to the Chi-
mjera in Lycia, which has now been burning for several thousands of
years (see above, p. 243-5, and note *) ; it is not a volcano, but a
fire-spring, difitusing to a great distance an agreeable odor (probably
from containing naphtha?). The Kuen-ldn, which, like me in the
Asie Ceritrak (t. i,, p. 127, and t. ii., p. 431), Dr. Thomas Thomson,
the learned botanist of Western Thibet (Flora Indica, 1855, p. 253),
describes as a continuation of the Hindu-Kho, which is joined from
the southeast by the Himalaya chain, approaches this chain at its west-
ern extremity to such a degree that my excellent friend, Adolph Schla-
gintweit, designates " the Kuen-liin and the Himalaya on the west side
of the Indus, not as separate chains, but as one mass of mountains." (Re-
port No. ix, of the Mwjnetic Survey in India, by Ad. Schlagintweit, 1856,
p. 61 .) In the whole extent toward the east, however, as far as 92° 20'
east longitude, in the direction of the starry lake, the Kuen-liin forms.

Vol. v.— S

410 COSMOS.

the volcano "of JoruUo. We are indebted for these remark-
able geognostic accounts from the Mantschurei to the indus-
try of W. P. Wassiljew {Geog. Bote, 1855, heft v., s. 31), and
to an essay by M. Semenow (the learned translator of Carl
Hitter's great work on Geology), in the 17th volume of the
Proceedings of the Imperial Russian Geographical Society.

In the course of the investigations into the geographical
distribution of volcanoes, and their frequent occurrence on
islands and sea-coasts ; that is to say, on the margins of con-
tinental elevations, the probable great inequality in the depth
to which the cru^t of the earth has hitherto been penetrated
has also been frequently brought under consideration. One
is disposed to believe that the surface of the internal molten
mass of the earth's body lies nearest to those points at which
the volcanoes have burst forth. But, as it may be conceived
that there are many intermediate degrees of consistency in
the solidifying mass, it is difficult to form a clear idea of any
such surface of the molten matter, if a change in the com-
prehensive capacity of the external firm and already solidified
shell be supposed to be the chief cause of all the subversions,
fissures, upheavals, and basin-like depressions. T[f we might
be allowed to determine what is called the thickness "of the
earth's crust in an arithmetical ratio deduced from experi-
ments drawn from Artesian wells and from the fusion-point
of granite — that is to say, by taking equal geothermal de-
grees of depth* — we should find it to be 20-^^^ geographical
miles, or ^^th of the polar diameter.f But the influences

as was shown so early as the 7th century of onr era, by minute descrip-
tions given under the Dynasty of Sai (Klaproth, Tableaux Historiques
de VAsie, p. 204), an independent chain running east and west, parallel
to the Himalaya, at a distance of about 7i degrees of latitude. The
brothers Hermann and Kobert Schlagintweit are the first who have had
the courage and the good fortune to traverse the chain of the Kuen-liin,
setting out from Ladak, and reaching the territory of Khotan, in the
months of July and September, 1856. According to their observations,
which are always extremely careful, the highest water-shedding mount-
ain chain is that on which is situated the Karakorum pass (18,304: feet),
which, stretching from southeast to northwest, lies parallel to the oppo-
site southerly portion of the Himalaya (to the west of Dhawalagiri).
The rivers Yarkland and Karakasch, which form a part of the great
water system of the Tarim and Lake Lop, rise on the northeastern slope
of the Karakorum chain. From this region of water-springs the trav-
elers arrived, by way of Kissilkorum and the hot springs (120° F.), at
the small mountain lake of Kiuk-kiul, on the chain of the Kuen-liin,
which stretches east and west (Report No. viii,, Agra^ 1857, p. 6).

* Cosmos, vol. i., p. 46, 174 ; see above, p. 37-40.

t Arago (Astron. Populaire, t. iii., p. 248) adopts nearly the same

#

TRUE VOLCANOES. 411

of the pressure and of the power of conducting heat exercised
by various kinds of rock render it likely that the geothermal
degrees of depth increase in value in proportion as the depth
itself increases.

Notwithstanding the very limited number of points at
which the fused interior of our planet now maintains an act-
ive communication with the atmosphere, it is still not unim-
portant to inquire in what ma^fc and to what extent the
volcanic exhalations of gas opera^ on the chemical composi-
tion of the atmosphere, and through it on the organic life de-
veloped on the earth's surface. We must, in the first place,
bear in mind that it is not so much the summit-craters them-
selves as the small cones of ejection and the fumaroles, which
occupy large spaces and surround so many volcanoes, that
exhale gases ; and that even whole tracts of country in Ice-
land, in the Caucasus, in the high land of Armenia, on Java,
the Galapagos, the Sandwich Islands, and New Zealand ex-
hibit a constant state of activity through solfataras, naphtha
springs, and salses. Volcanic districts, which are now reckon-
ed among those which are extinct, are likewise to be regard-
ed as sources of gas, and the silent working of the subterra-
nean forces, whether destructive or formative, within them is,
with regard to quantity, probably more productive than the
great, noisy, and more rare eruptions of volcanoes, although
their lava fields continue to smoke either visibly or invisibly
for years at a time. If it be said that the effects of these
small chemical processes can be but little regarded, for that
the immense volume of the atmosphere, constantly kept in
motion by currents of air, could only be affected in its primi-
tive mixture to a very small extent through means of such
apparently unimportant additions,* it will be necessary to

thickness of the earth's crust — namely, 40,000 metres, or about 22
miles ; Elie de Beaumont {Systhnes de Monta<jnes, t. iii., p. 1237) cal-
culates the thickness at about one fourth more. The oldest calcula-
tion is that of Cordier, in mean value 56 geof^raphical miles, an amount
which, according to Hopkins's mathematical theory of stability, would
have to be multiplied fourteen times, and would give between 688 and
860 geographical miles. I quite concur, on geological grounds, in the
doubts raised by Naumann in his admirable Lehrbuch der Geognosie
(vol. i., p. 62-64, 73-76, and 289), against this enormous distance of the
fluid interior from the craters of the active volcanoes.

* A remarkable example of the way in which perceptible changes of
mixture are produced in nature by very minute but continuous accu-
mulation is afforded by the presence of silver in sea-water, which was
discovered by Malaguti and confirmed by Field. Notwithstanding the
immense extent of the ocean and the trifling amount of surface pre-

412 COSMOS.

bear in mind the powerful influence exerted, according to the
admirable investigations of Percival, Saussure, Boussingault,
and Liebig, by tlu'ee or four ten-thousandth parts of carbonic
acid in our atmosphere on the existence of the vegetable
organism. From Bunsen's excellent work on the different
kinds of volcanic gas, it appears that among the fumaroles
of different stages of activity and local diversity some (as, for
example, at Hecla) yield J»|n 0*81 to 0-83 of nitrogen, and
in the lava streams of t^^ountain 0-78, with mere traces
(0*01 to 0-02) of carbonic acid ; while others in Iceland, as,
for instance, near Krisuvik, on the contrary, yield from 0*86
to 0-87 of carbonic acid, with scarcely 0*01 of nitrogen.*
We find likewise, in the important work on the emanations
of gas in Southern Italy and Sicily, by Charles Sainte-Claire
Deville and Bornemann, that there is an immense proportion
of nitrogen gas (0-98) in the exhalations of a fissure situated
low down in the crater of Vulcano, while the sulphuric acid
vapors show a mixture of 74-7 nitrogen gas and 18*5 oxygen,
a proportion which approaches pretty nearly to the composi-
tion of the atmospheric air. On the other hand, the gas
which rises from the spring of Acqua Santa,t in Catania, is
pure nitrogen gas, as was also the gas of the Volcancitos de
Turbaco at the time of my American journey.|

Are we to conclude that the great quantity of nitrogen
dispersed through the medium of volcanic action consists of
that alone which is imparted to the volcanoes by meteoric
water ? or are there internal and deeply-seated sources of
nitrogen ? It must also be borne in mind that the air dis-
solved in rain-water does not contain, like the atmosphere,
0*79 of nitrogen, but, according to my own experiments, only
0-69. Nitrogen is a source of increased fertility,§ by the form-

sented to it by the ships which traverse it, yet the trace of silver in the
sea-water has in recent times become observable on the copper sheath-
ing of ships.

* Bunsen, Ueher die ckemischen Prozesse der VulkaniscJien Gesteins-
bildungen, in Poggend,, Annalen, bd. Ixxxiii., s. 242 and 246.

t Comptes rendus de VAcad, des Sciences, t. xliii., 1856, p. 366 and
689. The first correct analysis of the gas which rushes with noise from
the great solfatara of Pozzuoli, and which was collected with great dif-
ficulty by M. Ch. St.-Claire Deville, gave the following results : Sul-
phurous acid(aaV/e sulfureux), 24-5 ; oxygen, 14'5 ; and nitrogen, 61*4.

X See above, p. 202, 208.

§ Boussingault, Economie Rurale (1851), t. ii., p. 724-726: "The
permanency of storms in the interior of the atmosphere (within the
tropics) is an interesting fact, being connected with one of the most
important questions in the physical histcrry of the globe, namely, that

TRUE VOLCANOES. 413

ation of ammonia, through the medium of the almost daily
electrical explosions in tropical countries. The influence of
nitrogen on vegetation is similar to that of the substratum of
atmospheric carbonic acid.

In analyzing the different gases of the volcanoes which lie
nearest to the equator (Tolima,.Furace, Pasto, Tuqueres, and
Cumbal), Boussingault. has discovered, along with a great
deal of aqueous vapor, carbonic acid and sulphureted hydro-
gen gas, but no muriatic acid, no nitrogen, and no free hy-
drogen.* The influence still exercised by the interior of our
planet on the chemical composition of the atmosphere in with-
drawing this matter, in order to give it out again under other
forms, is certainly but an insignificant part of the chemical
revolutions which the atmosphere must have undergone in
remote ages on the eruption of great masses of rock from open
fissures. The conjecture as to the probability of a very large,
portion of carbonic acid gas in the ancient aeriform envelope
is strengthened by a comparison of the thickness of the pres-
ent seams of coal with that of the thin coal-strata (seven lines
in thickness) which, according to Chevandier's calculations,
our thickest woods in the temperate zone would yield to the
soil in the course of one hundred years.f

In the infancy of geognosy, previous to Dolomieu's ingen-
ious conjectures, the source of volcanic action was not placed

of the fixation of the nitrogen of the air in organized beings. When-
ever a series of electric sparks passes through the humid atmosphere,
the production and combination of nitric acid and ammonia take place.
The nitrate of ammonia uniformly accompanies the rain during a
storm, and being by nature 'fixed it can not maintain itself in a state
of vapor ; carbonate of ammonia is found in the air, and the ammonia
of the nitrate is carried to the earth by the rain. Thus it appears, in
fact, to be an electric action which disposes the nitrogen of the atmos-
phere to become assimilated by organized beings. In the equinoxial
zone, throughout the whole year, every day, and pi-obably even every
moment, there is a continual succession of electric discharges going on.
An observer stationed at the equator, if he were endowed with organs
sufficiently sensitive, would hear without intermission the noise of
thunder." Sal ammoniac, however, together with common salt, are
from time to time found as products of sublimation, even in lava
streams — on Hecla, Vesuvius, and ^tna, in the volcanic chain of
Guatemala (the vo^no of Izalco), and, above all, in Asia, in the vol-
canic chain of thai^hian-shan. The inhabitants of the country be-
tween Kutsch, Turfan, and Hami pay their tribute to the Emperor of
China in certain years* in sal ammoniac (in Chinese, nao-sha, in Per-
sian nuskaden), which is an important article of internal trade. {Asie
Centrale, t. ii., p. 33, 38, 45, and 428.)

* Viajes de Boussingault (1849), p. 78.

t Cosmos, vol. i., p. 280-282.

414 COSMOS..

below the most ancient rock formations, which were then
generally supposed to be granite and gneiss. Resting on
some feeble analogies of inflammability, it was long believed
that the source of volcanic eruptions, and the emanations of
gas to which they for many centuries gave rise, was to be
sought for in the later upper silurian floetz strata, containing
combustible matter. A more general acquaintance with the
earth's surface, profounder and more strictly conducted geo-
logical investigations, together with the beneficial influence
which the great advances made by modern chemistry have
exercised in the study of geology, have taught us that the
three great groups of volcanic or eruptive rock (trachyte,
phonolite, and basalt), when viewed as large masses, appear,
when compared together, to be of different ages, and for the
most part widely separated from each other. All three, how-
ever, have come later to the surface than the Plutonic gi'an-
ite, the diorite, and the quartz porphyry — later than all the
Silurian, secondary, tertiary, and quartary (pleistocene) form-
ations; and that they frequently traverse the loose strata of
the diluvial formations and bone-breccias. A striking vari-
ety* of these intersections, compressed into a small space, is
exhibited, as we learn from Kozet's observations, in Auvergne.
AVhile the great trachytic mountain masses of the Cantal,
Mont-Dore, and Puy de Dome .penetrate the granite itself,
and at the same time inclose in some parts (for example, be-
tween Vic and Aurillac, and at the Giou de Mamon) large
fragments of gneissj and limestone, we find also the trachyte
and basalt intersecting as dikes the gneiss,- and the coal-beds
of the tertiary and diluvial strata. Basalt and phonolite,
closely allied to each other, as the Auvergne and the central
mountains of Bohemia prove, are both of more recent forma-
tion than the trachytes, which are frequently traversed in
layers by basalts.^ The phonolites are, on the other hand,

* Eozet, Mimoire sur les Volcans d' Auvergne, in the Memoires de la
Soc. G'eol de France, 2me Scrie, t. i., 1844, p. 64 and 120-130: "The
basahs (like the trachytes) have penetrated through the gneiss, the
granite, the coal formations, the tertiary formations, and the oldest
diluvian bed. The basalts are even frequently seen overlying masses
of basaltic bowlders ; they have issued from an infinite number of open-
ings, several of which are still perfectly recognizaWe. Many of them
exhibit cones of scorias more or less considerable, but nowhere do we
find craters similar to those which have given out streams of lava."

t Resembling the granitic fragments imbedded in the trachyte of
JoruUo. See above, p. 303.

X Also in the Eifel, according to the important testimony of the mine
director, Von Dechen. See above, p. 226.

TRUE VOLCANOES. 415

more ancient than the basalts; where they probably never
form dikes, but on the contrary dikes of basalt frequently in-
tersect the porphyritic schist (phonolite). In the chain of
the Andes belonging to Quito I found the basalt formation
a great distance apart from the prevailing trachytes ; almost
solely at the Rio Pisque and in the valley of Guaillabamba.*
As in the volcanic elevated plain of Quito every thing is
covered with trachytes, trachytic conglomerates, and tufas, it
was my most earnest endeavor to discover, if possible, some
point at which it might be clearly seen on which of the older
rocks the mighty cone and bell-shaped mountains are placed,
or, to speak more precisely, through which of them they had
broken forth. Such a point I was so fortunate as to dis-
cover in the month of June, 1802, on my way from Eio-
bamba Nuevo (9483 feet above the surface of the South Pa-
cific), when I attempted to ascend the Tunguragua, on the
side of the Cuchilla de Guandisava. I proceeded from the
delightful village of Penipe over the swinging rope-bridge
(pue?ite de maroma) of the Kio Puela to the isolated hacienda
of Guansce (7929 feet), where to the southeast^ opposite to
the point at which the Rio Blanco falls into the Rio Cham-
bo, rises a splendid colonnade of black trachyte resembling
pitch-stone. It looks at a distance like the basalt quarry at
Unkel. At Chimborazo, a little higher than the basin of
Yana-Cocha, I saw a similar group of trachytic columns of
greater height, but less regularity. The columns to the south-
east of Penipe are mostly pentagonal, only fourteen inches
in diameter, and frequently bent and diverging. At the foot
of this black trachyte of Penipe, not far from the mouth of
the Rio Blanco, a very unexpected phenomenon presents itself
in this part of the Cordilleras — greenish-white mica-slate with
garnets interspersed in it ; and farther on, beyond the shal-
low stream of Bascaguan, at the hacienda of Guansce, near

* See above, p. 313. The Rio de Guaillabamba floAvs into the Eio
de las Esmeraldas. The village of Guaillabamba, near which I found
the isolated oliviniferous basalt, is only 6480 feet above the level of the
sea. An intolerable heat prevails in the valley, v^'hich is still more
intense in the Valle de Chota, between Tusa and the Villa de Ibarra,
the sole of which sinks to 5288 feet, which is rather a chasm than a val-
ley, being scarcely 9G00 feet wide and 4800 feet deep (Humboldt, Rec.
d Observations Astronomiques, vol. i., p. 307). The rubbish-ejecting
Volcan de Ansango, on the descent of the Antisana, does not belong
to the basalt formation at all : it is an oligoclase trachyte resembling"
basalt (compare, for the distances, Antagonisme des Basaltes et des Tra-
chytes, my IJssai Geognostique sur le Gisement des Roches, 1823, p. 348
and 359, and generally, p. 327-336).

416 COSMOS.

the shore of the Rio Puela, and probably dipping below the
mica-slate granite of a middling-sized grain, with light red-
dish feldspar, a small quantity of blackish-green mica, and a
great deal of grayish-white quartz. There is no hornblende
nor is there any syenite. Thus it appears that the trachytes
of the volcano of Tungurahua, resembling those of Chimbo-
razo in their mineralogical condition, that is to say, consist-
ing of a mixture of oligoclase and augite, have here pene-
trated granite and mica-slate. Farther toward the south,
and a little to the east of the road leading from Riobamba
Nuevo to Guamote and Ticsan, in that part of the Cordille-
ras which recedes from the sea-shore, the rocks formerly called
primitive, mica-slate, and gneiss, make their appearance every
where, toward the foot of the colossal Altar de los Collanes,
the Cuvillan, and the Paramo del Hatillo. Previous to the
arrival of the Spaniards, even before the dominions of the
Incas extended so far to the north, the natives are said to
have worked metalliferous beds in the neighborhood of the
volcanoes. A little to the south of San Luis numerous dikes
of quartz are observed running through the greenish clay-
slate. At Guamote, at the entrance to the grassy plain of
Tiocaxa, we found large masses of rock, consisting of quartz-
ites very poor in mica, of a distinct linear parallel structure,
running i*egularly at an angle of 70 degrees to the north.
Farther to the south, at Ticsan, not far from Alausi, the
Cerro Cuello de Ticsan shows large masses of sulphur im-
bedded in a layer of quartz, subordinate to the neighboring
mica-slates. So great a diffusion of quartz in the neighbor-
hood of trachytic volcanoes appears at first sight somewhat
strange. The observations which I made, however, of the
overlying, or rather of the breaking forth of trachyte from
mica-slate and granite at the foot of the Tungurahua (aphe-«
nomenon which is as rare in the Cordilleras as in Auvergne),
have been confirmed, after an interval of forty-seven years,
by the admirable investigations of the French geologist Se-
bastian Wisse at the Sangay.

That colossal volcano, 1343 feet higher than Mont Blanc,
entirely destitute of lava streams (which Charles Deville de-
clares are also wanting in the equally active Stromboli), but
ejecting uninterruptedly, at least since the year 1728, a black,
and frequently brightly glowing rock, forms a trachytic isl-
and of scarcely eight geographical nailes in diameter,* in the

* Sebastian Wisse, Exploration du Volcan de Sangay, in the Comptes
rendusde I' Acad, des Sciences, t. xxxvi., 1853, p. 721 ; comp. also above,
p. 239.

TRUE VOLCANOES. 417

midst of beds of granite and gneiss. A totally opposite con-
dition of stratification is exhibited in the volcanic district of
Eifel, as I have already observed, both from the activity
which once manifested itself in the Maars (or mine-funnels)
sunk in the Devonian schist, and that shown in the raised
structures from which lava streams flow, as on the long ridge
of the Mosenberg and Gerolstein. The surface does not here
indicate what is hidden in the interior. The absence of tra-
chyte in volcanoes which were so active thousands of years
ago is a still more striking phenomenon. The augitiferous
scorias of the Mosenberg, which partly accompany the ba-
saltic lava stream, contain small burned pieces of schist, but
no fragments of trachyte, and in the neighborhood the tra-
chytes are absent. This species of rock is only to be seen in
the Eifel in a state of entire isolation,* far from the Maars
and lava-yielding volcanoes, as in the Sellberg and Quiddel-
bach, and in the mountain chain of Reimerath. The differ''
ent nature of the formations through which the volcanoes
force their way, so as to operate with power on the outer
crust of the earth, is geologically as important as the mate-
rial which they throw out.

The conditions of configuration in those rocky structures
through which volcanic action manifests itself, or has en-
deavored to do so, have at length been in modern times far
more completely investigated and described, in their often

According toBoussingault,the ejected fragments of trachyte brought
home by Wisse, and collected on the upper descent of the cone (the
traveler reached an elevation of 960 feet below the summit, which is
itself 485 feet in diameter), consist of a black, pitch-like fundamental
mass, in which are imbedded crystals of glassy (?) feldspar. It is a
very remarkable phenomenon, and one which up to the present time
seems to stand alone in the history of volcanic ejections, that, along
with these large black pieces of trachyte, small sharp-edged fragments
of pure quartz are thrown out. According to a letter from my friend
Boussingault, dated January, 1851, these fragments are no longer than
four cubic centimetres in bulk. No quartz is found disseminated in
the trachytic mass itself. All the volcanic trachytes which I have ex-
amined in the Cordilleras of South America and Mexico, and even the
trachytic porphyries in which the rich silver veins of Real del Monte,
Moran, and Regla are contained, to the north of the elevated valley
of Mexico, are entirely destitute of quartz. Notwithstanding this seem-
ing antagonism, however, of quartz and trachyte in still-active volca-
noes, I am by no means inclined to deny the volcanic origin of the
" trachytes et porphyres meulieres (mill-stone trachytes)" to which Beu-
dant first drew attention. The mode, however, in which these are
formed, being erupted from fissures, is entirely different from the form-
ation of the conical and dome-like trachyte structures.

* See above, p. 221-225.

S2

418 COSMOS.

very complicated variations, in the most distant quarters of
the globe than in the previous century, when the entire mor-
phology of volcanoes M'as limited to conical and bell-shaped
mountains. There are many volcanoes whose configuration,
altitude, and range (what the talented Carl Friedrich Nau-
mann calls the geotectonics)* we now know in the most sat-
isfactory manner, while we continue in the greatest ignorance
regarding the composition of their different rocks and the
association of the mineral species which characterize their
trachytes, and which are recognizable apart from the princi-
pal mass. Both kinds of knowledge, however — the morphol-
ogy of the rocky piles and the oryctognosy of their compo-
sition— are equally necessary to the perfect understanding of
volcanic notion ; nay, tlie latter, founded on crystallization
and chemical analysis, on account of the connection with
Plutonic rocks (porphyritic quartz, green-stone, and serpent-
ine) is of even greater geognostic importance. The little we
believe we know of what is called the volcanicity of the Moon
depends too, from the very nature of the knowledge, on con-
figuration alone.f

* The fullest information we possess on any volcanic district, found-
ed on actual measurements of altitudes, angles of inclination, and
profile views, is contained in the beautiful work of the astronomer of
Olmiitz, Julius Schmidt, on Vesuvius, the solfatara, Monte Nuovo, the
Astroni, Kocca Monfina, and the old volcanoes of the Papal territory
(in the Albanian Mountains, Lago Bracciano, and Lago di Bolsena).
See his hypsometrical work, Die Eruption des Vesuvs im Mai, 1855,
with Atlas, plates iii., iv., ix.

t The progressive perfection of our acquaintance with the formation
of the surface of the Moon as derived from numerous observers, from
Tobias Mayer down to Lohrmann, Madler, and Julius Schmidt, has
tended, on the whole, rather to diminish than to strengthen our belief in ,
great analogies between the volcanic structures of the earth and those
of the moon; not so much on account of the conditions of dimension and
the early recognized ranging of so many ring-shaped mountains, as on
account of the nature of the rills and of the system of rays which cast
no shadows (radiations of light) of more than 400 miles in length and
from 2 to 16 miles in breadth, as inTycho, Copernicus, Kepler, and
Aristarchus. It is remarkable, however, that Galileo, in his letter to
Father Christoph Grienberger, Sulle montuosita della Lima, should have *
thought of comparing annular mountains, whose diameters he consid-
ered greater than they actually are, to the circumvallated district of
Bohemia, and that the ingenious Robert Hooke, in his "Micography,"
attributes the type of circular formation almost universally prevalent
on the moon to the action of the interior of its body on the exterior
(vol. ii,, p. 701, and vol. iv., p. 496). With respect to the annular
mountain ranges of the moon, I have been of late much interested
with the relation between the height of the central mountain and that
of the circumvallation or margins of the crater, as well ns by the exist-

TRUE VOLCANOES.

419

If, as I would fain hope, what I here propound regarding
the classification of the volcanic rocks — or, to speak more

ence of parasitic craters on the circumvallation itself. The result of
all the careful observations of Julius Schmidt, who is occupied with
the continuation and completion of Lohrmann's Topography of the
Moon, establishes " that no single central mountain attains the height
of the wall of its crater, but that in all cases it probably even lies, togeth-
er with its summit, considerably below that surface of the moon from
which the crater is erupted." While the cone of ashes in the crater of
Vesuvius, which rose on the 22d of October, 1822, according to Brios-
chi's trigonometrical measurement, exceeds in height the Punta del
Palo, the highest edge of the crater on the north (618 toises above the
sea), by about 30 feet, and was visible at Naples, many of the central
mountains of the moon, measured by Madler and the Olmiitz astrono-
mer, lie fully 6400 feet lower than the mean margin of circumvalla-
tion, nay, even 100 toises below what may be taken as the mean sur-
face level of that part of the moon to which they respectively belong
(Madler, in Schumacher' s Jahrhuch fur 1841, p. 272 and 274; and Jul.
Schmidt, Der Mond, 1856, s. 62). In general the central mountains,
or central mountain masses of the moon, have several summits, as in
Theophilus, Petavius, and Bulliald. In Copernicus there are six cen-
tral mountains, and Alphonsus alone exhibits a true, central, sharp-
pointed peak. This state of things recalls to mind the Astroni in the
Phlegrsean Fields, on whose dome-formed central masses Leopold von
Buch justly lays much stress. "These masses," he says, "like those
in the centre of the annular mountains of the moon, did not break
forth. There existed no pernMwient connection with the interior — no
volcano, but they rather appeared like models of the great trachytic
unopened domes so abundantly dispersed over the earth's crust, such
as the Puy de Dome and Chimborazo." (Poggendorflf' s Annakn, bd.
xxxvii., 1836, p. 183.) The circumvallation of the Astroni is of an
elliptic form, closed all round, and rises in no part higher than 830
feet above the level of the sea. The tops of the central summits lie
more than 660 feet lower than the maximum of the southwestern wall
of the crater. The summits form two parallel ridges, covered with
thick bushes (Julius Schmidt, Eruption des Vesuvs, s. 147, and Der 3Iond,
s. 70 and 103). One of the most remarkable objects, however, on the
whole surface of the moon is the annular mountain range of PetaVius,
in which the whole internal floor of the crater expands convexly in
the form of a tumor or cupola, and is crowned besides with a central
mountairw The convexity here is a permanent form. In our terres-
trial volcanoes the flooring of the crater is only temporarily raised by
the force of internal vapors, sometimes almost to the height of the mar-
gin of the crater, but as soon as the vapors force their way through the
floor sinks down again.- The largest diameters of craters on the earth
are the Caldeira;de Pogo, according to Charles Deville 4100 toises
(4*32 geographic^ miles), and the Caldeira de Palma, according to
Leop. von Buch 3100 toises ; Avhile, on the moon, Theophilus is 50,000
toises, and Tycho 45,000 toises, or respectively 52 and 45 geographical
miles in diameter. Parasitic craters, erupted from a marginal wall
of the great crater, are of very frequent occurrence on the moon. The
base of these parasitic craters is usually empty, as on the great rent
margin of the Maurolycus ; sometimes, but more rarely, a smaller cen-

1:20 COSMOS.

precisely, on the arrangement of the trachytes according to
their composition — excites any particular interest, the merit
of this classification is entirely due to my friend and Sibe-
rian fellow-traveler, Gustav liose. His accurate observa-
tion of nature, and the happy combination which he possesses
of chemical, crystallo-mineralogical, and geological knowl-
edge, have rendered him peculiarly well qualified to promul-
gate new views on that set of minerals whose varied but fre-
quently recurring association is the product of volcanic ac-
tion. This great geologist, partly at my instigation, has with
the greatest kindness, especially since the year 1834, repeat-
edly examined the fragments which I brought from the slopes
of tlie volcanoes of New Granada, Los Pastos, Quito, and the
high land of Mexico, and compared them with the specimens
from other parts of the globe contained in the rich mineral
collection of the Berlin Cabinet. Before my collections were
separated from those of my companion Aime Bonpland, Leo-
pold von Buch had examined them microscopically with per-
severing diligence (in Paris, 1810-1811, between his return
from Norway and his voyage to TenerifFe). He had also at
an earlier period, during my residence with Gay-Lussac at
Rome (in the summer of 1805), as well as afterward in
France, made himself acquaintedr- with what I had noted
down in my traveling journal on the spot, in the month of
July, 1802, respecting certain volcanoes, and in general on
the alRnity between volcanoes and certain porphyries desti-
tute of quartz.* I preserve, as a memorial which I consider

tral mountain, perhaps a cone of eruption, is seen in them, as in Lo-
gomontanus. In a beautiful sketch of the crater system of -^tna,
which my friend Christian Peters, the astronomer (now in Albany,
North America), sent me from Fiensburg, in August, 1854, the para-
sitic marginal crater, called the Pozzo di Fuoco, which was formed in
January,' 1833, on the east-southeast side, and which had several vio-
lent eruptions of lava, is distinctly recognizable.

* The unspecific and indefinite term *' trachyte" (Rauhstein), which
is now so generally applied to the rock in which the volcanoes break
out, was first given to a rock of Auvergne in the year 1822, by Hauy, in
the second edition of his Traite de Mineralogie, vol. iv., p, 579, with a
mere notice of the derivation of the word, and a short description in
which the older appellations of granite chauffe en "j^acQ of Desmarets,
trap-porphyry, and domite are not even mentionedV It was only by
oral communication, originating in Ilauy's Lectures in the Jardin des
Plantes, that the term " trachyte" was propagated previous to 1822;
for example, in Leopold von Buck's treatise on basaltic islands and
craters of upheaval, published in 1818; in Daubuisson's Traite de Min-'
eralogie^ 1819; and in 'Beudant's important work. Voyage en Hongrie.
Prom letters lately received by me from M. Elie de Beaumont, I find

TRUE VOLCANOES. 421

invaluable, some sheets with remarks on the volcanic prod-
ucts of the elevated plateaux of Quito and Mexico, which
the great geologist communicated to me for my information

that the recollections of M. Delafosse, formerly Aide-Naturaliste to
Hauy, and now Member of the Institute, fix the application of the
term "trachyte" between the years 1813 and 1816. The publication
of the term " domite" by Leop. v. Bucli seems, according to Ewald, to
have occurred in the year 1809 ; it is first mentioned in the third let-
ter to Karsten {Geognost. Beohacht. mif JReisen durch Deutschland und
Italien, bd. ii., 1809, s. 244). "The porphyry of the Puy de Dome,"
it is there stated, "is a peculiar and hitherto nameless rock, consisting
of crystals of feldspar with a glassy lustre, hornblende, and small lam-
inae of black mica. In the clefts of this kind of rock, which I provi-
sionally term domite, I find beautiful drusic cavities, the walls of which
are covered with crystals of iron-glance. Through the whole length
ot the Puy cones of domite alternate with cones of cinders." The
second volume of the Travels, containing the letters from Auvergne,
was printed in 1806, but not published till 1809, so that the publication
of the name of domite properly belongs to the latter year. It is singu-
lar that four years later, in Leopold von Buch's treatise on the trap
porphyry, domite is not even mentioned. In referring to a drawing
of the profile of the Cordilleras, contained in the journal of my traveis
in the month of July, 1802, and included between the 4th degree north
and 4th degree south latitude, under the inscription " Aifinite entre le
feu volcanique et les porphyres," my only object was to mention that
this profile, which represents the three breakings through of the vol-
canic groups of Popayan, Los Pastos, and Quito, as well as the erup-
tion of the trap porphyry in the granite and mica-slate of the Paramo
de Assuay (on the great road from Cadlud, at a height of 15,526 feet),
led Leopold von Buch, too kindly and too distinctly, to ascribe to me
the merit of having first noticed "that all the volcanoes of the chain
of the Andes have their foundation in a porphyry which is a peculiar
kind of rock, and belongs essentially to the volcanic formations" (Ab-
liandiamjen der Akademie der Wissensch. zu Berlin, aus den Jahren
1812-1813, s. 131, 151, and 153). I may, indeed, have noticed the
phenomenon in a general way, but it had already, as early as 1789,
been remarked by Nose, whose merits have long been too little appreci-
ated, in his Orographical Letters, that the volcanic rock of the Siebenge-
birge is " a peculiarly Rhenish kind of porphyry, closely allied to ba-
salt and porphyritic schist." He says "that this formation is especial-
ly characterized by glassy feldspar," which he proposes should be called
Siinidine, and that it belongs, judging from the age of its formation, to
the middle floetz rocks (Niederrheinische Reise, th. i., s. 26, 28, and 47;
th. ii., s. 428). I do not find any grounds for Leopold von Buch's con-
jecture that Nose considered this porphyry formation, which he not
very happily terms granite porphyry, as well as the basalts, to be of
later date than the most recent floetz rocks. "The whole of this
rock," says the great geologist, so early removed from among us,
"should be named after the glassy feldspars (therefore sanidine por-
phyry), had it not already received the name of trap porphyry" {Abh.
der Berl. Akad. aus den Jahren 1812-13, s. 134). The history of the
systematic nomenclature of a science is so far of importance as the
succession of ijrevalent opinions is found reflected in it.

422

COSMOS.

more than forty-six years ago. Travelers, as I have else-
where* said, being merely the bearers of the imperfect knowl-
edge of their age, and their observations being deficient in
many of the leading ideas, that is to say, those discriminat-
ing marks which are the fruits of an advancing knowledge,
the materials which have been carefully collected and geo-
graphically arranged will almost alone maintain an enduring
value.

To confine the term trachyte, as is frequently done (on ac-
count of its earliest application to the rocks of Auvergne and
of the Siebengebirge, near Bonn), to a volcanic rock contain-
ing feldspar, especially AVerner's vitreous feldspar. Nose's and
Abich's sanidine, is fruitlessly to break asunder that intimate
concatenation of volcanic rock which leads to higher geo-
logical views. Such a limitation might justify the expres-
sion " that in ^tna, so rich in Labradorite, no trachyte oc-
curs." Indeed, my own collections are said to prov'e that " no
single individual of the countless volcanoes of the Andes con-
sists of trachyte ; that, in fact, the substance of which they
are composed is albite, and that therefore, as oligoclase was
at that time (1835) always erroneously considered to be al-
bite, all kinds of volcanic rock should be designated andesite
(consisting of albite with a small quantity of hornblende)."!
Gustav Rose has taken the same view that I myself adopted,
from the impressions which I brought back wdth me from my
journeys, on the common nature of all volcanoes, notwith-
standing a mineralogical variation in their internal composi-
tion ; on the principle developed in his admirable essay on
the feldspar groups, J in his classification of the trachytes, he
generalizes orthoclase, sanidine, the anorthite of Mount Som-
ma, albite, Labradorite, and oligoclase, as forming the feld-
spathic ingredient of the volcanic rocks. Brief appellations
which are supposed to contain definitions led to many ob-
scurities in orology as well as in chemistry. I was myself
for a long time inclined to adopt the expressions orthoclase

* Humboldt, Kleinere Schriften, bd. i., Vorrede, s, iii.-v.

t Leop. V. Buch, in Poggend., Annalen, bd. xxxvii., 1836, s. 188, 190.

X Gustav Rose, in Gilbert's Annalen, bd. Ixxiii., 1823, s. 173 ; and
Annates de CIdmie et de Physique, t. xxiv,, 1823, p. 16. Oligoclase was
first held by Breithaupt as a new mineral species (Poggendoi-fTs Anna-
len, bd. viii., 1826, s. 238). It afterward appeared that oligoclase >vas
identical with a mineral which Berzelius had observed in a granite
dike resting upon gneiss near Stockholm, and which, on account of the
resemblance in its chemical composition, he had called "Natron Spo-
dumen." (Foggcndorff s AnnciL, bd. ix., 1827, s. 281.)

TRUE VOLCANOES. 423

trachytes, or Labrador trachytes, or oligoclase trachytes, thus
comprehending the glassy feldspar (sanidine), on account of
its chemical composition, under the species orthoclase (com-
mon feldspar). The terms were at least well-sounding and
simple; but their very simplicity must have induced error;
for, thougli Labrador trachyte points to JEtna and to Strom-
boli, yet oligoclase trachyte, in its important two-fold com-
bination with augite and hornblende, would erroneously con-
nect the widely diffused and very dissimilar formations of
Chimborazo and the volcano of Toluca. It is the associa-
tion of a feldspathic element with one or two others which
here forms the characteristic feature, as it does in the forma-
tion of some mineral dikes.

The following is a view of the divisions into which Gustav
Rose, subsequently to the winter of 1852, distributes the tra-
chytes, in reference to the crystals inclosed in them, and
separately recognizable. The chief results of this work, in
which there is no confounding of oligoclase with albite, were
obtained ten years earlier ; when my friend discovered, in
the course of his geognostic investigations in the Riesenge-
birge, that the oligoclase there formed an essential ingredient
of the granite, and his attention being thus directed to the
importance of oligoclase as an ingredient of that rock, he was
induced to look for it likewise in other rocks.^ This exam-
ination led to the important result (Poggend., Ann., bd. Ixvi.,
1845, s. 109) that albite never forms a part in the mixed
composition of any rock.

jF'i7'st Division. — " The principal mass contains only crys-
tals of glassy feldspar, which are laminar, and in general
large. Hornblende and mica either do not occur in it at all,
or in extremely small quantity, and as an entirely unessential
admixture. To this division belongs the trachyte of the
Phlegraean Fields (Monte Olibano, near Pozzuoli), that of
Ischia and of La Tolfii, as also a part of the Mont Dore (the

* See Gustav Rose on the granite of the Riesengebirge, in Poggen-
dorfTs Ann., bd. Ivi., 1842, s. G17. Berzelius had found the oligoclase,
his "Natron Spodumen," only in a dike of granite ; in the treatise just
cited it is for the first time spoken of as an ingredient in the composi-
tion of granite (the mineral itself). Gustav Rose here determined the
oligoclase according to its specific gravity, the greater proportion of
lime contained in it as compared with albite, and its greater fusibility.
The same compound with which he had found the specific gravity to
be 2-682 was analyzed by Rammelsberg (Handtvdrterhuch der Miner-
alog., supplem. i., s. 104; and G. Rose, Ueber die zur Gronitgriippe
gehbrenden Gchirgsarten, in the Zeitschr. der Uentschen geol, Gescll-
'schaft, bd. i., 1849, s. 364).

424 COSMOS.

Grande Cascade). Augite is but very rarely found in small
crystals in trachytes of Mont Dore* — never in the Phlegraean
Fields together with hornblende ; nor is leucite, of which
last, howev.er, Hoffmann collected some pieces on the Lago
Averno (on the road to Cumae), while I found some on the
slope of the Monte Nuovoj (in the autumn of 1822). Leu-
cite ophyr in lo.ose fragments is more frequent in the island
of Procida and the adjoining Scoglio di S. Martino."

Second Division. — "The ground mass contains some de-
tached crystals of glassy feldspar, and a profusion of small
snow-white crystals of oligoclase. The latter are frequently
overspread with the glassy feldspar in regular order, and
form a covering about the feldspar, as is so frequently seen
in G. Rose's granitite (the principal mass of the Eiesenge-
birge and Isergebirge, consisting of granite with red feldspar,
particularly rich in oligoclase and magnesian mica, but with-
out any white potash mica). Hornblende and mica, and in
some modifications augite, occasionally appear in small quan-
tity. To this division belong the trachytes of the Drachen-
fels and of the Perlenhardt, in the Siebengebirge,j: near Bonn,

* Rozet, Sur les Montagues de I'Auvergne, in the Man. de la Soc.
Gcol. de France, 2me Serie, t. i., partie i., 1844, p. 69.

t Fragments of leucite ophyr, collected by me at the Monte Nuovo,
are described by Gustav Rose in Fried. Hoffmann's Geognostischen Beo~
hachtungen, 1839, s. 219.. On the trachyte of the Monte di Procida of
the island of the same name, and the rock of San Martino, see Roth,
Monographie des Vesuvs, 1857, s. 519-522, tab. viii. The trachyte of
the island of Ischia contains in the Arso, or stream of Cremate (1301),
vitreous feldspar, brown mica, green augite, magnetic iron, and olivin
(s. 528), but no leucite.

X The geologico-topographical conditions of the Siebengebirge near
Bonn have been developed with comprehensive talent and great exact-
ness by my friend H. von Dechen, director of mines, in the 9th annual
volume of the Verhandlungen des Naturhistorischen Vereines der Preuss,
Rheinlande, und Westphalens, 1852, s. 289-567. All the chemical analy-
ses of the trachytes of the Siebengebirge which have hitherto appeared
are there collected (p. 323-356) ; mention is also made of the trachytes
of the Drachenfels and Rottchen, in which, besides the large crystals
of sanidine, several small crystalline particles may be distinguished in
the fundamental mass, " These portions have been found by Dr. Bothe,
on chemica lanalysis in Mitscherlich's laboratory, to be oligoclase,
corresponding exactly with the oligoclase of Danvikszoll (near Stock-
holm) noticed by Berzelius." (Dechen, s. 340-346.) The Wolken-
burg and the Stenzelberg are destitute of glassy feldspar (s. 357 and 363),
and belong, not to the second division, but to the third ; they contain a
Toluca rock. That section of the geological description of the Sie-
bengebirge which treats of the relative age of trachyte conglomerate
and basalt conglomerate contains many new views (p. 405-461). " With
the more rare dikes of trachyte in the trachyte conglomerates, which

TRUE VOLCANOES. 425

and many modifications of the Mont Dore and Cantal ; some
trachytes also of Asia Minor (for which we are indebted to
that industrious traveler Peter von Tschichatscheff ), of J^fi-
un Karahissar (famous for the culture of the poppy) and Me-
hammed-kyoe in Phrygia, and of Kayadschyk and Donan-
lar in Mysia, in which glassy feldspar, with a great deal of
oligoclase, some hornblende, and brown mica, are mingled."
Third Division. — "The ground mass of this dioritic tra-
chyte contains many small crystals of oligoclase, with black
hornblende and brown magnesia mica. To this belong the
trachytes of ^gina,* of the valley of Kozelnik, near Schem-

prove that the formation of trachyte has still continued after the de-
posit of the conglomerate (s. 413), are associated a great number of ba-
salt courses (s. 416). The basalt formation extends decidedly into a
later basalt than the trachyte formation, and the principal mass of the
basalt is here more recent than the trachyte. On the other hand, a
portion of this basalt only, and not of all basalts (s. 323), is more re-
cent than the great mass of the brown-coal rocks. Both formations,
the basalt and the brown-coal rocks, run into each other in the Sie-
bengebirge, as well as in many other places, and must be considered
in the aggregate as contemporaneous." Where very small crystals of
quartz occur by way of rarity in the trachytes of the Siebengebirge, as
(according to Nog^erath and Bischof ) in the Drachenfels and in the
valley of Rhondorf, they fill up cavities and seem to be of later forma-
tion (p. 361 and 370) ; caused perhaps by efflorescence of the sanidine.
On Chimborazo I have on one solitary occasion seen similar deposits
of quartz, though very thin, on the internal sui;faces of the cavities of
some very porous, brick-red masses of trachyte at an elevation of
about 17,000 feet (Humboldt, Gisement des Roches, 1823, p. 336).
These fragments, which are frequently mentioned in my journal, are
not deposited in the Berlin collections. Efflorescence of oliglocase, or
of the whole fundamental mass of the rock, may also yield such traces
of disengaged silicic acid. Some points of the Siebengebirge still
merit renewed and persevering investigation. The highest summit,
the Lowenburg, represented as basalt, seems, from the analyses of
Bischof and Kjerulf, to be a doleritic rock (H. v. Dechen, s. 383, 386,
393). The rock of the little Rosenau, which has sometimes been called
sanido-phyre, belongs, according to G. Rose, to the first division of his
trachytes, and is very closely allied to many of the trachytes of the
Ponga Islands. The trachyte of the Drachenfels with large crystals
of glassy feldspar seems, according to Abich's yet unpublished investi-
gations, most nearly to resemble the Dsyndserly-dagh, which rises to a
height of 8526 feet, to the north of the great Ararat, from a formation
of nummuhtes under-dipped by Devonian strata.

* From the close propinquity of Cape Perdica, of the island of
iEgina, to the long famous red-brown Trozen-trachytes {Cosmos, see
above, p. 219) of the peninsula of Methana, and from the sulphur
s])rings of Bromolimni, it is probable that the trachytes of Methana, as
well as those of the island of Kalauria, near the small town of Poros,
belong to the same third division of Gustav Rose (oligoclase with horn-
blende and mica) (Curtius, Peloponnesos, bd. ii., s. 439, 446, tab. xiv.).

426 COSMOS.

nitz ;* of Nagyag, in Transylvania ; of JMontabaur, in the
Duchy of Nassau ; of the Stenzelberg and the Wolkenburg,
in the Siebengebirge, near Bonn ; of the Puy de Chaumont,
near Clermont, in Auvergne ; and of the Liorant, in Cantal ;
also the Kasbegk, in the Caucasus ; the Mexican volcanoes
of Tolucaf and Orizaba ; the volcano of Purace, and the
splendid columns of Pisoje, J near Popayan, though whether
the latter are trachytes is very uncertain. The domites of
Leopold von Buch belong likewise to this third division. In
the white fine-grained fundamental mass of the trachytes of
the Puy de Dome are found glassy crystals, which were con-
stantly taken for feldspar, but which are always streaked on
the most distinct cleavage surface, and are oligoclase ; horn-
blende and some mica are also present. Judging from the
volcanic specimens for which the royal collection is indebted
to Herr Mollhausen, the draughtsman and topographist of
Lieutenant Whipple's exploring expedition, the third division,
or that of the dioritic Toluca trachytes, also includes those
of Mount Taylor, between Santa Fe del Nuevo Mexico and
Albuquerque, as well as those of Cieneguilla, on the»westem
slope of the Rocky Mountains, where, according to the able
observations of Jules Marcou, black lava streams overflow the
Jura formation." The same mixture of oligoclase and horn-
blende which I saw in the Azteck highlands, in Anahuac
proper, but not in^the Cordilleras of South America, are
also found far to the west of the Rocky Mountains and of
Zuni, near the Mohave River, a tributary of the Rio Colorado
(see Marcou, Resume of a geological reconnaissance from the

* See the admirable geological map of the district of Schemnitz by
Bergrath, Johann von Peltko, 1852, and the Abhandlungen der k. k.
geolocfischen Reichsanstalt, bd. ii., 1855, abth. i., s. 3. '

t Cosmos, see above, p. 375-6.

X The basaltic columns of Pisoje, the feldspathic part of which has
been analyzed by Francis (Poggend., Annal., bd. Hi., 1841, s. 471), near
the banks of the Cauca, in the plain of Amolanga (not far from the
Pueblos of Sta. Barbara and Marmato), consist of a somewhat modi-
fied oligoclase in large beautiful crystals, and small crystals of horn-
blende. Nearly allied to this mixture are the quartz, containing dio-
ritic porphyry of Marmato, brought home by Degenhardt, the feld-
spathic part of which was named by Abich andesine — the rqck, desti-
tute of quartz, of Cucurusape, near Marmato, in Boussingault's collec-
tion (Charles Ste.-Cl. Deville, Etudes de Lithologie, p. 29); the rock
which I found twelve geographical miles eastward of Chimborazo, be-
low the ruins of old Riobamba (Humboldt, Kleinere Schriften^ bd. i.,
s. IGl) ; and, lastly, the rock of the Esterel Mountains, in the de-
partment of the Var (Elie de Beaumont, Explic. de la Carte Gcol. de
France, t. i., p. 473).

TRUE VOLCANOES. 427

Arkansas to California, July, 1854, p. -^-48. See also two
important French treatises — Besume explicatif dune Carte
Geologique des Etats-Unis, 1855, p. 113-116, and Exquisse
d\me ClassiJicatio7i des Chaines de Montagnes de VAmerique du
Nord, 1855 ; Sierra de S. Francisco et Mount Taylor, p. 23).
Among the trachytes of Java, for specimens of which I am
indebted to my friend Dr. Junghuhn, we have likewise rec-
ognized those of the third division in three volcanic districts ;
namely, Burung-agung, Tyinas, and Gurung Parang (in the
Batugangi district).

Fourth Division. — " The leading mass contains augite with
oligoclase — the Peak of Teneriffe,* the Mexican volcanoes
Popocatepetl! and Colima, the South American volcanoes

* The feldspar in the trachytes of TenerifFe was first recognized in
184:2 by Charles Deville, who visited the Canary Islands in the autumn
of that year ; see that distinguished geologist's Voyage Geologique aux
Antilles et mix lies de Tcncriffe et de Fogo, 1848, p. 14, 74, and 169;
also Analyse du Feldspath de Teneriff'e, in the Comptes rendus de I' Acad,
des Sciences, t. xix., 1844, p. 46. "The labors of Messrs. GustavRose
and H. Abich," he says, " have contributed in no small degree, both crys-
tallographically and chemically, to throw light on the numerous varie-
ties of minerals which were comprised under the vague denomination
of feldspar. I have succeeded in submitting to analysis care/nlli/ iso-
lated crystals whose densitv in different specimens was very uniformly
2-593, 2-594, and 2-586. *This is the first time that the oligoclase
feldspar has been indicated in volcanic regions, with the exception,
perhaps, of some of the great masses of the Cordillera of the Andes.
It was not detected, at least with any certainty, except in the ancient
eruptive rocks (Plutonic, granite, vsyenite, syenitic porphyry ....;)
but in the trachytes of the Peak of TenerifFe it plays a part analogous
to that of the Labrador in the doleritic masses of ^tna." Compare
also Rammelsberg, in the Zeitschr. der Jjeutschen Geol. Gesellschaft, bd.
v., 1853, s. 691, and the 4th Supplement of his Handworterbuchs der
Chem. Minei-alogie, s. 245.

t The first determination of height of the great volcano of Mexico,
Popocatepetl, is, so far as I am aware, the trigonometrical measure-
ment already mentioned (see above, p. 43, note f), executed by me on
the 24th of January, 1804, in the Llano de Tetimba. The summit
was found to be 1536 toiscs above the Llano, and as the latter lies bar-
ometrically 1234 toises above the coast of Vera Cruz, we obtain 2770
toises, or 17,728 English feet, as the absolute height of the volcano.
The barometrical measurements which have succeeded my trigono-
metrical calculation lead me to conjecture that the volcano is still
higher than I have made it in the Essai sur la Geographie des Plantes,
1807, p. 148, and in the Essai Politique sur la Nouvelle Espagne, t. i.,
1825, p. 185. William Glennie, who first reached the margin of the
crater on the 20th of April, 1827, found it, according to his own cal-
culation (Gazeta del Sol, published in Mexico, No. 1432), 17,884 feet,
equal to 2796 toises ; but, as corrected by the mining director, Burkart,
who has acquired so high a reputation in the department of American
hypsometry, and who compared the calculation in Vera Cruz with baro-

428 COSMOS.

Tolima (with the Baramo de Euiz), Purace near Popayan,

metrical observations taken nearly at the same time, it comes out fully
18,017 feet. On the other hand, a barometrical measurement by Sam-
uel Birbeck (10th of Nov., 1827), calculated according to the tables of
Oltmanns, gave only 17,854 feet ; and the measurement of Alex. Doig-
non (Gumprecht, Zeitschrift fiir Allg.Erdkunde, bd. iv., 1855, s. 390),
coinciding almost too precisely with the trigonometrical measurement
of Tetimba, gives 5403 metres, equal to 17,726 feet. The talented
Herr Von Gerolt, the present Prussian embassador in Washington,
accompanied by Baron Gros, likewise visited the summit of Popocate-
petl (28th of May, 1833), and found, by an exact barometrical meas-
urement, the Roca del Fraile, below the crater, 16,896 feet above the
sea. Singularly contrasted with these chronologically-stated hypso-
metrical results appears a carefully conducted barometrical measure-
ment by M. Craveri, published by Petermann in his valuable Mitthei-
lungen iiher wichtige neue Erforschungen der Geographic, 1855 (heft x.),
s. 358-361. The traveler found, in September, 1855, the height of
the highest margin of the crater, the northwest, compared with what
he considered the mean height of the atmospheric pressure in Vera
Cruz, only 5230 metres, or 17,159 feet, which is 555 feet (^^j of the
whole height under measurement) less than I found it by trigonomet-
rical measurement half a century previous. Craveri, likewise, makes
the height of the city of Mexico above the sea 196 feet less than Burk-
art and I have found it to be at very different times ; he reckons it at
only 2217 metres, or 7274 feet, instead of 2277 metres, or 7471 feet.
In Dr. Petermann's periodical above referred to, p. 479-481, I have
explained myself more particularly on the subject of these variations
plus or minus, as compared with the result of my trigonometrical
measurement, Which unfortunately has never been repeated. The 453
determinations of height which I made from September, 1799, to Feb-
ruary, 1804, in Venezuela, on the woody shores of the Orinoco, the
Rio de la Magdalena, and the River Amazon ; in the Cordilleras of
New Granada, Quito, and Peru, and in the tropical region of Mexico,
all of which, recalculated by Professor Oltmanns, uniformly accord-
ing to the formula of Laplace and the coefficients of Ramond, have
been published in my Nivellement Barometrique et Geologique, 1810 {Re-
cueil d'Obsei'V. Astron., t. i., p. 295-334), were performed without
exception with Ramsden's cistern barometers "a niveau constant,"
and not with apparatus in which several fresh-filled Torricellian tubes
may be inserted one after another, nor by the instrument, projected by
myself, described in Lametherie's Journal de Physique, t. iv., p. 468,
and occasionally used in Germany and France during the years 1796
and 1797. Gay-Lussac and I made use, to our mutual satisfaction, of
a portable Ramsden cistern barometer exactly similar in construction,
in the year 1805, during our journey through Italy and Switzerland.
The admirable observations of the Olmutz astronomer, Julius Schmidt,
on the margins of the crater of Vesuvius (Beschreibung der Eruption
im Mai, 1855, s. 114-116) furnish, from their similarity, additional
motives of satisfaction. As I never have ascended the summit of
Popocatepetl, but measured it trigonometrically, there is no foundation
whatever for the extraordinary criticism (Craveri, in Petermann's
Geogr. Mitthdlungen, heft x., s. 359), *' that the height of the mount-
ain as described by me is unsatisfactory, because, as I myself stated,

TRUE VOLCANOES. 429

Paste and Cumbal (according to specimens collected by Bous-

I had made use of fresli-filled Torricellian tubes." The apparatus
with several tubes ought never to be used in the open air, more espe-
cially on the summit of a mountain. It is one of those means which,
from the conveniences furnished by large towns, may be employed at
long intervals, when the operator feels anxious as to the state of his
barometer. For my own part, I have had recourse to it only on very,
rare occasions, but I would nevertheless still recommend it to travel-
ers, accompanied by a comparison with the boiling point, as warmly
as I did in my Observations Astronomiques (vol. i., p. 363-373): "As
it is better not to observe at all than to make bad observations, we
ought to be less afraid of breaking the barometer than of putting it
out of order. M. Bonpland and I having four different times trav-
ersed the Cordilleras of the Andes, the determinations which chief-
ly interested us were repeated at different times, as we returned to
the places which seemed doubtful. We occasionally employed the
apparatus of Mutis, in which Torricelli's primary experiment is per-
formed, by applying successively three or four strongly-heated tubes,
filled with mercury recently boiled in a stone-wai-e crucible. When
there is no possibility of replacing the tubes, it is perhaps prudent not
to boil the mercury in the tubes themselves. In -this way I have found,
in experiments made in conjunction with Lindner, Professor of Chem-
istry at the School of Mines in Mexico, the height of the column of
mercury in six tubes, as follows :

259-7 lines (old Paris foot) 259-9 lines (old Paris foot)
259-5 260-0

259-9 259-9

"The two last tubes alone had, by means of heat, been deprived of air
by Bellardoni, the instrument maker at Mexico. As the exactness of
the experiment depends partly on the perfect cleanlinei(i^of the inside
of the empty tubes, which are so easily carried, it is a good plan to
seal them hermetically over a lamp." As the angles of altitude can
not, in mountainous districts be taken from the sea-§hore, and the
trigonometrical measurements are of a mixed nature and to a consid-
erable extent (frequently as much as i or 1-2 -7 of the whole height)
barometrical, the determination of the height of the elevated plain in
which the base line may be measured is of great importance. As cor-
responding barometrical observations at sea are seldom obtained, or
for the most part only at too great a distance, travelers are too often
induced to take the results they have obtained from a few days' observa-
tions, conducted by them at different seasons of the year, as the mean
height of the pressure of the atmosphere on the elevated plain and at
the sea-shore. " In wishing to know whether a measurement made"
by means of the barometer possesses the exactness of trigonometrical
operations, it is only necessary to ascertain whether, in a given case,
the two kinds of measurement have been taken under equally favor-
able circumstances, that is to say, by fulfilling those conditions which
both theor}' and long experience have prescribed. The mathematical
experimenter dreads the effect of terrestrial refraction, while the phys-
ical experimenter has reason to fear the unequal and far from simul-
taneous distribution of the temperature in the column of air at the
extremities of which the two barometers are placed. It is probable

430 COSMOS.

sjngault), Rucu-Pichincha, Antisana, Cotopaxi, Chimborazo,*

enough that near the surface of the earth the decfease of caloric is
slower than at gi-eater elevations, and in order to ascertain with pre-
cision the mean density of the whole column of air it would be neces-
sary to ascend in a balloon so as to examine the temperature of each
successive stratum or layer of the superimposed air" (Humboldt, Re-
cueil d^ Observ. Astron., vol. i., p. 138 ; see, also, 371, in the appendix
on refraction and barometrical measurements). While the baromet-
rical measurement of MM. Truqui and Craveri gives only 17,159 feet
to the summit of Popocatepetl, whereas Glennie gives 17,889 feet, I
find that the lately-published measurement of Professor Carl Heller,
of Olmutz, who has thoroughly investigated the district surrounding
Mexico, as well as the provinces of Yucatan and Chiapa, corresponds
to within 32 feet of my own. (Compare mj Essay on the Height of the
Mexican Volcano Popocatepetl^ in Dr. Petermann's Mittheilimgen aus
Justus Perthes Geographischer Anstalt, 1856, s. 479-481.)

* In the Chiraborazo rock it is not possible, as in the JEtna rock, to
separate mechanically the feldspathic crystals from the ground mass
in which they lie, but the large proportion of silicic acid which it con-
tains, along with the fact connected therewith of the small specific
gravity of the rock, make it apparent that the feldspathic constituent
is oligoclase. The quantity of silicic acid which a mineral contains
and its specific gravity are generally in an inverse ratio ; in oligoclase
and Labradorite the former is 64 and 53 per cent., while the latter is
2-66 and 2*71. Anorthite, with only 44 per cent, of silicic acid, has
the great specific gravity of 2'76. This inverse proportion between
the quantity of silicic acid and the specific gravity does not occur, as
Gustav Kose remarks, in the feldspathic minerals, which are also iso-
morphous, but with a different crystalline form. Thus feldspar and
leucite, for instance, have the same component parts — potash, alumina,
and silicic acid» The feldspar, however, contains 65, and the leucite
only 56 per ^^t. of silicic acid, yet the former has a higher specific
gravity, namely, 2*56, than the latter, whose specific gravity is only 2-48.

Being desirous, in the spring of 1854, to obtain a fresh analysis of
the trachyte of'Chimborazo, Professor Rammelsberg kindly undertook
the task, and performed it with his usual accuracy. I here give the
results of this analysis, as they were communicated to me by Gustav
Eose, in a letter in the month of June, 1854. He says: "The Chim-
borazo rock, submitted to a careful analysis by Professor Rammels-
berg, was broken from a specimen belonging to your collection, which
you had brought home from the narrow rocky ridge at' a height of
more than 19,000 feef above the sea."

Rammelsberg'' s Analysis.
(Height, 19,194 English feet; specific gravity, 2*806.)

Oxygen.

Silicic acid 59-12 ... 30-70 2-33

Alumina 13-48 ... 6-30]

Protoxyd of iron 7-27 1-61^ ... I ,

Lime 6-50 1-85 ... ' ^

Magnesia 5-41 2-13 V 6-93

Soda 3-46 0-89

Potash 2-64 0-46.

97-88 -

TRUE VOLCANOES. 431

Tunguragua, and trachyte rocks, which are covered by the

AhicK's Analysis.
(Height, 16,179 English feet; specific gravity, 2685.)

^ Oxygen.

Silicic acid 6o^09 .., 33-81 2-68

Alumina 15-58 ... 7-27^

Oxydofiron 3-83 ... 1-16

Protoxyd 1-73 ... 0-39 |

Lime 261 ... 0-73

Magnesia 4*10 ... 158

Soda 4-46 ... 1-14

Potash 1-99 ... 0-33,

Chlorine, and loss by heat... 0-41
99-80

In explanation of tlifese figures it must be observed that the first se-
ries gives the ingredients in a percentage, the second and third give
the oxygen contained in them. The second space shows only the
oxygen of the stronger oxyds (those which contain one atom of oxy-
gen). In the third space this is recapitulated, so as to offer a compar-
ison with that of the alumina earth (which is a weak oxyd) and of the
silicic acid. The fourth space gives the proportion of the oxygen of
the silicic acid to the oxygen of the aggregate bases, which latter are
fixed = l. In the trachyte of Chimborazo this proportion is — 2*33 : 1.
"The differences between the analyses of Rammelsberg and of
Abich are certainly important. Both analyzed minerals from Chim-
borazo, from the relative heights of 19,194 and 16,179 feet, which
wer6 broken off by you, and were taken from 3'our geological collec-
tion in the Royal Mineral Cabinet at Berlin. The mineral from the
lower elevation (scarcely 400 feet higher than the summit of Mont
Blanc), which Abich has analyzed, possesses a smaller specific gravity,
and in correspondence therewith a greater quantity of silicic acid,
than the mineral talien from a point 2918 feet higher, analyzed by
Rammelsberg. Assuming that the argillaceous earth belongs only to
the feldspathic ingredient, we may reckon in the analysis of Rammels..
berg :

Oligoclasc 58-66

Augite 34-14

Silicicacid 4-08

As thus, by the assumption of oligoclase, a portion of silicic acid re-
mains over uncombined, it is probable that the feldspathic ingredient
is oligoclase, and not Labradorite. The latter does not occur Avith un-
combined silicic acid, and if we were to suppose Labradorite in the
rock, a greater quantity of silicic acid would remain over."

A careful comparison of several analyses for which I am indebted to
the friendship of M. Charles Sainte-Claire Deville, to whom the valu-
able geological collections of our mutual friend Boussingault are ac-
cessible for chemical experiment, shows that the quantity of silicic
acid contained in the fundamental mass of the trachytic rocks is gen-
erally greater than in the feldspars which they contain. The table
kindly communicated to me by the compiler himself in the month of
June, 1857, contains only five of the great volcanoes of the chain of
the Andes :

432 COSMOS.

ruins of Old Riobamba. In the Tunguragua, besides the

Names of the
Volcanoes.

Structure and Color of the
Mass. **

Silicic Acid in
the whole Mass.

Chimborazo

Antisana

Cotopaxi

Pichincha
Purace

Guadaloupe
Bourbon

semi-vitrified, brownish gray

semi-vitreous and black

crystalline, compact, gray....
gray-black

vitreous and brownish

granulated

bjack, vitreous

nearly bottle-green ....

65-09 Abich
63-19 Deville
62-66 Deville
64-26 Abich
62-23 Abich
69-28 Abich
63-98 Abich
67-07 Abich
68-80 Deville

58-26
58-26

55-40

gray, granulated, and cellular
crystalline, gray, porous

57-95 Deville
50-90 Deville

"These differences, as far as regards the relative richness in silica
of the ground mass (and the feldspar)," continues Charles Deville,
"will appear still more striking when it is considered that, in analyz-
ing a rock en masse, there are included in the analysis, along with the
basis properly so called, not only/ragments of feldspar similar to those
Avhich have been extracted, but even such minerals as amphibole, pyr-
oxene, and especially peridote, which are less rich in silica than the
feldspar. This excess of silica manifests itself sometimes by the pres-
ence of isolated grains of quartz, which M. Abich has detected in the
trachytes of the Drachenfels (Siebengebirge, near Bonn), and which I
have 'myself observed with some surprise in the trachytic dolerite of
Guadaloupe."

"If," observes Gustav Hose, " we add to this remarkable synopsis of
the silicic acid contained in Chimborazo the result of the latest anal-
ysis, that of Kammelsberg in May, 1854, we shall find that the result
obtained by Deville occupies exactly the mean between those of Abich
and Kammelsberg. Thus :

Clihiiborazo Rock.

Silicic acid 65*09 Abich (specific gravitv, 2-685)

63-19 Deville

62-66 do.

59-12 Kammelsberg (specific gravity, 2-806)."

In ih^Echo du Pacifique, of the 5th of January, 1857, published at
San Francisco, in California, an account is given of a French travel-
er, named M. Jules Remy, having succeeded, on the 3d of November,
1856, in company with an Englishman, Mr. Brencklay, in reaching the
summit of Chimborazo, which was, "however, enveloped in a cloud, so
that we ascended without perceiving it." He observed, it is stated, the
boiling point of water at 171 °'5 F., with the temperature of the air at
31°-9 F. On calculating, upon these data, the height he had attained,
by a hypsom'etrical rule tested by him in repeated journeys in the Ha-
way Archipelago, he was astonished at the result brought out. He
found, in fact, that he was at an elevation of 21,467 feet; that is to

TRUE VOLCANOES. 433

aiigites there occur also separate blackish-green crystals of
uralite, of from half a line to five lines in length, with a per-
fect augite form and the cleavage of hornblende (see Rose,
Reise nach dern Ural, bd. ii., s. 353)." I brought a similar
fragment, with distinct uralite crystals, from the slope of the
Tunguragua at an elevation of 13,260 feet. GustavRose con-
siders this specimen strikingly different from the seven frag-
ments of trachyte from the same volcano which are contained
in my cabinet. It recalls to mind the formation of green-
slate (schistose augitic porphyry) which we have found so*
diffused on the Asiatic side of the Ural {Ibid., s. 544).

Fifth Division, — "A mixture of Labradorite* and au-

say, at a height differing by only 40 feet from that given by my trig-
onometrical measurement at Riobamba Nuevo, in the elevated plaiTi
of Tapia, in June, 1803, as the height of the summit of Chimborazo
— namel}^, 21,426 feet. This correspondence of a trigonometrical
measurement of the summit with one founded on the boiling point is
the more surprising as my trigonometrical measurement, like all
measurements of mountains in the Cordilleras, involves a barometrical
portion ; and from the want of coiTesponding observations on the
shore of the South Sea, my barometrical determination of the height
of the Llano de Tapia, 9484 feet, can not possess all the exactness
that could be desired. (For the details of my trigonometrical meas-
urement, see my Recueil d Observations Astron., vol. i., p. 72 and 74).
Professor Poggendorff kindly undertook to ascertain what result, un-
der the most probable hypotheses, a rational mode of calculation would
produce. He found, reckoning under both hypotheses, that, the pre-
vailing temperature of the atmosphere at the sea being 81''*5 F., or
790.7 Y.^ and the barometer marking 29^922 inches, with the thermom-
eter at the freezing point, the follov/ing result is obtained by Reg-
nault's table: the boiling point at the summit at 171°-5F. answers to
12*677 inches of the barometer at 32'^ temperature ; the temperature
of the air may therefore be taken at 35° '3 F. =34°"7 F. According to
these data, Oltmann's tables give, for the height ascended, under the
first hypothesis (81°-5),=:7328"'-2, or 24,043 feet; and under the sec-
ond (79''.7), = 7314'"-5, or 23,998 English feet, showing an average of
777"^, or 2549 English feet more than my barometrical measurement.
To have corresponded with this, the boiling point should have been
found about 2°'25cent. higher, if the summit of Chimborazo had act-
ually been reached. (Poggendorff s Annakn, bd. c, 1857, s. 479.)

* That the trachytic rocks of jEtna contain Labradorite was demon-
strated by Gustav Rose in 1833, when he exhibited to his friends the
rich Sicilian collections of Friedrich Hoffmann in the Berlin Minera-
logical cabinet. In his treatise on the minerals known by the names
of green-stone and green-stone porphyry (Poggend., AnnaL, bd.
xxxiv., 1835, p. 29), Gustav Rose mentions the lavas of ^tna, which
contain augite and Labradorite (compare Abich, in his interesting
treatise on the whole feldspathic family, Poggend., Annul., 1840, bd.
1., s. 347). Leopold von Buch describes the rock of ^tna as analo-
gous to the dolerite of the basalt formation (Poggend,, Annal, bd.
xxxvii., 1836, s. 188).

Vol. v.— T

434 COSMOS.

gite,* a doleritic trachyte : .^iEtna, Stromboli ; and, according
to the admirable works on the trachytes of the Antilles by
Charles Sainte-Claire Deville, the Soufriere de la Guade-
loupe, as well as the three great cirques which surround the
Pic de Salazu, on Bourbon."

Sixth Division. — " The ground mass, often of a gray color,
in which crystals of leucite and augite lie imbedded, with
very little olivin: Vesuvius and Somma; also the extinct
^volcanoes of Vultur, Rocca Monfina, the Albanian Hills, and
Borghetto. In the older mass (for example, in the wall and
paving stones of Pompeii) the crystals of leucite are more
considerable in size and more numerous than the augite. In
the present lavas, on the contrary, the augites predominate,
and the leucites are, on the whole, very scarce, although the
lava stream of the 22d of April, 1845, has furnished them
in abundance. t Fragments of trachytes of the first division,

* Sartorius von Waltershausen, who has for many years carefully
investigated the trachytes of -^tna, makes the following important
observations: "The hornblende there belongs especially to the older
masses — the green-stone veins in the Val del Bove, as well as the
white and red trachytes, which form the ground mass of vEtna in
the Serra Giannicola. Black hornblende and bright yellowish-green
augite are there found side by side. The more recent lava streams,
from 166i) (especially those of 1787, 1809, 1811, 1819, 1832, 1838, and
1842), show augite, but no hornblende. The latter seems to be gen-
erated only after a longer period of cooling" (Waltershausen, Ueber
die vulkanischen Gesteine von Sicilien und Island, 1853, s. 111-114). In
the augitiferous trachytes of the fourth division, in the chain of the
Andes, along with the abundant augites, I have indeed sometimes
found none, but sometimes, as at Cotopaxi (at an elevation of 14,068
feet) and- at Rucu-Pichincha, at a height of 15,304 feet, distinct black
hornblende crystals in small quantities.

t See Pilla, in the Comptes rendus de VAcad. des Sc, t. xx., 1845,
p. 324. In the leucite crystals of the Rocca Monfina, Pilla has found
the surface covered with worm tubes {serpulm), indicating a submarine
volcanic formation. On the leucite of the Eifel, in the trachyte of
the Burgberg, near Rieden, and that of Albano, Lago Bracciano, and
Borghetto, to the north of Rome, see above, page 224, note *. In the
centre of large crystals of leucite, Leopold von Buch has generally
found the fragment of a crystal of augite, round which the leucite
crystallization has formed, "a circumstance which, considering the
ready fusibility of the augite, and the infusibility of the leucite, is
somewhat singular. More frequently still are fragments of the funda-
mental mass itself inclosed like a nucleus in leucite porphyry." Oli-
vin is likewise found in lavas, as in the cavities of the obsidian which
I brought from the Cerro del Jacal, in Mexico {Cosmos, vol. i., p. 266,
note %), and yet, strange to say, also in the hypersthene rock of Elf-
dal (Berzelius, Sechster Jahresbericht, 1827, s. 302), which was long
considered to be syenite. A similar contrast in the nature of the
places where it is found is exhibited by oligoclase, which occurs in the

TRUE VOLCANOES. 435

containing glassy feldspar (Leopold von Buch's trachyte
proper), are imbedded in the tufas of Monte Somma ; they
also occur detached in the layer of pumice which covers
Pompeii. The leucite ophyr trachytes of the sixth division
must be carefully distinguished from the trachytes of the
first division, although leucites occur in the westernmost
part of the Phlegrasan Fields and on the island of Procida,
as has been already mentioned."

The talented originator of the above classification of vol-
canoes, according to the association of the simple minerals
which they present, does not by any meanfe suppose that he
has completed the grouping of all that are found on the sur-
face of the earth, which is still, on the whole, so very im-
perfectly investigated in a scientifically geological and chem-
ical sense. jVjEodifications in the nomenclature of the asso-
ciated minerals, as well as additions to the trachyte forma-
tions themselves, are to be expected in two ways, both from
the progressive improvement of mineralogy itself (in a more
exact specific distinction both with regard to form and chem-
ical composition), and from the increased number of col-
lections, which are for the most part so incomplete and so
aimless. Here, as in all other cases where the governing
law in cosmical investigations can only be discovered by a
widely-extended comparison of individual cases, we must
proceed on the principle that every thing which, in the pres-
ent condition of science, we think we know is but a small
portion of what the next century will bring to light. The
means of early acquiring this advantage lie in profusion
before us, but the investigation of the trachyte portion of
the dry surface of the earth, whether raised, depressed, or
opened up by fissures, has hitherto been very deficient in the
employment of thoroughly exhaustive methods.

Though similar in form, in the construction of their frame-
work, and their geotectonic relations, volcanoes situated very
near each other have frequently a very different individual

trachytes of still burning volcanoes (the Peak of TenerifFe and Coto-
paxi), and yet at the same time also in the granite and granitite of
Schreibersau and Warmbrunn, in the Silesian Riesengebirge (Gustav
Rose, in the minerals belonging to the granite group, in the Zeit-
schriflen d. Deutsch. geol. Gesellsch., zu Berlin, bd. i., s, 364). This is
not the case with the leucite in the Plutonic rocks, for the statement
that leucite has been found disseminated in the mica-slate and gneiss
of the Pyrenees, near Gavarnie (an assertion which even Hauy has
repeated), has been found erroneous, after many years' investigation,
by Dufrenoy {TraiU de Miniralogie, t. iii., p. 399).

436 COSMOS.

character in regard to the composition and association of
their mineral aggregate. On the great transverse fissure
which, extending from sea to sea almost entirely in a direc-
tion from west to east, intersects a chain of mountains, or,
more properly speaking, an uninterrupted mountainous swell,
running from southeast to northwest, the volcanoes occur in
the following order: Colima (13,003 feet), Jorullo (4265
feet), Toluca (15,168 feet), Popocatepetl (17,726 feet), and
Orizaba (17,884 feet). Those situated nearest to each other
are dissimilar in the composition which characterizes them,
a similarity of tr*achyte occurring only alternately. Colima
and Popocatepetl consist of oligoclase, with augite, and con-
sequently have the trachyte of Chimborazo or Teneriffe;
Toluca and Orizaba consist of oligoclase with hornblende,
and consequently have the rock of ^gina,and Kozelnik.
The recently-formed volcano of Jorullo, which is scarcely
more than a large eruptive hill, consists almost alone of
scoriaceous lavas, resembling basalt and pitch-stone, and
seems more like the trachyte of Toluca than that of Colima.
In these considerations on the individual diversity of the
mineralogical constitution of neighboring volcanoes, we find
a condemnation of the mischievous attempt to introduce a
name for a species of trachyte, derived from a mountain
chain, chiefly volcanic, of more than 7200 geographical miles
in length. The name of Jura limestone, which I was the
first to introduce,* is unobjectionable, because it is taken
from a simple unmixed rock — from a chain of mountains
whose antiquity is characterized by its containing organic

* In the course of a geological tour which I made, in 1795, through
the south of France, western Switzerland, and the north of Italy, I
had satisfied myself that the Jura limestone, which Werner reckoned
among his muschel-kalk, constituted a pecuhar formation. In my
treatise on subterranean gases, pubUshed by my brother, Wilhelm von
Humboldt, in 1799, during my residence in South America, this
formation, which I proA'isionally designated as Jura limestone, was for
the first time mentioned (s. 39). This account of the new formation
was immediately transferred to the Oberbergrath Karsten's mineral-
ogical tables, at that time so generally read (1800, p. 64, and preface,
p. vii.). I named none of the petrifactions which characterize the
Jura formation, and in relation to which Leopold von Buch has ac-
quired so much credit (1839) ; I erred likewise in the age .ascribed by
me to the Jura formation, supposing it to be older than muschel-kalk,
on account of its propinquity to the Alps, which were considered older
than Zechstein. In the earUest tables of Buckland, on the Superpo-
sition of Strata in the British Islands, the Jura limestone of Humboldt^
is reckoned as belonging to the upper oolite. Compare my Essai
G^oyn. sur le Gisement des Bodies, 1823, p. 281.

TRUE VOLCANOES. 437

remains. It would in like manner be unobjectionable to
designate trachyte formations after particular mountains —
to make use of the expression TenerifFe trachyte or iEtna
trachyte for decided oligoclase or Labradorite formations.
So long as there was an inclination among geologists to find
albite every where among the very different kinds of feldspar
which are peculiar to the chain of the Andes, every rock in
which albite was supposed to exist was called andesite. I
first meet with the name of this mineral, with the distinct
definition that "andesite is composed of a preponderating
quantity of albite and a small quantity of hornblende," in
the important treatise written in the beginning of the year
1835, by my friend Leopold von Buch, on " Craters of up-
heaval and volcanoes.''* This tendency to find albite every

* The name of andesite first occurs in print in Leopold von Buch's
treatise, read on the 26th March, 1835, at the BerHn Academy. That
great geologist Hmits the appellation of trachyte to those cases in
which glassy feldspar is contained, and thus speaks in the above
treatise, which was not printed till 1836 (Poggend., Annal., bd.
xxxvii., s. 188-190): "The discoveries of Gustav Rose relating to
feldspar have shed a new light on volcanoes and geology in general,
and the minerals of volcanoes have in consequence presented a new
and totally unexpected aspect. After many careful investigations in
the neighborhood of Catanea and at -^tna, Elie de Beaumont and I
have convinced ourselves that feldspar is not to be met with on ^tna,
and consequently there is no trachyte either. All the lava streams,
as well as all the strata in the interior of the mountain, consist of a
mixture of augite and Labradorite. Another important difference in
the minerals of volcanoes is manifested when albite takes the place of
feldspar^ in which case a new mineral is formed, which can no longer
be denominated trachyte. According to G. Rose's (present) investi-
gations, it may be considered tolerably certain that not one of the al-
most innumerable volcanoes of the Andes consists of trachyte, but
that they all contain albite in their constituent mass. This conjecture
seems a very bold one, but it loses that appearance when we consider
that we have become acquainted through Humboldt's journeys alone^
with one half of these volcanoes and their products in both hemi-
spheres. Through Meyen we are acquainted with these albitiferous
minerals in Bolivia and the northern part of Chili ; through Poppig,
as far as the southernmost limit of the same country ; through Erman,
in the volcanoes of Kamtschatka. Their presence being so widely
diffused and so distinctly marked, seems sufficiently to justify the
name of andesite, under which this mineral, composed of a prepon-
derance of albite and a small quantity of hornblende, has already been
sometimes noticed." Almost at the same time that this appeared,
Leopold von Buch enters more into the detail of the subject in the
addenda with which, in 1836, he so greatly enriched the French edi-
tion of his work on the Canary Islands. The volcanoes Pichincha,
Cotopaxi, Tungurahua, and Chimborazo, are all said to consist of an-
desite, while the Mexican volcanoes were called genuine (sanidinifer-

438 COSMOS.

where lasted for five or six years, until renewed investiga-

ous) trachytes {Description physique des Iks Canaries, 1836, p. 486,
487, 490, and 515). This lithological classification of the volcanoes
of the Andes and those of Mexico shows that, in a scientific point of
view, such a similarity of mineralogical constitution and the possibili-
ty of a general denomination ilerived from a large extent of country,
can not be thought of. A year later, when Leopold von Buch first
made mention in PoggendoriF's Annalen, of the name of andesite,
which has been the occasion of so much confusion, I committed the
mistake myself of making use of it on two occasions — once in 1836,
in the account of my attempt to ascend Chimborazo, in Schumacher's
Jahrbuch, 1837, s. 204, 205 (reprinted in my Kkinere Schriften, bd. i.,
s. 160, 161); and again in 1837, in the treatise on the highland of
Quito (in Poggend., Ann., bd. xl., s. 165). "Recent times have
taught us," I observed, already strongly opposing my friend's con-
jecture as to the similar constitution of all the Andes volcanoes,
"that the different zones do not always present the same (mineral-
ogical) composition, or the same component parts. Sometimes we
find trachytes, properly so called, characterized by the glassy feldspar,
as at the Peak of Teneriffe and in the Siebengebirge near Bonn, where
a little albite is associated with the feldspar — feldspathic trachytes,
which, as active volcanoes, exhibit abundance of obsidian and pumice ;
sometimes melaphyre,and doleritic mixtures of Labradorite andaugite,
more nearly resembling the basalt formation, as at .^tna, Stromboli,
and Chimborazo ; sometimes albite with hornblende prevails, as in the
lately so-called andesites of Chili, and the splendid columns, described
as dioritic porphyry, at Pisoje, near Popayan, at the foot of the vol-
cano of Purace, or in the Mexican volcano of JoruUo ; finally, they
are sometimes leucite ophyrs, a mixture of leucite and augite, as in
the Somma, the ancient wall at the crater of elevation of Vesuvius."
By an accidental misinterpretation of this passage, which shews many
traces of the then imperfect state of geological knowledge (feldspar
being still ascribed to the Peak of Teneriffe instead of oligoclase,
Labradorite to Chimborazo, and albite to the volcano of Toluca), that
talented investigator Abich, who is both a chemist and a geologist, has
erroneously attributed to myself the invention of the term andesite as
applied to a trachytic, widely-dispersed rock rich in albite (Poggend.,
Ann., bd. li., 1840, s. 523), and has given the name of andesine to a
new species of feldspar, first analyzed by him, but still somewhat enig-
matical in its nature, " with reference to the mineral (from Marma-
to, near Popayan) in which it was first observed." The andesine
(pseudo-albite in andesite) is supposed to occupy a middle position
between Labradorite and oligoclase; at the temperature of 55°'7 its
specific gravity is 2*733, while that of the andesite in which the ande-
sine occurred is 3*593. Gustav Rose doubts, as did subsequently
Charles Deville (Etudes de Liihologie, p. 30), the individuality of
andesine, as it rests only on a single analysis of Abich, and because
the analysis of the feldspathic ingredient in the beautiful dioritic por-
phyry of Pisoje, near Popayan, brought by me from South America,
■which was performed by Francis (Poggend., bd. lii., 1841, s. 472) in
the laboratory of Heinrich Rose, while it certainly shows a great re-
semblance to the andesine of Marmato, as analyzed by Abich, is, not-
withstanding, of a different composition. Still more uncertain is the

TRUE VOLCANOES. 439

tions of a more profound and less prejudiced character led to
the recognition of the trachytic albites as oligoclase.* Gus-

andesine in the syenite of the Vosges (from the Ballon de Servance,
and Coravillers, which Delesse has analyzed). Compare G. Rose, in
the already often-cited Zeitschift der Deutschen geologischen Gesell-
schaft, bd. i., for the year 1849, s. 369. It is not unimportant to re-
mark here that the name andesine, introduced by Abich as that of a
simple mineral, appears for the first time in his valuable treatise, en-
titled Beitrag zur Zemitniss des Feldspaths (in Poggend., Ann., bd. I.,
s. 125, 341; bd. li., s. 519), in the year 1840, which is at least five
years after the adoption of the name andesite, instead of being prior
to the designation of the mineral from which it is taken, as has been
sometimes erroneously supposed. In the formations of Chili, which
Darwin so frequently calls andesitic granite and andesitic porphyry,
rich in albite {Geological Observations on South America, 1846, p. 174),
oligoclase may also very likely be obtained. Gustav Rose, whose
treatise on the nomenclature of the minerals allied to green-stone and
gi*een-stone porphyry (in Poggendorflf's Ann., bd. xxxiv., s, 1-30) ap-
peared in the same year, 1835, in which Leopold von Buch employed
the name of andesite, has not, either in the treatise just mentioned or
in any later work, made use of this term, the true definition of which
is, not albite with hornblende, but in the Cordilleras of South Amer-
ica, oligoclase with augite. The now obsolete account of the desig-
nation of andesite, of which I have perhaps treated too circumstan-
tially, helps to show, like many other examples in the history of the
development of our physical knowledge, that erroneous or insufficient-
ly grounded conjectures (as, for instance, tlie tendency to enumerate
varieties as species) frequently turn out advantageous to science, by
inducing more exact observations.

* So early as 1840, Abich described oligoclase trachyte from the
summit rock of the Kasbegk and a part of the Ararat ( Ueber die Natur
tind die Zusammensetzung der VuUcan-BUdungen, s. 46), and even in
1835 Gustav Rose had the foresight to say that though *'he had not
hitherto in his definitions taken notice of oligoclase and pcricline, yet
that they probably also occur as ingredients of admixture." The be-
lief formerly so generally entertained, that a decided preponderance
of augite or of hornblende might be taken to denote a distinct species
of the feldspar family, such as glassy orthoclase (sanidine), Labradorite,
or oligoclase, appears to be very much shaken by a comparison of the
trachytes of the Chimborazo and Toluca rocks, belonging to the fourth
and third division. In the basalt formation hornblende and augite
often occur in equal abundance, which is by no means the case in the
trachytes ; but I have met with augite crystals quite isolated in Toluca
rock, and a few hornblende crystals in portions of the Chimborazo,
Pichincha, Purace, and Teneriffe rocks. Olivins, which are so very
rarely absent in the basalts, are as great a rarity in trachytes as they
are in phonolites ; yet we sometimes find in certain lava streams oli-
vins formed in great abundance by the side of augites. Mica is, on
the whole, very unusual in basalt, and yet some of the basaltic sum-
mits of the Bohemian central mountains, first described by Reuss,
Freiesleben, and myself, contain plenty of it. The unusual isolation
of certain mineral bodies, and the causes of their legitimate specific
association, probably depend on many still undiscovered causes of

440 COSMOS.

tav Rose has come to the general conclusion that it is very
doubtful whether albite occurs at all among the minerals as
a real and essential element of commixture ; consequently,
according to the old conception of andesite, this mineral
would actually be wanting in the chain of the Andes.

The mineralogical condition of the trachytes is imperfectly
recognized if the porphyritically inclosed crystals can not be
separately examined and measured, in which case the inves-
tigator must have recourse to the numerical proportions of
the earths, alkalies, and metallic oxyds which the result of
the analysis furnishes, as well as to the specific gravity of the
seemingly amorphous mass to be analyzed. The result is
obtained in a more convincing and more certain manner if
the principal mass, as well as the chief elements of the mix-
ture, can be singly investigated both mineralogically and
chemically. This is the case with the trachytes of the Peak
of Teneriffe and those of ^tna. The supposition that the
principal mass consists of the same small, inseparable com-
ponent parts Avhich we recognize in the large crystals appears
to be by no means well grounded, for, as we have already
noticed, as shown in Charles Deville's work, the apparently
amorphous principal ijiass generally furnishes more silicic
acid than would be expected from the nature of the feldspar
and the other visible commixed elements. Among the leu-
cite ophyrs, as Gustav Kose observes, a striking contrast is
exhibited, even in the specific difference of the prevailing
alkalies (of the potash containing interspersed leucites) and
the almost exclusively natroniferous principal mass.*

But along with these associations of augite with oligoclase,
augite with Labradorite, and hornblende with oligoclase,

pressure, temperature, fluidity, and rapidity in cooling. The specific
differences of the association are, however, of great importance, both
in the mixed rocks and in the masses of mineral veins ; and in geo-
logical descriptions, noted down in the open air, in sight of the object
described, the observer should be careful not to make any mistake as
to what may be a prevailing, or at least a rarely absent member of
the association, and what may be sparingly or only accidentally com-
bined. The diversity which prevails in the elements of a mixture —
for instance, in the trachytes — is repeated, as I have already noticed,
in the rocks themselves. In both continents there exist large tracts
of country in which trachyte formations and basalt formations, as it
were, repel each other, as basalts and phonolites ; and there are other
countries in which trachytes and basalts alternate with each other in
tolerably close proximity (see Gustav Jenzsch, Monogi-aphie der boh-
miscken Phonolithe, 1856, s. 1-7).

* See Bischof, Chemische und Physikalische Geologie, bd. ii., 1851,
s. 2288, 2297; Roth, Monograpliie des Vesuvs, 1857, s. 305.

TRUE VOLCANOES. 441

which are referred to in our classification of the trachytes,
and which especially characterize them, there exist likewise
in each volcano other easily recognizable, unessential ele-
ments of commixture, whose presence in large quantities or
total absence in different volcanoes, often situated very near
to each other, is very striking. Their occurrence, either in
frequent abundance, or else at long and separate intervals,
depends probably, in one and the same natural laboratory, on
various conditions of the depth from which the matter origin-
ally came, the temperature,' the pressure, the fluidity, or the
quicker or slower process of cooling. The fact of the specific
occurrence or the absence of certain ingredients is opposed to
certain theories, such as the derivation of pumice from glassy
feldspar or from obsidian. These views, which have not been
altogether lately adopted, but originated as early as the end
of the 18th century from a comparison of the trachytes of
Plungary and of Teneriife, engaged my attention for several
years in Mexico and the Cordilleras, as my journals will
testify. From the great advancement which lithology has
undeniably made in modern times, the more imperfect defini-
tions of the mineral species made by me during my journey
have, through Gustav Rose's careful mineralogical elabora-
tion of my collections, been improved and accurately certified.

• Mica.

Black or dark-green magnesian mica is very abundant in
the trachytes of the Cotopaxi, at an elevation of 14,470 feet
between Suniguaicu and Quelendana, as also in the subterra-
nean pumice-beds of Guapulo and Zumbalica at the foot of
Cotopaxi,* but sixteen miles distant from the same. The
trachytes of the volcano of Toluca are likewise rich in mag-
Ncsian mica, which is wanting in the Chimborazo.f In the
Continent of Europe micas have shown themselves in abund-
ance: at Vesuvius (for example, in the eruptions of 1821—
1823, according to Monticelli and Covelli); in the Eifel, in
the old volcanic bombs of the Lacher Lake ;f in the basalt

* Cosmos, see above, p. 323.

t It is almost superfluous to mention that the term wanting signifies
only that, in the investigation«of a not inconsiderable portion of volca-
noes of large extent, a particular sort of mineral has hitherto been
vainly sought for. 1 wish to distinguish between what is wanting (not
being found), being of very rare admixture, and what, though more
abundant, is still not normally characteristic.

X Carl von Oeynhausen, ErU. der geogn. Karte des Lacher Sees^
1847, s. 38.

T2

442 COSMOS.

of the Meronitz, of the marly Kausawer Mountain, and espe-
cially of the Gamayer summit* of the central Bohemian
chain ; more rarely in the phonolite,| as well as in the dole-
rite of the Kaiserstuhl near Freiburg. It is remarkable that
in the trachytes and lavas of both continents not only no
white (chiefly bi-axal) potash mica is observable, but that it
is entirely dark-colored (chiefly uni-axal) magnesian mica,
and that this exceptional occurrence of the magnesia mica is
extended to many other rocks of eruption and Plutonic rocks,
such as basalt, phonolite, syenite, syenitic slate, and even
granitite, while the granite proper contains at one and the
same time white alkaline mica and black or brown magnesia
mica4

Glassy Feldspar.

This kind of feldspar, which plays so important a part in
the action of European volcanoes, in the trachytes of the first
and second division (for example, on Ischia, in the Phlegraean
Fields, or the Siebengebirge near Bonn), is probably entirely
wanting in the New Continent, in the trachytes of active vol-
canoes. This circumstance is the more striking, as sanidine
(glassy feldspar) belongs essentially to the argentiferous, non-
quartzose Mexican porphyries of Moran, Pachuca, Villalpan-
do, and Acaquisotla, the first of which are connected with
the obsidians of Jacal.§

* See the Bergmannisches Journal, von Kohler imd Hofmann, 5ter
Jahrgang, bd. i., 1792, s. 244, 251, 265. Basalt rich in mica, as on
the Gamayer summit in the Bohemian centre mountains, is a rarity.
I visited tliis part of the Bohemian central range in the summer of
1792, in company with Carl Freieslehen, afterward my companion in
my Swiss tour, who has exercised so great an influence over my geo-
logical and mining education. Bischof doubts all production of mica
by the igneous method, and considers it a metamorpbic product by the
moist method. See his Lehrhuch der Chem. und Physikal. Geologic,
bd. ii., s. 1426, 1439.

t Jenzsch, Beitrage zur Kenntniss der PhonoUthe, in der Zeitschrift
der Deutschen Geologischen Gesellsckqft, bd. viii., 1856, s. 36.

X Gustav Rose, Ueber die zur Granitgruppe gehorigen Gebirgsarten, in
derselben Zeitschrift, bd. i., 1849, s. 359.

§ The porphyries of Moran, Real del Monte and Regla (the latter
celebrated for the rich silver mines of the Veta Biscayna, and the vi-
cinity of the obsidians and pearl-stones ©f the Cerro del Jacal and the
Messerberg, Cerro de las Navajas), like almost all the metalliferous
porphyries of America, are quite destitute of quartz (on these and oth-
er analogous phenomena in Hungary, see Humboldt, PJssai Gcognos-
tique sur le Gisement des Roches, p. 179-188, and 190-193). The por-
phyries of Acaquisotla, however, on the road from Acapulco to Chil-
panzingo, as well as those of Villalpando to the north of Guanaxuato,

true volcanoes. 443

Hornblende and Augite.

In this account of the characteristics of six different divi-
sions of the trachytes, it has been already observed how the
same minerals which occur as essential elements of commix-
ture (for example, hornblende in the third division, or the
Toluca rock) appear in other divisions in a separate or spo-
radic condition (as in the fourth and fifth divisions, in the
rock of Pichincha and of uSCtna). I have found hornblende,
though not in large quantities, in the trachytes of the volca-
noes of Cotopaxi, Rucu-Pichincha, Tungurahua, and Anti-
sana, along with augite and oligoclase, but scarcely ever
along with these two minerals on the slope of the Chimbo-
razo up to a height of more than .19,000 feet. Among the
many specimens which I brought from Chimborazo, horn-
blende is recognized only in two, and even then in small
quantity. In the eruptions of Vesuvius in the years 1822

which are penetrated by auriferous veins, along with the sanidine con-
tain also grains of brownish quartz. The small inclosures of grains of
obsidian and glassy feldspar being, on the whole, rare in the volcanic
rocks at the Cerro de las Navajas, and in the Valle de Santiago, so
rich in basalt and pearl-stone, which is traversed in going from Valla-
dolid to the volcano of Jorullo, I was the more astonished at finding
at Capula and Pazcuaro, and especially near Yurisapundaro, all the
ant-hills filled with beautifully shining grains of obsidian and sanidine.
This was in the month of -September, 1803 {Nwellement Barometr.,
p. 327, No. 366, and Essai Gcognostique sur le Gisement des Roches,
p. 356). I was amazed that such small insects should be able to drag
the minerals to such a distance. It has given me great pleasure to find
that an active investigator, M. Jules Marcou, has observed something
exactly similar. " There exists," he says, " on the high plateaux of
the Rocky Mountains, and particularly in the neighborhood of Fort
Defiance (to the west of Mount Taylor), a species of ant which, instead
of using fragments of wood and vegetable remains for the purpose of
building its dwelling, employs only small stones of the size of a grain
of maize. Its instinct leads it to select the most brilliant fragments
of stones, and thus the ant-hill is frequently filled with magnificent
transparent garnets and very pure grains of quartz." (Jules Marcou,
Resume expUcat'ifd''une Carte Geogn. des Etats Unis, 1855, p. 3.)

Glassy feldspar is very rare in the present lavas of Vesuvius, but this
is not the case in the old lavas ; for instance, in those of the eruption
of 1631, where it occurs along with crystals of leucite. Sanidine is
also found in abundance in the Arso lava stream, from Cremate toward
Ischia, of the year 1301, without any leucite ; but this must not be
confounded with the older stream, described by Strabo, near Montag-
none and Rotaro ( Cosmos, see above, p. 252, 399). Glassy feldspar is
not only rare in the trachytes of Cotopaxi and other volcanoes of the
Cordilleras generally, but it is equally so in the subterranean pumice
quan-ies at the foot of the Cotopaxi. What was formerly described as
sanidine are crystals of oligoclase.

444 COSMOS.

and 1850, augite and crystals of hornblende (these nearly
nine Parisian lines in length) were contemporaneously formed
by exhalations of vapors on fissures.* .The hornblende of
JEtna, as Sartorius von Waltershausen observes, belongs es-
pecially to the older lavas. That remarkable mineral, so
widely diffused in Western Asia and at several points of
Europe, which Gustav Rose has denominated Uralite, being
allied in structure and crystalline form to hornblende and
augite,t I here once more gladly point attention to the first
occurrence of uralite crystals in the New Continent ; they
were recognized by Rose in a piece of trachyte which I ab-
stracted from the slope of the Tungurahua, 3200 feet below
the summit.

ILeucites.

Leucites, which in Europe belong exclusively to Vesuvius,
the Rocca Monfina, the Albanian Mountains near Rome, the
Kaiserstuhl in the Breisgau, and the Eifel (in the western en-
virons of the Lachar Lake in blocks, and not in the con-
tiguous rock, as in the Burgberg near Rieden), have never
yet been found in volcanic rocks of the New Continent, or
the Asiatic portion of the Old. Leopold von Buch discov-
ered them round an augite crystal as early as the year 1798,
and described in an admirable treatise their frequent forma-
tion.{ The augite crystal, round •which, according to this
great geologist, the leucite is formed, is seldom wanting, but
appears to me to be sometimes replaced by a small grain or
morsel of trachyte. The unequal degrees of fusibility be-
tween the grain of trachyte and the surrounding mass of leu-
cite raise some chemical difficulties to the explanation of
the mode in which the integumental covering is formed.
Leucites, partly detached, according to Scacchi, and partly
mixed with lava, were extremely abundant in the recent
eruptions of Vesuvius in 1822, 1828, 1832, 1845, and 1847.

Omvin.
Olivin being very abundant in the old lavas (3f Vesuvius §

* Roth, Monographie des Vesuvs, s. 267, 382.

t See above, p. 434, note * ; Rose, Heise nach dem Ural., bd. ii., s.
369; Bischof, Chem. und Physik. Geologk, bd. ii., s. 528-571.

X Gilbert's AnnaJen der Physik., bd. vi., 1800, s. 53 ; Bischof, Geolo-
gic, bd. il., s. 2265-2303.

§ The recent lavas of Vesuvius contain neither olivin nor glassy feld-
spar; Roth, Mon. des Vesuvs., s. 139. According to Leopold von Buch,
the lava stream of the Peak of Teneriffe of 1704, described by Viera and

TRUE VOLCANOES.

^*^oftLibq^

(especially in the leucite ophyrs of the Somma), in the Arso
of Ischia, in the eruption of 1301, mixed with glassy feld-
spar, brown mica, green augite, and magnetic iron, in the
volcanoes of the Eifel, which emit lava streams (for example,
in the Mosenberge, westward of Manderscheid),* and in the
southeastern portion of TenerifFe, in the lava eruption of
Guimar in the year 1704, I have also searched for it very
dihgently, but in vain, in the trachytes of the volcanoes of
Mexico, New Granada, and Quito. Our Berlin collections
contain sixty-eight specimens of trachyte of the four volca-
noes, Tungurahua, Antisana, Chimborazo, and Pichincha
alone, forty-eight of which were contributed by me and twen-
ty by Boussingault.f In the basalt formations of the New
World olivin, along with augite, is as abundant as in Europe ;
but the black, basaltic trachyte of Yana Urcu, near Calpi, at
the foot of the Chimborazo, J as well as those enigmatical tra-

Glas, is the only one which contains olivin (Descr. des Iks Canaries,
p. 207). The supposition that the eruption of 1704 was the first which
had taken place since the conquest of the Canary Islands, at the
end of the 15th century, has been shown by me in another place (£x~
amen Critique de VHistoire de la Gcographie, t. iii., p. 143-146) to be
erroneous. Columbus saw the eruption of fire on TenerifFe, at the
time of his first voyage of discovery, on the nights of the 21st to the
25th of August, when he went in search of Dona Beatriz de Bobadilla,
of the Gran Canaria. It is thus noticed in the admiral's journal, un-
der the Rubric of " Jueves^ 9 de Agosto," which contains notices np to
the 2d of September — " Vieron salir gran fuego de la Sierra de la Isla
de Tenerife, que es muy alta en gran manera" — "they saw a great
deal of fire rising with a grand appearance out of the mountain of the
island of Teneriffe, which is very high ;" Navarrete, Col. de los Viages
de los Espanoles, t. i., p. 5. The lady above named must not be con-
founded with Dona Beatriz Henriquez of Cordova — the mother of his
illegitimate son, the learned Don Fernando Colon, the historian of his
father — whose pregnancy in the year 1488 so materially contributed to
detain Columbus in Spain, and to lead to the discovery of the New
World beiijg made on account of Castile and Leon, and not for
Portugal, France, or England (see my Examen Critique, t, iii., p. 350,
and 367). * Cosmos, see above, p. 222.

t A considerable portion of the minerals collected during my Ameri-
can expedition has been sent to the Spanish Mineral Cabinet, ^to the
King of Etruria, to England, and to France. I do not refer to the geo-
logical and botanical collections which my worthy friend and fellow-
laborer, Bonpland, possesses, with the two-fold right of self-collection
and self-discovery. This extensive dispersion of the material (which,
from the very exact account given of the places in which they origin-
ated, does not prevent the maintenance of the groups in their geograph-
ical relations) has this advantage, that it facilitates the most compre-
hensive and exact definition of those minerals whose substantial and
habitual association characterizes the different kinds of rocks.

X Humboldtj Kleinere Schrijlen, bd. i., s. 139.

446 COSMOS.

chytes called La reventazon del Volcan de Anzango* contain
no olivin. It was only in the great brown-black lava stream,
with a crisp, scoriaceous surface raised like a cauliflower,
whose track we followed in order to reach the crater of the
volcano of JoruUo, that we met with small grains of olivin
imbedded.! The prevailing scarcity of olivin in the modern
lavas and the greater part of the trachytes seem less striking
when we recollect that, essential as olivin appears to be for
basalt in general, yet (according to Krug von Nidda and Sar-
torius von Waltershausen) in Iceland and in the German
Rhone Mountains the basalt destitute of olivin is not dis-
tinguishable from that which abounds in it. The former it
has been the custom from the earliest times to call trap and
wacke, the latter we have in modern times denominated Ane-
masite-X Olivins, which sometimes occur as large as a man's
head in the basalts of Rentieres, in the Auvergne, attain
also in the Unkler quarries, which were the object of my
first youthful researches, to the size of six inches in diameter.
The beautiful hypersthene rock of Elfdalen, in Sweden, much
employed for ornamental purposes, § a granulated mixture
of hypersthene and Labradorite, which Berzelius has described
as syenite, likewise contains olivin, || as does also (though
more rarely) the phonolite of the Pic de Griou, in the Can-
tal.^ While, according to Stromeyer, nickel is a very con-
stant accompaniment of olivin, Rumler has, on the other hand,
discovered arsenic in it,** a metal which has been found in
the most recent times widely diffused in so many mineral

* Humboldt, Kleiner eSchrif ten, s. 202; and Cbs772os,'see above, p. 222.

t Humboldt, Kl. ScJir., vol. i., p. 344. I have also found a great
deal of olivin in the tezontle (cellular lava, or basaltic amygdaloid? — in
Mexican, tetzontli, i. e., stone-hair, from tetl, stone, and tzontli, hair)
belonging to the Cerro de Axusco, in Mexico.

X Sartorius von Waltershausen, Phi/sisch-geograpkisch&, JSkizze von
Island, s. 64.

[§ It is there cut into vases, sometimes of a considerable size, and
other ornamental objects. From the high polish it takes, and the
contrast of its colors, it is one of the most beautiful stones in exist-
ence.— Tr.]

II Berzelius, Sechster Jahreshencht, 1827, p. 392 ; Gustav Kose, in
Poggend., Ann., vol. xxxiv., 1835, p. 14.

f Jenzsch, Phonolithe, 1856, p. 37 ; and Senft, in his important work,
Classification der Felsarten, 1857, p. 187. According to Scacchi, olivin
occurs also, along with mica and augite, in the lime blocks of the Som-
ma. I call these remai'kable masses erupted blocks, not lavas, for the
Somma appears never to have ejected the latter.

** Poggend., Annal, bd. xlix., 1840, s. 591, and bd. Ixxxiv., s. 302;
Daubr^e, in the Annales des Mines, 4me Serie, t. xix., 1851, p. 669.

TRUE VOLCANOES. 447

springs and even in sea-water. The occurrence of olivin in
meteoric stones* and in artificial scoriae, as investigated by
Seifstrom,t I have already mentioned.

Obsidian.

As eai'ly as in the spring and summer of 1799, while I
was preparing in Spain for my voyagS to the Canary Isles,
there prevailed generally among the mineralogists in Madrid
— Hergen, Don Jose Clavijo, and others — the opinion that
pumice was entirely derived from obsidian. This opinion
had been founded on the study of some line geological collec-
tions from the Peak of TeneriiFe, and a comparison of them
with the phenomena which Hungary furnishes, although the
latter were at that time explained chiefly in accordance with
the Neptunian views of the Freiberg school. Doubts of the
correctness of this theory of formation, awakened at an early-
period in my mind by my observations in the Canary Isles,
the Cordilleras of Quito, and in the range of Mexican volca-
noes, J impelled me to direct my most earnest attention to
two groups of facts : first, the different nature of the inclosures
of obsidians and pumice in general ; and, secondly, the fre-
quency of the association or entire separation of them in well
investigated active volcanic structures. My journals are filled*
with notices on this subject, and the specific definition of the
imbedded minerals has been ascertained by the most varied
and most recent investigations of my ever-ready and obliging
friend, Gustav Rose.

Both glassy feldspar and oligoclase occur in obsidian as
well as in pumice, and frequently both of them together. As
examples may be cited — the Mexican obsidians of the Cerro
de las Navajas, on the eastern slope of the Jacal, collected by
me — those of Chico, with many crystals of mica — those of
Zimapan, to the S.S.W. of the capital of Mexico, mixed with
small distinct crystals of quartz, and the pumice of the Rio
Mayo (on the naountain road from Fopayan to Pasto), as
well as those of the extinct volcano of Sorata. near Popayan.
The subterranean pumice quarries near Lactacunga§ contain
a large quantity of mica, oligoclase, and (which is very rare
in pumice and obsidian) hornblende also*; the latter, how-
ever, is also found in the pumice of the volcano of Arequipa,

* Cosmos, vol. i., p. 131, and vol. iv., p. 225.

t Ibid., vol. i., p. 267, note *.

i Humboldt, Personal Narrative, vol. i., p. 113 (Bohn's edition).

§ See above, p. 322.

448 COSMOS.

Common feldspar (orthoclase) never occurs in pumice along
with sanidine, nor is augite ever present. The Somma, not
the cone of Vesuvius itself, contains pumice, inclosing earthy
masses of carbonate of lime. It is by this remarkable vari-
ety of a calcareous pumice that Pompeii was overwhelmed.*
Obsidians are rare in genuine lava-like streams ; they belong
almost solely to the Peak of Teneriffe, Lipari, and Volcano.

Passing now t;o the association of obsidian and pumice in
one and the same volcano, the following facts appear. Pi-
chincha possesses large pumice fields, and no obsidian. Chim-
borazo, like ^tna, whose trachytes, however, have a totally
different composition (containing Labradorite instead of oligo-
clase), shows neither obsidian nor pumice ; this same defi-
ciency I observed on my ascent of the Tungurahua. The
volcano Purace, near Popayan, has a great deal of obsidian
mixed in its trachytes, but has never yielded any pumice.
The immense plains out of which rise the Ilinissa, Carguai-
razo, and Altar are covered with pumice. The subterra-
nean pumice quarries near Lactacunga, as well as those of
Huichapa, southeast of Queretaro ; and the accumulations
of pumice at the Rio Mayo,t those near Tschegem in the
Caucasus,^ and near Tollo § in Chili, at a distance from act-
ive volcanic structures, appear to me to belong to the phe-
nomena of eruption from the numerous fissures in the level
surface of the earth. Another Chilian volcano, that of An-
tuco II (of which Poppig has given a description as scientific-
ally important as it is agreeably written), produces, like Ve-
suvius, ashes, triturated rapilli (sand), but gives out no pum-
ice, no vitrified or obsidian-like mineral. Without the pres-
ence of either obsidian or glassy feldspar, Ave sometimes meet
with pumice in trachytes of very dissimilar composition, al-
though in many cases it is not present. Pumice, as Charles
Darwin observes, is entirely wanting in those of the Archi-

* Scacchi, Osservazioni criticJie sulla maniera comefu sepdlita Vantica
Pompei, 1843, p. 10, in opposition to the theory proposed by Carmine
Lippi, and afterward shared by Tondi, Tenore, Pilla, and Dufrenoy,
that Pompeii and Jlerculaneum were not overwhelmed by rapilli and
ashes direct from the Somma, but that they were conveyed there by
water. Eoth, Mono^. des Vesuvs, 1857, s. 458 ; see above, p. 401.

t Nivellement Barovietrique, in Humboldt, Ohservat. Astron., vol. i.,
p. 305, No. 149. t See above, p. 324.

§ For an account of the pumice hill of Tollo, at a distance of two
days' journey from the active volcano of Maypu, which has itself never
ejected a fragment of such pumice, see Meyen, Beise um die £rde, th.
i., s. 338 and 358.

II Poppig, Rdse in Chile und Peru, bd. i., s. 42 G.

TRUE VOLCANOES. 449

pelago of the Gallapagos. We have already remarked in
another place that cones of cinder^ are wanting in the mightj
volcano of Maiina Loa, in the Sandwich Islands, as well as
in the volcanoes of the Eifel,* which once emitted lava
streams. Though the island of Java contains a series of
more than forty volcanoes, of which as many as twenty-three
are still active, yet Junghuhn was only able to discover two
points in the volcano of Gunung Guntur, near Bandong and
the great Tengger Mountains,! in which masses of obsidian
Lave been formed. These do not appear to have given occa-
sion to the formation of pumice. The sand lakes of Dasar,
which lie about 6828 feet above the mean level of the sea,
are not covered with pumice, but with a layer of rapilli, de-
scribed as being obsidian-like, semi-vitrified fragments of ba-
salt. The cone of Vesuvius, which never emits pumice, gave
out, from the 24th to the 28th of October, 1822, a layer
eighteen inches thick of sand-like ashes, consisting of pulver-
ized trachytic rapilli, which has never been mistaken for
pumice.

The cavities and air-holes of obsidian, in which crystals of
olivin, probably precipitated from vapors, have formed — as,
for example, in the Mexican Cerro del Jacal — are sometimes
found, in both hemispheres, to contain another kind of in-
closures, which seem to indicate the manner of their origin
and formation. In the wider portions of these long-extended,
and for the most part very regularly parallel cavities, frag-
ments of half-decomposed earthy trachyte are found imbeds
ded. Beyond these the cavity runs on in the form of a tail,
as if a gas-like elastic fluid had been developed by volcanic
heat in the still soft mass. This phenomenon particularly
attracted the attention of Leopold von Buch when he visited
the Thomson collection of minerals at Naples, in company
with Gay-Lussac and myself, in the year 1805.$ The infla-
tion of obsidian by the operation of fire, which did not escape
attention iji the early period of Grecian antiquity,§ is cer-
tainly caused by some such development of gas. According
to Abich, obsidians pass the more easily into cellular (not
parallel-porous) pumice, the poorer they are in silicic acid

* See above, p. 367dMid notes, p. 302-304.

t Franz Junglmhn, Jnva, bd. ii., s. 388, 592.

X Leopold von Buch, in the Abhandl. der Akademie der Wiss. zu
Berlin, for the years 1812-1813 (Berlin, 1816), s. 128.

§ Theophrastus de Lapidibus, s. 14 and 15 (Opera ed. Schneider, t. i.,
1818, p. 689 ; t. ii., p. 426; and t. iv., p. 551), ^ays this of the "lipa-
rian stone" Q^iTrapalog),

450 COSMOS.

and the richer they are in alkalies. It remains, however,
very uncertain, according to Rammelsberg's researches,*
whether the tumefaction is to be ascribed to the volatiliza-
tion of potash or hydrochloric acid. It is probable that
similar phenomena of inflation in trachytes rich in obsidian
and sanidine, in porous basalts and amygdaloids, in pitch-
stone, tourmalin, and that dark-brown flint which loses its
color, may have very different causes in the diiFerent mate-
rials themselves. An investigation which has now been long
looked for in vain, founded on accurate experiments, exclu-
sively directed to these escaping gaseous fluids, would lead to
an invaluable extension of our knowledge of the geology of
volcanoes, if at the same time attention were paid to the
operation of the sea-water in subterranean formations, and
to the great quantity of carbureted hydrogen belonging to
the commingled organic substances.

The facts which I have brought together at the end of this
section, the enumeration of those volcanoes which produce
pumice without obsidian, and those which yield a great deal
of obsidian and no pumice — the remarkable, not constant,
but very diversified association of obsidian and pumice with
certain other minerals, early led me, during my residence in
the Cordilleras of Quito, to the conclusion that the forma-
tion of pumice is the result of a chemical process, which may
be verified in trachytes of very heterogeneous composition,
without the necessity of a previous intervention of obsidian
(that is to say, without its pre-existence in large masses).
The conditions under which such a process is performed on a
large scale are perhaps founded (I would here repeat) less on
the diversity of the material than on the gradation of heat,
the pressure determined by the depth, the fluidity, and the
length of time occupied in solidification. The striking, though
rare, phenomena presented by the isolation of immense sub-
terraneous pumice quarries, far from any volcanic structures
(conical and bell-shaped mountains), lead me at the same
time to conjecture! that a not inconsiderable — perhaps even,
in regard to volume, the greater — number of the volcanic
rocks have been erupted, not from upraised volcanic struc-

* Rammelsberg, in Pogpend., Anna!., M. ]|xx., 1850, s. 464, and
fourth supplement to his Chemiscke Handwdr^huch, s. 169 ; compaue
also Bischof, Geol, bd. ii., 2224, 2232, 2280.

t See above, p. 291, 311, 312-316, 322-325. For particulars re-
specting the geographical distribution of pumice and obsidian in the
tropical zone of the New Continent, see Humboldt, Essai Geognostiqfte
sur k Gisement des Roches, etc., 1823, p. 340-342, and 344-347.

TRUE VOLCANOES. 451

tures, but from a net-work of fissures on the surface of the
earth frequently covering over in the form of strata a space
of many square miles. To these probably belong those masses
of trap of the lower silurian formation of the southwest of
England, by the chronometric determination of which my
worthy friend, Sir Roderic Murchison, has so greatly increased
and heightened our acquaintance with the geological con-
struction of the globe.

INDEX TO VOL. V.

ABtoa on. volcanic phenomena in Ghilan,jAntisana,tlie colossal moun^in, described,
169; his views on the Caucasian mount- • 311; its dikes, 312; lakes, 313.
ain system, 201, 383 ; analysis of the Antuco, volcano of, 273.

Chimborazo rock, 431,

Aconcagua, volcano of, measurement of,
273.

Acoata on the volcancitos of Turbaco, 205.

Adams, Mount, a volcano, 390.

.<Enaria, the island of Apes, 252.

^olu3, residence of, on Strongyle, 244.

-(Etna, eruptions of, usually occur within a
space of six years, 243; periods of its
greatest activity, 244; height to which
ejected matters attain, 251; its trach-
yte?, 434.

Africa, determination of the magnetic
equator in, by Sabine, 102 ; its transla-
tion, 104; snowy mountains in, 333 ; vol-
canoes in, 332; their smairnumber, 334.

African magnetic node, its varying posi-
tion, 102.

Agaschagokh, island of, 348.

Agreeable odor diffused from certain vol-
canoes, 219.

Agua, Volcan de, described, 262.

Airy, density of the earth determined by,
35; on terrestrial magnetism,- 79.

Alaid, great eruptions of the volcano on
the isle of, 349.

Albite, 438.

Aleutian islands, numerous volcanoes in,
347.

Alps, temperature of springs in the, 184.

America. See Central Amei'ica, Chili,
Mexico, Northwest America, Peru and
Bolivia, Rocky Mountains, South Sea.

Ampdre on the cause of earthquakes, 162.

Ampolletas, 57.

Amsterdam, volcanic island of, 360.

Auahuac, series of volcanoes of, 266.

Anaxagoras, maxim of, verified, 11.

Andaman isles, volcanic phenomena in the,
359.

Andes, large spaces in the chain of, desti-
tute of volcanoes, 267 ; groups and dis-
tances, 268 ; special direction of the
three Cordilleras, 276.

Andesite, 437, 439.

Andrea Bianco, his early charts exhibit
the magnetic variation, 55,

Anemasite, 446.

Annular valleys, 221.

Ansango, lake of, 313.

Ansogorri, Father Joaquin, his description
of the rise of the volcano Jorullo, 292.

Ant-hills in the Rocky Mountains, their
remarkable construction, 443.

Antilles, Little, volcanoes of the, described,
394.

Aphron^ the northern pole of the mag-
netic needle, 54.

Apparatus employed by Humboldt for hia
453 determinations of height in the New
Worid, 428.

Arabia, lava eruptions in, 336.

Arago on magnetic inclination, 105; hia
series of magnetic observations, 75.

Aiarat, as a volcano, 339.

Arare, crater of, 368.

Arequipa, volcano of, 270.

Argaeus, the volcano, 237.

Arimer, country of the, 252.

Aristotle on the fundamental principles of
nature, 9; volcanic phenomenon upou
Hiera described by, 219.

AiTan, volcanic phenomena in, 329.

Artesian wells, Walferdin's observations
on, 38.

Ascension, volcanic phenomena of the isl-
and of, 331.

Asia, situation of the principal volcanoes
in, 281 ; volcanoes of the western and
central parts, 334; of Kamtschatka, 340 ;
of the islands of Eastern Asia, 344 ; of
the islands of Southern Asia, 354; of the
Indian Ocean, 358.

Atlantic Ocean, volcanoes of the islands
of the, 330; presumed submarine vol-
cano, 332.

Atlantis of Solon, 173.

Atolls, or lagoon reefs, 363.

Attraction of the magnet known to the
Greeks and Romans, 51.

Augite, 443.

Aurora Borealis, 147 ; observations of the
black segment, 148; colors observed in
high latitudes, 149 ; accompanying fleecy
clouds, 150 ; influence on terrestrial mag-
netism, 152 ; observations at Berlin and
at Edinburgh, 153.

Auvergne, extinct volcanoes of, 227, 263.

Azores, craters of elevation in the, 217;
the volcano of Pico, 236.

Azufral de Quindiu, Humboldt's visit to
the, 211; change of temperature ob-
sei-ved by Boussingault, 212.

Baily on the density of the earth, 34.

Baker, Mount, a volcano, 390.

Banda, a volcanic island, 357.

Barba, the volcano, described, 259.

Barile, earthquake at, 167.

Barrancos on the slopes of volcanoes,

287.
Barren Island, one of the Andamans, ap-

454

INDEX.

pearance of, as described by Horsburgh,
S55.

Basalt-like columns of Pisoje, 426.

Beaufort, Admu-al, the Chimsera described
by, 244.

Beauvais, Vincent of, on the magnetic
needle, 54.

Belcher, Sir E., magnetic observations by,
111.

Bell-shaped volcanic mountains, 218.

Berg, Albert, his description of the burn-
ing spring Chimsera, 244.

Berlin, aurora%ob8erved at, by Humboldt,
153.

Bessel, determination of the size and fig-
ure of the earth, 18, 29.

Biot, pendulum measurements by, 26.

Bolivia. See Peru.

Borda, his services in equipping the expe-
dition of La Perouse, 62.

Borneo, the Giava Maggiore of Marco Polo,
355; doubtful whether volcanoes exist
there, 355; great number of volcanoes
in its vicinity, 355.

Bo-shan, eruption of the volcano, 409.

Bouguer's experiments on the deviation of
the plummet, 33; on the pumice-quar-
ries of Lactacunga, 322.

Bourbon, volcanoes of the isle of, 359.

Boussingault's method of determining the
mean temperature, 42 ; on the cause of
earthquakes, 164 ; on the matters eject-
ed fi-om volcanoes, 315 ; on gases, 413.

Bove, Val del, on ^tna, 215, 230.

Bramidos de Guanaxuato, 172.

Bravais on Artesian wells, 40; on the
black segment of the aurora, 148.

Brisbane, Sir Thomas, liis observatory at
Makerston, 120.

British isles, volcanic phenomena in the,
329, 450.

Bromo, a volcano in Java, its crater-lake,
285

Brooke, Rajah, on the volcanic appear-
ances in Borneo, 356.

Brooks of cold water said to be converted
into thermal springs, 296.

Brown, Mount, a volcano, 390.

Buch, Leopold von, his work on basaltic
islands and craters of elevation, 216;
on the erupted matters of Vesuviua, 224;
on the trachytes of ./Etna, 437.

Buddhist fancy as to the cause of earth-
quakes, 170.

Bunsen on fumaroles, 396.

Burkart, his visit to JoruUo, 300.

Calabria, earthquake in, in 1783, 166.
Calamatico, el, an ancient name for the

magnetic pole, 57.
Calbuco, Volcan de, 274.
Caldron-like depressions of volcanoes,

221.
California, list of the volcanoes of, 389.
Callaqui, volcano of, 274.
Canary Islands, eruptions in the, 445.
Capac-Urcu, an extinct volcano, 267.
Cape of Good Hope, magnetic observations

at, 111.
Carbonic acid gas, considerations on, 413.
Carbonic acid gas, jets of, 193.

Cascade mountain range, in California, 388.

Castillo, Fray Bias del, explores the crater
of Masaya, 247.

Catalans, advanced state of navigation
among the, 54, 55.

Caucasus, volcanic phenomena of the, 199 ;
a continuation of the Thian-schan, 338;
its extinct volcanoes, 338.

Cavanilles, his account of the earthquake
of Riobamba, 166.

Celebes, volcanoes of, 357.

Central America, linear volcanoes of, 255,
258 ; number of volcanoes in, 259 ; rec-
ommended for farther examination, 263.

Chacani or Charcani, volcano of, 270.

Chahon-a, the crater of, on the Peak of
Teneriffe, 249.

Chatham Island, its position, 376.

Chili, group of volcanoes in, 272; their
greatest elevation, where attained, 280.

Chilian, Volcan def, 273.

Chiloe, submai-ine volcano near, 272.

Chimsera, in Lycia, not a volcano, but a
perpetual burning spring, 203, 244; an-
alogous phenomenon in the Kuen-liin,
409.

Chimborazo, majestic dome, form of, 419;
ascent of, 432; considerations on the
height of the mountain, 432.

Chimborazo rock, Rammelsbei^'s analysis
of, 430; Abich's, 431; remarks on the
differences between them, 432.

Chinal, volcano of, 274.

Chinese, early acquainted with the polari-
ty of the magnet, 52 ; rope-boring, 209 ;
early maps of the, 405.

Chuapri, volcano of, 272.

Cinders, cones of, wanting in several vol-
canoes which once emitted lava streams,
449 ; thickness of the layers of, on San-
gay, 251.

Circumvallations, volcanic, 220; that of
Oisans, in France, its great extent, 220 ;
of Mont Blanc, 220.

Coal strata, 413.

Coan, the missionary, on the basin of Kil-
auea, 368.

Coast Range mountains, in California, old
volcanic rocks of the, 389.

Cofre de Perote, Humboldt's ascent of, 307.

Columbus determines astronomically a lin«
of no variation, 55; notice of an eruption
on Teneriffe, by, 445.

Comangillas, Aguas de, a hot spring, 189.

Commotion, waves of, in earthquakes, 165;
theory of, 166; attempts to explain the
rotatory shocks experienced in Calabria,
166.

Commotions of the earth in earthquakes
often confined within narrow limits, 175.

Comoro Islands, burning volcano in the,
360.

Compass. See Mariner's Compass.

Compression, polar, 32.

Conchagua, a volcano, 261.

Conical volcanic mountains, 228.

Conseguina, eruption of, 260.

Copiapo, destruction of the town of, 272.

Coquimbo, volcano of, 272.

Coral islands, number of, in the Pacific,
according to Dana, 365.

INDEX.

455

Corcovado, Volcan de, 274.

Cordilleras. See Andes.

Corea, volcanoes of, 353.

Cosiraa, small elevation of the volcano of,

234.
Costa, Colonel A., his experiments on

mean annual temperature, 43.
Cotopaxi, mineralogical composition of,

322.
Craters of elevation, 215; distinguished

from true volcanoes, 217. See, also,

Volcanoes.
Crozet's group, traces of former volcanic

action in, 362.
Crust of the earth, considerations on its

varying thickness, 410.
Crystallized minerals of the Maars, 224;
.. greater number found on Vesuvius, 224.
Cueva de Antisana, 312.
Cyclades, volcanic phenomena in the, 254.

Dana, James, his valuable researches in
the Pacific, 364; his grouping of the ba-
saltic and coral islands, 365 ; and the vol-
canoes of the Sandwich Islands, 367.

Darwin, Charles, his enlarged views on
earthquakes and eruptions of volcanoes,
272 ; general acknowledgment of obliga-
tions of science to, 364.

Dasar, sand lakes of, 440.

Dechen, H. von, on volcanic phenomena
in thp Eifel, 226.

Declination. See Magnetism.

Degree, table of the inci-ease in length of
the, from the equator to the pole, 21.

Demavend, volcano of, 335; question of
its altitude, 334.

Density of the earth, experiments to de-
temiine, 33 ; Airy's results, 35.

Detritus dikes, 811.

Deville, on the structure and color of the
mass in certain volcanoes, 432.

Devonian slate, 221.

Diablo, Monte del, in California, 389.

Diamagnetism, its discovery by- Faraday,
51, 77.

Dio Cassius on the eruptions of Vesuvius,
399.

Diodorus Siculus on the Phlegraean Fields,
400.

Disturbances, magnetic, table of, 130.

Djebel el Tir, a volcano, 334.

Dome-shaped and bell-shaped mountains,
peculiar aspect given by, to the land-
scape, 218.

Domite, origin of the term, 421.

Dry fog of the summer of 1783, 393.

Duperrey, his observations on the mag-
netic equator, 103.

Earth, its size, configuration, and density,
14,35; interior heat, 37, 234 ; magnetic
activity, 50 ; magnetic storms, 1.37 ; po-
lar light, 146; reaction of the interior on
the surface, 157 (see, also. Earthquakes,
Volcanoes); thickness of the crust of,
probably very unequal, 163.

Earthquakes, variety of views as to their
cause, 162; the impulse, 162; trans-
latory movements, 167; subterranean
noises, 171 ; velocity of propagation, 172;

distinguished, but improperly, as Plu-
tonic and Volcanic, 174; three groups
of phenomena which indicate the exist-
ence of one general cause, 176; list of
memorable examples of these phenome-
na, 176.

Earth-waves in volcanic phenomena, 165.

Eastern Asia, volcanoes of the islands of,
344.

Edgecombe, Mount, a volcano, 255, 391 ;
another, in New Zealand, 372.

Edinburgh, beautiful aurora observed at,
153.

Edrisi on the land of Gog and Magog, 337.

Eifel, extinct volcanoes of the, 221; two
kinds of volcanic activity distinguish-
able, 222 ; Mitscherlich on the minerals,
224 ; Ehrenberg on the infusoria, 227.

Elburuz, as an extinct volcano, 339.

Elevation, question of the influence of, on
magnetic dip and intensity. 111 ; craters
of, distinguished from true volcanoes,
217.

Elias, Mount, a volcano, 239, 391.

Elliot, Captain, on the magnetic equator,
104.

Ellipticity of the earth, speculations of the
ancients on the, 29 ; Bessel's detennina-
tion, 29.

El Nuevo, a volcano, 260.

El Viejo, a volcano, measurements of, 260.

El Volcancito, now a mountain of ashes,
302.

Emanations from fumaroles, their nature,
396.

Enceladus. See Typhon.

England, volcanic phenomena in, 329, 450.

Equator, magnetic. See Magnetic Equa-
tor.

Erebus, Mount, the volcano, 101, 237.

Erman on the magnetic equator, 103 ; his
researches on the volcanoes of Kamt-
schatka, 340.

Erupted blocks, 446.

Eruption, masses of, considerations on,
215; craters of, 216.

Eruptions of volcanoes, considerations on
the general laws of, 243; varying heights
to which matters are cast, 251.

Euboea, Strabo's description of an earth-
quake in, 215.

Europe, active volcanoes of, 328; extinct
volcanoes and volcanic phenomena, 221,
227, S29, 450.

Fairweather, Mount, a volcano, 391.

Faraday's discoveiy of the paramagnetic
force of oxygen, 78; important results
expected from it, 81, 98; on diamagnet- '
ism, 51, 78.

Feldspar, variety of minerals comprised
under the denomination of, 427, 442.

Ferdinandea, the volcanic island, 328.

Figure of the earth, attempts to solve the
problem, 18; detei-minations of Bessel,
19 ; earlier observations, 20.

Fissures caused *by earthquakes, 166 ; vol-
canic, 216, 218; volcanoes upheaved on
fissures, 252. See Volcanoe,^.

Fitzroy's magnetic observations, 71.

Floods in rivers, prognostication of, 180.

456

INDEX.

Fogo, volcano of the Ilha do, 249.

Forbes, on the conductive power of differ-
ent rocks, 41.

Formosa, the tnrning-point of the lines of
volcanic elevation in the islands of East-
ern Asia, 346 ; its volcanoes, 353.

Foucault's apparatus for demonstrating the
rotation of the earth, 28.

France, extinct volcanoes of, 22T, 263.

Franklin on frozen earth in the northwest
of America, 50; his Arctic voyages, 65;
search for him, 65.

Franklin's Bay, volcano of, more properly
a salse, 391.

Fredonia, near Lake Erie, springs of in-
flammable gag at, 204.

Fremont's hypsometrical investigations in
Northwest America, 383.

Fremont's Peak, 388.

French Alps, highest summit of the, 220.

Frozen earth, its geographical extension,
48.

Fse-7mn. a Chinese magnetic apparatus,
52.

Fuego, Volcan de, described, 262.

Fumaroles, various classes of, 396; Bun-
sen on their products, 396.

Fummarole of the Tuscan Maremma, 202.

Fused interior of the earth, 234.

Galapagos, the, countless cones and ex-
tinct craters, 374; pumice not found
there, 375.

Galera Zamba, teiTible eruptions of flames
and terrestrial changes at, 208.

Gandavo, Fray Juan de, explores the cra-
ter of Masaya, 247.

Gas, volcanic exhalations of, inquiry into.
412. See, also, Sprin.^s.

Gauss, his theory of terrestrial magnetism.
63.

Gay-Lussac on the chemical causes of vol-
canic phenomena, 163 ; on waves of com
motion and oscillation, 165.

Gemellaro, his estimate of the height to
which erupted bodies ascend from ^t-
na, 251.

Geographical distribution of volcanoes,
393; an abnormal phenomenon in, no-
ticed, 405.

Geological terms, origin of some, 421.

Geysers, the, of Iceland described, 191.

Gilbert, William, lays down comprehen
aive views on the magnetic force of the
earth, 58.

Glassy feldspar. See Feldspar.

Godivel, Lac de la, an extinct volcano.
227.
' Gog and Magog, Oriental myth of, 337.

Gold, believed to be found in volcanoes,
248 ; descent into Masaya, in search of
it, 248.

Graham, his observation of the hourly va-
riations of the magnetic force, 61.

Graham Island, temporary formation of,
328.

Grand Ocean, a term for the basin of the
South Sea, objected to, 378.

Granite, Mitscherlich's experiments on the
melting point of, 234.

Greece, has frequently suffered from earth-

quakes, 170 ; great number of thermal

springs, 170.
Grenelle, the Artesian "Well of, 38.
Ground temperature, observations on, 182.

See, also. Frozen Earth.
Guadeloupe, the Soufri6re of, described,

395.
Guagua-Pichincha, its meaning, 231.
Gualatieri, volcano of, 271.
Guanacaure, a volcano, 260.
Guanahuca (Guanegue ?) volcano of, 274.
Guettard's observations on extinct volca-
noes, 310.
Gunung, the Javanese term for mountain,

282.
Gunung Tengger, a volcano in Java, vast

size of its crater, 284.
Guyot of Provins, his mention of the mag-*

netic needle, 54.

Hair glass, a volcanic product, 367.

Hall, Captain Basil, experiments to de-
termine the mean temperature of places
within the tropics, 42 ; measurement of
the volcanoes of Old Guatemala, 262;
his admirable description of Sulphur Isl-
and, S53.

Halley's theory of four magnetic poles, 59.

Hallmann, his classification of springs,
196.

Hansteen on the magnetism of the earth,
66.

Harton, pendulum experiments at, relative
to the density of the earth, 35.

Hawaii, the volcanoes of, described, 369,

Heat, distribution of, in the interior of our
globe, 37 ; hypothesis of the depth of the
fused interior of the earth below the
present sea-level, 234.

liecla, the volcano, its aspect, 232; in-
frequency of its eruptions, 243 ; how
classified by Waltershausen, 330.

Helena, St., volcanic phenomena of, 331.

Helen's, St., Mount, a volcano, 390.

Hell, the cold, of the Buddhists, 189.

Hephaestos, Volcano, the holy isle of, 244.

Herefordshire, sedimentary rocks of, 221.

Hesse, on the volcanoes of Central Ameri-
ca, 258.

Hiera, volcanic phenomena upon, de-
scribed by Aristotle, 219.

Himalayan chain, four highest mountains
of the, 271 ; known to the Greeks as the
elongated Taurus, 406.

Hobarton, magnetic observations at, 99.

Ho-cheu, a volcano, also called Turfan,
835.

Hood, Mount, an extinct volcano, 389.

Hooker, Joseph, on the hot springs of Mo-
may, 891.

Hopkins on earthqtiakes, 162, 165, 168.

Horaiy variation of the declination not
ascribable to the heat of the sun, 81;
maxima and minima, at various mag-
netic stations, 107.

Hornblende and augite, 443.

Hornitos, low volcanic cones, 176 ; farther
notices of them, 298, 303.

Homos or Hornitos. See Hornitos.

Horsburgh, description of Barren Island
by, 359.

INDEX.

457

Ho-schan and Ho-tsing^ of Eagtem Asia,

209.
Humboldt, Alexander von, observations of
temperature in Mexico and Peru, 43;
magnetic observations by, 95; his de-
temiination of the magnetic equator,
102; observations of polar bands, 150;
visit to the scene of the earthquake of
Kiobamba, 166 ; observations of the phe-
nomena of an eruption of Vesuvius, 1T4;
barometrical measurements of the same
mountain, 236 ; his definition of the term
"•volcano," 272; his visit to Jorullo,
295, 301 ; the name Jura limestone in-
troduced by, 436; apparatus employed
by, in the New World, 429 ; his miner-
alogical collections, 445; on the forma-
tion of pumice, 450
Humboldt, Alexander von, -vvorks by, cited
in the text or notes :
Asie Centrale, 52, 100, 114, 143, 144,
ITO, 199, 202, 209, 210, 238, 279, 316
334, 336, 338, 349, 353, 409.
Atlas Geogi-aphique et Physique de la
Nouvelle Espagne, 228, 235, 250.
291, 404.
Essai Geognostique sur le Gisement
des Roches, 212, 302, 415, 425, 436,
442, 450.
Essai sur la Geographic des Plantes,

239, 427.
Essai Politique sur la Nouvelle Es-
pagne, 44, 189, 2C5, 277, 293, 294.
307, 379, 380, 391, 427.
Examen Critique de I'Histoire i3e la
Geographie, 52, 116, 123, 173, 233
248.
Fragmens de Geologic et de Climato-

logie Asiatiques, 349, 354.
Kleinere Schrifteu, 165, 205, 228, 275,

315, 316, 321, 422, 446.

Recueil d' Observations Astronomiques,

43, 103, 139, 212, 239, 265, 297, 307.

Relation Historique du Voyage aux

Regions Equinoxiales, 96, 110, 113,

Invariable temperature, stratum of, 41.

Ischia, 252.

Island of Desolation. See Kerguelen' s Isl-
and.

Islands, temporary, enumerated, 828.

Islands and the shores of continents, great
number of volcanoes found on, 403.

Islands of the Pacific, Dana's classification
of, 365.

Isluga, volcano of, 271.

Izalco, volcano of, described, 248 ; its erup-
tions, 261.

Iztaccihuatl, a volcano, meaning of the
name, 228.

Jacob, valley of, on Ararat, 230.

Jakutsk, mean annual temperature of, 46 ;
extreme variations, 47.

Jan Mayen, volcanoes of the island of, 330.

Japan, notice of the volcanoes of, commu-
nicated by Siebold, 350.

Jaques de Vitry, his mention of the mag-
netic needle, 54.

Java, large number of volcanoes in, 281 ;
their comparatively low elevation, 282 ;
direction of the principal axis, 284 ; vast
craters of some, 284; ribbed formation,
286; lava streams, 288; salses of, and
mofette grottoes, described by Jung-
huhn, 210 ; tertiary formations, 281.

Javanese names of mountains explained,
290.

Jefferson, Mount, 389.

Jesso, island of, 349 ; its numeroua Vol-
canoes, 850.

Jorullo, rise of the volcano, 266, 291 ; de-
scription of, by eye-witnesses, 292 ; vis-
it of Humboldt to, 295, 300; visit of
Buckart, and changes noticed by him,
300.

Juan Jayme, his scientific voyage, 56.

Julia, the volcanic island, 328.

Julius, the proconsul, 188.

381, 899.
Vues des Cordilleres, 208, 228, 230, 235.
Hypersthene rock, its employment for or-
namental purposes, 446.
Hypsometry of volcanoes, first group, 235;
second gi'oup, 285; third group, 236;
fourth group, 238 ; fifth group, 239.

Iceland, the Geysers of, 190 ; mud springs,

203; volcanoes, 880.
Ilha do Fogo, one of the Cape Verd Isl-
ands, so called, 249.
Impulse in volcanic phenomena, summary

of views on, 162.
Inarima, 253.
Inclination, magnetic, 100; maxima and

minima, 107 ; secular variation, 109.
Indian Ocean, volcanoes of the, 858, 363.
Infusoria, universal diffusion of the, 26.
Intensity of the magnetic terrestrial force,

68, 61, 87.
Interior of the earth, its reaction on the

surface, 157. See, also. Earthquakes,

Volcanoes.

Vol. v.— U

Jumnotri, hot well of, 190.

Junghuhn, his researches in Java,''210, 281.

116, 167, 168, 179, 237, 238, 286, 394^ Jura limestone, name introduced by Hum-

896, 399. TI boldt, 436.

Views of Nature, 248, 343,

Kaimenes, upheaval of the three, 328.

Kamtschatka, the loftiest volcano of Asia
found in, 284; described, 340.

Kerguelen' s Island, extinct craters of, 363.

Kilauea, the great crater of, not a solfa-
tara, 367.

Kina IBailu, a lofty mountain of Borneo,
356.

Kirghis Steppe, fonner water-courses of
the, 408.

Kljutschewsk, the highest Asiatic vol-
cano, 284.

Korai. See Corea.

Kotzebue on the volcanic island of Um-
nack, 220.

Krafto. See Saghalin.

Ki-apf, discovery of a volcano in Easteni
Africa by, 883.

Krasnajazarki, polar bands observed by
Humboldt at, 150.

Kreil on the magnetism of the moon, 85.

Krusenstern on a presumed submarine vol-
cano, 332,

458

INDEX.

Kuen-liin, fire-springs of the, 409; the

chain visited by the hrothei-s Schlagint-

weit, 409.
Kuopho on the magnetic needle, 52.
Kupffer on the frozen soil of Northern

Asia, 50.
Kurile islands, active volcanoes of the,

349.

La Berarde, remarkable position of the vil-
lage of, 220.

Lactacunga, repeated destniction of the
town of, 322 ; subterranean pumice quar-
ries of, 321, 447.

Ladrone islands, volcanoes of, 3T0.

Lagoni of the Tuscan Maremma, 202.

Lament deduces the law ot the period of
alterations of declination, S3.

Lancerote, destruction of the islands of,
213.

Lava, recent, often peifectly similar to the
oldest formations of eruptive roqk, 216 ;
important conclusion di-awn thei'efrom,
216.

Lava fields, various names for, 305.

Lava streams rare in the volcanoes of the
Cordilleras of Quito, 263; discovered in
the eastern chain of the Andes, 2T9 ; also
in Java, 2SS; their essential charactei-,
2S9; of Auvergne, 311; of ^tna, 434;
of Hecla, 231; of Temate, 357.

Lazarus, St., Mount, volcano, 255.

Lelantus, in Eubcea, eruption at, 215.

Lemnos, destruction of the mountain Mo-
sychlos in, 328.

Letronne on earthquakes in Egypt, 171.

Leucite, 435, 444.

Liraari, volcano of, 272.

Linschoten, notices tlie volcanoes of Ja-
pan, 351.

Lipara, the volcano, question of its iden-
tity, 243.

Lipari, the ancient Meligunis, 243; lava
stream found in, 320.

Llandeilo strata, volcanic fragments found
in the, 329.

Llanquihue, volcano of, 274.

Log, ship's, introduction of the, an im-
portant era in navigation, 57.

Lombok, volcano on the isle of, 357.

Lucia, St., the volcano of, 395.

Lunar-diurnal magnetic variation, 75.

Liitke, Admiral, on the volcanoes of Kamt-
schatka, 341.

Luzon, active volcano in, 232,

Maars, in Germany, 221; in Auvergne, 227.

Macas. See Sangay.

M'Laughlin, Mount, its height, 389.

Madagascar, volcanic indications in, 360.

Madeira, volcanic phenomena of, 330.

Magnet, attraction, but not polarity of the,
known to the Greeks and Romans, 51 ;
variations of the, early known to the
Chinese, 53; variation charts, 55; ho
rary periodical variations, 61.

Magnetic disturbances, table of, 131.

Magnetic equator, its position and change
of form, 101 ; Humboldt's determina-
tions, 102; Duperrey's observations, 103;
Elliot's, 104.

Magnetic Jptensity, 61 ; the knowledge of,
rl 1,. to Borda. 62; inclination chart, 62.

Magnetic nee(' c, early known to the Chi-
nese, 52; 116 introduction to Europe,
54 ; declination, 55.

Magnetic obsei-vatories, 63. ,

Magno'^' storms, 130.

Magne^-.- wagon, the, of the Chinese, 52.

Magnetism, early researches in, 56, 58;
increased activity of observation in the
19th century, 62 ; table of magnetic in-
vestigations, 64 ; influence of the moon,

Magnetism of mountain masses, 154.

Makerston, Sir Thomas Brisbane's ob-
servatiii^ at, 120, 121.

Malpais, a tenn applied to lava fields,
289.

Mandeira, the volcano, 259.

Mantschurei, extinct volcano in, 409.

Marco Polo, date of his travels, 54; the
mariner's compass known in Europe be-
fore his time, 54.

Marcou, on the ant-hills in the Rocky
J )untains, 445.

Maribios, los, a line of six volcanoes, 260.

Mariner's compass known in Europe in the
12th century, 55; English ships guidei
by it in 1345, 57.

Marion's Island, traces of fonner volcanio
action on, 362.

Martinique, recent volcanic action in the
island of, 395.

Masaya, volcano of, described, 245; de-
scent into the crater of, 247.

Mauna Roa, a volcano of the Sandwich
Islands, 238; its height greatly exag.
gerated, 238 ; meaning of the name, 234;
described, 366; the largest volcano of
the South Seas, 366; called also Mouna
Loa, 366; its lava lake of Kilauea, 368.

Maypu, volcano of, 273.

Medina, volcano of, 334.

Meligunis, See Lipari.

Methone, volcanic phenomena of the pen-
insula of, 218.

exico, list of elevations of the table-land
of, 382; volcano of, 376; considerations
on the mountain chains, 379. See, also,
New Mexico,

Mica, 441.

Micuipampa, mean annual temperature of,
44.

Middendorf s two Siberian expeditions, 45 ;
on the frozen toil ol Northern Asia, 49. .

Minchinmadom, volcano of, 274.

Mines, observations in, on magnetic dip
and intensity, 114.

Mitscherlich on the minerals of the Eifel,
224; on the melting point of granite,
234.

Mofette grottoes of Java, described by
Junghuhn, 210.

Momay, hot springs of, 189.

ilomobacho, the volcano, 259.

Momotombo, the volcano, 260.

Monkwearmouth, the coal mine at, 39.

Mont Blanc, the Grand Plateau of, 220.

Mont Pelvoux, the highest summit of th»
French Alps, 220.

Monte del Diablo, in California, 389.

' 9li

INDEX.

459

Moon, extent of our acquaintance with thejOrthoclase, 448.

surface of the, 418; volcanoes and para- 1 Osomo, volcano of, 274.

sitic craters, 419 ; Kreil on the magnet-
ism of the, 84 ; investigation of the sub-
ject by General Sabine, 84.

Mormons, Great Salt Lake of the, 383.

Mortero, Cerro del, 302.

Mosenberg, the, an extinct volcano, 222,
227.

Mosychlos, the mountain, destruction of,
328.

Mouna Loa. See Manna Roa.

Mountain masses, magnetism of, 154.

Mountain peaks, comparison of, with the
bulging of the earth's surface, 31.

Mousart (corruption of Muztag), equivalent
to Sierra Nevada, 405.

Moya cones of Pelileo, 166, 207.

Mud springs of Iceland, 203.

Mud volcanoes, 207, 255.

Murchison, Sir R., on eruptive trap masses,
329, 451.

Muriatic acid fumaroles, 397.

Mutis, apparatus of, 428.

Naphtha springs, 199.

Negropont. See Euboja.

Neptune, connection of, with earthquakes,

173.
New Britain, volcanoes of, 371
New Caledonia, volcanic action absent

from, 372.
New Guinea, volcanoes of, 371.
New Mexico, barometric levelings in, 380 ;

list of heights, 382.
New Zealand, geology of, 371 ; volcanoes,

372.
Niphon, recorded volcanic eruptions in, 350.
Nodes, magnetic, their changes of position,

102, 104.
Noises from volcanoes, differences observed

in, 250; extraordinaiy distances at which

heard, 251.
Norman, Robert, determines the inclina-
tion of the magnetic needle in London,

58.
Northwest America, volcanoes of, 377;

hypsometry of, 382.
No variation (magnetic), points and lines

of, 55, 59

Obsidian, 447; its cavities and air-holes,

449.
Oerafa, in Iceland, fearful eruptions of,

330.
Oeynhausen, temperature of the salt spring

at, 39."
Oisans, natural amphitheatre of, its vast

extent, 220.
Oligoclase, 439.

Olot, extinct volcanoes of, 405.
Olympus, Mount, in America, 390.
Omato, Volcan de, 271.
Ometepec, an active volcano, 259.
Orinoco, high temperature of its waters at

certain seasons, 179.
Orizaba, a volcano, measurement of the

peak of, 239 ; lava field of, 305.
Oron, fresh-water lake of, seals found in

the, 408.
Orosi, the volcano, 259.

Overweg's researches on volcanic phenom-
ena in Africa, 334.

Ovid, volcanic phenomena clearly described
by, 219. '

Owliyhee. See Hawaii.

^caya, eruptions of, 262.
Jfcific Ocean, the term '' Grand Ocean"
improperly applied to the, 378 ; compar-
atively small number of active volcanoes,
364; grouping of its islands by Dana,
365. See, also. South Pacific Ocean,
South Sea.
PanguipuUi, Volcan de, 274.
Papagayos, remarkable storms so called,

257.
Paramagnetism exhibited by oxygen gas,
51 ; importance of the discovery, 78, 81,
98.
Paramos, their elevation and vegetation,

278.
Parasitic craters of the moon, 419.
Parinacota, volcano of, 271.
Passuchoa, the extinct volcano of, 317.
Patricius, the bishop, his theory of central

heat, 188.
Paul, St., volcanic island of, 360.
Pele's hair, volcanic glass so called, 367.
Pelileo, eruption of the Moya of, 166, 207.
Pendulum, vibrations of the, applied to
determine the figure of the earth, 23;
Sabine's expedition, 26 ; other obsei-vei-s,
26 ; the form of the earth not exactly
determinable by such means, 29 ; Airy's
experiments at Harton, 35.
Pentland, his discovery of lava streams in

the eastern chain of the Andes, 279.
Perlite, 323.

Pertusa, hot springs of, 188.
Peru and Bolivia, series of volcanoes of,

276.
Peshan, volcano of, 335, 406.
Petermann's notices from Overweg, of vol-
canic phenomena in Africa, 334.
Peteroa, volcano of, 273.
Phaselis, flame of the Chimajra, near, 203.
Philippines, volcanoes of the, 232.
Phlegraian Fields, ancient descriptions of

the, 400.
Pic de Nethou, the highest summit of the

Pyrenees, 220.
Pic of Timor, formerly an evei"-active vol-
cano, 358.
Pichincha, remarkable form of, 230; ascent
of, by Humboldt, 231 ; visited by Wisse,
231; its height, 238.
Pichu-Pichu, Volcan de, 271.
Pico, the volcano, 236 ; eruptions of other
volcanoes in the Azores apparently de-
pendent on, 330.
Piedmont, trembling of the earth in, 17G.
Pilla, on theleucite crystals of Rocca Mon-

fina, 434.
Pisoje, basalt-like columns of, 426.
Pithecusse, Bokh on the, 253.
Pitt, Mount, in America, 389.
Plato, on the Pyriphlegethon, 37, 254 ; on

the magnetic chain of rings, 51.
Polar li"rht. Seo Aurora.

4G0

INDEX.

Polarity, the force of, unknown to the
Greeks and Komans, 51.

Polea, magnetic, traditions regarding, 56 ;
Ilalley's variation chart, 60.

Polybius, hi.s knowledge of St^ongyle, 244.

Polynesia and similar divisional terms, ob-
jected to, 304.

Pomarape, volcano of, 271.

Popocatepetl, a volcano, 239 ; meaninggB
the name, 228; determinations of Wte
height of, 427.

Porphyries of America, 443.

Porphyiy of the Puy de Dome, its peculiar
character, 421.

Porto Cabello, hot springs of, 190.

Pozzuoli, eruption from the solfatara of,
395.

Procida or Pi'ochyta, 252.

Proclus on earthquakes, 173.

Pulu Batu, lava streams of, 358.

Pumex Pompejanus, 402.

Pumice not found at JoruUo, 301; abun-
dant in Lipari, 320 ; the pumice quarries
of Lactacunga, 321; of Cotopaxi, 322;
isolated eruptions of, 323; found in
Madagascar, 360; and in the island of
Amsterdam, 301; Humboldt's view of
its formation, 450.

Pumice eruption of the Eifel, 226.

Punhamuidda, volcano of, 274.

Pusambio, the river, acidified by sulphur,
194.

Pyrenees, highest summits of the, 220, 221.

Pyriphlegethon, Plato's geognostic myth,
37, 254.

Quelpaert's island, a volcano, 353,

Que.^altenango, Volcan de, 262.

Quetelet on daily variations of tempera-
ture, 41.

Quindiu. See Azufral de Quindiu.

Quito, observations on the older rocks of
the volcanic elevated plains of, 415.

Quito and New Granada, the group of vol-
canoes of, 200.

Rainier (or Regnier) Mount, an active vol-
cano, 390.

Rains, regions of summer, autumn, and
winter, ISO.

Raking of mountain chains explained,
278.

Rammelsberg's analysis of the Chimborazo
rock, 431.

Ranco, volcano of, 274.

Rapilli, 223.

Raton Mountains, extinct volcanoes of the,
380.

Regnier, Mount, an active volcano, 391.

Rehme, the Artesian well at, 39.

Reich's experiments to determine the dens-
ity of the earth, 34; the subject more
lately investigated by Aiiy, 35.

Results of observations in the telluric por-
tion of the physical description of the
universe, 13.

Revillagigedo, volcanic islands of, 200.

Ribbed formation of the volcanoes of the isl-
and of Java, 286 ; analogous phenoraona
of the mantle of the Somma of Vesuvius,
288.

Richer, observations on the pendulum, by,

Rigaud, Professor, on the proportion of
water and terra firma, 303.

Rindjani, a volcano, its height, 357.

Riobamba, terrible earthquake at, 101, 166,
107.

Rio Vinagre, described, 194.

Rock-debris, 311.

Rocky Mountains, the chain described,
385; traces of ancient volcanic action,
387 ; parallel coast ranges, still volcanic,
388.

Ronquido and bramido^ distinguished, 250.

Rope-boring of the Chinese, 209.

Rose, Gustav, his classification of volcanic
rocks, 420, 423.

Ross, Sir Jame^ Clark, his Antarctic voy-
age, 75, 141.

Ross, John, his Polar voyages, 65.

Rucu-Pichincha, its meaning, 231.

Ruido, el (jran^ 166.

Sabine, Major-General, his pendulum ex-
pedition, 26 ; on the horary and annual
variations, 81; on the influence of the
moon on terrestrial magnetism, 84.

Sacramento Butt, an extinct crater, 389.

Saghalin, called Krafto by the Japanese,
345.

Silhama, Volcan de, 271.

Salses and naphtha springs, 199.

Salt Lake, Great, of the Mormons, 383.

San Bruno, rotatory motion of the obelisks
before the nionasteiy of, in Calabria, 166.

San Clemente, volcano of, 274.

Sandwich Islands, a volcanic Archipelago,
366 ; the volcanoes, 233 ; height of some
greatly exaggerated, 238.

Sangai or Sangay, the volcano, 239; its
position, 239 ; the most active fof the
South American volcanoes, 249 ; its erup-
tions observed by Wisse, 175.

Sanidine, 443.

San Miguel Bosotlan, a volcano, 261.

San Pedro de Atacama, Volcan de, 273.

San Salvador, a volcano, eruptions of, 261.

Santa Cruz, volcano of, 309,

Santorin, volcanic eruption of, 219.

San Vicente, a volcano, eruptions of, 261.

Saragyn, hot springs of, 325.

Sawelieff on magnetic inclination. 111,

Schagdagh, the perpetual fires of the, 201.

Schergin's shaft, at Jakutsk, 45.

Schiwelutsch, a volcano, its peculiar form,
237. •

Schlagintweit, the brothers, observations on
springs, 183; traverse the Kuen-tun, 410.

Schrenk on the frozen soil in the country
of the Samojedes, 48.

Sea, distance of volcanic activity from the,
statements of, examined, 404; volcanic
eruption observed in the, 354.

Seals found in the Caspian Sea and the Sea
of Baikal, 408 ; also in the distant fresh-
water lake of Oron, 408.

Secular variation of the magnetic inclina-
tion, 109.

Semi-volcanoes, 390.

Senarmont, his preparation of artificial
minerals, 195,

INDEX.

461

Seneca on volcanoes, 216.

Seriarga, volcano of, 370.

Shasty Mountains, basaltic lavas found in
the, 3S9.

Siebengebirge, trachyte of the, 226; geo-
logical topography, 424.

Siebold on the volcanoes of Japan, 349.

Sierra Madre, erroneous notions regarding
the, 379, 383 ; east and west chains, 384.

Silla Veluda, volcano of, 2T3.

Silurian and Lower Silurian formations,
eruptive trap-masses of the, 329, 450.

Silver in sea-water, its presence how mani-
fested, 411.

Sitka or Baranow, 45, 255.

Smyth, Captain, on the Columbretes, 329 ;
detennination of the height of .(Etna, 237.

Society Islands, the geology of, recom-
mended for investigation, 373.

Soconusco, the great volcano of, 263.

Soffioni, the, of Tuscany, 202.

Soil, frozen, in Northern Asia, 44 ; its ge-
ographical extension, 48.

Solfatara, the term inapplicable to the cra-
ter of Kilauea, 367.

Solo islands, character of the, 355.

Solomon's islands. See Sesarga.

Soufriere de la Guadeloupe, the, described,
395.

South Pacific Ocean, great number of vol-
canoes of the, 403. *

South Sea, volcanoes of the, 364; its isl
ands incon-ectly described as scattered
364; the tema "Gi'and Ocean" objected
to, 378.

Southern Asia, volcanoes of the islands of,
354.

Spain, extinct volcanoes of, 404. ■

Spartacus and his gladiators, their en-
campment on Vesuvius, 399.

Special results of observation in the do-
main of telluric phenomena, 5.

Springs, rise of temperature in, during
eartliquakes, 169 ; difficulty of classify-
ing into hot and cold, 178 ; method pro-
posed, 178; considerations on tempera-
ture, 180; heights at which they are
found, 183; boiling springs I'are, 189;
the Geyser and Strokkr, 190; gases,
193; Kallmann's classification, 196; va-
por and gas springs, salses, 198.

Stokes, on the density of the earth, 35.

Stone streams distinguished from lava
streams, 289.

Strabo, on the figure of the earth, 30; on
4ava, 216; on a double mode of produc-
tion of islands, 252.

Strokkr, the, of Iceland, described, 191.

Sti-omboli, description of, 243; periods of
its greatest activity, 244.

Strongyle, described by Polybius, 244.

Strzelecki, Count, on tlie basin of Kilauea,
368.

Styx, the waters of, 194; visits to their
source, 195.

Submarine volcano, presumed, in the At
lantic Ocean, 332 ; one observed in the
Pacific, near Chiloe, 272.

Subterranean noises, 171 ; attempts to de
termine the rate of their transmission
•172.

Sulphur Island, described by Captain Basil

Hall, 353.
Sulphureted hydrogen, question as to its

existence in certain fumaroles, 397.
Sumatra, the Giava Minore of Marco Polo,

355.
Sumbava, violent eruption of the volcano

of, 357.
Sun, magnetism of the, 84.
Sunda islands, volcanoes of the, 356, 357.
Swalahos, Mount, an extinct volcano, 390.

Taal, active volcano of, its singular po-
sition, 232 ; fcmall elevation, 233.

Table-land of South America, of Mexico,
and Thibet, 380; list of elevations, 382.

Tacora, Volcan de, 271.

Tafua, the peak of, 373.

Tahiti, the geology of, recommended for
investigation, 373.

Tajamulco, the volcano of, 262.

Taman, mud volcanoes of the peninsula of,
207.

Taranaki, a volcano in New Zealand, 372.

Taurus, elongated, the Thian-shan, includ-
ing the Himalayas, known as the, to the
Greeks, 405.

Tazenat, Gouffre de, an extinct volcano,
227.

Telica, Volcan de, described, 260.

Telluric phenomena, special results of ob-
servation in the domj,in of, 5.

Temboro, a volcano, its violent emption in
1815, 357.

Temperature, invariable, stratum of, 41;
mean annual, how determined in the
tropics, 42; observations of, in Mexico
and Peru, by Humboldt, 46 ; frozen soil
in Northern Asia, 44; Schergin's shaft,
45. See Interior of the Earth.

Temperature, rise of, in springs, during
eartliquakes, 169.

Teneriffe, the feldspar of the trachytes of,
427 ; notice of an eruption on, by Colum-

• bus, 444.

Ternate, violent eruptions andlava streams
in, 357.

Tertiaiy formations in Java, 281.

Thermal springs, their connection with
earthquakes, 170.

I'hian-schan, the volcanic mountain chain
of, 337 ; peculiarity of the position of the
volcano, 405; the chain known to the
Greeks as the elongated Taurus, 405.

Thibet, hot springs of, 189 ; geyser, 191.

Tierra del Fuego, volcanoes of, 280.

Timor, Pic of, fonnerly an ever-active vol-
cano, 368.

Tollo, the paadce hill of, 448.

Tonga Islanal^ active volcanoes of the, 369.

Toronto, magnetic observations at, 99.

Trachyte, origin of the word, 421; fre-
quently used in too confined a sense,
422 ; farther remai-ks, 437.

Tractus chalyboeliticos, what, 60.

Translatory movements in earthquakes,
167.

Trap, masses of. Sir R. Murchison on, 329,
451.

Trass fonnation, 225.

Trincheras, hot springs of, 180.

462

INDEX.

Tristan da Cunha, a volcanic island, 331.
Tshashtl Mountains, basaltic lavas of the,

3S9.
Tucapel, volcano of, 273.
Tupnngato, measurement of the peak of,

273.
Turbaco, the Volcancitos of, 204.
Tuscan Maremma, volcanic phenomena of

the, 202.
Typhon, fable of, 253.

Umnack, volcanic island of, 220.

Unalavquen, volcano of, 274.

Under currents of cold water in the tropics,
186.

United States scientific expeditions, bene-
fits to natural history from the, 378.

Uvillas or Uvinas, Volcan de, 271.

Val del Bove, on -^tna, remarkable infer-
ence regarding, 215.

Valleys of elevation, what, 193.

Vancouver, Mount, 389.

Vapor and gas springs, 212.

Variation charts, their early date, 55.

Vegetation, limit of, in Northern Asia, 45.

Vesuvius, phenomena of an eruption of, as
observed by Humboldt, 174 ; barometri-
cal measurements by the same, 235;
lengthened series of eruptions of, 398;
described by Strabo, 398 ; by Dio Cassius,
399; by Diodorus Siculus, 400; by Vi-
truvius, 400 ; difference of constitution
of the old and the recent lavas, 444 ; en-
campment of Spartacus and his gladia-
tors on, 399.

Vesuvius, valley furrows on the mantle of
the Somma of, 288.

Vidua, Count Carlo, his melancholy death,
357.

Vilcanoto, peak of, 279.

Villarica, Volcan de, 274.

Vincent, St., the volcano of, 394.

Vincent of Beauvais, his mention of the
magnetic needle, 54.

Virgenes, las, extinct volcanoes in Old Cal-
ifornia, 389.

Vitruvius, notice of Vesuvius by, 400.

Vivarais, extinct volcanoes of the, 263.

Volcan Viejo, a crater in Southern Peru.
271.

Volcancitos of Turbaco, described, 204.

Volcanic districts, different aspects pre-
sented by, 214.

Volcanic islands in the South Atlantic
Ocean, 332.

Volcanic reaction, bands of, 170.

Volcano, what intended imder the term,
by Humboldt, 272.

Volcano, the island styled " the holy isle
of Hephsestos," 244.

Volcanoes, considered according to the dif-
ference of their foi-mation and activity,
214; definite language of modern sci-
ence, 217 ; number of, on the earth, 393 ;
their great number in the Eastern Ar-
chipelago, 355 ; hypsometiy of, 235 ; lin-
ear arrangement of, 254 ; table of differ-
ences in structure and color of the mass
in certain, 432; the Mexican system,
264; sequence, latitude, and elevation,
26C ; particulars of the five groups of, in
the New Continent, 270; list of active,
263; geography of active, examined, 328;
geographical distribution of, 402; -open
in historical periods, list of, 330; semi-
volcanoes, 396.

Volcanoes of the moon, 418.

Vulcanicity, definition of, 158.

Wales, volcanic phenomena in, 329.

Walferdin on Artesian wells, 38.

Waltershausen, his classification of the
volcanoes of Iceland, 330; his remarks
on the period of recurrence of eruptions
in various volcanoes, 243; on the tra-
chytes of -^i^tna, 433.

Wilkes, Captain, commander of the Ameri-
can expedition, 102, 364.

Wislizenus, positions in Northwest Ameri-
ca ascertained by, 381.

Wisse, his observations of the eruptions of
the volcano of Sangai, 175, 251 ; his visit
to Pichincha, 231.

Yana-Urcu, a volcanic hill, 1S5.
Yanteles (Yntales), volcano of, 274.

Zapatera, extinct crater of the island,

259.
Zohron, the southern pole of the magnetic

needle, 54.
Zone of volcanic activity, 170.
Zuni, petrified forest near, 387.

THE END.

Illi