Full text of "The Valley of Ten Thousand Smokes"

See other formats

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project
to make the world's books discoverable online.

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject
to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books
are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover.

Marks, notations and other marginalia present in the original volume will appear in this file - a reminder of this book's long journey from the
publisher to a library and finally to you.

Usage guidelines

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we have taken steps to
prevent abuse by commercial parties, including placing technical restrictions on automated querying.

We also ask that you:

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for
personal, non-commercial purposes.

+ Refrain from automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the
use of public domain materials for these purposes and may be able to help.

+ Maintain attribution The Google "watermark" you see on each file is essential for informing people about this project and helping them find
additional materials through Google Book Search. Please do not remove it.

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner
anywhere in the world. Copyright infringement liability can be quite severe.

About Google Book Search

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web

at http : //books . google . com/|

The Library

MusentD of Comparative Zoology

Harvard University

Digitized by

Google

Digitized by

Google

115

(?7

Digitized by

Google

THE VALLEY OF
TEN THOUSAND SMOKES

GRIGGS

Digitized by

Google

Digitized by

Google

Digitized by

Google

Digitized by

Google

Digitized by

Google

Digitized by

Google

Digitized by

Google

Digitized by

Google

The Valley of
Ten Thousand Smokes

By
ROBERT F. GRIGGS, Ph.D.

With 9 Maps and 233 Illustrations

WASraNGTON

The National Geographic Society
1922

Digitized by

Google

115
Gl

By The National Geographic Society

Published, December, 1922

MCZ

(u ::.: ■?93

Digitized by

Google

To the Inspiration
of my life and work —

My Wife

Digitized by

Google

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

TABLE OF CONTENTS

FOREWORD BY GILBERT GROSVENOR XV

CHAPTER l: THE ERUPTION AND ITS SIGNIFICANCE.

Eruption felt over all northwestern America — A similar eruption in New York
City would be seen in Albany, heard in Chicago, tarnish brass in Denver, prostrate
Philadelphia, destroy Greater New York — Crater largest in the world — No lives
lost — Geological bearings; eruption of a new type; aid in understanding minerali-
zation; may throw light on mechanism of volcanoes — Revegetation has important
agricultural bearings — Discovery of one of the greatest wonders of the natural
world, the Valley of Ten Thousand Smokes — Safe for tourist travel — Can be
made readily accessible.

CHAPTER II : THE ERUPTION AS EXPERIENCED AT KODIAK .

Introductory — Premonitory symptoms unknown; ashfall began without warn-
ing — Kodiak not situated in volcanic district — Ignorance of seat of eruption —
Darkness which obscured lanterns held at arm's length — People collected in ware-
house — Rescue of crew of tug; people taken aboard the Manning; suffering from
overcrowding — Tribute to bravery of inhabitants — Suffering due more to terror
that to danger.

CHAPTER III: RECORDS OF THE ERUPTION FROM OTHER LOCALITIES 15

Identification of the volcano by steamer Dora; other vents supposed to have
been responsible; final determination of seat of eruption not made until 191 5 —
Reports of eyewitnesses from Katmai Village — Statement of chief of Savonoski na-
tives — Experiences of people at Kaflia Bay, where the ashfall was three feet — Pre-
liminary symptoms as observed at outlying stations — Maximum violence for 60
hours — Gradually subsiding activity continued all summer — Records of earth-
quakes accompanying eruption; records of explosions as heard at distant stations;
explosions not heard at Kodiak, perhaps because of dust screen — Sulphuric acid
rains destroy vegetation at Seward, La Touche and Cordova, more than 300 miles
away — Brass tarnished and clothes destroyed in the Puget Sound region, 1,500
miles away — Ashfall a foot deep or more over an area one and one-half times as
large as Delaware — Dust fell 1,500 miles away — Records from Puget Sound —
Ejecta from eruption totalled over six cubic miles, a quantity equivalent to the
output of the stone crushers of the United States for over 500 years.

CHAPTER IV: CLIMATIC EFFECTS OF THE ERUPTION 33

Haze from volcano held responsible for cold summer of 1912; lowest point ever
reached in terrestrial temperatures followed eruptions of 1 783-1 785; the "dry
fog" of 1783; "the year without a summer*' followed eruption of Tomboro, 1815
— Haze from Kral^toa caused remarkable sunsets for two years; haze from Kat-
mai appeared in Wisconsin before the eruption was over, in Virginia two days later;
peculiar haze persisted all summer in Algeria — Interference with astronomical
observations — Dust cloud absorbed 10 per cent of the sun's heat; whole North-
em Hemisphere cooler because of eruption — Slight changes in mean tempera-
tures of great practical importance — Volcanic dust shown to be capable of produc-
ing glacial climates — Water liberated in eruptions probably of vast importance
in economy of nature; all life perhaps dependent on continuance of volcanic action.

Digitized by

Google

viii THE VALLEY OF TEN THOUSAND SMOKES

CHAPTER V: THE RECOVERY OF VEGETATION AT KODIAK 45

Remarkable recovery of vegetation; not due to any fertilizer in the ash; ash
forms a very poor soil; composition similar to pulverized granite; beneficial action
due to effect as a mulch — Revegetation result of recovery of old plants; plants
resurrected after three years' burial — Seedlings slow to start; revegetation re-
tarded by sand blast — Importance of horsetail as a soil binder — Bizarre ap-
pearance of ash-laden trees overgrown with moss — Network of moss on forest
floor — Seedlings starting under protection of trees — Ash rapidly eroded; earth-
worms active in mixing it with old soil — Little damage to fields.

CHAPTER VI : THE ALASKAN VOLCANOES 57

Aleutian volcanic chain of Asiatic, rather than American affinity — Volcanic
festoons fringing Asia, from East Indies to Kamchatka — The Aleutian chain,
traversing mountain plateaus and ocean deeps, evidently a more fundamental
feature than ordinary geologic structures — Cones of Iliamna and Redoubt —
Explosions of Augustine — The volcanoes of Unimak, Shishaldin and Pogromni
— Birth of a new island, Bogoslof ; upheaval of Grewingk, companion to old
Bogoslof; formation of Metcalf Peak and McCulloch Peak; eruption of 1909;
present condition of Bogoslof — Importance of further study of Aleutian Vol-
canoes.

CHAPTER VII : CARRYING THE WORK INTO THE VOLCANIC AREA — GEOGRAPHY —

CHRONOLOGY — PERSONNEL 67

Remoteness of the region — Dangers of navigation — General sketch of geo-
graphic features — Broad valleys key to geography of district; Katmai Valley;
Katmai Canyon; Second Katmai Valley; X-shaped valley on Naknek slope —
Lesser geographic features — Geographic order of description adopted in prefer-
ence to chronological narrative — Previous attempts to reach the crater — Expe-
dition of 1915 — Expedition of 1916; Walter Metrokin, the one-armed bear
hunter of Kodiak; area traversed in 1916 — Organization of expedition of 1917;
ground covered; creation of Katmai National Monument — Considering pack
horses and airplanes for transportation — All supplies carried in on man-back —
Only the most concentrated foods could be used — Sending in the expedition of
1918 from Bering Sea — Objectives of expedition of 1919 — Cooperation by
Geophysical Laboratory of Carnegie Institution — Organization in 1919; person-
nel.

CHAPTER VIII: THE FIRST JOURNEY TO THE VOLCANO DEVASTATION OF

KATMAI VALLEY 8I

Weird appearance of desolated shores — Dread of casting loose in such an out-
landish country — Difficulty in choosing a landing — Supply of drinking water
assured — Sticky mud everywhere — Luxuriant poplar groves — Comparative
damage to different plants: poplars growing from dormant buds; alder and birch
exterminated; willows suffered less than other trees because of capacity for forming
adventitious roots — Exploring in a dust storm — Starting up the valley —
First view of volcanoes — Discovery of a new volcano, Mt. Martin — Climbing
slides of loose ash — Fording Soluka Creek, a mile of quicksand — Straining
drinking water out of pumice — Pressing forward for fear of bad weather; fear of
caverns formed by snow melting away beneath the ash — Upper Katmai Valley;
awe-inspiring desolation — All animal life destroyed — Comparative verdure of
lower valley — Surprising alDsence of damage by fire — Progress up the valley;
stopped by Katmai Canyon — Katmai mudflow — First crossing Katmai River
with aid of poles; later experiences in crossing; use of ropes in fording; folding
boat with a "G" string — Waiting for sight of volcanoes through the clouds —
Second new volcano, named Trident — First sight of Katmai — Volcanoes obvi-
ously piles of lava rather than cinder cones.

CHAPTER IX: THE KATMAI FLOOD IO3

One of the greatest of recorded floods, in which the expedition of 191 5 narrowly
missed disaster — Katmai Bay full of wreckage brought down by flood — Excep-
tional height of flood evidenced by condition of Katmai Village; houses entirely
submerged — River formerly used by schooners, choked with ash — A maze of

Digitized by

Google

CONTENTS IX

quicksands five miles uide — Speculations as to the cause of the flood; could not
have been tidal; confined to Katmai River — Forests uprooted and heeled over
before the waters — Trees sheared off at ground for miles at a stretch — A 250,000-
pound bowlder picked up by the flood — Temporary Niagara over obstructing
ridge — A river whose bottom was higher than its banks — Freak of the flood
— Thankful for the lucky chance that had delayed our landing until after the
flood — Later exploration shows that flood was caused by downpour of water im-
pounded by landslide — Slide occurred at a natural dam site; impounded lake of
950 acres — Probable condition of dam — Katmai Canyon filled many feet deep
with cobbles — Thirty-five times as great as Johnstown Flood — Destructive
power a million times that of a Mississippi flood.

CHAPTER X: THE CANYON AND SECOND VALLEY OF THE KATMAI I23

Habitual underestimation of sizes of physical features; grandeur of Katmai
Canyon not appreciated at first — Beautiful waterfalls near foot of canyon —
Katmai Canyon almost as deep as Grand Canyon — Prospect Point — Descent
into canyon — Contrast with other canyons of district; pre-glacial rather than
post-glacial — Proof that last lava flows from Katmai were post-glacial — Gravel
banks set rolling by wind — Second Valley of the Katmai — Beaches of former
*' flood lake" — A possible explanation of the Second Valley — Curious physiog-
raphy; numerous regular conical mounds; all wood buried in debris; drainage very
irregular; bowlders in river bed sharp and angular — Noisy Mountain in action:
rock falls; mud avalanches, intermediate between landslides and overloaded
streams; key to other unusual physiographic features — Upper portion of Second
Valley; Princess Glacier — Passes at head of Second Valley — Attempt to find
route to sea — Discovery of hot spring — Supperless night in the rain.

CHAPTER XI: THE MAGEIK LANDSLIDE 135

Head of Martin Creek strewn with rocks, some as big as a house — Evidence
that they could not have been thrown out from a volcano — Resemblances to glacial
deposit — Broken bushes intermingled with debris — M uch peat in terminal
portion — Lateral edges show slumping back of material from highest level
reached — Terrane forked halfway down — Original plants still growing on
bowlders — Indications as to date; "infantile" physiography; buried fragments
of wood not decayed; covered with ash; contemporaneous with eruption — Evi-
dently some sort of landslide — Surface studded with remarkable cones; geological
structure favorable for landslides — Dimensions of slide — Surprising liquidity
manifested by slide — Description of similar slide in motion — Mageik Slide com-
pared with others: Elm, Frank, Rossberg, Katmai, Noisy Mountain, Falling
Mountain; the greatest recorded slides: Gohna and Bandai San — Rank of
Mageik Slide.

CHAPTER XII: KATMAI VALLEY IN LATER YEARS 147

General appearance of Katmai Valley in 191 6 similar to that in 191 5, but much
changed in detail — Instability the key to the abnormal physiography — Shifting
surface most striking characteristic of devastated country; cut and fill at Observa-
tion Mountain; frequent avalanches from the overloaded mountains — Snowdrifts
buried by ash persist for years — Wind erosion of ash slopes; bark of trees cut away
by sand blast — Fickle Creek shifts its course 1,000 feet in a year — Antics of
pumice-choked streams — Caricatures of aggrading rivers building fans, dikes and
deltas — Observations on depth and character of ash deposits — Irregularities in
ashfall due to variable winds — Three major strata, with a fourth near the volcano
and a cap-layer of fine mud — Plants that lay dormant for three years until un-
covered by flood wash — Seedlings first started in numbers in 1916 — Seeds,
scattered mostly by wind, fail to lodge on pumice flats — Lupines the most im-
portant pioneer plants; abundance of root tubercles; a few plants growing in situa-
tions extremely poor in nitrogenous compounds — A few small fish on the quick-
sands — Trout in a brook tributary to Katmai River — A perverted run of
salmon — Remarkable increase in insects in 191 6 — Flies dying of starvation —
Insects mostly predacious or parasitic — Speculations as to source of insects —
First mammals carnivorous — Notable increase of herbivorous animals in 191 9 —
Return of Herbivora not correlated with increased food supply — Similar condi-
tions observed at Kodiak.

Digitized by

Google

X THE VALLEY OF TEN THOUSAND SMOKES

CHAPTER XIII : KATMAI CRATER I67

Reconnoitering: the lower slopes — Upper slopes plastered with soft, slippery
mud — Scenery from upper levels superb; clouding up as rim is approached —
Crater rim a knife edge of loose unstable ash, the same layers as those below —
Waiting for clouds to lift — First view of crater; an abyss of inconceivable size —
Second ascent — Ash-covered glaciers — Magnitudes too great for senses to
grasp — Second view of crater — Glaciers on crater rim — Ascents in 191 7 —
Comments of new members — Perfect proportions of crater — No changes during
four years — Weird effect as crater is walled in by clouds — Dimensions of crater;
would hold water enough to supply New York City over four years; skyscrapers
of New York would not begin to fill crater — The largest active crater in the
world — Comparison with Kilauea.

CHAPTER XIV: MOUNT MARTIN I8I

Inaccessibility by reason of position and cloudiness — First attempt to climb —
Mount Martin a subordinate lava cone on flank of Mageik — Despite present
activity, Martin seems not to have thrown out ash in great eruption — All vegeta-
tion in vicinity kept down by acid rain — Second attempt — Third climb, from
Baked Mountain — "A real glacier** — Waiting for a break in the clouds —
Climbing a 60® dirt slope at the end — Descending into the crater in thick steam —
Returning — Fear of being caught on glacier by a cloud.

CHAPTER XV : DISCOVERY OF THE TEN THOUSAND SMOKES 1 89

Reconnaissance up Mageik Creek; the Mageik '* mudflow ** — Ascent to Katmai
Pass — Discovery of first fumarole in the pass — First sight of the Valley of Ten
Thousand Smokes — Character of the vents — Entrance to the main valley —
First sight of Novarupta — An interpretation of the fumaroles — Return to
camp — Reflections on night of discovery.

CHAPTER XVI : EXPLORATIONS IN THE TEN THOUSAND SMOKES IN I917 I97

Impatience to reach the Valley again — No change in the Smokes in a year's
interval — No probability of recurrence of explosive condition — Effect of Valley
on uneducated native — Crossing the mountains against a hail of pumice —
Comments of new members of party — Overawed by wonderful valley — Fear of
cave-ins and fumes — Learning to travel safely — Effect on shoes — Cooking at a
fumarole — Temperatures too high for our instruments — A steam- heated tent

— Therapeutic possibilities of the Valley — Katmai Pass a notorious weather
breeder — A few signs of animal life — Surveying the Valley — Collecting
gases for analysis — Practically all plant life destroyed — Testimony of asso-
ciates: L. G. Folsom; Paul R. Hagelbarger; J. D. Say re; Clarence F. Maynard;
James S. Hine; J. W. Shipley; D. B. Church.

CHAPTER XVIi: THE TEN THOUSAND SMOKES IN I919 219

Fumaroles unaffected by seasonal variations in streams — Some indications of
slackened activity — Melting lead and zinc in the fumaroles — Kindling a fire by
plunging stick into water — A red-hot fumarole — Valley becoming green in
spots with moss and algae — Bacteria invading soil — Bears investigate Valley;
one of them digs a new fumarole of his own — Baked Mountain Camp — Acid
from cookstove eats the pots — Cookstove on roof of a bridged-over fissure;
any desired temperature available — Com pone k la fumarole — Frying bacon
in superheated steam — Varied menus at Baked Mountain — Trouble with storms

— Protective measures — Final wreck of Baked Mountain Camp; apparatus and
sleeping tents torn up; "grub" tent begins to weaken; extra tarpaulin tears away;
hail of pumice; holding down the last comer of the tent; waiting for dawn; gale
picks me clear up off ground; fleeing before the wind; relief at Ukak; camp moved
to Mount Cerberus.

Digitized by

Google

CONTENTS XI

CHAPTER XVIII : A DESCRIPTIVE ACCOUNT OF THE VALLEY OF TEN THOUSAND

SMOKES 233

Valley shaped like great Y — Underlain by sedimentary, not by volcanic rocks —
Fissure fumaroles — Craters — Steaming mud areas — Steam rising through
undisturbed sandstone strata — Hot springs — Mud pots — The new volcano,
Novarupta — Falling Mountain; cataclysm began with great slide from Falling
Mountain; fumarole action on Falling Mountain — Baked and Broken Moun-
tains — Fissure Lake; warm water from snowdrifts — Drainage water nearly all
evaporated by heat of Valley — Bewildering complex of odors — Chemistry of
fumaroles: all emanations 99 per cent steam; no relation between temperature
and composition of gases, — ammonia — unparalleled abundance of fluorine —
sulphur — arsenic — corundum — potassium alum — iron; pyrite; Venetian red;
magnetite — petroleum-like emanations — Wonderfully colored deposits; shoes
daubed with brilliant pigments — Color photographs.

CHAPTER XIX : THE GREAT HOT SANDFLOW OF THE VALLEY 253

The great mass of tuff filling the Valley — Valley floor remarkably flat — Vol-
ume of tuff about a cubic mile — Valley surrounded by conspicuous "high water
mark" — Comparison with frozen pond — Fissures suggest sagging of original
surface — Proof that activity of Valley began before explosion of Katmai —
Relation to forest showed that tuff originally flowed like a liquid — Every stick
touched by sandflow reduced to charcoal — Search for origin of sandflow — Hot
sand must have broken through floor of Valley itself — Sandflow an eruption of
new type; gives volcanologist new viewpoint.

CHAPTER XX: ANTECEDENT CONDITIONS 265

Little information concerning antecedent conditions available — Petrof 's de-
scription of Katmai district; Katmai Pass; Katmai Village — Importance of
Katmai Pass revived by discovery of Nome — Spurr's account of original condi-
tion of country and his experience in crossing Katmai Pass — Katmai Pass prob-
ably in unstable condition before eruption — Difficulty of judging weather across
the pass — Travelers give no account of Katmai, although it must have been
dominant peak of district — Only information concerning Katmai derived from
U. S. Coast Survey Chart — Comparison with present stub.

CHAPTER XXI : A HYPOTHETICAL CHRONICLE OF THE ERUPTION 277

An account pieced together from evidence left behind — In the beginning a
beautiful green valley — Great landslide from Falling Mountain opens the cata-
clysm — Valley floor cracks open and seething magma rushes out — Impenetrable
black clouds probably veil the scene — Lava froths up into fiery foam — Roaring
fires sweep up the mountain side; misery of entrapped animals quickly ended —
Fumes must have been indescribable — Valley becomes seething fiery torrent —
An analogy from Hawaii — Red-hot ash rolls down Valley like liquid — Ex-
plosion of Novarupta — Openmg of Katmai and the other volcanoes — Summary
of events of eruption.

CHAPTER XXII : WHAT BECAME OF THE TOP OF KATMAI? 287

Interpretation of subterranean disturbances the real problem of the volcanol-
ogist — Rock of original mountain disappeared utterly — Rock of old mountain
distinguishable from new lava at a glance — Could the old mountain top have
been swallowed up in the bowels of the earth? No evidence of subsidence —
Much of old mountain eaten away by new magma — All stages in penetration of
old rock by new magma shown in fragments ejected; ash at Kodiak carries many
crystals from old mountain — Much of old rock permeated with gases from new
magma — Temperature of new magma low, less than 1,000® — All the ash could
be packed into crater — A portion of old mountain probably sank into depths —
Old rock mixed with lava of Novarupta may have been derived from Katmai —
Explosions appear to have started at summit of mountain, not in depths — Lava
formed molten lake in crater; frozen foam from lake surface; chilled scum —
Magma did not explode immediately on reaching surface; explosions not a simple
physical process like popping of carbonated liquors — What is the character of the
magma chamber beneath the region? Sill or batholith? — Decision on theoreti-
cal questions best deferred.

Digitized by

Google

xii THE VALLEY OF TEN THOUSAND SMOKES

CHAPTER XXIII : GEOGRAPHICAL AND BIOLOGICAL OBSERVATIONS IN THE SUR-
ROUNDING COUNTRY 305

Northern end of park not desolated by eruption — Coniferous forest border
migrating westward — Naknek Lake extremely beautiful — Majestic mountains
rising out of lake — View from Mount Katolinat — Discovery of Lakes Grosve-
nor, Coville and Brooks — Resemblance to Finger Lakes of central New York —
Trout 32 inches long — One of the greatest of spawning grounds for sockeye
salmon — Salmon Falls — Enormous numbers ascending — Remarkable homing
instinct of salmon — Countless waterfowl on lakes and ponds — Professor Hine's
collections of birds and insects — Big game returning — Alaskan brown bear, the
largest of carnivorous animals, abundant — Necessity for protection — Bear
trails everywhere — Bears inspect our trails — Going into interior unarmed,
despite bears — Feeding habits — First sight of bear — Encounter with sleeping
bear — Going after Bruin with a pistol — Problem of finding a harbor — Possibil-
ities of Kukak and Kinak Bays — Discovery of ideal harbor, named for National
Geographic Society — Geographic Harbor fitting entrance to Valley of Ten
Thousand Smokes — Feasible route for road into interior — Strategic importance
of road across Alaska Peninsula — Sixty miles of automobile road would open up
whole district — More accessible than Yellowstone at time of its discovery —
Predictions of the future.

PRESIDENTIAL PROCLAMATION ESTABLISHING THE KATMAI NATIONAL MONU-
MENT 331

ACKNOWLEDGMENT 333

INDEX 335

Digitized by

Google

COLOR PLATES

I.
II.
III.

IV.
V.

VI.

VII.

VIII.

IX.

XI.

XII.

XIII.

XIV.

XV.

XVI.

KODIAK FROM THE CHANNEL FACING PAGE 80

A CANYON IN THE MUDFLOW AT THE HEAD OF KATMAI VALLEY '* " 8 1

ALPINE GLOW ON MOUNT MARTIN *' '* 96

FUMAROLES AT THE FOOT OF FALLING MOUNTAIN '' " 97

MOUNT MAGEIK, WITH CERBERUS CROUCHING AT ITS FOOT,

STANDS GUARD AT THE HEAD OF THE VALLEY *' ** 1 92

THE OVEN WHERE CORN BREAD WAS BAKED BY NATURE'S

FIRES ** " 193

BRILLIANT INCRUSTATIONS FROM THE FUMAROLES *' " 208

A ROW OF GORGEOUS "STOVE HOLES" " " 209

THE CAVERNOUS MOUTH OF ONE OF THE FUMAROLES '* " 224

A RED-BURNED FISSURE ON THE WEST SIDE OF THE VALLEY,

LOOKING TOWARD THE GLACIERS OF MAGEIK *' '* 225

MOSAICS OF ALUM DEPOSITED IN AN AREA OF DIFFUSE

ACTIVITY " '* 240

THE GAUDY MUDS OF THE VALLEY MAKE GOOD PIGMENTS FOR

A PAINTER *' ** 241

HARD BAKED CRUST AROUND A SMALL FUMAROLE " " 288

THE INTERIOR OF A FUMAROLE LAID BARE " " 289

TAKING THE TEMPERATURE OF MOTHER EARTH " '* 304

PEACEFUL FOREST AND PELLUCID LAKE WITHIN A DOZEN

MILES OF THE SMOKES '' " 305

MAPS

BASE OF ALASKA PENINSULA AND ADJOINING ISLANDS IN FRONT COVER

ASHFALL FROM THE ERUPTION OF KATMAI FACING PAGE I

VOLCANIC AXES AND STRUCTURE LINES IN EASTERN ASIA AND ALASKA ** *' 56

VOLCANOES AND HOT SPRINGS OF THE ALEUTIAN REGION *' " 60

CENTRAL PORTION OF THE KATMAI NATIONAL MONUMENT '' 68

MOUNT KATMAI BEFORE THE ERUPTION ** 27O

MOUNT KATMAI AS IT NOW STANDS " 270

A RECONSTRUCTION OF MOUNT KATMAI AS IT WAS BEFORE THE

ERUPTION " 272

TOPOGRAPHIC MAP OF THE KATMAI NATIONAL MONUMENT IN BACK COVER

Digitized by

Google

Digitized by

Google

FOREWORD

The following pages tell the story of the discovery by
the National Geographic Society expeditions of one of the
wonders of the world, **The Valley of Ten Thousand
Smokes," and the prolonged study of one of the most
gigantic volcanic eruptions in history.

This remarkable exploration of Mt. Katmai by Pro-
fessor Griggs and his associates, which continued for five
years, was financed by the 750,000 members of the National
Geographic Society, each of whom, millionaire and college
professor, captain of industry and clerk, had an equal
share in its support. Every member of the organization
may thus derive considerable satisfaction that he or she
has assisted to bring about such important additions to
our knowledge of the young and active planet upon which
we live.

To Dr. Griggs, the leader of the National Geographic
Society's expeditions to the Katmai region, and to each of
his able assistants, the congratulations and thanks of the
Society's membership are herewith extended. Their in-
trepid, intelligent and faithful services achieved results of
historic value and reflect great credit on all participants.

GILBERT GROSVENOR

President, National Geographic Society

Digitized by

Google

Digitized by

Google

THE ERUPTION AND ITS SIGNIFICANCE

The Story of the Valley of Ten Thou-
sand Smokes begins with the great vol-
canic eruption of 1912 which brought it
into existence. Exploration, undertaken
primarily for the scientific study of the
effects of the ashfall, led to the discovery
of one of the great marvels of the natural
world, a place unseen and unsuspected
by white man and native alike until
entered by a National Geographic Society
Expedition.

Over an area of fifty square miles the
ground is all broken open, and hot gases
from the molten material below are even
now everywhere pouring out, forming the
several millions of fumaroles that con-
stitute the Valley of Ten Thousand
Smokes.

No one suspected that such a cata-
clysm was in progress until the 6th of
June, when volcanic ash suddenly began
falling over all of northwestern America.
Kodiak, in southwestern Alaska, was
reported as the chief sufferer. But as
Kodiak is one hundred miles from any
volcano, the actual seat of the disturb-
ance was uncertain. Katmai* volcano
was mentioned in the first dispatches,
yet it was not settled for a long time
that it was Katmai which had blown up.

The eruption was one of the dozen
greatest of historic times, yet it attracted
comparatively little attention, because
of the remoteness of the region concerned.
Katmai was, in fact, merely a name on
the map, unknown except to local pilots
along the unfrequented coast of Shelikof
Strait. It may aid in giving some con-
ception of the magnitude of the outbreak
if we can picture to ourselves the effects
a similar eruption would produce if
centered in a better known region, say
New York City.

^ Pronounced K&t-mai' ; & as in cat, ai as in aisle.

* More properly termed magma, for it had not

given up its surcharge of gases. Strictly speak-

If such an eruption should occur on
Manhattan Island, the column of steam
would be conspicuous as far as Albany.
The sounds of the explosions would be
plainly audible in Chicago. The fumes
would sweep over all the states east of
the Rocky Mountains. In Denver they
would tarnish exposed brass, and even
linen hung out on the line to dry would
be so eaten by the sulphuric acid content
as to fall to pieces on the ironing board.
As far away as Toronto the acid rain-
drops would cause stinging burns wher-
ever they fell on face or hands.

Ash would accumulate in Philadelphia
a foot deep. To add to the terrors of the
catastrophe, that city would grope for
sixty hours in total darkness — darkness
blacker than anything imaginable, so
thick that a lantern held at arm's length
could not be seen.

As for the horrors that would be en-
acted along the lower Hudson, no de-
tailed picture may be drawn. There
would be no occasion for rescue work,
for there would be no survivors. The
whole of Manhattan Island, and an
equal area besides, would open in great
yawning chasms, and fiery fountains of
molten lava* would issue from every
crack.

This, disrupted by the escaping gases,
would be changed into red hot sand,
which, consuming everything it touched,
would run like wildfire through the town.
The flow of incandescent sand would
effectually destroy all evidence of the
former city. In its deepest parts the
near-molten sand would probably over-
top the tallest skyscrapers, though the
tip of the Woolworth tower might pro-
trude, if its steel supports could endure
the fiery furnace surrounding them. It
is doubtful, indeed, if there would be any

ing, there were no lava flows in connection with
the eruption of Katmai.

Digitized by

Google

•s

I •-
I 1

A c/) c

>. o

s I

^ §

S .2

U 3

to .=

o •>

2 "S
" I

3
O

Digitized by

Google

THE ERUPTION AND ITS SIGNIFICANCE

considerable ruins left behind to mark
the site of that great city.

In any case, for several months no one
could approach the place nearer than
Paterson, New Jersey. In addition to
this, a hole would have been blown in
the ground big enough to hold all the
buildings of Greater New York several
times over.

With such a frightful comparison in
mind, the reader will be prepared for an
account of suffering and devastation un-
paralleled in the annals of volcanic dis-
asters. But of this he may be relieved
at the outset. Terror of the most fright-
ful sort there undoubtedly was; but of
actual destruction of life there was none:
not a single person perished in the erup-
tion.

Mankind may thus be congratulated
on having one of the greatest opportu-
nities ever afforded to study the manifold
problems of volcanism, without having
to pay the terrible toll of human misery
that usually accompanies such cata-
clysms.

The flow of incandescent sand, alluded
to above, constitutes an outburst of a
new type, different from all recorded
eruptions. The recognition that out-
pours of this nature probably occurred in
the past may help to explain many de-
posits now puzzling the geologists.

The phenomena here occurring at the
surface suggest what must have taken
place in the formation of many a heavily
mineralized region. A study of the
fumaroles of this area bids fair to throw
light on the formation of metal-bearing
veins in the depths of the earth. If we
can gain a satisfactory knowledge of the
activity here going on, it will help greatly
toward an understanding of the deposi-
tion of metalliferous ores generally.

The explosions, too, constitute a prob-
lem of great interest and import. Simi-
lar outbursts in the past have left little
clue to the processes responsible for the
liberation of the enormous amounts of
energy involved. Up to the present no
carefully reasoned theory of volcanic
explosions has been proposed. Here

also Katmai, by reason of certain acci-
dental circumstances, gives an unusual
insight into the events preceding the
great explosions, and seems likely to per-
mit us to come nearer to an understand-
ing of them.

No less important than the questions
of volcanism are the problems of revegeta-
tion of the devastated region, both in
their practical and theoretical bearings.
The extent of the area covered by ash
gives an unprecedented opportunity for
study of the means by which £tsh is con-
verted into a soil fit for the growth of
higher plants. The processes by which
such a mass of raw mineral debris is
transformed into a soil rich in organic
matter have the largest bearing on cer-
tain problems of practical agriculture.

The Valley of Ten Thousand Smokes
is far beyond other volcanic districts in
the number of its vents and in the quan-
tity of steam given off. It is well within
the truth to say that, except during
periods of dangerous eruption, the sum
total of the emanations from all the other
volcanoes of the New World, from the
Aleutians to Patagonia, is less than is
given off continuously, within the range
of one's vision, from the Valley of Ten
Thousand Smokes.

To find a parallel to this Valley we
must search the records of geology, for
here was a volcanic outburst such as the
geologist finds recorded in the rocks of
past ages but has never before had
opportunity to observe in the world of
the present.

The violently explosive eruptions char-
acteristic of the*opening stages of the
activity have, moreover, entirely passed
away. The innumerable vents function
as a gigantic safety valve, through which
the titanic forces gathered beneath find
avenue for harmless escape.

There is, therefore, no likelihood of
return to a dangerous condition. The
observer may indeed travel in security
anywhere among the volcanoes, with no
fear of a sudden change in conditions.
He may even with perfect safety peer
over the edges of the orifices from which

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by W. J. Erskine
THE END OF THE ASH FALL AT KODIAK

At first the ash was comparatively soft and fluffy. Later it settled down into a hard compact layer

ten inches thick.

Photograph by W. J. Erskine
ASH PILE FROM A ROOF, KODIAK

It looks very much like snow as it lies on the ground but, being in reality pulverized rock of great

weight, it did much damage.

Digitized by

Google

THE ERUPTION AND ITS SIGNIFICANCE

Photograph by W. J. Erskine
HOUSE IN KODIAK HALF FILLED WITH ASH BY AVALANCHE FROM THE HILLS
The level of the ash within may be seen through the window at the left.

the gases issue, and look fiar down into
the depths through swirling currents of
superheated steam that rush forth at his
very elbow.

So many natural phenomena of inter-
est to the general public, as well as to the
scientist, are exhibited in this remarkable
region that President Wilson paused in the
midst of the World War to set aside the
whole district as the Katmai National

Monument, which embraces over a
million acres.

Finally, a new harbor, discovered by
the expeditions, permits the opening of
the region to travel. A few miles of road
connecting the harbor with the volcanic
area will render the Valley of Ten
Thousand Smokes as readily accessible
as the geysers of Yellowstone National
Park.

Digitized by

Google

e
o

z t2

2 «

S g

s a

< S
uT a

o •&

• 3

Digitized by

Google

THE ERUPTION AS EXPERIENCED AT KODIAK

The chronicle of the eruption does not
begin with a description of the volcano
and the premonitory symptoms that
must have preceded the great explosions,
for all these passed unnoticed by reason
of the remoteness of the region from
human habitations.

No one suspected that anything un-
usual was in prospect until the full force
of the explosions had broken loose.
Many hours before any notice was taken
of the eruption, disturbances must have
occurred which, in a thickly populated
territory, would have been heralded
around the world in flaming headlines.
Not until the lives and property of the
people of Kodiak, a hundred miles from
the volcano, were threatened, was it
realized that a great eruption was in
progress.

ASH BEGAN FALLING WITHOUT WARNING

The weather was fair; there was a
strong northwest breeze, which in that
region betokens a high barometer, and
there were none of those signs of impend-
ing disaster which one is wont to associate
with volcanic eruptions. Suddenly, and
without other warning than the gradual
approach of a peculiar dark cloud from
the northwest, gray ash began falling at
Kodiak at about five o'clock in the
afternoon, Thursday, June 6th.

The only abnormal manifestations in
advance of the falling ash were thunder
and lightning; but, as these are almost
unknown phenomena on the rainy north-
west coast of America, their occurrence
excited more than usual comment, es-
pecially among the natives, many of
whom had never before heard thunder.
Not even thunder and lightning were
observed, however, at the neighboring
town of Afognak.

In the beginning, the falling ash was
hailed as a great curiosity, and the people

carefully swept up samples for mementoes
of an event such as they had never wit-
nessed before and were not likely to see
again.^

KODIAK NOT IN A VOLCANIC DISTRICT

For it must be remembered that
Kodiak is not situated in a volcanic dis-
trict. There are no volcanic formations
anywhere on the island ; and the nearest
of the dormant volcanoes of the distant
mainland was a hundred miles away,
which under ordinary conditions is an
altogether safe distance.

Kodiak was, therefore, almost as much
surprised as New York would be if one
of the extinct volcanoes of Connecticut
should suddenly awake and shower it
with ash.

Indeed, the people of Kodiak had but
little more cause to expect serious dam-
age from an eruption than has New York ;
there was no known reason for alarm.
Neither did they possess that careless
confidence born of familiarity with vol-
canoes and frequent experience with
minor eruptions, which often lulls the
inhabitants of a volcanic district into the
belief that nothing really serious is going
to happen, until they fail to heed warn-
ings which everyone agrees afterward
should have been sufficient to drive the
people out of the danger zone.

Beginning gradually, the ashfall in-
creased until it was regarded as ** heavy"
at 6 p.m. By 6.30 it had become so
thick as to blot out the daylight, cover-
ing the country with total darkness at a
season when, in that latitude, daylight is
almost continuous for the whole twenty-
four hours.

The ash steadily accumulated through
the night, but the fall began to slacken

* There is, however, a record that in March,
1867, half an inch of volcanic ash fell at Kodiak
from some distant eruption (see page 62).

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

at about 3, and ceased at 9.10, on the
morning of the 7th.

This fall constitutes the bottom layer
of relatively coarse gray ash, as it now
lies on the ground. As it originally fell,
this stratum was about five inches deep
at Kodiak, but it has since settled down
to about four-fifths of that thickness.

During the night the people retained
their composure to a large degree; and
when, in the morning, the ashfall gradu-
ally diminished and ceased, they believed
that it was over. They had, however,
no inkling as to the seat of the eruption.

Then, and for many days there-
after, they were completely isolated
by the failure of the wireless serv-
ice. The Kodiak Radio Station
on Wood Island had been struck
by lightning and burned down at
some time during the eruption.
The wireless apparatus of the
United States Revenue Cutter
Mannings which was coaling at
Kodiak at the time, was useless
because of the electrical condition
of the air, which was excessively
charged with ''static.**

The report of Captain K. W.
Perry of the Manning furnishes at
once the most vivid and most
authoritative account of the experi-
ences of those who went through
those trying days together. From
it, all the following quotations not
otherwise credited are taken.*

TERROR BEGAN WITH THE SECOND
FALL

"All streams and wells had now
become choked, about five inches
of ash having fallen, and water
was furnished the inhabitants by
the Manning and the schooner
Metha Nelson, lying at the end of
the dock. We started evaporators
to provide drinking water and con-
tinued this for several days.

*' At noon ashes had begun to fall
again, and at 12.30 were increasing
until at I p.m., it was impossible
to see over fifty feet. Deep con-
cern was visible on every countenance,
and the advisability of the Manning's
getting to sea was discussed.

"However, this was out of the ques-
tion, as few of the inhabitants wished to
leave, and the voice of the ship's com-
pany was 'take all or none.*"

With the beginning of this second phase
of the eruption the confidence which had
carried the people through the first fall
deserted them. When it was realized

* Perry, Captain K. W. Report of June 21.
Annual Report Revenue Gutter Service for 1912,
pp. 11$ et seq.

Digitized by

Google

THE ERUPTION AS EXPERIENCED AT KODIAK

THE GREEK CHURCH AT KODIAK DURING THE ERUPTION

One can only imagine the feelings of the inhabitants as they groped their way to church through the
utter blackness when the bells boomed out the call to prayer on the morning of the third day.

that Kodiak was experiencing not merely
such a passing shower of ash as might
befall almost any community on the
Pacific, but was in the grip of a tremen-
dous eruption, one of perhaps unprece-
dented violence, the worst fears were
naturally engendered.

Once the affair had passed beyond the
limit of reasonable probability, no one
knew what to expect, and a spectre of
abject terror took possession of the place.
This was kept from degenerating into
panic only by the character of the men
in command of the situation. Captain
Perry goes on :

"During the afternoon I visited the
two saloons in the place and, finding
considerable drinking, requested the pro-
prietors to close. This they readily did,
agreeing with me that it was a time for
every man to keep his mind clear.

"At 2 p.m. pitch darkness had shut in;
heavy static disturbances were observed,
and our radio was dumb. A few refugees

were on board, and the night of the 7th
was spent in anxious watching. We got
little sleep, and the dawn of the 8th,
which we anxiously awaited, failed to
appear.

"While all ashes of the previous day
had been removed, decks, masts, and
yards were loaded, as were also the ship's
boats. The ashes now were fine dust
and flakes, and of a yellowish color.
Sulphurous fumes came at times in the
air, and many thought and spoke of the
destruction of Pompeii. Avalanches of
ashes on the neighboring hills could be
heard, and these sent forth clouds of
suffocating dust and ashes.

"All hands were on duty from 7 a.m.
Men often collided in working about
decks, as the feeble glow of the electric
lights and lanterns failed to dispel the
awful darkness for any distance. The
crew kept constantly at work with shov-
els, and four streams of water from the
fire mains were playing incessantly in

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

THE MANNING

Photograph by R. F. Griggs
LYING AT THE DOCK IN KODIAK

When we first visited Kodiak a year after the eruption, the Manning was again coaling alongside
the warehouse which had sheltered the population before she took them aboard.

what at times seemed a vain effort to
clear the ship of its horrible burden.

'* The bells of the Greek church boomed
out in the blackness, and few, if any, of
its followers there were who did not grope
their way to the call to prayer."

DARKNESS SO THICK A LANTERN COULD
NOT BE SEEN AT ARM'S LENGTH

None of those who went through those
days of terror fail to mention, in recount-
ing their experiences, the awful darkness
which is universally described as some-
thing so far beyond the darkness of the
blackest night that it cannot be compre-
hended by those who did not experience
it. Many have told me that it was im-
possible to see a lantern at arm's length,
so thick was the cloud of ash which
enveloped everything.

The sensations produced by the falling
ash are vividly described by J. E.
Thwaites, Mail Clerk of the steamer
Dora, which crossed the area of falling
ash: 3

** And now began the real rain of ashes;
it fell in torrents; it swirled and eddied.
Gravity seemed to have nothing to do
with the course of its fall. The under
side of the decks seemed to catch as much
ash as the sides or the decks under our
feet. Bright clusters of electric light
could be seen but a few feet away, and
we had to feel our way about the deck.

"The officers of the deck had to close
the windows of the pilot-house tightly,
and even then it was with the greatest
difficulty that the man at the wheel could
see the compass, through the thick dust
that filled the room.

* National Geographic Magazine, vol. 24: p. 154. 1913.

Digitized by

Google

THE ERUPTION AS EXPERIENCED AT KODIAK

Photograph by Lieut. J. F. Hahn. U.S.R.S.
REFUGEES FROM KODIAK ON BOARD THE ''MANNING"

The suffering caused by overcrowding was far worse than that due to the eruption. On to a vessel
capable of accommodating at most a hundred people in comfort were crowded more than 500 men,
women, and children, all suffering from nervous shock and in fear of their lives. There was hardly
room enough on the decks for all the people to lie down at once.

"In the meantime, lurid flashes of
lightning glared continuously round the
ship, while a constant boom of thunder,
sometimes coinciding with the flash, in-
creased the horror of the inferno raging
about us. As far as seeing or hearing
the water, or anything pertaining to
earth, we might as well have been miles
above the surface of the water. And
still we knew the sun was more than two
hours above the horizon.

**In the saloon everything was white
with a thick layer of dust, while a thick
haze filled the air. The temperature
rose rapidly, and the air, what there
was left of it, became heavy, sultry, and
stifling. Below deck conditions were
unbearable, while on deck it was worse
still. Dust filled our nostrils, sifted
down our backs, and smote the eye like

a dash of acid. Birds floundered, crying
wildly, through space, and fell helpless
on the deck.**

THE PEOPLE ARE GATHERED TOGETHER

When daylight failed to appear on the
morning of the third day (8th) it began to
be felt that the people of Kodiak should
be brought together in order to facili-
tate the execution of any measures that
might be decided upon for their safety.
Accordingly, Captain Perry continues :

** At about 8.30 a.m. at my request, the
storehouse on the wharf was opened, and
after a hurried consultation with some
of the chief citizens I sent a message to
the priest at the church that all the people
could be cared for on the Manning and
in the storehouse. Prior to this many
had sought refuge on board.

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

THE DORA,

Photograph from J. E. Thwaites
PARTAKER IN MANY A WILD ALASKAN ESCAPADE

With uncanny luck the Dora has had a share in almost every adventure in southwestern Alaska.
No man can count the number of times she has been on the rocks, but until recently she was still ply-
ing those dangerous waters. It was she that located the seat of the eruption and brought the first news
to the outside world. Sixteen years earlier it was she that rescued Spurr's party (see Chapter xx).

*'I then called together a committee
meeting of the officers of the ship and
several citizens, among whom were local
pilots, and it was agreed that, as every
landmark was obliterated, it was impos-
sible to see from the bridge, and that as
the chances were vastly against a ship
making the narrow channel without
striking, it was better to stay where we
were and take what most of us believed
to be only a fighting chance. . . .

'*At 2.30 p.m. of the 8th the fall of ashes
decreased, the skies assumed a reddish
color, and finally objects became dimly
visible."

The ashes that fell during this period
constitute the second layer of the de-
posit at Kodiak, a stratum of fine ash
of a bright terra cotta color, now 45 inches
thick.

When the light thus came again, the
people looked about them to find that
'*all clothed and festooned with ashes.

nothing looked familiar.*' This interval
of cessation in the fall was considered an
opportunity to escape, for —

"As similar phenomena had occurred
the previous day, and frequent seismic
disturbances were still felt, much fear
existed that worse was still to come.

"I hastily summoned a committee of
citizens and, after hearing various opin-
ions, decided that to stay might mean
death, and there would be a chance of
life if the ship could get to sea.

** Consequently all hands were taken
on board except United States Deputy
Marshal Armstrong (who decided that
duty required him to remain) and three
other men. While Marshal Armstrong re-
mained ashore, he sent his wife on board.
This action furnishes its own comment.
Only brave men offer such a sacrifice.

**At 5.30 p.m. the ship cast off and^
with two leads going, Capt. Brown, an
old and skillful Kodiak pilot, conned her

Digitized by

Google

THE ERUPTION AS EXPERIENCED AT KODIAK

through the narrow channel. At 5.55
p.m. we anchored in the outer harbor,
having secured a bearing on Woody
[Wood] Island that practically assured a
safe passage to sea, even though darkness
again shut in.

**As soon as the ship anchored, the
motor boat Norman stood in to Woody
Island and brought off all the inhabitants,
103 in number, many of whom were
nearly famished for food and water, while
others demanded the attention of doc-
tors and nurses. This day food and
water were furnished to 486 people, out-
side our own crew, quarters to 414, and
72 were quartered on board the tug
Printer, which came out of harbor with
the Manning and moored alongside.*'

The suffering during those hours when
the cutter lay at anchor awaiting her
doom, as she supposed, was intense.
Everyone was worn out from the strain
of the ordeal and the work of relief, but
there was little rest for those on board.
The crew gave up their hammocks to the
women and children, but with more
than 500 people of all ages and conditions
aboard a vessel not capable of accom-
modating one-fourth that number in
comfort, hardly a place to lie down could
be found on the crowded decks.

THIRD FALL OF ASH ON THE NIGHT OF
THE EIGHTH

During the night of the 8th ash began
falling again and continued until some
unrecorded hour in the morning. This
fall constitutes the top, third layer of
fine gray ash, which amounted to an inch
and a half, after being packed down.

Before the air finally cleared, after this
last fall, Kodiak had experienced two
days and three nights of practically un-
broken darkness.

Notwithstanding the confidence in the
possibility of escape on the Manning, the
congestion on the little craft was so great
as to be intolerable, and when the morn-
ing of the 9th dawned clear and bright,
murmurs to be landed began to be heard.

Accordingly, in the afternoon the in-
habitants of Wood Island were taken
ashore, since it was possible to care for
them all in comfort at the Baptist Mission
there. The Kodiak people remained
aboard until the morning of the loth,
when they also were gradually trans-
ferred to their homes.

Everyone agrees that the real and
unavoidable discomfort resulting from
the eruption was slight as compared to
the suffering occasioned by terror and
by the overcrowding of those days on the
Manning, If the people could have
known in advance what was to come,
they might have remained in their homes
throughout the ashfall in perfect safety
and in comparative comfort. It was
the vivid descriptions of the '*Last Days
of Pompeii" and the fear of being buried
alive that made the experience so ter-
rible.

If some one had remembered that
Pompeii was only five miles from Vesu-
vius, whereas Kodiakwas a hundred miles
from any volcano, and had assured the
terror-stricken people that nothing worse
than they were already experiencing
could possibly come to Kodiak, he might
have greatly alleviated conditions.

Digitized by

Google

^ 5

Dh o
i4

Digitized by

Google

• III

RECORDS OF THE ERUPTION FROM OTHER LOCALITIES

Among the people of Kodiak impris-
oned on the Manning, speculation was
rife as to the source of their misfortune.
But it was several days before they were
to be enlightened.

When the ashfall finally ceased and
the outlook began to improve, Captain
Perry dispatched one boat to Seward,
the terminus of the cable, to communi-
cate with the outside world, and another
to ascertain how the neighboring settle-
ments had fared, and offer assistance. It
was not until the I2th that information
was received in a message from one of
these boats that the eruption had come
from Mount Katmai, a peak which until
that time was practically unknown.

THE IDENTIFICATION OF THE VOLCANO

The determination of the seat of the
eruption was first made by C. B. McMul-
len, Captain of the steamer Dora, '*that
notorious partaker in many a wild Alas-
kan escapade,** which, with characteris-
tic luck, happened to be passing through
Shelikof Strait at the time. Captain
McMullen's description of his experience
is as follows: ^

* ' Left Uyak at 8.45 a.m., June 6 ; strong
westerly breeze and fine clear weather.
At I o'clock p.m., while entering Kuprea-
nof Straits, sighted a heavy cloud of
smoke directly astern, raising from the
Alaska Peninsula. I took bearings of
same, which I made out to be Katmai
Volcano, distance about 55 miles away.
The smoke arose and spread in the sky,
following the vessel, and by 3 p.m. was
directly over us, having traveled at the
rate of 20 miles an hour.'*

Captain McMullen's determination,
although corroborated by other mariners
and by the testimony of the natives

* This portion of our account is almost alto- pedition sent to the scene by the National
gether abstracted from the excellent narrative Geographic Society. For fuller details see the
of George C. Martin, who directed the first ex- National Geographic Magazine, February, 1913.

quoted below, was not, however, un-
challenged.

The press dispatches named Iliamna
and Redoubt as equal partners in the
disturbance. This uncertainty was, how-
ever, due largely to the fact that there
were no traditions of previous activity
from Katmai. That mountain was, in-
deed, almost unknown. The Juneau
Alaska '* Dispatch,'* in commenting on
the news, remarked that Katmai was not
shown on the maps. J. L. McPherson,
of the Seattle Chamber of Commerce,
who had spent a summer prospecting
within 50 miles of Katmai stated cate-
gorically that there was no Mount Kat-
mai, and supplied pictures of Peulik, on
Becharof Lake, which he suspected might
be the seat of the eruption. Other less
known peaks were also mentioned.

As a matter of fact, the identity of the
volcano was not established to the satis-
faction of the people of Kodiak until
three years later, when the first of the
expeditions of the National Geographic
Society, under the leadership of the writer,
reached the scene and proved that the
ash had come from Katmai and not from
some other vent.

As there were no considerable settle-
ments anywhere near the volcano, there
is so little testimony from eyewitnesses as
to the events of the eruption that only
a very fragmentary account can be given.

Of the preliminary disturbances which
must have preceded the great explosion
practically nothing is known. This is
the more surprising since, as subsequent
exploration has shown, it was during
this period, before the explosion of Kat-
mai, that the Valley of Ten Thousand
Smokes was opened up and its enormous
flows of incandescent sand poured forth.

Digitized by

Google

I6 THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs
A ROOF CRUSHED BY THE WEIGHT OF ASHES
Picture taken on Long Island, over a hundred miles from Mount Katmai, a year after the eruption.

Photograph by W. J. Erskine
ASH FALL COVERED SHIPS AT KODIAK, JUNE 9, I912

Digitized by

Google

RECORDS OF THE ERUPTION FROM OTHER LOCALITIES

Surely we would have supposed that
the tremendous cataclysm necessary to
open the great fissures of the Valley
would make itself felt far and wide, but
as a matter of fact there is no definite
record of events during this period.

ACCOUNT OF THE ERUPTION BY NATIVES
OF KATMAI VILLAGE

Of those nearest the volcano, the in-
habitants of Katmai Village, all but six
had been away fishing at Kaflia Bay for
some time before the eruption. These
two families remained on the ground
until June 4, when they were frightened
away by the earthquakes and by other
signs of impending disaster, but they
gave no circumstantial account of the
cause of their flight.

They were still camped along the coast
in the vicinity of Cape Kubugakli within
sight of Mount Katmai when it blew up,
and so reported when they reached Cold
Bay, where their accounts were tran-
scribed to the diaries of two white men
resident at that place. C. L. Boudry
wrote, under date of June 8th:

"Two families arrived from Katmai
scared and hungry and reported the vol-
canoes up 15 mile from Katmai (village)
to the left of Toscar trail and that 3^ the
hill blun up and covered up everything
as far as they could see also that small
rock were falling for 3 or 4 miles at sea
but could not say more of it as every
thing es closed up with smoke. These
people had left Katmai village June 4
and camped between Kashvik and Alin-
chak bays. The rocks an fall in their
boat are of the size of coarse rice and of
the pumice stone formation. No one es
lost as far as they know but all the
natives are east of Katmai and the main
flow of smoke go there as a strong S W
wind blowing so they are in a bad
shape."

Jack Lee wrote: "They report the top
of Katma Mountain blun of. There was
a lot of Pummy stone in their dory when
they got here and the say Hot Rock was
flying all eraund them.**

The only other persons within sight of
the volcano at the time of the explosion
were the natives in the village of Savon-
oski at the head of Naknek Lake, 21
miles (34 km.) from Katmai Crater. The
ravages of the eruption caused these
people to flefe in terror to Naknek, aban-
doning their village at Savonoski and
settling near Naknek, where they were
interviewed by P. R. Hagelbarger, of the
National Geographic Expedition of 191 8.

"American Pete," chief of the village,
had barabaras* (houses) both at Savono-
ski and at Ukak, up the Valley near the
present foot of the sandflow. Warned
by the earthquakes preliminary to the
outburst, he was in process of removing
his belongings from Ukak to Savonoski
when the eruption began. His account
of what followed runs thus:

STORY OF

AMERICAN PETE,
SAVONOSKI

CHIEF OF

"The Katmai Mountain blew up with
lots of fire, and fire came down trail from
Katmai with lots of smoke. We go fast
Savonoski. Everybody get in bidarka
(skin boat). Helluva job. We come
Naknek one day, dark, no could see.
Hot ash fall. Work like hell. Now I
go back every year one month maybe
after fish all dry and kill bear. Too bad.
Never can go back to Savonoski to libe
again. Everything ash. Good place,
too, you bet. Fine trees, lots moose,
bear and deer. Lots of fish in front of
barabara. No many mosquitoes. Fine
church, fine house. Naknek no good."

The fire which American Pete saw
coming down the trail from Katmai was
probably started by the great Incandes-
cent Sandflow, which was already rolling
down the Valley before the explosion of
Katmai.

Realizing that this man and his party
were the only human beings in Ukak
Valley at the time of the bursting forth
of the Valley of Ten Thousand Smokes,
Mr. Hagelbarger endeavored by ques-
tioning to elicit further details, but none
were to be obtained. The sole interest

* Pronounced Bi rS' bSra.

Digitized by

Google

C
3

•w

(X4

♦-

1-4

4-i

C/5

«•«

4-i

«3

•*rf

Digitized by

Google

RECORDS OF THE ERUPTION FROM OTHER LOCALITIES

of the natives had evidently been to flee
from the coming wrath of the eruption.
The only additional information that
could be obtained concerned the condi-
tion of the Valley before the eruption,
which was described as having clumps
of trees along the stream clear up under
the glaciers of Mount Mageik, where
travelers camped before going over the
Pass. The tundra in the upper part of
the Valley was firm, and it was a great
place for caribou.

EXPERIENCES OF PEOPLE AT KAFLIA BAY

The only other persons anywhere near
the volcano were the natives at the fish-
ing station in Kaflia Bay. They were,
to be sure, about 30 miles (47 km.) from
the crater and ^ so situated as to be
screened from the volcano by intervening
mountains, which, while protecting them,
prevented their observing the eruption.
Their ordeal, while similar to that of
the people of Kodiak, was more terrible,
being both much more intense and more
prolonged.

Unfortunately, there appears to be
no other record of their experiences than
this very interesting letter, written by
one of them :

"Kaflia Bay, June 9, 1912.
My dear Wife Tania:

First of all I will let you know of our un-
lucky voyage. I do not know whether we
shall be either alive or well. We are awaiting
death at any moment. Of course do not be
alarmed. A mountain has burst near here, so
that we are covered with ashes, in some places
10 feet and 6 feet deep. All this began on the
6th of June. Night and day we light lamps.
We cannot see the daylight. In a word it is
terrible, and we are expecting death at any
moment, and we have no water. All the rivers
are covered with ashes. Just ashes mixed
with water. Here are darkness and hell,
thunder and noise. I do not know whether it
is day or night. Vanka will tell you all about
it. So kissing and blessing you both, good-bye.
Forgive me. Perhaps we shall see each other
again. God is merciful. Pray for us.

Your husband.

Ivan Orloff.

The earth is trembling; it lightens every
minute. It is terrible. We are praying."

That there were no casualties among
these people is sufficiently conclusive
evidence that there is no great danger of
being buried alive by a fall of volcanic
ash. Their survival makes it perfectly
clear that the people of Kodiak were at
no time in any real danger, and that if
they could have been spared their terror
they might have endured the ordeal with
comparative ease.

These fragmentary statements con-
tain all the testimony available from
points within 30 miles of the crater. We
are, therefore, dependent for further
information as to the character of the
eruption upon inferences drawn from
conditions found later around the vol-
canoes. These may be more appropri-
ately taken up in the discussion of the
exploration than at this point and are
accordingly deferred to Chapters xxi,
and XXII.

PRELIMINARY SYMPTOMS OBSERVED AT
OUTLYING STATIONS

Earthquakes began at Katmai Village
at least five days before the explosion,
while on June 4 and 5 there were more
severe shocks, sufficiently strong to be
felt at Kanatak,65 miles (100 km.) south-
west of the volcano, at Uyak, 58 miles
(93 km.) southeast, and at Nushagak, 130
miles (210 km.) northwest (see map, op-
posite page i).

A considerable eruption of some sort
probably occurred in the evening of the
5th, since observers at Cold Bay, 40
miles (65 km.) southwest, noted that
the northern sky, i.e., in the direction of
the Valley of Ten Thousand Smokes,
looked "black and storming'* late that
night in spite of the fact that there was
clear weather on the coast.

During the morning of the 6th explo-
sions were heard at Seldovia, 150 miles
(240 km.) northeast of Katmai and at
Nushagak, 130 miles (210 km.) north-
west, but were not noted elsewhere.

Digitized by

Google

Photograph by George C. Martin
ASH-LADEN TREES NEAR KODIAK, AUGUST 26, I9I2
The branches bending to the ground made a tent under every tree.

Digitized by

Google

RECORDS OF THE ERUPTION FROM OTHER LOCALITIES

At I p.m., on June 6, the terrific explo-
sions and earthquakes which ushered in
the period of violent activity were noted
at Cold Bay, while at the same time the
Dora first observed the black column
rising from the volcano (see quotation
page 15). The same cloud of ash and
steam was seen also from Lake Clark
and from Cook Inlet.

PERIOD OF MAXIMUM ACTIVITY

From this time on the volcano con-
tinued in a state of prodigious activity
for more than sixty hours, until some
time after midnight on the 8th. During
this period especially violent explosions,
corresponding roughly, at least, to the
three major layers of ash, occurred at
3 p.m. on the 6th, at ii p.m. on the
same day, and at 10.40 p.m. on the 7th.

After the third of these explosions, the
volcano subsided a little, but it remained
in a state of violent activity for several
days more, as indicated by the letter of
Ivan Orloff (quoted above), which,
though dated the 9th, gives no hint of a
cessation of the eruption. The explo-
sions continued to be heard through the
9th at Seldovia, 150 miles (240 km.)
northeast.

**On the morning of the 9th observers

* Quotation from Martin, op. cit. p. 158.

• The only record of this period of waning ac-
tivity extant is given in the diary of C. L. Boudry
of Cold Bay. From it I glean the following:

"9th. Still shaking and fumes are getting
strong, ashes are falling steady . . .; loth.
Still shaking . . . fumes not so bad . . .
tremors are not so steady and not so strong
. . .; nth. Shaking some . . .; 12th. Still
shaking 7 a.m., and all day, only few hard shake;
fumes are not so bad; 13th. Still shaking but
not so hard. No more shakes after 12 o'clock
and fumes are few; 14th. Not shaking since
7 or 8 o'clock last night . . . start shaking
at I p.m. but not so hard, about 6 or 7 shocks till
9 p.m. 15th and i6th same conditions; 17th. Lots
of smoke and fumes are pretty strong. Shaking
hard and after 12 o'clock. Can hear the roar
every once a while. After each roar a strong
shake occur . . .; 21st. To-day shake is the
hardest since the nth of June; 23rd ... at
3 p.m. could see like a large reflection of a large
fire; the heaven was very red so was the fog or
smoke right over the hills.

"July 2 1 St. At 2 p.m. fog clear up and could

at Naknek, 80 miles northwest of the
volcano, saw a beautiful illuminated
funnel-shaped cloud which rose straight
into the air to a considerable altitude
and, as there was no wind, retained its
shape. It afterward assumed different
colors and dissolved into cloud banks,
being illuminated all the time. A similar
cloud was observed from Iliamna, 115
miles north-northeast of the volcano and
135 miles from Naknek, at the same time,
the description differing from the above
only in the statement that in losing its
funnel-shape form it assumed the shape
of a ship." 2

GRADUALLY SUBSIDING ACTIVITY CON-
TINUED ALL SUMMER

All accounts agree that the volcano
was very active for a long time after the
great explosions, but details are lacking.
Apparently the freely eruptive condition
continued for many weeks. This ap-
pears to have been followed by a gradual
subsidence of activity until the present
quiescent condition was reached.'

Turning now to the record of the erup-
tion preserved in the layers of ash, we
find that even such activity as was re-
corded by Mr. Boudry on June 17th and
23rd and on July 21st was not accom-

see smoke from new crater, not a steady flow but
puff every 2 or 3 minutes or so some small and
other, hard, one smoke far above cloud; . . .
at 5 p.m., the smoke close of the top of hillock
was like over a fire as the color is reddish. . . .
All smoke from new crater is like steam escaping
or else light yellow. At 10 p.m. a small shock,
not over 6 seconds duration, direction E. & W.
One hard shock at 11.45, ^ard enough to throw-
hat off table, duration about i minute to i Vs
min. Rock broke loose on cliff on beach."

On November 23 Mr. Boudry described the
condition of the volcano as follows:

" The volcano are still raising cane. ... I
will try to go there in winter or spring but
can't make it now. I try 3 time. Cannot see
not ting for smoke and after you are 10 or 12
miles the acid raise hell . . . the acid bume
the close you got on and raise blister on your
hands. The worst shake we got was on the 20th
Sept. Twestie the houses bad and knock the
toiling from the houses down — ^also I was out-
side, just throw me as ef I ad received a good
rock on the side — Not much dogs, they get
blind from acid."

Digitized by

Google

C/3

Q
O

U 00
O &

^ 2.

1-4

In
Q

u
o

C/5

:^
u
<

Digitized by

Google

RECORDS OF THE ERUPTION FROM OTHER LOCALITIES

panied by the extrusion of any great
quantity of solid matter.

In Upper Katmai Valley a fourth yel-
lowish layer of ash overlies the three
primary strata. This was probably
thrown out immediately after the more
widely distributed layers, probably on
the 8th or 9th. It was later capped by
the stratum of very finely divided mud,
which marks the top of the ejecta in
Katmai Valley (see Chapter xii).

RECORDS OF EARTHQUAKES ACCOMPANY-
ING THE ERUPTION

The earthquakes accompanying the
great explosions were severe enough to
be recorded by seismographs at great
distances, as evidenced by the following
information supplied by Mr. Napier
Dennison, of Victoria, B. C, who writes:

'*I may also say that on the 6th and
7th of June, 1912, our Milne seismograph
recorded a very large number of quakes
and most of them appeared to have had
their origin in the Katmai district.
Though some of these quakes run into
each other, ten are registered on the 6th
and twenty on the 7th (Pacific time).**

The first shock was recorded by the
University of Washington (Seattle) seis-
mograph at 2.41 p.m., Thursday, June
6th, according to the daily press. The
Georgetown University (Washington,
D. C.) seismograph also recorded shocks
almost continuously, beginning at 5 p.m.
on the 7th and continuing through the
8th, some of them being of extraordinary
violence.

Notwithstanding the occurrence of
shocks strong enough near the volcano
to knock a man down, there are no
reports of damage from earthquakes.
Whether the record would be the same
if the country were thickly populated
is doubtful.

There are certain evidences, however,
that the earthquakes were not so violent
as might have been expected. On the
mountains around Katmai are large
numbers of ** balanced** rocks standing
in positions of such apparently unstable
equilibrium as to suggest that they were

never subjected to any very heavy earth-
quakes.

EXPLOSIONS HEARD 75O MILES AWAY

The explosions that marked the period
of maximum activity were so violent
that they were distinctly heard over an
enormous extent of country, having
been recorded at several places on Iliamna
and Clark lakes, 100 miles (160 km.)
away; at Iliamna Bay, 115 miles (185
km.) northeast, where "the sounds were
accompanied by a sudden quick motion
of the clouds that would start and stop ** ;
at Seldovia, 150 miles (240 km.) northeast,
they continued from the morning of the
6th until the loth; at Cordova, 360 miles
(580 km.) northeast, beginning at 7 p.m.,
June 6th, they continued till 10 p. m.,
June 7th; at Copper Center, 450 miles
(725 km.) northeast, they were heard on
the 7th ; at Fairbanks, 500 miles (800 km.)
northeast; at Chicken, 600 miles (1000
km.) northeast, they were recorded from
2.30 p.m., June 6th, until 6 a.m., June
9th; at Dawson, 650 miles (1,050 km.)
northeast ; at Katalla, 410 miles (660 km.)
northeast, the detonations "sounded like
the explosions of dynamite in the near-by
hills. '* Juneau, 750 miles (i ,200km.) east,
appears to have been the most distant
place at which they were recorded.

It should be pointed out in this con-
nection that the detection of such sounds
at any place depends largely on the acci-
dent of their attracting the attention of
some observer, for in this noisy modern
world most people take no notice of dis-
tant thunder or of explosions that would
be distinctly heard by anyone listening
for such sounds.

No one at Kodiak observed the sound
of the explosions. After the beginning
of the ashfall there was, to be sure, so
much thunder and lightning that the
later explosions, even if heard, were not
distinguished from the peals of thunder.
But the afternoon of the 6th is described
as quiet and uneventful until the ap-
proach of the ash cloud, after 4 o'clock.
It is not easy to understand why the
explosions that occurred prior to that

Digitized by

Google

l«

•o

b«

*— ;

fcq

C/3

■*-•

4-)

>«

*5<>

ffi

Digitized by

Google

RECORDS OF THE ERUPTION FROM OTHER LOCALITIES

time were not heard at Kodiak and Afog-
nak, since they were noticed at Chicken,
600 miles (1,000 km.) away.

Curiously enough, a similar phenome-
non was observed in the case of
Krakatoa, where no sounds were heard
in the vicinity, after the beginning of the
period of maximum activity, although
they were clearly heard (not as one
sound but as many) at Rodriguez, 3,000
miles (5,000 km.) away. This was be-
lieved by the Royal Society's commission,
that gathered the data concerning Kraka-
toa, to be due to the screen of solid matter
in the air.

But this explanation appears not to
apply to the present instance, inasmuch
as there was no such screen of solid mat-
ter around Kodiak until several hours
after the explosions began.

SLT.PHURIC ACID RAINS AT GREAT DIS-
TANCES

The fumes from the eruption pene-
trated to even greater distances than the
sounds. Throughout the area of heavy
ashfall disagreeable sulphuric odors were
common for many months after the erup-
tion. The gases from the volcanoes,
moreover, frequently combined with the
moisture of the clouds to produce acid
rains which were both disagreeable and
destructive. Such acid rains occurred at
great distances.

An Associated Press dispatch from
Seward, 250 miles (400 km.) northeast of
the volcano, on June nth, read: ''A
slight rain fell at Seward today. The
water was charged with sulphuric acid,
which is killing all vegetation and eating
into the metal work of buildings."

On the same day the following dis-
patch came from Cordova, 360 miles
(580 km .) northeast : ' * The first real danger
here was made apparent when a heavy
rain began to fall. The water mixing
with the ash in the air formed sulphuric
acid which burned painfully whenever it
came in contact with the unprotected
parts of the bodies of persons in the
street. Before the cause was realized
many persons received painful burns in

the eyes, although none was severely
injured.'*

At Latouche, also, in Prince William
Sound, 300 miles (480 km.) northeast of
the volcano, Mr. F. R. Van Campen, then
superintendent of the mines, in a private
letter states that, following the eruption,
the rain was so acidified by the fumes as
to cause stinging burns wherever it
touched the flesh. He had his chemist
analyze this rain, and found that the
trouble was caused by sulphuric acid,
which was present in considerable quan-
tity. Unfortunately the analysis giving
the precise concentration of the acid has
been lost.

This acid rain did serious injury to
plants. The tender annuals of the gar-
dens were completely destroyed, and the
leaves of many of the native perennial
plants were so burned that they dropped
off.

At Cape Spencer, about 700 miles
(i , 100 km.) east of the volcano, the fumes
were so strong that "brass tarnished 15
or 20 minutes after being polished*' on
the decks of the steamship Admiral
Sampson,

The extreme limit reached by the
fumes was, however, very much more
distant, for their presence was recorded
from several places in Washington and
in British Columbia, about 1,500 miles
(2,400 km.) from the volcano. It will
be advisable, however, to defer citation
of authority for this statement until the
discussion of the ashfall.

A FOOT OF ASH OVER AN AREA HALF
AGAIN AS LARGE AS DELAWARE

The descriptions of the eruption at
different points have already given some
conception of the enormous ashfall. A
more comprehensive idea of its amount
and distribution is furnished by the map
opposite page i, on which the varying
depths of the deposit are indicated by the
contours. This map must not, however,
be construed as supplying exact informa-
tion. Because of the uninhabited charac-
ter of the country, it was not possible to
secure sufficient data to permit drawing

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by George C. Martin
INTERIOR OF A BARABARA SHOWING THE CONSTRUCTION OF THE NATIVE HUTS
Compare with exterior, page 24. The ash sifted even into such structures.

the contours with any great precision
except in the small sector between Kod-
iak and the volcano.

Accepting, however, the data on the
map as the best obtainable, we may list
the areas within the different contours.

An area of about 930 square miles
(2,500 sq. km.) was covered with three
feet (90 cm.) or more of ash. The area
covered by a fall of two feet (60 cm.)
includes 2,500 square miles (6,400 sq.
km.). Ash to the depth of a foot (30
cm.) or more fell over an area of 3,000
square miles (7,800 sq. km.), half again
as large as Delaware. An area of 29,600
square miles (77,000 sq. km.) was covered
with one inch (25 mm.) or more. The
quarter-inch contour (6 mm.) incloses a
territory of 42,000 square miles (109,000
sq. km.), equal to the area of Tennessee.

In addition to this, dust, having an
enormous aggregate volume, fell far

beyond the quarter-inch mark on our
map. It is recorded, for example, that
•* about one-fourth inch of ash fell" at
Katalla, which is 410 miles (660 km.)
northeast of the volcano, twice as far as
that depth of fall is indicated on the map.

DUST FELL TO A DISTANCE OF I,500 MILES

Small but appreciable quantities also
fell at much greater distances,* as shown
by the following records: Ruby and Inno-
ko, approximately 400 miles (650 km.)
north; Rampart and Fairbanks, 500
miles (800 km.) northeast; a point 90
miles southwest of Eagle, 550 miles
(875 km.) northeast, on the loth of June;
Dawson, 650 miles (1,050 km.) northeast,
on the nth; Chicken, 600 miles (1,000
km.) northeast; Juneau, 750 miles (1,200

^ These data are partly from Martin's article
and partly from Kimball, Monthly Weather Re-
view vol. 41, pp. 153-59. 1913-

Digitized by

Google

RECORDS OF THE ERUPTION FROM OTHER LOCALITIES

km.) east; Loring,in the southeastern cor-
ner of Alaska, 900 miles (1,450 km.) east.

Outside of Alaska the ashfall was
noticed generally for a long distance, but
definite records are scanty until the
thickly populated Puget Sound country
is reached. The experience of a steamer
in this area was described in a press
dispatch, dated Monday, June loth, as
follows:

"Vancouver, B. C, June 10. — Passen-
gers on the steamer Camosun, which
arrived this morning from Prince Rupert,
witnessed a wonderful phenomenon all
the way down until within 100 miles of
Vancouver. A steady shower of ashes,
presumably from the volcanic eruptions
in Alaska, began falling into the sea and
on to the vessel's deck soon after clearing.

'*The shower kept up steadily from
Saturday morning until last night at mid-
night. * It was the strangest spectacle I
ever saw,* said Dr. U. W. Ewing, the
chief surgeon of the Grand Trunk Pacific
Medical Service, who was a passenger.
The entire heavens were filled with fine
dust, which fell steadily.

"The ashes obscured the sun and
made everything hazy. Every foot of
the steamer was covered, and extra men
were assigned to sweeping it away. No
sooner was it removed than a fresh sup-
ply fell from the sky. We were not
aware of the volcanic disturbances in
Alaska until we steamed into Van-
couver."

Definite records of falling dust were
made by a number of observers in the
Puget Sound region, which is approxi-
mately 1,500 miles (2,400 km.) southeast
of the volcano.

At Victoria, B. C, Mr. Napier Denni-
son. Superintendent of the Dominion
Meteorological Service in British Co-
lumbia, vouches for the following quota-
tion from the ** Victoria Colonist " of June
1 1 , 191 2, adding that he himself furnished
most of the information :

** Volcanic ash from the Alaskan crater
has been falling in Victoria for the past
thirty hours, the effect being especially
noticeable on brass work on motor cars

and elsewhere, due to sulphuric acid in
the ash.

"In connection with these remarkable
after-effects of the great Alaskan erup-
tion, when thousands of tons of fine
ashes were thrown up to the upper air to
drift to many quarters of the globe, Mr.
Napier Dennison furnishes the * Colonist*
with some interesting notes. Yesterday
afternoon he received the following
report from Mr. C. D. Rouge, at Malbat
Station (on Vancouver Island 20 miles
north of Victoria) :

'* 'I observed a peculiar haze which
settled upon us Sunday, 9th, about 4 p.m.
and continued till the time of writing
(10 a.m. Monday). This haze is appar-
ently heavily charged with sulphuric
acid. It is thin in appearance but varied
in mass as it crossed the sun. Thinking
perhaps it might be sulphuric acid gas,
which it resembles both in appearance
and smell, I thought perhaps our higher
location might render it more noticeable.*

"The presence of sulphuric acid gas or
other injurious fumes was noticed in the
city (Victoria) by motorists, particularly
when driving on the higher apd wind-
exposed roads. The effect was to dis-
colour the polished brass work of the
cars.

"Mr. Dennison and Mr. C. C. Pem-
berton also found the windward sides of
the exposed trees on the Gonzales rocks
covered with a fine white ash, samples
of which have been secured for analyses."

The corrosive effects of the acid on
brass were reported more circumstan-
tially from Port Townsend, Washington,
as follows :

"Port Townsend, June 11. — ^Sulphuric
acid is present in the air in quantities
sufficient to blacken the brass work on
automobiles. The owners of cars noticed
the presence of a peculiar dull-colored
corrosion on all brass work. The brass
work was repeatedly cleaned but the
deposit would quickly return. The de-
posit is not hard to remove, being in the
form of thin dust, but when removed,
leaves the brass work slightly colored,
requiring some little trouble to polish."

Digitized by

Google

Photograph by G<H)rge C. Martin
OLD FOREST SPRUCE, WITH MOSS-COVERED TRUNK AND BRANCHES LOADED WITH ASH,

NEAR KODIAK, SEPTEMBER 4, I9I2

Digitized by

Google

RECORDS OF THE ERUPTION FROM OTHER LOCALITIES

More injurious effects of the acid were
reported in a press dispatch from Van-
couver the next wash day :

''Vancouver, B. C, June 12. — Scores
of housewives in Vancouver were suffer-
ers because acid, presumably from the
Alaskan eruptions, permeated their fine
white clothes hanging on the lines after
the Monday washing. Linen goods were
apparently the only ones attacked.
When this white wear came to be ironed
it simply fell to pieces, and not until
many complaints had reached dry goods
houses today were the women able to
solve the problem."

The ashfall was likewise detected at
Olga and at Tatoosh Island, Washington.
At the latter place J. E. Hissong, the
weather observer, reported, on June loth :

** Light deposit of fine white ash found
on southwest side of storm-warning
tower at noon. Evidently carried by
wind from volcanic explosions in south-
western Alaska"; and three days later:
** During the evening of the 13th the air
was filled with a very fine ash and
smoke, a light deposit being noticed on
the office furniture on the morning of the
14th."

Beyond this point the effects of the
dust clouds in the upper air belong more
properly to the meteorological effects of
the eruption, and will be discussed in the
next chapter (see page 32).

ASH AND SAND TOTAL SIX CUBIC MILES

It will be of interest at this point to
attempt a rough estimate of the quantity
of material thrown out by the volcano.

Martin, summing up the volumes of
the deposits of the several thicknesses
shown on his map, arrives at an estimate
of 4.9 cubic miles as the volume of meas-
urable ashfall. Since his work was
done, a considerable amount of informa-
tion on matters connected with the cish-
fall has become available. Part of this

*The reader may be inclined to inquire if
another liberal allowance should not be made for
the dust in the upper atmosphere which was re-
sponsible for the meteorological effects treated in
the next chapter. Humphreys, however, has

tends to increase the estimate, part to
decrease it.

Exploration has shown that the ashfall
to the northeast of the volcano was not
nearly so heavy as Martin supposed.
Although details are still exceedingly
meager, we have ventured to trace con-
tours of ashfall in this region. From the
contours of the map, including those in
the deeper deposits close to the volcano
(which were unknown to Martin), I
have estimated the ashfall from Katmai
as 4.75 cubic miles.

In this estimate I have assumed that
the amount of ash that fell outside the
quarter-inch contour was equivalent to
one-sixteenth of an inch over a circle of
500 miles radius. This seems conserva-
tive, in view of the quarter-inch fall re-
ported at Katalla, 410 miles away and
out of the track of the heaviest clouds of
ash.^

In addition to the ash from Katmai
itself, we should consider here the eject a
of other vents thrown out at the same
time, for all belong to the same eruption.

In the first place, there is the material
composing the great Incandescent Sand-
flow of the Valley of Ten Thousand
Smokes, as described in Chapter xix.
Though it is impossible to estimate this
with any degree of satisfaction, its mag-
nitude is of the order of one cubic mile.

Second, there is the ash and pumice
thrown out by Novarupta in the period
of its explosive activity. This covers a
circle approximately seven miles in ra-
dius to a maximum thickness of about
100 feet. Allowing for the taper char-
acteristic of volcanoes this may be esti-
mated conservatively as a cone having a
radius of seven miles and a height of 50
feet, the volume being, therefore, ap^
proximately half a cubic mile.

Summing up these various items, we
have in round numbers a total mass of

estimated that one fifteen-hundredth part of a
cubic mile of finely divided dust would be suf-
ficient to produce the meteorological effects ob-
served. This is much too small a quantity to be
considered in our estimate, which is subject to
errors of very much larger proportions.

Digitized by

Google

•s
I

2 X

(ft

•M .

?«

•^ s

^5.2

3 w

o a

O w

c^*^

Cl cn

:§

q >»

^Cd

c B

— c

2 a

(S4

"oJ 8

«= c

^ —

'6^

CO o

Q (A

u.Si

& c

c 5.

a> cr

(14

£ a>

fcS

•^ 2

w r

E."2

(2h

•^a

•s

Digitized by

Google

RECORDS OF THE ERUPTION FROM OTHER LOCALITIES

volcanic ejecta of 6.25 cubic miles thrown
out from the interior of the earth during
the eruption.

This figure is so stupendous that it
quite passes the possibility of compre-
hension. As it has been my privilege to
work over these deposits, season after
season, I have found my conception of
their magnitude continually growing as
my acquaintance with them has broad-
ened. Not because I have gained any
concrete conception of their size, but
rather because of an increasingly vivid
realization that their msiss is beyond the
grasp of the mind.

EJECTA EQUIVALENT TO OUTPUT FROM

STONE CRUSHERS OF UNITED STATES

FOR OVER 500 YEARS

In attempting to form some idea of
the bulk of the ejecta, it may be helpful
to seek an equivalent in terms of human
labor, which is our chief standard of
measurement. Since the ash is essen-
tially pulverized rock, the volcano may
be compcired to a gigantic stone crusher,
except that Katmai reduced a large
percentage of its product to dust, thus

•This estimate is arrived at by multiplying
the volume of the deposits by their specific
gravity. A number of samples of the ash as it
lies on the ground have been found to average

doing its work much more thoroughly
than a crusher. The analogy is not
quite just, because the ash was formed
by the disruption of molten magma
rather than by the crushing of solid
rock.

Nevertheless, the energy expended in
the eruption probably far exceeded the
power necessary to crush an equivalent
amount of hard rock and, in any case,
comparison with the crushed stone
production of the country will give a fair
idea of the quantity of material thrown
out.

The output of broken stone from all
the quarries of the United States is
about 50 millions of tons annually. The
total weight of the ash was about 28,700
millions of tons.' It is thus evident that
Katmai ''working*' only one week, pro-
duced as much crushed rock as could be
turned out by all the crushers of the
United States during a period of 575
years. That is to say, to catch up with
Katmai they would have had to begin
operations one hundred and fifty years
before the discovery of America and kept
up steadily at their present rate ever
since.

a little greater in density than water. The weight
of the ejecta may be computed, therefore, as
equivalent to that of an equal volume of water.

Digitized by

Google

1880

1890

1900

1910

f 1

PER
CEN1

102
96
90

NUM
BER

30
60
90

F.
+03'

PER
CENT

—

- ^

—

— H

— w

^ —

102 -V f'

i v^ "*

Si.

\ i^ '

^s^^

-M I/I 1 1

96 \ t ~

T-"^

<,'

^jfz '

^ RADIATION

INTENSITY (PYRH

ELIOMETRIC

VALUES)

rf^

NUM

° y-

--^

■^

.t^

^ L /.

i ^

f .

30^ ^ -

" A ^

L__

^ r

t ^^ v^

60 ^^y

^^^^

t\^\

^c>

SUN-SPOT NU

ERS = S 1

X-

.,^

■^

—i ^^

"^. ^

-''

1 "

" ^ ^

f "

" u ^^

I t "

" A v^

V-

3 "

1 .^

^:;^

P«S COMBINED

.***%•

♦0.9

♦0.6?
♦03*

-0.3'
-0.6*

A .«>•

•0.6r

^B K—

*^o®

> t ^^

«<X3

' 5 ^'^

)

o^^

^ ^^

\ jt

^^^Z

-03" ^r- -.^i

■v:?

°-^ V / TEI

V«P LRATURE DEPA

RTl

JR^S - (Ml

LANS)

-06* ^-^ "

aife

==l==l=!

1 i

I m H

_^^^^ii

^!>J«L^

^^^

.lif^

s^,

KRAKATOA Tara^vera BooosLOf Awoe Pele' Santa Coli ma Katmai
Banoaisan Maria

CURVES SHOWING THE RELATIONS OF TEMPERATURE, SUNSPOT NUMBERS AND RADIA-
TION INTENSITY (PYRHELEOMETRIC VALUES) AS AFFECTED BY VOLCANIC ERUPTIONS

(After Humphreys.)
32

Digitized by

Google

CLIMATIC EFFECTS OF THE ERUPTION

As intimated in the last chapter, the
effects of the eruption were felt far be-
yond the zone of ashfall. They were in
fact world-wide.

The fine dust carried into the upper
atmosphere formed a haze which so re-
duced the intensity of sunshine as to
cause the cold summer of 1912, through-
out the northern hemisphere. Many
readers will probably be surprised to learn
that a single localized eruption can have
any measurable effect on the weather of
so large an area. Nevertheless, an ex-
amination of the records shows that all
great volcanic explosions have been fol-
lowed by pronounced drops in tempera-
ture, the world over.

The lowest point ever reached in the
temperature of the world, since the
beginning of reliable records, occurred in
the years 1 783-1 785, following the erup-
tions of Asama-yama in Japan, and
Skaptar JokuU in Iceland, which are
reported as having been among the most
frightful known.

Asama hurled enormous rocks in all
directions. One of these, said to have
measured 264 by 120 feet, fell into a river
forming a miniature island . Many towns
and villages were buried. The activity
of Skaptar Jokull largely took the form of
immense lava flows, the greatest within
historic times; but there were also ex-
plosions producing ashes and fumes that
affected a large area, even destroying
crops in Scotland, 600 miles away.

These eruptions were followed by a
remarkable "dry fog" which appeared
simultaneously on June 17, 1783, all
over Europe. It was world-wide, being
recorded in Asia, Africa, and America,
as well as Europe. It was so thick that

^ This material is summarized from the Report
of the Krakatoa Committee of the Royal Society
pp. 388-392. 1888. The observations of twi-

in some places '* large houses 2 miles off
could hardly be seen." The sun was
invisible for some time after rising and
before setting. At noon it was blood-
red, and could be observed easily with
the unprotected eye. In the middle of a
moonless night, the twilight, due to reflec-
tion from the dust cloud illuminated the
earth as much as a full moon. Strong
sulphurous fumes, which accompanied
this haze, damaged many plants, were
unpleasant to the eyes, and caused some
persons inconvenience in breathing.*

The effect of the same fog in America
was thus described by Benjamin Frank-
lin, who at the time was ignorant of its
cause.^

"the year without a summer"

''During several of the summer months
of the year 1783, when the effects of the
sun*s rays to heat the earth in these
northern regions should have been the
greatest, there existed a constant fog
over all Europe, and great part of North
America. This fog was of a permanent
nature; it was dry, and the rays of the
sun seemed to have little effect toward
dissipating it, as they easily do a moist
fog arising from the water. They were
indeed rendered so faint in passing
through it that, when collected in the
focus of a burning-glass, they would
scarce kindle brown paper. Of course,
their summer effect in heating the earth
was exceedingly diminished.

*' Hence the surface was early frozen.

"Hence the first snows remained on it
unmelted, and received continual addi-
tions.

"Hence perhaps the winter of 1783-4

light at midnight were probably made in northern
Europe above latitude 50".

•As quoted by Humphreys, Physics of the
Air, p. 570. 1920.

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

»«S S g t >

%- T

t 241

i IT;!

i \.

' • "^

sp ^''~

-k ^

1 >•

- :C

1 i "

- H-

? ^li..

Si f^^

? ^

'' %

I /^

! "*:

i -t|

1 -f :.

-4-'/t

^ i-

^ 5%' i

' 'f-^

1 ../.•

1 IV

i .^J-^

1 , .,( ■

8 ^ •. .

i ^^^""

« ! 8 « «

■a

C
CO

<r5

90
00

rt5

O u
5 ♦J

i5 Stf>

O
c/)

b
O

en
O

C C

c <«

o ^*

S 2"

c
o

was more severe than any that happened
for many years."

In 1815 occurred the eruption of Tom-
boro, in the East Indies. This, **if the
published reports are to be credited,
appears to have exceeded all other known
eruptions. It caused darkness lasting
for three consecutive days at a distance
of over 300 miles. Ash fell to a depth of
two feet more than 850 miles away.
Dust fell over an area of 1,000,000 square
miles. The explosions were heard at a
distance of 1,000 miles. The material
ejected has been variously estimated at
28.6, and even at 50 cubic miles."' In
this eruption 56,000 people are reported
to have lost their lives.

After Tomboro, streaky skies, hazi-
ness, and brilliantly colored sunsets were
conspicuous in Europe. At the same
time the temperature dropped about 1.1°
C. below the normal, producing such
marked effects on plant growth that the
year 1816 is known as "the year without
a summer," ** poverty year," or ** eight-
een hundred and froze-to-death."

REMARKABLE ATMOSPHERIC EFFECTS OF
THE ERUPTION OF KRAKATOA

The best known instance of world-
wide effects from an eruption is that of
Krakatoa. The eruption occurred in
August, 1883, but scant notice of it was
taken in Europe until months afterward,
when its remarkable atmospheric effects
compelled attention.

The dust from the volcano caused
such brilliant sunsets that they became
matters of universal comment. The
dust particles in this case were blown so
high that it required two years for them
to settle down to the level of the highest
clouds. During all this time remarkable
sunset colors, haze effects, and abnor-
mally long twilights occurred through-
out the whole world.

The amount of material thrown put in
this eruption, 3.2 cubic miles, was only
about half as much as from Katmai. But
in violence, the explosion far exceeded

» Martin, G. C, National Geographic Maga-
zine, 1913. P- 165.

Digitized by

Google

CLIMATIC EFFECTS OF THE ERUPTION

Katmai, for it was heard at much greater
distances.

In the Andaman Islands, 1,500 miles
from the volcano, the ''report was heard
as of a distant signal gun. A steamer
was sent round the coast to search for
the vessel supposed to be firing as signals
of distress.'** At a point in South
Australia 2,023 miles from Krakatoa,
people were awakened at midnight by
the explosion, which resembled blasting.
The extreme limit at which the sounds
were detected was, however, nearly
twice as far away. At Rodriguez, 2,968
miles distant, the detonations sounded
like '*the distant roar of heavy guns."

In addition to the tremendous sounds,
the explosion of Krakatoa set up another
type of atmospheric disturbance marked
by a fall in air pressure, as recorded by
self-registering barometers all over the
world. This remarkable air wave pro-
ceeding from the volcano spread entirely
round the earth to the antipodes; was
there reflected and passed back again to
the volcano, where it was again reflected
back to the antipodes; was again sent
back to the volcano; and so on, until it
had circled the globe seven times before
it finally became too faint to be detected.

After the eruption of Krakatoa the
intensity of the sunshine was reduced

* Quotation from Report to Royal Society, op.
cit., pp. 84-87.

* The data concerning this haze and its effects
have been collected by Kimball, Humphreys,
Abbot, and Fowle, from whose contributions I
gather the facts summarized below. For more
detailed accounts the reader is referred to the
original papers:

Abbot, C. G. Do Volcanic Explosions Affect
Our Climate? National Geographic Magazine,
vol. 24, pp. 181-198. 1913.

Abbot, C. G. and Fowle, F. E. Volcanoes and
Climate. Smithsonian Miscellaneous Collec-
tions, vol. 60, No. 29, 1913.

Humphreys, W. J. Volcanic Ehist and Other
Factors in the Production of Climatic Changes
and Their Possible Relation to Ice Ages. Bulle-
tin Mt. Weather Observatory, vol. 6, pp. 1-34,

1913.

, Physics of the Air, Philadelphia, 1920.

Kimball, H. H. The Dense Haze of June lo-
II, 1912. Bulletin Mt. Weather Observatory,
vol. 5, pp. 161-165, 1912-1913.

very much as in the other instances cited,
but in this case the greatest eflfect,
amounting to a 12 per cent loss in the
sun's power, was not reached until
nearly two years after the eruption.
These sky phenomena were accompanied
by a marked drop in the world's tempera-
ture — 0.45° C. (see page 32).*

VOLCANIC HAZE FROM KATMAI APPEARED

IN WISCONSIN BEFORE THE

ERUPTION WAS OVER

With these records of previous erup-
tions before us, we are prepared to accept
reports of world-wide effects from the
explosion of Katmai that would seriously
tax our credulity if they stood alone.

Within the two days after the begin-
ning of the eruption, June 8, a curious
haze different from anything ever before
seen by the observer appeared at Madi-
son, Wisconsin.* It assumed '*the form
of streaks, rapidly moving wave systems,
and clotted forms, observed from 11
a.m. until late in the afternoon.'** On
the next two days the "high haze con-
tinued but became denser, so that definite
forms were more difficult to identify."

By June loth the haze had spread to
Virginia, where it was so noteworthy on
the loth and nth as to induce Professor
H. H. Kimball to publish a special paper

, The Effect of the Atmospheric Tur-
bidity of 1912 on Solar Radiation Intensities
and Skylight Polarization. Ibid., pp. 295-312,

1913.

, The Effect upon Atmospheric Trans-
parency of the Eruption of Katmai Volcano.
Monthly Weather Review, vol. 41, pp. 153-159,

1913-

, The Relation between Solar Radi-
ation Intensities and Temperatures of the Air
in the Northern Hemisphere in 1912-1913; Bul-
letin Mt. Weather Observatory, vol. 6, pp.
205-220, 1914. (With full citations to other
literature).

, Volcanic Eruptions and Solar Radia-
tion Intensities. Monthly Weather Review, vol.
46, pp. 355-356, 1918.

See also Arctowski, Henryk. Volcanic Dust
Veils and Climatic Variations. Annals N.Y.
Academy of Science, vol. 26, pp. 149-174, 191 5.
(Also literature there cited.)

•Monthly Weather Review, 1913, vol. 41, p.
154.

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

CCNT

-10
-20

CtNT

100
90
80
C.

±0'

-O-

POLARIZATION.

■ 1 1 I

—

\
\

%
\

^0^

JL,

'7)-

*o-

S)-

I INSOLATI

1 1 •

ON.

%
\

-<r

—

%
\

^M^

. X

/<r

■^

-ft-

-XT

_V -

ER/

^"^

m^\-

.^v-

^?"

■*\

■>

SDL.

■^

*a-

•0.a
-1.0

-ty*

^li/^

1912

913

MAY

JUN

JUL

AUG SEP

OCT

NOV

DEC

JAN

FEB

MAR

APR

MAY

JUN

JUL

AUG

SEP

OCT

NOV

DEC

JAN

CURVES SHOWING THE EFFECT OF THE DUST FROM KATMAI ON:

I. Skylight polarization at Mount Weather, Vii^inia.

II. Insolation at Mount Weather, Virginia (O) and Madison, Wisconsin (X).
III. Temperature Departures in the Northern Hemisphere, for the period following the eruption.
(After Kimball.)

devoted to it, although at the time of
writing he had no inkling of its cause.

"On the 8th and 9th the atmosphere
was unusually clear for June."^ But on
the loth "the upper atmosphere was
filled with a white veil of haze," although
the lower air "was still clear so that
mountains 30 miles distant were dis-
tinctly visible." . . . "Just before noon
the polarization of skylight, which was 67
per cent at 9 a.m. of the 9th had decreased
to 14 per cent, and the atmospheric trans-
mission coefficient, which was 0.756
shortly before noon of the 9th, had
decreased to 0.589."

'H. H. Kimball in Bulletin Mt. Weather
Observatory. 191 2-13, vol. 5, pp. 162-3.

Abbot and Fowle, who were engaged
in studying the intensity of solar radia-
tion, the former in Algeria, the latter in
California, soon afterward found their
observations interfered with by the same
peculiar haze, which, needless to say,
was most unexpected in the desert regions
which they had chosen especially because
of their freedom from clouds. Abbot
describes his experience thus:*

PECULIAR HAZE PERSISTED ALL SUMMER
IN ALGERIA

"While observing on June 19, 1912, I
noted streaks resembling smoke lying

' National Geographic Magazine, February,
1913, p. 181.

Digitized by

Google

CLIMATIC EFFECTS OF THE ERUPTION

along the horizon, as if there were a forest
fire in the neighborhood of the station.

"These appearances continued and
were soon joined by others more notice-
able. After a day or two we began to see
peculiar mottled figures like those of the
mackerel sky, although absolutely no
clouds were present. The phenomenon
became so marked that we ceased en-
tirely our observations of solar radiation,
as the sky seemed to be quite too poor
for such work.

"About the 25th of June a cloudy
period began, with rain, and, after this,
of course, we expected that all these curi-
ous sky phenomena would have passed
away ; but not so, for when the clouds had
quite passed by, about the 29th of June,
we found that the whole sky was filled
with haze, and this state of affairs con-
tinued even more pronounced until the
expedition left Algeria, about Septem-
ber 10."

Unusual haze was noticed during this
summer at many other stations, both in
the United States and in Europe, but
few made the precise instrumental ob-
servations that give especial significance
to the accounts above quoted.

The dust veil interfered also with the
work of the astronomical observatories.
At the Yerkes Observatory "all observers
using the Bruce Spectrograph noted that
exposure times had to be lengthened." •
The Director of the Dominion Observa-
tory, Ottawa, reported: "Photographs
of the solar spectrum show a marked de-
crease in intensity (apparently not less
than 25 per cent) ." • P. G. Nutting, work-
ing at the Bureau of Standards, Wash-
ington, states: "We detected a change
in color of direct sunlight by spectro-
photometric comparison with acetylene.
First noted on June 11," • the next day
after it was observed in Virginia.

In the train of this haze was a more or
less marked increase in the intensity of
sunset colors, somewhat like that which
followed the eruption of Krakatoa in
1883, but so much less pronounced as not

• Monthly Weather Review, vol. 41, p. 157,
1913.

to have attracted general notice, being
recorded only by a few careful obser^'^ers.

THE DUST CLOUD ABSORBED TEN PER
CENT OF THE SUN's HEAT

The haze had the effect of intercepting
a very considerable fraction of the sun's
rays and reflecting them back into space,
materially reducing the intensity of the
sunshine.

Abbot and Fowle's measurements
summarized in the table given below ^'^

Table Showing Diminishbo Hbating Power of Sun-
light Due to the Dust from Katmai

Total radiation

Values in
calories

Ratio to
early June

Percentage
depletion

A Mount WUson. California

Early June . . .
Late June ....

July

August

Haziest Days.

1.49

1. 00

1.48

.99

1.33

.89

1.24

.83

1. 13

.76

II
17
24

B. Bassour, Algeria

Early June . . .

Late June

July

August

Haziest Days.

- 1.39

1. 00

1.37

.99

1. 14

.8a

1. 14

.83

.99

.71

18
18
29

show "that the uncommon haziness of
the sky during the summer of 191 2
produced a very marked decrease in the
direct solar radiation in all parts of the
spectrum, and reached nearly 20 per
cent at high sun for the total heat.*'"

There was, however, some compensa-
tion for this great loss in an increase of
the brightness of the sky, due to the
fact that dust particles reflect a portion
of the light down to the earth's surface
as well as off into space. The effect of
this diffuse reflection was to increase the
twilight both in duration and intensity.
Taking everything into consideration,
Abbot concluded **that the dust of
Katmai diminished the heat available to

^^ From Smithsonian Miscellaneous Collections,
vol. 60, No. 29, p. 6, 1913.

^i National Geographic Magazine, February,
1913, pp. 184-186.

Digitized by

Google

Photograph by B. B. Fulton
MASSES OF MOSS ON THE TREES NEAR KODIAK

These moss balls held quantities of the falling ash, which have since been consolidated and bound in

place by new growth.

Digitized by

Google

CLIMATIC EFFECTS OF THE ERUPTION

warm the earth in the North Temperate
Zone by about lo per cent during the
summer of 1912.

THE WHOLE NORTHERN HEMISPHERE WAS
COOLER BECAUSE OF THE ERUPTION

**In accordance with the laws of heat
and radiation, this might produce a fall
of 7° C. [12.6° F.] in the temperature of
the earth as a whole, if it was effective
for a long enough period of time, pro-
vided that there were no counteracting
influences." The records show that the
cooling of the earth postulated by Abbot
actually occurred.

As summarized by Kimball*^ they re-
veal a deficiency in temperature amount-
ing to 0.9° C. [1.6° F.], which reached its
maximum in September.

None of the eruptions since the be-
ginning of the record has produced so
striking an irregularity in the tempera-
ture curve as Katmai, for the reason that
it occurred at a time when terrestrial
temperatures were at a maximum,*^ so
that the depression it caused appears as
a notch out of the top of the curve shown
on page 32, whereas the other great
eruptions happened to come at periods
when their depressing effect on the
temperature curve is not so conspicuous.

A reduction of the earth's temperature
by a degree or two will doubtless appear
to most people a trivial matter, without
any bearing on the affairs of everyday
life: this is a great mistake.

We are all so used to variations of sev-
eral degrees from hour to hour and from
day to day that we are likely not to real-
ize that, although short-period daily va-

" "Temperatures below the normal commenced
to prevail throughout Alaska, the United States,
and Mexico in June, 1912; over Canada, the
Atlantic Ocean, Europe, and northern Africa in
July; over Asiatic Russia to longitude 105** E. in
August; and over the eastern part of Asia in
September. Low temperatures persisted gen-
erally in low latitudes during the remainder of
191 2 and throughout the summer of 191 3, with
a maximum deficiency for the whole hemisphere
of 0.9° C. [1.6° P.] in September, 1912. There
was an excess of temperature in high latitudes
[due perhaps to increased sky light] during the
winter of 1912-13, with a maximum departure

riations compensate each other, the cu-
mulative effect of a slight reduction of
temperature continued over a long time
becomes very important. A change of
even i° F. continued through the season
has a marked effect on plant growth. A
deficiency of that amount entails a shift
of fully 80 miles in the limits at which
crops can be grown. That is to say,
other things being equal, the occurrence
of an eruption like that of Katmai means
that wheat would fail to mature over a
belt 80 miles wide along the northern bor-
der of its range.

In addition to changing the area over
which crop production is possible, such
changes in temperature may greatly
affect the time required for plant devel-
opment. Thus, a deficiency of only 1 .26**
F. has been known to postpone the ma-
turing of sugar-cane by a full year!

VOLCANOES CAPABLE OF CAUSING ICE
AGES

An eruption like that of Katmai would
lower the temperature of the world very
much more than 1°, if the dust remained
in the air long enough for its full effect
to be felt. Abbot and Humphreys agree
that the dust from Katmai would have
reduced atmospheric temperature 6** or
7*" C. (about 12° F.), if the dust had re-
mained in the air long enough. for the
earth to adjust itself fully to. the de-
creased radiation and no compensating
factors had intervened to counteract its
effect.

Such a lowering of terrestrial tempera-
ture, *'if long enough continued, would
be more than sufficient to produce

for the hemisphere of 0.7° C. [1.3° F.] in Decem-
ber. The average daily deficiency of tempera-
ture for the whole Northern Hemisphere for the
period of June, 1912, to October, 1913, inclusive,
was 0.16° C. [0.29° F.]" Bulletin Mt. Weather
Observatory vol. 6, pp. 214-15, 19 13.

*' Terrestrial temperatures vary inversely with
the spots on the sun's disc, reaching a maximum
when sunspots are fewest and vice versa. The
discrepancies between the temperature and sun-
spot curves, T and S (page 32) occurred "in
every important case simultaneously with violent
volcanic eruptions." Humphreys. Bulletin Mt.
Weather Observatory vol. 6, p. 24, 1913.

Digitized by

Google

Photograph by Robert F. Griggs
A DEVIL CLUB GROWING THROUGH THE ASH

There are two sets of roots, one just beneath the surface, the other in the old soil. The three layers
of ash are well shown. The ashfall in Lower Katmai Valley averaged 20 inches.

Digitized by

Google

CLIMATIC EFFECTS OF THE ERUPTION

glaciation equal to the most extensive of
any ice age.**^*

Whether volcanic explosions actually
brought on the great ice ages to which
the earth has been subjected repeatedly
in the past is, of course, another question.
Because of the ephemeral character of
their products, it is difficult to ascertain
whether such explosions were frequent
enough at any given time in the past to
produce the glacial periods recorded.
But, even in historic times violent erup-
tions have been so frequent that they
have markedly reduced the earth's
temperature throughout that period.
Humphreys concludes on this point:

** Since the beginning of reliable rec-
ords, say 160 years ago, the average tem-
j>erature of the earth has been percepti-
bly lower, possibly as much as i** F.,
than it would have been if, during all
this time, there had been no volcanic
explosions violent enough to put dust
into the isothermal region of the atmos-
phere. Similarly, on the other hand, if,
during this period, violent volcanic ex-
plosions had been three or four times
more numerous than they actually were,
our average temperatures probably
would have been at least 2° F. to 3° F.
lower, or low enough, if long continued,
to bring on at least a moderate ice age.'* ^^

WATER LIBERATED IN ERUPTIONS PROB-
ABLY OF VAST IMPORTANCE IN THE
ECONOMY OF NATURE

It would not be proper to bring this
chapter to a close without brief allusion,
at least, to one other way in which the
eruption exerted a world-wide influence
of far-reaching significance.

It is recognized that the chief agent in
the tremendous explosions was the ex-
pansion of gases imprisoned in the mol-
ten ms^ma, as the lava is called while
yet in the bowels of the earth. We ought

^ Humphreys, Physics of the Air, p. 590.

>* Bulletin Mt. Weather Observatory, 1913,
vol. 6, pp. 33-34-

w As quoted by F. W. Clarke, Data of Geo-
chemistry, 4th ed., p. 209, 1920. This work gives
authority for my other statements relative to the
contributions of volcanoes to the atmosphere.

to make some inquiries concerning the
quantity of the gas thus liberated, its
source, and its effect on the atmosphere.

Analysis shows that the ash has practi-
cally the same composition as granite.
If the magma had not been extruded, it
would have become granite. Now it
has been demonstrated that granite, if
heated to the temperature prevailing in
a volcano, gives off water and gases
identical with those of an eruption.

The celebrated French volcanologist,
Gautier,^' estimated that a cubic kilo-
meter of granite heated to 1,000** C.
would give off 160,000 millions of cubic
meters of steam and 28,000 millions of
cubic meters of other gases. The magma
from which the ash and pumice came,
before expansion, amounted to some-
thing like ten cubic kilometers. Hence, if
these data represent conditions in Kat-
mai, the ash and pumice thrown out
would have liberated in round numbers
56,000,000 million cubic feet of steam
and 10,000,000 million cubic feet of other
gases.

As a matter of fact, however, a highly
explosive magma like that in the recent
eruption probably contains far more gas
than the average granite. Our estimates
are, therefore, presumably much too low.
Besides, they include only the gases set
free by the ash thrown out, and take no
account of the vast quantities of steam
which have been given off continuously
ever since the eruption, without further
extrusion of solid matter. The numerical
value given is, therefore, too small, but
the estimates serve to suggest the enor-
mous quantity of steam given off in a
great eruption.

Even this quantity of steam, repre-
senting only a fraction of the total out-
put of the volcano, would form, when it
condensed, about 300 million long tons
of water.^^

*^ It must be recognized that this figure, great
as it is, is relatively minute when compared with
either the stock of water in the ocean or the rain-
fall of the world. Its significance lies not in the
size but in the fact that it is ?tew water, a per-
manent addition to the ocean.

Digitized by

Google

PQ
Q

■M

»
O

H
Ou

8
(/)

C bfi

c o

^*-

** C

o «

■M U

•|8

a
o

Digitized by

Google

CLIMATIC EFFECTS OF THE ERUPTION

Whence came this great volume of
water?

Some have maintained that it was due
to the percolation of sea water into the
molten magma in the depths. But this
view does not stand up well under close
examination. It is more probable that
it is ** juvenile water'* brought up from
the interior of the earth for the first
time, forming, therefore, a permanent
addition to the waters of the earth's
crust.

In the words of that great master,
Exluard Suess,^^ "Volcanoes are not fed
by the infiltration of the sea but the
waters of the sea are increased by every
eruption." If this view is correct, vol-
canoes play a r61e of supreme importance
in the economy of nature, for their erup-
tions are the principal means by which
the oceans and all of the waters of the
world have been separated from the solid
crust. If there had been no volcanic
eruptions in the past, the earth would be
an uninhabitable desert on which life
could not have developed.

ALL LIFE PERHAPS DEPENDENT ON THE
CONTINUANCE OF VOLCANIC ACTION

One other product of eruptions re-
quires consideration: carbon dioxide.
This gas is an absolute essential to green
plants; it is the indispensable basis from
which all human foods are built up.
Moreover, it has been locked up in coal
and in limestone in such enormous
quantities during past ages that the
original source of supply has always been
a puzzle to geologists.

It may be seen how vital this problem
is when it is stated that the amount of
this essential compound that has been
withdrawn from the atmosphere and
stored away in coal and limestone *'is
about 30,000 times as much as is now
contained in the atmosphere.*' ^'

But it is not probable that any such
quantity of carbon dioxide was present
at one time in the atmosphere of the
primeval world. If it were all liberated

*• Geographical Journal, vol. 20, p. 520, 1902.
» F. W. Clarke, op. cit. p. 48.

at once, it would raise the pressure of the
air to several hundreds of pounds per
square inch and make impossible the life
of any such organisms as we are ac-
quainted with today. No sort of animal
life can endure even a moderately high
concentration of carbon dioxide.

The abundance of animals throughout
the whole period when carbon was being
laid away in the rocks indicates a uni-
formity of conditions throughout geolog-
ical history and argues against any great
concentration at the beginning. Indeed,
it is largely by the activity of organisms
of one sort or another that all carbon de-
posits, whether in coal or limestone, were
produced.

The facts seem to require us to suppose
that carbon dioxide has been steadily
supplied from some source at about the
same rate at which it was consumed
throughout the geological ages. No
terrestrial source other than the gradual
additions known to be brought in by
volcanoes has been suggested.

We do not know the history of vol-
canic eruptions well enough to enable us
to be certain whether they have contrib-
uted enough in the past to have main-
tained the supply. But we do know
that the amount of this indispensable
substance which they bring up from the
interior of the earth is very great. There
is no means of estimating how much may
have been liberated by Katmai. But an
ordinary quiescent volcano may give oflF
tremendous quantities. It has been
stated, for example, that Cotopaxi pours
forth more carbon dioxide than a city
like Paris (100 million cubic feet per
day).2o

Although our data are not sufficient to
permit assurance, it is evident that we
are here dealing with one of the most
fundamental questions in the whole realm
of nature and one of extreme importance
to mankind. In the words of F. W.
Clarke: 2* ** Obviously, if the volcanic
hypothesis be true, the cessation of vol-
canism would signify the end of life on

^ F. W. Clarke, op. cit. p. 47.
« Ibid., p. 56.

Digitized by

Google

44 THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggt
SLEEK GALLOWAY CATTLE BELONGING TO THE EXPERIMENT STATION AT KODLAK

After the eruption the station herd had to be taken to "the States*' for the first two years; but
their pastures made such a remarkable recovery that they were soon returned. A stranger would
hardly suspect that this country had been buried under a foot of ash only four years previously.

the globe. It would be followed by the Suggestive as are speculations of this
consumption of all available carbon diox- nature, their pursuit is beyond the scope
ide, so that plant life, and consequently of this book. We must return to a con-
animal life, could no longer be sup- sideration of our specific subject — the
ported/ ' more immediate effects of the eruption.

Digitized by

Google

THE RECOVERY OF VEGETATION AT KODIAK^

The effect of the ash blanket at Kodiak,
coming as it did just when the plants
were putting forth their spring growth,
was to strike down all weaker plants,
giving "Green Kodiak" the appearance
of a pine barren devoid of vegetation, ex-
cept for the trees and bushes which pro-
truded through the ash uninjured. To
everyone who visited Kodiak during the
first two seasons after the eruption, the
damage done to vegetation seemed ir-
reparable.

It was during this period that I first
saw Kodiak, in June, 1913, almost
exactly a year after the eruption. It
was indeed a bleak and desolate prospect.
Outside the forest the country had the
appearance of a desert, whose gray-
brown slopes were relieved only here and
there by spots of green where some alder
or willow pierced the ashy blanket
(see page 22).

Lupines, fireweeds, and other strong-
stemmed perennials had, to be sure, come
up through the ash in places, but they
were not abundant enough to lend much
color to the landscape. Within the
forest the prospect was less desolate.
Yet even here the undergrowth was gone,
and the branches were still heavily laden
and bent down to the ground with ash.

The officials of the Experiment Station
and of the Kodiak Baptist Orphanage
were attempting to grow a crop of oats,
to provide ensilage enough to keep their
cattle alive over the winter. We felt it
our duty to encourage them in this effort,
but in reality the prospect seemed very
gloomy.

I well remember debating with one
of the citizens whether the country would
ever come back to its original condition.

' For a more detailed discussion of this aspect of
the eruption^ with citations to the literature, the
reader is referred to a paper by the writer under
the same title in the series of "Scientific Results

With the vividness of his memory of
things as they had been before the erup-
tion, he was pessimistic; but believing
that the ash would probably be bene-
ficial after it was incorporated with the
soil, I reassured him with the prediction
that in ten years vegetation would begin
to come back in some abundance.

THE REMARKABLE RETURN OF
VEGETATION

The recovery of the area was, however,
extremely rapid. During the second and
third years the old roots of the stronger-
growing plants sent up new growth
through the ash layer in such profusion
as to upset completely even the most
optimistic of predictions.

When I landed in June, 1915, I could
not, despite the reports I had received,
believe my eyes. It was not the same
Kodiak I had left two years before. The
mountains were everywhere green with
their original verdure. The character of
the change is indicated by the picture on
page 42 better than by any description.
Where before had been barren ash there
was now rich grass as high as one's head.
Everyone agrees that the eruption was
*'the best thing that ever happened to
Kodiak." In the words of our hotel
keeper, "Never was any such grass be-
fore, so high nor so early. No one ever
believed that the country could grow so
many berries nor so large, before the
ash'* (see page 51).

I had come to study the revegetation,
but I found my problem vanished in
an accomplished fact. The revegetation
which I had hoped to study at Kodiak
had given place to a remarkable recovery
of the antecedent plants, transferring the
problem of revegetation proper to the

of the Katmai Expeditions of the National Geo-
graphic Society,*' Ohio Journal of Science, vol.
19, pp. 1-57, 1919.

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs
ASH DRIFT LODGED BEHINT) FIREWEEDS

The drifts have every appearance of snow banks. The fireweed endured such burial for three years,

but then succumbed.

more deeply buried country near the cerning the fertility of volcanic soils-
volcano. Some basic lavas have all the ingredients

of a rich soil. Some acid rhyolites are

THE LUXURIANT GROWTH NOT DUE TO ANY , ^ «* j a

almost as poor as quartz sand. As

shown by the subjoined analysis, the ash

There is a widespread idea that the from Katmai is in the latter class.* It

luxuriance of growth was due to some has in fact practically the composition of

fertilizer added to the soil by the ash, .

which stimulated the plants. This is er- rl^"^^/^^ ^jS^t- ^ f'^'^u ^n^^^

, t- r •! •! Chemist of Washmgton, for the United States
roneous. The ash forms as stenle a soil Department of Agriculture. Sample was col-
as can well be imagined. lected at Kodiak and consisted of all three layers
The belief that volcanic ash forms mixed so as to give a fair average of the condition
good fertilizer rests largely on miscon- encountered by plant roots.

ception. It is commonly supposed that sil^^"(s!o'^/^" 2 '^6'^''

all lavas are very much alike. The fer- ^'J^^ ^\^^ (Fe^O,)..'; . . . ^1%

tility of soils derived from the weathering Manganese oxide (MnO) 0.41

of basaltic lava flows is well known. It is Titanium oxide (TiOt) trace

assumed that volcanic ash has the same Alumina (Al»Oj) 13 85

property. There is, however, a very li""^ ^^^^?xV,w^^ ^!?

*^ *V J • 'j- ' ' 1 T-u Magnesia (MgO) 0.47

great diversity in igneous rocks. The g^^ (^^^j 3 86

content of inert silica may be as much as Potash (KjO) 2 .43

77.0 per cent or as little as 35.0 per cent. Sulphuric add (SOj) 0.20

Potash varies from 8.6 per cent to 0.04 Phosphoric acid (P,0,) Q-36

percent; lime from 15.5 to 0.2 per cent; ^^^^j 101.05%

phosphoric acid from 2.6 per cent to ^ , ^ ,...,, . ...

nth tr Remarks: The ash is highly magnetic; in all

^ J™'^^' . re . probability some of the iron present is magnetic.

These figures are quite sufficient to of more interest in the present connection is an

show the fallacy of generalizations con- analysis made by Prof. C. W. Foulk, of the Ohio

Digitized by

Google

THE RECOVERY OF VEGETATION AT KODIAK

Photograph by R. F. Griggs

HORSETAIL SENDING RUNNERS INTO BARE ASH

This plant has occupied large areas which otherwise would have become dunes of drifting ash.
importance in the revegetation of the country cannot be overestimated.

Its

pulverized granite. And granite noto-
riously forms a poor soil.

In order to put the fertility of the ash
to a practical test, a number of experi-
ments were set up in which it was used as
a soil, either alone or in various mixtures.

One series of these pot cultures is
shown on page 50. The plants in the
lower row of glasses were grown in the
ash ; in the upper row in the purest quartz
sand. To each series were added ferti-
lizing salts as indicated by the symbols
on the labels. Whether in sand or in ash,
the plants turned red and died of starva-
tion as soon as the food stored in the seed
had been consumed, unless they were
supplied with a complete fertilizer.

Of the different fertilizer salts, nitrates
were more beneficial than any others
used singly. This accords with field
experience at Kodiak. With the use of
nitrate fertilizers very fair crops of oats
were grown for ensilage the first year
after the eruption. Curiously enough.

State University, to determine the quantity of
iK'ater-soluble salts, for plants can use only the
soluble constituents of the soil. He found the
available potash (KtO) 0.05 per cent, and the

it was not possible to duplicate these
results in succeeding seasons, perhaps
because the small quantities of other
essential salts present were leached out of
the ash during the first year.

BENEFICIAL EFFECT OF ASH DUE TO ITS
ACTION AS A MULCH

The cause of the improvement in the
vegetation of Kodiak is to be sought not
in any chemical added to the soil by the
ash, but in its action as a mulch.

It destroyed the smaller plants and
gave the quick-growing species, which
are of most interest to man, more favor-
able opportunities for growth than they
had ever enjoyed under normal condi-
tions. It helped the surviving plants
very much as weeding improves a
garden.

Probably the most striking example of
this fact is furnished by the trees, whose
rate of growth increased remarkably
after the eruption. A section of a tree

phosphoric acid he denominated as a "small
amount." This, according to Professor Foulk,
was rather more than a trace, yet it was so small
that it would have been difBcult to weigh it.

Digitized by

Google

THE VALLEY OF TEX THOUSAND SMOKES

Photograph by George C. Martin
INSPECTING AN EXHUMED GARDEN AFTER THE ERUPTION

Captain Perry and Mr. Learn of the Kodiak Baptist Orphanage speculating on the possibilities of
growing vegetables after the eruption. The result may be seen by a comparison with the picture re-
produced on the opposite page.

from the ash-covered district pictured on
page 53 gives clear testimony of the ex-
tent to which the ashfall bettered grow-
ing conditions.

For twenty years this particular tree
had been struggling along, growing very
slowly. During all that time it had
reached a diameter of only 0.6 inch (16
mm.). Then came the eruption. The
record of that year appears as a ring of
extreme thinness, scarcely perceptible in
the photograph. In 1913 the tree re-

covered and did a little better than in any
previous year. But the next year there
came a great jump. Five times as much
new wood was produced as before. This
greatly increased growth continued there-
after till the tree was cut. In four years,
it had more than doubled its diameter.
The trunk selected for figuring was an
extreme case — a suppressed tree from a
former bog. Its reaction to the ashfall
was, however, typical. In habitats orig-
inally more favorable the percentage of

Digitized by

Google

THE RECOVERY OF VEGETATION AT KODIAK

Photograph by B. B. Fulton
THE SAME GARDEN THREE YEARS LATER
All anxieties concerning the size of the crop have long since vanished.

increase was naturally not so great; yet
it is plainly evident in trees from all sorts
of situations. Trees of all ages and of all
species were, moreover, similarly affected.
Even if the ash were rich in fertilizing
salts, the good it did the trees by destroy-
ing the grass would probably outweigh
any possible fertilizing effect. It has
been shown by experiment * that com-
petition among closely crowded plants,
whether of the same sort like the grasses
of a meadow, or of different sorts like
trees and undergrowth in a forest, stunts
growth to a surprising extent.

' This will be discussed in detail with citations
to the literature in a special paper devoted to the
subject.

Before the ashfall the trees were in the
condition of an orchard that has been
allowed to run to grass. Removal of the
sod from a neglected orchard will often
more than double the growth. The ash-
fall had the same effect on these trees
that cultivation has on an apple orchard.

REVEGETATION DUE TO RECOVERY OF OLD
PLANTS

It was assumed by all who visited
Kodiak during the first two seasons fol-
lowing the eruption that the smaller
plants had been practically exterminated,
except for the few individuals so situated
that they could readily grow through the
ash layer. Consequently, when the re-

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

SOIL VALUE OF KATMAI ASH AS COMPARED WITH SAND

Wheat seedlings in the upper row were grown in quartz sand, the lower in ash from Kodiak. The
pots at the left were untreated. To the others fertilizers were added as indicated by the symbols on
the labels (N, Nitrogen; K, Potassium; P, Phosphorous). Unless a complete fertilizer was supplied,
the plants turned red and died when the nutriment stored in the seed had been exhausted.

markable growth of succeeding seasons
was observed, it was natural to suppose
that the new vegetation must consist of
new plants which had started in the ash
from seed.

Field study at once showed this to be
incorrect. Even the most superficial
observation indicated that at the begin-
ning of the fourth season (191 5) there
were practically no seedlings, most of the
new plants being directly traceable to the
old roots.

An excellent demonstration of this
fact was furnished by the condition of a

field on the Frye-Bruhn ranch, south of
Kodiak, which had been plowed before
the eruption. Where cultivation had
destroyed the weeds, no new vegetation
appeared for five years, but the plants of
the uncultivated land all around came up
in undiminished vigor and tx)mpletely
covered the ground. The contrast be-
tween the bare field and the luxuriant
growth surrounding it was conspicuous
at a distance of five miles (see page

54).

Excavation around the underground
parts of the new vegetation always re-

Digitized by

Google

THE RECOVERY OF VEGETATION AT KODIAK

Photograph by M. G. Dickman

A BRANCH OF SALMON-BERRIES, INDICATING THE PROFUSION OF WILD BERRIES AT

KODIAK SINCE THE ERUPTION

These berries are somewhat like the persimmon, in that they have an astringent taste that disap-
pears only when they are dead ripe. They have, however, a distinctive and extremely delicate flavor,
and when served with sugar and cream equal or surpass any other berry with which the author is
acquainted.

vealed its connection with old roots in the
original soil.

A number of species of trees and
shrubs were found to have developed
striking "two-storied" root systems by
putting out new roots from their stems
just below the surface of the ash (see
page 133). It was noteworthy that the
new roots were confined to the top of the
ash. The intermediate zone between
the surface and the old soil was as bare
as though still growing in air.

PLANTS RESURRECTED AFTER THREE
years' BURIAL

If the recovery of vegetation had oc-
curred during the first season after the
eruption, it would have occasioned no
particular surprise. It was the delay in

*For details, see the appropriate technical
papers in the Scientific Results of the Katmai

the reappearance of the plants that upset
all predictions.

Everyone assumed that only that
small percentage of the old plants which
reappeared within a few weeks had sur-
vived. But, as demonstrated by the
event, practically all of the members of
the plant society had the capacity of re-
covering after having lain dormant be-
neath the ash for an astonishingly long
period.

Both at Kodiak and on the mainland
it was proved beyond question* that the
old roots could send up new growth after
a period of enforced dormancy of three
years. In a number of cases, where
there could be no doubt as to the dura-
tion of the dormant period, the new
growth was as vigorous as though it had

Expeditions, Ohio Journal of Science, vol. 19, pp.
32. 195.

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by Robert F. Griggs
WHERE NO PLANTS COULD COME THROUGH ON
THE LEVEL

Horsetail sometimes penetrated the reduced thickness of
ash in the bottoms of the gullies.

been buried only a week or two (see
also page 153).

The bulk of the vegetation that
wrought such a change in the appearance
of Kodiak after two years consisted of the
same species of strong-growing plants that
had appeared from the first in smaller
numbers. After these had reappeared in
full force a number of the smaller plants,
such as the little ''Kostinika" berry
(Rubus pedatus), the baked-apple berry
{Rubus chamaemorus), and the mountain
cranberry {Vitis-Idaea) , also began to
grow up from the old roots, under con-
ditions that made it certain that they
had lain dormant through the interim.

SEEDLINGS SLOW TO START

These luxuriant hold-over plants stand
in strong contrast to the feeble seedlings

which in places came up in the
ash. At the Experimental Farm
at Kalsin Bay a plot was seeded
down with timothy soon after the
eruption. Although the seeds
had sprouted well and given a 100
per cent stand, the plants were
mostly only about 3 inches tall
at the end of three years (see
page 54). The contrast between
this grass and that shown on page
42, which in places penetrated 20
inches of ash, requires no com-
ment.

THE SAND BLAST

The importance of surviving
plants in the re vegetation of
Kodiak cannot be overestimated.
Where there was no plant cover
to protect the surface of the
ground, the loose ash was picked
up in clouds by every passing
wind, forming a sand blast that
was very hard on the plants ex-
posed to it. All of them were
lopped over before the wind, and
their lower leaves were either cut
to pieces by the sharp sand or
buried beneath it (see pages 46
and 59).

The particles of ash are all very
sharp, sharper than ordinary sand. In-
deed, volcanic ash forms thd basis of nu-
merous commercial scouring agents. The
ash is also finer and much lighter than
shore sand, and it is more easily carried
by the wind. Consequently, this sand
blast is a very different thing from the
drifting sand common among beach
dunes.

One might suppose that the frequent
rains which characterize the climate of
the region would hold the sand blast in
check, but it is surprising how quickly it
starts again after the rain stops. We
found that after a day of soaking rain the
sand was blowing early the next morn-
ing, although only the very surface had
dried off.

This blowing .ash lodged behind any
obstruction, like snow. Among the

Digitized by

Google

THE RECOVERY OF VEGETATION AT KODIAK

weeds at the edge of cultivated fields and
along the fence rows, drifts two feet deep
were formed. On mountain tops and in
other places where there was no vegeta-
tion to catch the ash, it formed dunes
like those on a seashore (see page 59).

One could see no prospect of revegeta-
tion in such situations until the ash sur-
face should become stable enougli neither
to bury the young plants before they
were well started nor to blow away the
soil and expose their roots.

IMPORTANCE OF HORSETAIL AS A SOIL
BINDER

It was of the utmost importance that
the ground be covered with vegetation,
regardless of the value of the plants
making the cover.

Of all the native plants the one which
could grow through the deepest ash and,
once through, spread most rapidly on the
bare surface was the field horsetail
{Equisetum arvense). This is a common
weed of railway embankments and like
places in the United States. In Kodiak
scattered individuals were frequent be-
fore the eruption, though they formed no
noticeable element in the landscape.
But it has come up everywhere through
the ash and has spread out on the surface,
forming in many places a beautiful
greensward, where hardly anything else
could come through.

Its present abundance contrasts so
greatly with its former state that, ac-
cording to Mr. Snodgrass of the Experi-
ment Station, some of the natives thought
it must have '*come with the ash," and
could only be convinced to the contrary
when he dug out the rootstocks and
showed that they originated in the old
soil beneath the ash.

A deposit of 10 or 12 inches would
have been fatal to most plants, if it had
not been for the contraction cracks in the
ash; yet the horsetail in many places
came through from 30 to 36 inches of ash !
This was especially well shown in the
deep deposits on the mainland, where,
with the ash four or five feet deep, nothing
at all could come through on the level;

THE EFFECT OF THE ASHFALL ON TREE
GROWTH

The slow gro^'th before the eruption is recorded
in the thin central rings. The great increase in
vigor since the ashfall is registered by the hea\'y
outer rings. Twice natural size (see text p. 48).

but little gullies in the ash were green
with a carpet of horsetail (see picture
opposite).

The description given thus far applies
largely to the open country. In the
forest the thick trees greatly modified
conditions by protecting the soil from the
wind and increasing the humidity of the
air. Their branches still retain a good
deal of ash in places (see p. 28).

In the deepest, most humid parts of
the forest great masses of moss grew
on the branches. These have since
grown out through the ash and consoli-
dated it into heavy balls, which give
the trees a very bizarre appearance (see
page 38).

THE NETWORK OF MOSS ON THE FOREST
FLOOR

More striking than anything in the
trees, however, was a remarkable net-
work of moss which developed on the
forest floor. When the ash dried up
after the heavy rains following the erup-
tion, deep cracks appeared, like the mud
cracks in a dried puddle (see page 58).

The cracks were, of course, quickly

Digitized by

Google

54 THE VALLEY OF TEN THOUSAND SMOKES

Photograph by D. B. Church

A PLOWED FIELD, PART OF WHICH WAS CULTIVATED JUST BEFORE THE ERUPTION

New plants were unable to start in place of the weeds destroyed by cultivation — ^illustrating the

importance of residual vegetation.

Photograph by R. F. Griggs
A TIMOTHY PLAT THREE YEARS OLD

Sowed in the ash soon after the eruption, the seed came up well and most of the plants are still alive
but have made no growth. Contrast with the grass come up from old roots, shown on page 42.

Digitized by

Google

THE RECOVERY OF VEGETATION AT KODIAK

filled up by drifting ash, but their posi-
tion was preserved by a heavy growth of
moss, which formed a most curious net-
work all over the ground. The micro-
scopic spores carried by the wind had
settled into the cracks. The same moss
was to be found around fallen sticks or
other objects that would catch wind-
borne spores.

SEEDLINGS OF ALL SORTS STARTED IN
THE FOREST

In the forest the trees protected the
ground from the wind, and insured a
stable surface on which new plants could
start. It was several years after the
eruption, however, before seedlings made
their appearance in any numbers, even in
the most protected situations. None
were observed in 191 3, and in the begin-
ning of the season of 1915 they were few
and far between.

But during the latter part of 19 15 they
began to appear in numbers. The seed-
lings included representatives of all the
important members of the flora. Al-
though they grew very slowly, still they
persisted, and some of them survived the
winter.

But seedlings had assumed no im-
portant place in the vegetation even five
years after the eruption.

It is already clear that the recovery at
Kodiak is permanent. For with the
demonstration that seedlings can start in
the ash, it is evident that any gaps which
may develop in the ranks of the old vege-
tation will be filled promptly by new
plants starting from seed.

The unstable conditions caused by the
shifting sand will be, for the most part, of
short duration. Where the ash layer
was only a foot thick originally it will
not require a very long period for the
wind to remove the whole deposit.

Indeed, in many exposed situations
the ash is already completely gone.
And in any case, it will be only a few
years before the larger part of the ash
from the mountains has been blown out
to sea. Even on the level some places
have already been nearly cleared. Thus

the field at the Frye-Bruhn ranch, pre-
viously spoken of, retains at present only
enough ash to veil the black soil be-
neath, and the sand blast is almost a
thing of the past.

In places sheltered from the wind,
erosion by water has proceeded rapidly,
and where the land is even moderately
steep, nearly all the ash has been washed
away. On the average grass-covered
mountain sides, the present covering of
ash amounts to only an inch or two and
this is so mixed with plant stems and
roots as to form a very indefinite layer.
By 1919 it appeared almost incredible to
a stranger that the ash had stood a foot
deep only seven years before.

EARTHWORMS WORKING OVER ASH

Another factor which is tending to
destroy the identity of the ash as a
separate layer is the action of earth-
worms. Ever since the classic researches
of Charles Darwin it has been recognized
that the constant working over of the
soil by these creatures is of great impor-
tance, by reason of the continual additions
of new material which they bring up
from below and void on the surface.

At Kodiak their castings are abundant
on the ash surface. From the character
of the castings it would appear that the
worms are confining their activity largely
to the ash itself; but, even so, their action
will serve to bury increasing quantities
of vegetable debris. And where they
bring up the old soil from beneath the
ash, they will thoroughly mix ash and
soil till the ash layer becomes hardly
recognizable.

On the whole, we must recognize that
the experience of Kodiak is decidedly re-
assuring. The damage to vegetation by
an eruption is not likely to be so great as
at first appears. Where the ashfall is a
foot or less in thickness, no permanent
injury to agricultural interests is to be
expected. There is no occasion for the
people in a region similarly affected to
abandon their property and go else-
where, as some were inclined to do at
Kodiak.

Digitized by

Google

c
.2

3
O

Q "5

<; CO

g ^

Q Z
w .S

2 So

o 2

3
O

c
cd

Digitized by

Google

Digitized by

Google

Digitized by

Google

THE ALASKAN VOLCANOES

The eruption of Katmai was not ac-
companied, as some eruptions have been,
by other sympathetic activity at a dis-
tance from the main focus. Yet it
should be recognized that Katmai is not
an isolated vent but one member of a
long chain. In a larger view the chain
itself is a single unit whose uplifting
and activities are the expression of a set
of fundamental stresses that have devel-
oped deep down within the earth's inte-
rior. Any particular eruption is, there-
fore, only a trivial incident in the drama.

Nothing could be more fascinating
than to unravel the history and discuss
the forces underlying the formation of
volcanic chains. But that would carry
us too far afield. Nor do we possess the
knowledge requisite for the purpose.
Yet we cannot properly take up the
problem of Katmai without giving some
idea of its setting and of the Asiatic vol-
canic complex of which it is a part, for,
as will appear, the relations of the Alas-
kan volcanoes are not American but
Asiatic.

It should be observed that all the vol-
canoes of Alaska except two ^ belong to
the Aleutian chain. This, although
reaching many miles into the mass of the
American continent, has an unmistak-
able affinity with the Asiatic volcanic
belts, while it shows no connection with
other American volcanoes.

THE VOLCANIC FESTOONS OF ASIA

An examination of eastern Asia (see
map opposite) reveals a very curious
succession of volcanic chains stretching
down the coast, like a series of garlands
draped along the edge of the continent.
Each member of the series is a gigantic

* Mt. Wrangell in the interior and Mt. Edge-
cumbe near Sitka.

* The data concerning these volcanic festoons
and the structure lines are largely summarized

crescent, convex toward the ocean.
Each trends southwesterly, beginning
with a southerly course and bending to-
ward the west as it continues along the
coast. Each joins the next lower seg-
ment approximately at right angles. The
crescentic arrangement recurs with re-
markable regularity five times in a series
of arcs stretching from Alaska to Suma-
tra.2

Directly in line with the prolongation
of the Aleutian chain are the Command-
er Islands (volcanic, though without any
modern vents), carrying our chain, the
first arc of the series, over to the shores of
Kamchatka, where it meets the second
crescent, a line of very lofty and notable
volcanoes running down through the
Kuril Islands into northern Japan.

The third member of the series begins
with the Island of Rishiri, in northern
Japan (with a theoretical extension
northward on a great fault along Sakha-
lin Island), and stretches south and south-
west to the western extremity of the
Japanese Islands, where, like the Aleu-
tian and Kuril chains, it meets the next
member of the series nearly at right
angles.

The only break in the regularity of
these festoons of volcanoes occurs in this
Japanese arc, which is intersected in the
middle by the great transverse fault of
Central Japan, along which an inde-
pendent series of volcanoes, including
the celebrated Fujiyama, stretches out
at right angles toward the Bonin Islands
far to the south. The general relations
of this chain suggest that it is a member
of the same series of festoons as the
others, but it appears to end in deep
water, without curving around to join
the next at its southern extremity.

from Hobbs* reviews of the work of Suess and
Richtofen. See American Geologist, vol. 34, p.
69, 1904.

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by B. B. Fulton
THE MESH WORK OF MOSS ON THE FOREST FLOOR
Perpetuating the position of the contraction cracks, see page 22 and text page 53.

The volcanoes of southern Japan are
developed along a fourth great line of
weakness, which comes down the east
coast of Korea and curves around
through the Lu-Chu Islands to northern
Formosa. Here again a fifth curving
line of volcanoes, also convex seaward,
carries the system through the Philip-
pines and Borneo into the great volcanic
complex of the East Indies.

REMARKABLE PARALLELISM OF VOLCANIC
ARCS

Not only is there a striking similarity
in the general position of these volcanic
festoons, but there is a very interesting
parallelism, perhaps only accidental, but
none the less remarkable, between the
physiographic features of the several
units.

In each case the crescent shuts off a
considerable sea. Nowhere else in the
world is there anything to be compared
with the series of enclosed basins con-

sisting of Bering Sea, Okhotsk Sea,
Japan Sea, Eastern Sea (Tunghai), and
China Sea.

The parallelism is further emphasized
by the remarkable fact that in three of
the five arcs the northeastern end con-
sists of a large peninsula — the Alaska
Peninsula, Kamchatka, and Korea —
while the distal portion is made up of a
narrow chain of small islands — the
Aleutians, Kurils, and Lu-Chu Islands.

It is of great interest to observe further
that the coastal region of Asia is cut off
from the interior by two older series of
similar crescents stretching from the
Arctic into Central Asia. The two arcs of
the Stanovoi Mountains and the Great
Khingan Mountains are conspicuous even
on ordinary political maps of Asia. Far-
ther south are four other crescents, not so
readily traced in the atlases because of the
rugged character of the country they
traverse. Theirposition as delineated by
Hobbs is shown on the accompanying

Digitized by

Google

THE ALASKAN VOLCANOES

Photograph by D. B. Church

A DUNE OF WIND-BLOWN ASH: WOMEN'S PENINSULA, NEAR KODIAK

This blowing ash lodged behind any obstruction, like snow. Among the weeds at the edge of cul-
tivated fields and along the fence rows drifts two feet deep were formed. On mountain tops and
in other places where there was no vegetation to catch the blowing ash it formed dunes like those on
a seashore.

map (facing page 56). Between these
and the island festoons is another in-
complete series along the coast line.

A study of Eastern Asia discloses the
fact that the same type of fractures has
dominated cnistal deformation through-
out the area since the beginning of the
geologic record.

The inner crescents are not at present
the seat of active volcanism, but it can
hardly be doubted that their formation is
due to the same cause as the volcanic
festoons along the coast. The regular
repetition of such striking lines of cres-
centic fractures over so large a segment
of the earth's crust cannot be thought of
as accidental. The cause responsible
for such great and long continued move-
ments must indeed be very deep-seated,
inherent in the very constitution of the
globe.

The fundamental character of the
forces which fix the location of the lines
of volcanoes is emphasized by the fact
that the course of the volcanic chains
shows in places a complete independence
of all surface features of the earth's to-
pography.

ALEUTIAN CHAIN CUTS ACROSS CONTI-
NENTAL AREAS AND OCEAN DEEPS

The Aleutian chain in particular holds
its course through continental strata and
across ocean deeps with a constancy that
is very significant.

The northern end of the chain parallels
the Chigmit Mountains, a much crumpled
mountain system of typical character.
Beyond this, in the Katmai district, the
volcanoes are built up on a thick plat-
form of sedimentary rocks, whose strata
lie almost as they were laid down on the

Digitized by

Google

I I
§ ^

C
CO

o «

H S3 O ■

O a '

< I

•T3
u

c
cd
u

.5 I

Digitized by

Google

Digitized by

Google

Digitized by

Google

THE ALASKAN VOLCANOES

shallow sea bottom, having been raised
above sea-level with scarcely any de-
formation. This whole area, both the
folded mountains and the eroded plateau
of our district, lies in a distinctly conti-
nental region, for Cook Inlet and the
waters around Kodiak are notably shal-
low, forming in fact considerable banks,
like the Grand Banks of Newfoundland.

In the strongest possible contrast to
this is the line of islands which stretches
500 miles beyond the tip of the penin-
sula. These are the crest of a very steep-
sided ridge rising directly out of water
more than a mile deep. Indeed, the
ridge between the thousand fathom
marks averages only about 90 miles wide.

Only a few miles south of the islands
lies one of the deepest depressions on the
ocean floor, a narrow groove running
closely parallel to the course of the vol-
canoes. Within a hundred miles of
peaks whose heights exceed 8,000 feet,
soundings of more than 24,000 feet have
been made. Here is one of the greatest
contrasts in elevation on the whole globe.

The basin of the Bering Sea on the
north is not so much depressed, yet the
depth attained here also is more than
12,000 feet, — ^as great as the average
ocean depth. There can be no doubt,
therefore, that the Aleutians stretch out
into a truly oceanic basin.

VOLCANIC CHAINS AMONG THE MOST
FUNDAMENTAL FEATURES OF THE GLOBE

It is unnecessary to point out the sig-
nificance of this fact, for as is well known
the dogma of the "permanence of ocean
basins and continental masses" has been
one of the most universally accepted
dicta of geology for fifty years, and an
inspection of a physical map of the world
makes it evident enough that most of the
larger structural features of the globe
parallel the boundary of ocean and conti-
nent.

It is not only unusual but significant
for the Aleutian axis, starting in the in-
terior of the continental mass, to reach

• Martin and Katz, U. S. Geological Survey
Bull. 485, p. 94, 1912.

out into the ocean deeps. That the line
of volcanoes should continue its course
straight through such diverse features as
folded mountains, elevated plateaus, con-
tinental shelves, and ocean deeps, sug-
gests that the causes underlying its
formation are among the most funda-
mental of terrestial forces.

The Aleutian chain is one of the
longest and straightest single lines of
volcanoes on the globe. Although situ-
ated in a sparsely settled region where
few records of eruptions have been pre-
served, ent)ugh is known to demonstrate
that it is likewise one of the most active
of volcanic belts.

But many of its peaks are so little
known that it is impossible to ascertain
whether or jiot they are volcanoes or com-
posed of volcanic rocks. Even the geo-
graphical position of some of them is un-
certain to the extent of several miles.
It was discovered in 1920, for example,
that the charted position of Bogoslof,
one of the best known vents, was in error
by six miles.

Beginning near the head of the Cook
Inlet, the active vents in the Aleutian
chain stretch out in a southwesterly di-
rection for more than 1,200 miles. There
is not space for any detailed account of
the 50 or 60 volcanoes which compose
this long chain. All that can be done is
to single out for brief comment a few of
the more noteworthy of them (see map
opposite).

Everyone who enters Cook Inlet has
occasion to admire the great snowy cones
of Redoubt and Iliamna, each standing
more than 10,000 feet above the water.

The perfection of their conical forms
is evidence enough that neither has ever
suffered a great eruption, yet both have
at different times showered the country
with ashes for many miles round about.

On January 18, 1902, an eruption of
Redoubt which "sent showers of dust
over the country from Lake Clark to the
Squentna Valley." ' It has been stated^
that Iliamna was the source of half an

* Davidson, George, Science, vol. 3, p. 286,
1884.

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph from Geo. C. Martin, by J. E. Thwaites

ISANOTSKY PEAKS, UNIMAK ISLAND

So many craters occur on Unimak Island that there is much confusion as to the location of the
various eruptions reported. Isanotslcy is credited with a violent eruption within the last century,
but there is no confirmation. Its appearance suggests a long period of erosion rather than frequent
activity.

inch of ash which fell atKodiak in March,
1867, but there is otherwise no record of
this eruption.

THE EXPLOSION OF AUGUSTINE VOLCANO

Augustine Volcano affords a good ex-
ample of what happens to a volcano in a
moderately violent eruption. Captain
Cook and others of the early explorers
describe it as a perfect cone, presenting
nearly the same appearance from every
point of view.

But some years ago its top blew off,
leaving behind a great cavity with a
jagged rim, resembling, on a small scale
the crater of Katmai (see page 60).
This eruption was described by David-
son as follows: ^

"About eight o'clock of the morning
of October 6, 1883, the weather being
beautifully clear and the wind light from
the southwestward (compass), and the
tide at dead low water, the settlers and
fishing parties at English Harbor [Port
Graham] (on the Eastern Shore of Cook
Inlet) heard a heavy report to windward
. . . when the heavy explosion was
heard vast and dense columns of smoke
were seen rolling out of the summit of St.

Augustine. . . . Fine pumice dust
soon began to fall. . . .

** Twenty-five minutes after the great
eruption a great 'earthquake wave* es-
timated as from twenty-five to thirty
feet high came upon Port Graham like a
wall of water. It carried off all the fish-
ing boats from the point and deluged the
houses. This was followed at intervals
of about five minutes by two other large
waves estimated at eighteen and fifteen
feet ; and during the day several large and
irregular waves came into the harbor.
The first wave took all the boats into the
harbor, the receding wave swept them
back again to the inlet and they were
finally stranded. Fortunately it was low
water or all the people at the settlement
must inevitably have been lost. The
tides rise and fall about fourteen feet."

For several months *' flames*' could be
seen issuing from the summit from a dis-
tance of 50 to 60 miles at night, while
vast columns of smoke appeared by day.

THE VOLCANOES OF UNIMAK, SHISHALDIN,
AND POGROMNI

Unimak Island, at the end of the pen-
insula, is a focus of volcanic activity

' Science, vol. 3, p. 186, 1884.

Digitized by

Google

THE ALASKAN VOLCANOES

Photo and copyright by J. E. Thwaites

POGROMNI VOLCANO STANDING SENTINEL OVER UNIMAK PASS

Pogromni is said to have "burst" in 1795. Other violent eruptions are reported in the years 1827

to 1830.

little inferior to the Katmai district.
Grewingk's map shows no less than
eleven craters within its comparatively
narrow limits. Because of the multi-
plicity of vents there is much uncertainty
as to which ones were concerned in the
numerous eruptions recorded.

The two best known peaks are Pog-
romni on the west and Shishaldin on the
east. The latter is one of the most
perfect cones in the world, seeming to
float high above the cloud-girt base,
recognized by all who have seen its white
symmetry as a formidable rival of the
celebrated Fujiyama.

THE BIRTH OF A NEW ISLAND, BOGOSLOF

One of the most interesting members
of this chain, and indeed one of the most
remarkable volcanoes in the world, is
Bogoslof , which lies about 40 miles north-
west of Unalaska. Here in 1796, a new
volcano suddenly appeared above the
surface of the ocean. In 1883 another
eruption occurred and a companion was
bom, which was named Grewingk in
honor of the great authority on the early
history of Alaskan volcanoes.

Since the first advent of the white man,

the locality had been marked by a single
pinnacle, doubtless the remnant of an
earlier eruption, known as Ship Rock,
standing high out of the water. Captain
Cook sailed close to it on October 29,
1779.

In March, 1906, activity was resumed
on a grand scale and two new peaks were
succe^ively pushed up in the general
position formerly occupied by Ship Rock.
These were so close together that the
first, Metcalf Peak, was half destroyed
by the rise of the second, McCulloch
Peak.

"It would be difficult to imagine a more
extraordinary spectacle"* than that pre-
sented by the island at this time. "As
a land mass above tide water it was a
continuous island about two miles long
with the two new steaming cones in the
middle. . . . Grewingk, flat-topped
and cliff -encompassed, guards the north
end of the island, while Castle Rock
[old Bogoslof], pinnacled and jagged,
rises in the south."

McCulloch Peak between these two
was cone-shaped, nearly 450 feet high,
and appeared like a steaming heap of
bowlders. The top was flattish but

•Quoted from Jaggar, Bull. American Geographical Society, vol. 40: pp. 385-400, 1908.

Digitized by

Google

J 6

•g 0)

go.

00 ce

.SK

<u d

CO «

•8-s

s ^

"S c

.S U
TJ.i-

S tin
(fi

-St

8H
.22 r4

•;: o o

Digitized by

Google

THE ALASKAN VOLCANOES

lumpy, made of hard rock and without a

crater. It was nearly enclosed by a

sickle-shaped lagoon of hot salt water

(temperature 90®).

If the loose stones ** could have been

washed away the rock core would have

looked like a gigantic potato lifting its

limipy protuberances above the waves.
>>

EXPLOSIONS OF THE PEAKS OF BOGOSLOF

On September i, 1907, about six
months after the discovery of McCulloch
Peak, the inhabitants of Unalaska saw a
dense black cloud rising in the west, and
the air was full of sulphur fumes, fol-
lowed by rain and distant rumbling. Ash
and sand descended on the town in the
evening to the depth of a quarter of an
inch. When Bogoslof was visited on
October 15 it was found that ''Mc-
Culloch Peak was absolutely gone, a
steaming lagoon in its place and the
rest of the island piled high with fallen
debris."

The half of Metcalf Peak was still
"standing in grim silence as a headstone
at the grave of the departed peak."
"The lagoon where McCulloch Peak had
been was a deep bay open to the west,
half a mile in diameter, with steam ris-
ing in considerable quantities from the
surface of the water.'*

Renewed activity in the bay between
Castle Rock and Grewingk is reported in
September, 1909, and in June, 1910, it
was found that a new peak with two
summits held risen to a height of 178
feet in the center of the island.

On September 18 of that year the
Coast Guard Cutter Tahoma, which hap-
pened to be in the vicinity, reported :

"Forked lightning in the direction of
Bogoslof was seen before daylight, and
when Bogoslof was sighted the new cen-
tral peak was seen to be in a state of
eruption. Immense clouds of vapor,

smoke, and ashes issued from the peak,
and enveloped the entire island. Flames
were reported at the peak, and lightning
followed by thunder appeared in the
cauliflower cloud of smoke and volcanic
dust which rose to a height of several
thousand feet above the island. The
eruption lasted during the several hours
the steamer remained in the vicinity and
two days later the central peak was ob-
served to be still steaming.** ^

From that time until the present the
activity has gradually subsided. The
cutter McCulloch reported a hot pond as
late as 1916, but when visited by the
cutter Boihwell August 9, 1920, the rocks
and water were all cold. Considerable
changes in the configuration of the island,
probably due entirely to erosion, had
occurred, however.

No man may predict how long the
present quiescent condition will continue
on Bogoslof. From past experience it is
evident that a new upheaval is likely to
occur at any time.

It is more probable, however, that the
next great eruption will come not from
Bogoslof or Katmai but from some one of
half a hundred unknown peaks. The
history of other regions shows that most
of the great eruptions have come from a
reawakening of vents that previously
had attracted little attention, many of
them being considered extinct.

Among the Aleutian volcanoes, where
only the most conspicuous and readily
accessible peaks are known, it is evident
that the probability of an eruption from
an unknown volcano is much greater
than in most regions.

It is very much to be hoped that before
the next great eruption occurs the Aleu-
tian region will have been studied suffi-
ciently to give us definite knowledge
of antecedent conditions — information
which is sadly lacking in the case of
Katmai.

' Quoted from Sidney Powers, Geographical Review, vol. 2, pp. 218-221, 1916.

As this goes to press, R. H, Sargent and W, R. Smith, of the Geological Survey, re-
port the discovery on the Alaska Peninstda of a gigantic crater six miles across, similar
to the basin of Crater Lake in Oregon,

Digitized by

Google

o
c

O -Q

o 6

Digitized by

Google

VII

CARRYING THE WORK INTO THE VOLCANIC AREA;

PRELIMINARY ACCOUNTS OF GEOGRAPHY,

CHRONOLOGY, AND PERSONNEL

Although the great distance at which
the eruption was felt gives a vivid con-
ception of the tremendous forces in-
volved, the narrative of the effects of the
ashfall at Kodiak and elsewhere leaves
us without much idea either of the vol-
cano itself or of the succession of events
during the outbreak.

The knowledge that the eruption was
of the first magnitude naturally aroused
a desire to learn more about the volcano
from which the ashes had come, and to
find out something of the nature of the
forces that had broken loose or, at least,
to ascertain their effect on the country
immediately surrounding the vent.

The Katmai volcanic district is so re-
mote from the conveniences of civiliza-
tion that its exploration involved
considerable difficulties. Comfortable
ocean steamers carry one cis far as Kodiak,
and there one finds boarding-houses and
well-stocked stores. But from thatpoint
forward it is necessary to charter a boat
and to provide all supplies, for the
country is an uninhabited wilderness
affording the explorer only water and
fuel.

DANGERS OF NAVIGATION IN THE DISTRICT

From the time we set sail, the enter-
prise presented serious obstacles. The
waters around the Kodiak Islands are
very difficult to travel. The coast is not
only imperfectly charted but has many
reefs and outlying rocks, which compel
the utmost vigilance on the part of the
mariner. The narrow channels are,
moreover, full of heavy tide-rips and the
region is subject to fogs, which render
navigation unusually hazardous.

Shelikof Strait, which separates the
islands from the Alaska Peninsula, has

an especially bad reputation, being
counted one of the most treacherous
bodies of water in the world.

Some of our experiences fully justified
the worst that could be said of it. Three
separate times during the expeditions a
gale, descending without warning from
the most promising skies, caught us as we
were in the midst of loading our supplies
and came near putting an end to our
work.

GENERAL SKETCH OF THE GEOGRAPHIC
FEATURES OF THE KATMAI DISTRICT

Before proceeding to a circumstantial
account of the changes wrought in the
volcanic area, and its many phases of
unique interest — the immense deposits
of ash, the dead forests, the great land-
slides, the tremendous flood, the in-
numerable f umaroles — it will be desirable
to sketch some of the more prominent
topographic features of the district.
This will help us keep our bearings as we
deal with details, now here, now there.
It will best serve our purpose if we can
imagine ourselves viewing the country
from an airplane skimming over its
surface.

The dominant feature of the region in
point of altitude is the line of old vol-
canoes running from northeast to south-
west and dividing the drainage of the
Pacific Slope from that of Bering Sea.
The volcanoes are not, however, the most
striking feature of the district.

The attention of an observer in an air-
plane would first be attracted by several
large open valleys in a country otherwise
covered with rugged mountains. The
most prominent of these is the broad flat
of Katmai River, stretching back north-
westward at right angles to the coast for

Digitized by

Google

Digitized by

Google

GEOGRAPHY, CHRONOLOGY AND PERSONNEL

20 miles and furnishing a sea-level route
into the very heart of the mountains.

This valley hcis two important tribu-
taries, one coming from the northeast,
the other from the west. Soluka Valley,
on the northeast, connects by a low pass
with Geographic Harbor at the head of
Amalik Bay. Martin Valley in the
west, extending up to the foot of the vol-
cano from which it takes its name, is oc-
cupied by the detritus of the great land-
slide which forms the subject of Chapter
XI (see map in back cover).

Katmai River is, further, the dividing
line between two types of physiography
which differ notably from each other.
To the south is an open country with
broad valleys and mountains of the mesa
type — a country readily traversed in any
direction. On the north the valleys are
narrow V*s, crowded in between close-set,
precipitous mountains — a. country of
great scenic beauty, but difficult to ex-
plore.

Although six miles wide at its mouth,
Katmai Valley narrows rapidly up-
stream, until, under Katmai Volcano,
less than 25 miles from the coast, it is
constricted into a narrow canyon 4,000
feet deep.

Beyond the volcano it expands again,
however, into a second, broad, open
basin. But this Second Katmai Valley
trends at right angles to the lower flat,
giving the impression that its drainage
has not always found outlet through the
canyon (see map opposite).

Connecting with Katmai Valley,
through the pass of the same name, is
another very wide basin, now the main
arm of the Valley of Ten Thousand
Smokes, stretching at right angles to the
volcanic axis from Mageik ^ Volcano
northward to the head of Naknek ^ Lake.
Curiously enough, this is crossed by the
broad valley of Knife Creek, also full of
fumaroles in its upper end, the two form-
ing a gigantic X which connects the two
ends of Iliuk Arm of Naknek Lake with
the two dominant volcanoes of the main
chain, Katmai and Mageik.
* Pronounced Ma-ge€k' — g hard as in get.

Indeed Katmai and Mageik may al-
most be thought of as standing athwart
these great valleys. The two parts of
Katmai Valley can, without any great
stretch of the imagination, be considered
as the extensions of the two arms of the
great X-shaped valley across the range.
The Second Katmai Valley appears al-
most like the extension of Knife Creek
Valley projected beyond Mount Katmai,
and the lower Katmai Valley bears a
somewhat similar relation to the main
arm of the Valley of Ten Thousand
Smokes. The alignment in the latter
case is not so close, but these two are
brought more nearly into physical con-
nection by the pass which joins them.

Katmai Pass is the key to the former
geographic importance of the district.
Joining two open valleys, it formed an
easy avenue for passage through a
mountainous country difficult of pene-
tration elsewhere. This route across the
peninsula, saving a thousand miles of
hazardous voyage into Bering Sea, was
much used for the transfer of supplies.

To the north of the volcanic area lies a
wide stretch of country occupied by alter-
nating series of mountain ranges and
broad depressions scooped out several
hundred feet below the level of the sea.
These are filled with a number of large
lakes, which, with the portages between
them, allow ready access to any part of
this section.

GEOGRAPHIC ORDER ADOPTED IN PREFER-
ENCE TO CHRONOLOGICAL

The features of the Katmai district
were so manifold, as they gradually came
to light during the progress of the explo-
ration, that it was necessary for the Na-
tional Geographic Society to send no less
than five expeditions into the country be-
fore even a preliminary account of its
phenomena could be sketched with any
degree of completeness.

In bringing together the accumulated
results of these expeditions, it would
be manifestly confusing to narrate the
chronological development of work cov-

' Pronounced Nik'-nftk.

Digitized by

Google

o
B

< >

■£

Digitized by

Google

GEOGRAPHY, CHROxNOLOGY AND PERSONNEL

ering somewhat the same area year by
year. It has seemed better, rather, to
follow a geographical sequence, carrying
the description up Katmai Valley, across
the Pass into the Valley of Ten Thousand
Smokes, and finally into the forested
country around the lakes.

It is recognized that in thus piecing
together the experiences of different
seasons some confusion as to the activi-
ties of the various expeditions is likely to
develop in the mind of the reader. In
order to permit the historical develop-
ment of the work to be followed it will be
well, therefore, to summarize at the out-
set the personnel and results of each of
the expeditions in turn.

PREVIOUS ATTEMPTS TO REACH THE
CRATER

Before 191 5 only two parties had at-
tempted any exploration of the main-
land. Dr. George C. Martin, gathering
material for his account of the eruption
published in the National Geographic
Magazine of February, 1913, landed at
Katmai Village in August, 1912, but was
unable to penetrate inland or to see the
volcano.

In 1913, William A. Hesse, of Cordova,
accompanied by M. A. Horner, of Sew-
ard, attracted by the possibility of
securing motion pictures of the volcano,
endeavored to reach the crater. They
were not adequately equipped for the
undertaking, but after suffering great
hardships they managed to secure good
photographs of Mount Martin. Post-
cards from their negatives, labelled
** Katmai Volcano,** are still for sale in
Seward.

Hesse and Horner themselves had,
however, no illusions as to the identity
of the volcano they had ascended. Both
have told me that they saw and recog-
nized the real Katmai from a distance
but were utterly unable to reach it. As
it was, they told me, they traveled con-
tinuously for twenty-three hours the day
they made their ascent of Martin.

They agree that they were able to
reach only the edge of the volcanic

district; that if they could have gone
farther they would have seen '*much
more than they did.**

They gave no clear account of what
they had seen in the distance, but as a
part of the Valley of Ten Thousand
Smokes is visible from the summit of
Mount Martin I am satisfied that they
must have had a glimpse of its fuma-
roles.

All the information Hesse and Horner
have been able to give me indicates that
the volcanoes of the district had already
settled down to their present steady
activity when they yisited them in 1913.
Their pictures show that Mount Martin,
at any rate, was in very much the same
condition then as now.

THE EXPEDITION OF I9I5

No one realized in advance the prob-
ability that a study of the seat of the
eruption would yield results of very
great interest. Our instructions in 191 5
were " not to attempt more than a hasty
reconnaissance on the mainland,'* for it
was believed that the return of plants to
the devastated district, which was our
sole problem at first, could be studied
better at Kodiak than elsewhere.

Accordingly we arranged for only a
brief trip to the mainland. That year
we examined the north side of Katmai
Valley and reached the lower slopes of the
volcano, but had neither time nor equip-
ment to do more.

There were three of us the first year,
B. B. Fulton, Entomologist, now of the
Oregon Experiment Station, L. G. Fol-
som, then manual training teacher at
Kodiak, and myself.

THE EXPEDITION OF I916

What we saw of Katmai Valley the
first year impressed us so deeply with
the importance of a thorough exploration
of the immediate environs of the volcano
that we felt it imperative to make a
study of the center of the volcanic dis-
trict and especially to examine the crater
the next year.

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

BASE CAMP IN I915

Photograph by B. B. Fulton

The flat was covered with a layer of white pumice deposited by the waters of the receding flood. It
was the only place we could find dry enough for a camp. From left to right, Fulton, Griggs and
Folsom.

The first season's work had vronvinced
us that the south side of the valley was
preferable to the north, but we were un-
able to find a satisfactory landing to the
south and perforce used the north side
once more, crossing the river near the
head of the valley. We made the first
ascent of the volcano on July 19th. On
July 3 1st we crossed Katmai Pass and dis-
covered the Valley of Ten Thousand
Smokes. The results of the exf)editions
of 1915 and 1916 furnished material for
an article in the National Geographic
Magazine of January, 1917.

This year D. B. Church became the
photographer (see page 73). He was a
youngster whom I had singled out from
my classes by reason of his habit of
turning up after a field trip with a set of
unusual pictures, which he had snapped
surreptitiously en route. His genius
for seeing pictorial values made him in-
valuable.

The experience of 191 5 hsid shown the
necessity of employing a packer the
second year. Here we met one of our
most difficult problems. The natives
were afraid of the volcano and could not
be induced to go to the mainland. When
we broached the matter to the chief, he
said at once very positively, *'Me no
Katmai," and we learned later that he
had advised his followers, ''Life better
than money."

THE ONE-HANDED BEAR HUNTER OF
KODIAK

The problem was most happily met,
however, when we thought of Walter
Metrokin, a famous one-handed bear
hunter of Kodiak. He agreed without
hesitation, and stuck to his promise, al-
though, as we found afterward, the other
natives used every means to dissuade
him (see page 74). Already a hero
among his fellows because of his many ex-

Digitized by

Google

GEOGRAPHY, CHRONOLOGY AND PERSONNEL

Photograph by R. F. Griggs
FOLSOM, WALTER (METROKIN) AND CHURCH ON THE TRAIL IN UPPER KA.TMAI VALLEY

Since the country was completely devastated, it was necessary to carry everything we had; if any
essential thing had been forgotten the expedition would have been stumped.

pioits as a hunter, he was doubly so when
he returned safely, having actually
looked into '*The Hole*' out of which
had come the devastating blsist. Even
Walter, however, was very nervous on
the crater rim, keeping sheltered behind
a rock a good share of the time and shift-
ing about uneasily as he watched us
work, finally remarking, when he thought
we had overstayed our time, ** Can't
make nothin' up here.**

Walter was one of those strong
characters found among all clcisses, who
stand out superior to their fellows. De-
prived of his right hand by a hunting
accident in his youth, he so overcame the
handicap that with his one hand he
could accomplish more than most men
with two. We found nothing he could
not do, even to tying knots and rolling
cigarettes.

But when it came to handling a boat,
I supposed that he would have to give it

up. Not so, however, for in a flash he
had somehow lashed one oar to his stub
and was rowing along like a sailor.

He proved a very entertaining com-
panion, as well as a capital assistant, and
the stories of hunting exploits he told in
his quaint Russian brogue furnished no
end of amusement around the camp fire.
The boys particularly enjoyed one tale
of a bear hunt "with two white men and
an Englishman," accompanied by a
demonstration of the Englishman's use
of his monocles, of which according to
Walter, he carried at least half a dozen of
different sizes.

THE EXPEDITION OF I917

During the first two years very little
organization was necessary, but the ex-
ploration of the Valley of Ten Thousand
Smokes was another matter.

The physical obstacles to be overcome
were so great that success could not have

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by L. G. Folaom

GRIGGS AND WALTER RESTING ON THE TRAIL

been attained without careful planning
in advance. Perfecting the details of
supplies and equipment, and seeing that
they were on the ground, consumed a
large share of my time and energy for six
months before the departure. But the
result justified the labor, for the outfit
proved adequate to the strain put upon
it and enabled us to carry out the work
substantially as planned.

The base camp in 19 17 was established
on the tundra near Kashvik Bay, and
supplies were ferried across the lagoon
and carried up the south side of Katmai
River to the Pass. The Valley of Martin
Creek, with the Mageik landslide at its
head, and the Second Valley of the
Katmai with Noisy Mountain were ex-
plored that year.

The work of 191 7 furnished the ma-
terial for the first adequate description
of the Valley of Ten Thousand Smokes,
which was published in the National
Geographic Magazine for February, 1918.
As a result of this report, the Katmai

National Monument, embracing more
than a million acres, was created by a
proclamation signed by President Wilson,
September 26, 191 8.

All told, there were ten of us in 191 7.
Besides the director, Folsom, Church,
and Walter, veterans of 1916, the party
included Clarence F. Maynard, to-
pographer, previously with the Peruvian
Expeditions of The Society; Professor
James S. Hine, of the Ohio State Univer-
sity, zoologist; Dr. J. W. Shipley, of the
University of Manitoba, chemist; Jasper
D. Sayre, assistant botanist; Paul R.
Hagelbarger, assistant botanist; and
Andrean Yagashof, packer (see page

76).

The complete devastation of the
country made the physical problem of
maintaining the expeditions a matter of
serious concern. In such a country we
were absolutely dependent on our own
resources. There was no possibility of
living off the country, nor of replenish-
ing supplies or obtaining aid or succor of

Digitized by

Google

GEOGRAPHY, CHRONOLOGY AND PERSONNEL

any kind, for the nearest settlements were
far beyond reach. One most crucial
problem was that of transportation. If
we could secure some adequate means of
carrying supplies into the country from
the coast, we could be assured of all the
things necessary for our work and a
considerable number of comforts besides.

CONSIDERING PACK-HORSES AND AIR-
PLANES FOR TRANSPORTATION

Accordingly, we considered very care-
fully the advisability of using pack-
horses. While there was no forage for
them in the desert, yet they might have
been kept outside in the undevastated
country and returned thither as soon as
unloaded.

But horses would have experienced
great difficulty with the quicksands, and
where the land was badly gullied it would
have been very hard to use them. Never-
theless, we would certainly have tried
them if any could have been secured in
the vicinity.

It would have been necessary, how-
ever, to ferry them across Shelikof
Strait and land them through the surf.
As this would have entailed not only a
heavy expense and the loss of much valu-
able time, but also a large chance of los-
ing the whole outfit in landing it on those
treacherous shores, it was decided that
the risk was too great to justify the at-
tempt.

The possibility of using an airplane
was very alluring. Often as we labored
over the rough ground our minds turned
enviously to the thought of skimming
swiftly overhead, with no care for the
gullies through which we were toiling.

There are unquestionably many land-
ing places available both on the pumice
flats and on the numerous ponds which
dot the valleys on the Bering Sea side of
the range. But an airplane has a deli-
cate mechanism, which cannot be ex-
pected to work continuously without
repairs which can be made only in a ma-
chine shop. In 1919 we did attempt to
use a motorcycle, and spent a great deal
of time building a trail for it. But just

when we got ready to use it, a small
spring went wrong in the magneto and
we were unable to fix it. The probability
of some such accident to an airplane was
so great that I have never been sorry we
did not attempt to use one.

ALL SUPPLIES CARRIED BY MAN-BACK

We were, therefore, thrown back on
the most primitive of all methods
of transportation — man-back packing.
This had the great merit of being certain
in operation, for wherever men could go,
there could supplies be transported.
But the amount of freight that a man
can carry on his back is at best pitifully
small. A load of "grub** which is a
crushing burden for the packer vanishes
quickly into the hungry mouths of an
active party.

To minimize the quantity of food con-
sumed by the packers themselves and to
reduce the amount of duflfel carried back
and forth, we adopted a system of relay
camps.

We would proceed half a day*s march
from our base, then establish a new camp
at the terminus, returning each night to
the starting point, until enough pro-
visions had been accumulated to permit
another move forward. In this way the
packers drew their sustenance altogether
from the rear camp, leaving the advance
supplies intact for future use.

Once our system of relay camps was
established, we made a practice of parcel-
ing out our spare bedding and extra
clothes among the different camps in
such fashion that a small party could
travel rapidly anywhere along the chain
without concerning itself about baggage
or supplies.

ONLY THE MOST CONCENTRATED FOODS
USED

Our extremely limited facilities for
transportation prohibited the use, except
at the base camp, of any of the ordinary
sorts of canned food, or of any provisions
containing a large percentage of water.
Our chief reliance was the trusted stand-
by of the prospector, bacon and flapjacks.

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by D. B. Church

MEMBERS OF THE EXPEDITION OF I917

From left to right: Jasper D. Sayre, botanist; Clarence F. Maynard, topographer; D. B. Church,
photographer; Lucius G. Folsom, assistant to the director; Robert F. Griggs; James S. Hine, zoolo-
gist; J. W. Shipley, chemist; Paul R. Hagelbarger, botanist; Andrean Yagashoff, packer, and Walter
Metroken, packer.

Next came sugar, rice, oatmeal, beans,
soluble coffee, dried apples, and other
dehydrated fruits and vegetables such
as apricots, cherries, cranberries, pears,
raspberries, blueberries, strawberries,
string beans, spinach, and potatoes.

Lunch consisted of hardtack, cheese,
raisins, kippered salmon and milk choco-
late, for we could take time to cook only
morning and night.

As an experiment I included tinned
butter among our supplies. This proved
a boon, for it added greatly to the enjoy-
ment of flapjacks and served as a con-
centrated substitute for evaporated milk
on oatmeal and rice. I was inclined to
regard condensed milk as a luxury not
worth the trouble of carrying, but I
found the men generally enjoyed it so
much that as a rule they were quite
ready to carry it, despite its weight.

At this time, before the importance of
vitamines in the diet was generally recog-
nized, I did not understand the reason for
the unusual desire for milk and butter.
But when it was shown how essential
vitamines are to a suitable dietary, an

explanation seemed to be offered, for, as
is now well known, butter and milk are
especially rich in these substances. I am
inclined to believe that they, together
with the fruits, are the most important
factors in preventing that craving for a
change of diet so likely to develop
when men are kept long on concentrated
rations.

This sort of provisions proved very
satisfactory. The members of the par-
ties always kept in excellent condition
throughout the field season and most of
the boys actually gained heavily in
weight, despite the continuous hard
work. For the whole party the gain in
flesh averaged about ten pounds, while
individual members took on as much as
twenty-five pounds during the summer.

The demand of the system for energy-
carrying foods was very great. I soon
found that even the small amount of
lean meat in bacon was more than was
desired. It was not long before bacon
began to be used exclusively for its fat.
In later years **crisco" was substituted
for a part of the allowance of bacon.

Digitized by

Google

GEOGRAPHY. CHRONOLOGY AND PERSONNEL

Photograph by Mrs. E. C. Kolb
THE LARGEST GROUP OF THE EXPEDITION OF I9I9 EVER ASSEMBLED IN ONE PLACE

Three men — ^Jones, Jacob and Ralph Hagelbarger — had left before the others came together. From
left to right are Helt, Sayre, Paul Hagellwirger, Kolb, Yori, Griggs, Folsom, Fenner, Allen, Miller,
Zies, Hine, Wallace, Basinger, Stone, and Henning.

Sweets of all kinds were also greatly
desired. A few jars of strawberry jam
were provided as delicacies, but it would
have been utterly impossible to provide
enough jam to satisfy the crowd. The
consumption of granulated sugar aver-
aged half a pound per man a day.

SENDING IN THE EXPEDITION OF I918
FROM THE BERING SEA

When the war made it advisable to
drop the extensive expedition that had
been projected for 1918, Sayre and
Hagelbarger of the 19 17 expedition were
dispatched to the field alone to keep
watch over the fumaroles and to begin a
study of the temperatures.

Theirs was no easy task. Their mis-
sion was not merely to continue the work
with which they were familiar, but to
break new ground and solve unknown
problems. They did not go in from
Kodiak, as previous expeditions had
done. Instead they circled the penin-

sula into the Bering Sea and ascended
Naknek River and Lake to the foot of the
Valley of Ten Thousand Smokes (see
map, front cover).

They surveyed the country traversed,
securing the data for the topographic
map, and made the first accurate
measurements of the temperatures of the
fumaroles, bringing back the information
that some of them were hot enough to
melt zinc.

OBJECTIVES OF THE EXPEDITION OF I9I9

While the exploration carried out in
191 7 was sufficient to supply the public
with evidence of the reality of the fuma-
roles, the hasty observations which alone
had been possible under the conditions
served only to arouse in us a keener de-
sire to investigate the scientific problems
presented. For we were convinced that
nowhere else in the world had there been
presented in our time such an opportunity
to study the great problem of volcanism.

Digitized by

Google

< I

X B
^ -3

S .o

2 <*>
I

Digitized by

Google

GEOGRAPHY, CHRONOLOGY AND PERSONNEL

All members of the expedition agreed
that the pictures of the Valley brought
back in 1917, while striking enough to
excite favorable comment, fell so far
short of the reality as to convey no ade-
quate conception of the thing itself.

As an experiment, I had purchased a
small motion picture outfit, and Church
had learned to operate it, so that we se-
cured incidentally a few hundred feet of
films. These had a most checkered
career. While awaiting shipment they
were ** taken by mistake'* from the ware-
house at Kodiak by an old prospector
and carried down to the south end of the
island, whence after being dumped out
in the surf and left in the dampness of the
native huts for months, they were finally
returned to us. Despite the rough
usage to which they had been subjected,
upon development they portrayed the
real character of the fumaroles so much
better than any still pictures that I de-
cided at once we could not consider the
district adequately made known until a
representative series of motion pictures
had been secured.

With these two very diverse objects in
mind, we were anxious to return and
carry forward the work. But the World
War held us up until 191 9.

When finally the opportunity came to
carry out our plans, we tried to equip
ourselves in such fashion that we would
be enabled to sit down and study our
problems at leisure, and to provide some
sort of laboratory in which preliminary
chemical work could be carried on. The
difficulties in the way of the realization
of this project were enormous.

The scientific problems encountered in
the region were so numerous that we
could not resist the temptation to lay out
more work than could possibly be ac-
complished in the time available, even
with the very generous appropriation
which the Research Committee of the
the National Geographic Society had
granted. But although many important
questions were still unanswered when
the season closed, we had nevertheless
accomplished a great many things which

could not have been attempted on pre-
vious expeditions.

COOPERATION BY THE GEOPHYSICAL
LABORATORY

The various aspects of volcanism de-
scribed by the earlier expeditions were so
exceptional that I felt our findings were
too important to stand alone without
corroboration by other observers.

After a survey of the field it was de-
cided that no other organization was so
well equipped to handle the many-sided
volcanic problems as the Geophysical
Laboratory of the Carnegie Institution,
which was accordingly invited to send a
party along with the expedition of 1919
under a cooperative agreement whereby
the Geographic Society undertook to
assume the field expenses of the party,
while the Geophysical Laboratory agreed
to work up the results on the return.
The party selected consisted of two
chemists and a petrologist.

ORGANIZATION OF THE EXPEDITION OF
I919

In view of the risk and difficulty in-
volved in landing a large outfit through
the surf, and carrying it by man-back
over Katmai Pass, it was decided to send
in the bulk of the outfit by Bering Sea,
whence it could be taken in power dories
up Naknek River and Lake almost to the
foot of the Valley of Ten Thousand
Smokes.

Since there is scant means of public
conveyance into those parts, we should
have been helpless in this matter without
the kindness of the Naknek Packing
Company, which undertook to carry up
our supplies and the advance party.

Say re, who knew the ground from his
experience in 1917 and 1918, was sent in
by this route with three of my students:
A. J. Basinger, Richard E. Helt, and
Harry E. Jacob. Their mission was to
provision a base camp at the head of
Naknek Lake and form a junction with
the second party, which I led in from
Kodiak to establish a series of relay
camps across the peninsula. By this

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

means we hoped to make it possible for
the specialists, who were to come later,
to begin work on their respective prob-
lems immediately upon arrival.

PERSONNEL IN I919

There were nineteen of us in all: the
Director; the Geophysical party consist-
ing of Dr. E. T. Allen, chemist, Dr. C. N.
Fenner, petrologist, and Dr. E. G. Zies,
assistant chemist; Professor J. S. Hine,
of the Ohio State University, zoologist;
J. D. Sayre, topographer; Paul R. Hagel-
barger, topographer; Emery C. Kolb, of
the Grand Canyon, motion picture man;
Frank I. Jones, of Portland, Oregon, color
photographer; Lucius G. Folsom, assist-
ant to the Director; A. J. Basinger,
Ralph Hagelbarger, Richard E. Helt,
William L. Henning, Harry E. Jacob,
August E. Miller, Julius Stone, Jr., H. N.
Wallace and Charles Yori, assistants.

It would be hard for anyone who was
not along to realize how these men worked
for the success of the expedition. No
task, however difficult or disagreeable,
was too great whenever it was recog-
nized that it was for the good of the expedi-
tion. Many of the jobs for which there

were volunteers in plenty would have
been altogether unreasonable in the eyes
of any except men who were there for the
love of overcoming difficulties.

In addition to the regular members of
the expedition, visitors were received for
the first time. The first "tourist** to
visit the Katmai National Monument
was Rodney L. Glisan. of Portland, Ore-
gon. Later in the summer the wives of
the three members of the expedition
whose families were at Kodiak, Mrs.
Griggs, Mrs. Folsom and Mrs. Kolb, also
came across and visited the Valley, thus
proving that despite the necessarily
primitive conditions, the place is by no
means impossible for women.

It was hoped that the preliminary ex-
ploration of the district could be brought
reasonably near completion in 1919, but
very much remains to be done. There
are several square miles of the most
interesting of the fumaroles which have
been examined only in the most cursory
fashion, a large part of the tributary ter-
ritory remains unmapped, and of the
more detailed problems of revegetation
and geology scarcely a beginning has
been made.

Digitized by

Google

Natural Color Photograph by Frank I. Jones

KODIAK FROM THE CHANNEL
Who would suppose that this verdant country was so recently smothered under a foot of ash?

Digitized by

Google

A CANYON IN THE MUDFLOVV AT THE HEAD OF KATMAI VALLEY
The blue waters of Katmai Bay may be made out in the distance.

Digitized by

Google

VIII

THE FIRST JOURNEY TO THE VOLCANO— THE
DEVASTATION OF KATMAI VALLEY

As the shores of Katmai Bay gradually
appeared out of the blue distance when
we approached them for the first time
(July 11,1915), they looked weird beyond
description. Over the sky was drawn
apall of finevolcanic dustwhich obscured
everything above a thousand feet, cut-
ting off the volcanoes which we wished
so much to see, and by its curious, dif-
fused light heightening the unearthly
aspect of the landscape, giving it some-
what of that ominous effect that raises
one's forebodings at the approach of a
storm.

The whole scene looked like the en-
trance to another world, so foreign was
it to anything within our experience.
The sea was full of fresh pumice. We
wondered whether this and the cloud of
dust might betoken a fresh eruption
from one of the volcanoes. The line
of breakers was continuous. We could
not tell where we might best pierce it.

It was even more difficult to guess
where we should find the best way back
into the interior. Trying to make the
right decision, I lay propped upon my
elbows studying the shore with a field
glass, for I was too weak from seasickness
to stand.

THE FIRST LANDING

First we tried the mouth of the river,
where the schooners used to land. Since
the eruption it had been entered by a
boat larger than ours. But it was out
of the question now; so much ash had
been washed down that there was no
possibility of getting inside. (Later we
found the river could not be ascended
half a mile, even with a row boat.)
Then, as the western side looked low
and swampy, we turned east, landing
under the cliffs of Mount Pedmar at the
end of the beach.

Along the base of the cliff was a white
gleaming waterfall, promising to settle
one of our problems, for in the uncer-
tainty as to the condition of the country
I had made up my mind that I would
not allow our boat to leave us until we
had at least assured ourselves of a supply
of drinking water.

Everywhere the ground was covered
ankle-deep with soft sticky mud from
the flood, described in Chapter ix, whos^
waterswere just subsiding. Fresh tracks
of fox and of the huge Alaskan brown
bear indicated that we were not alone on
the beach, and suggested the advisability
of carrying guns.

But the omnipresent mud gave us
more concern, for it covered the whole
valley, clear up to the cliffs. Between
the precipitous mountains and the muddy
flat no place was left for us to pitch our
tent except a mound of avalanche detri-
tus, which we felt was too dangerous, for
bowlders were rolling off the mountains
round us every few minutes.

Finally we selected a bed of clear white
pumice. Although there was very wet
mud a few inches below, the porous sur-
face was fairly dry, and there we pitched
our tent (see page 72).

Our camp was surrounded by good-
sized balsam poplar trees. We had ex-
pected no such luck as this, for the best
information we had been able to secure
in advance indicated that Katmai was
well beyond the limit of trees, so that we
might expect difficulty, at least in the
country deeply covered with ash, in
finding even enough sticks for firewood.

In view of this probability we had
provided jointed poles for the tent and a
small blue-flame oil stove for cooking.
We were most happy to be relieved of the
necessity of packing the stove with a
supply of oil on our backs, but finally did

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

AN ASH SLIDE : SOLUKA CREEK

' Some of these slides spread out into gigantic fans more than a thousand feet high. Standing at the
critical angle, their slopes are very hard climbing. We soon ground our finger-nails to the quick in
their sharp sand

carry the tent poles clear up to the head
of the Valley in the groundless fear that
we could not find poles there to serve our
purpose.

Near the coast the poplars were in
scattered groves and somewhat stunted
in the more exposed situations, but far-
ther inland there were forests, whose
trees sometimes exceeded four feet
(135 cm.) in diameter and commonly
reached a height of 70 to 80 feet.

COMPARATIVE DAMAGE TO DIFFERENT
PLANTS^

The desolation of the country beggared
description. All the trees had perished,
except those favored by some special cir-
cumstance, such as proximity to the pro-
tecting mountain sides.

Inoneway the trees and bushes suffered
more seriously than the herbage, for
wherever the ground had been swept bare
of ash the old roots of the herbaceous

^ For a more detailed discussion of the effect
of the eruption on mainland vegetation the reader
is referred to a special paper devoted to the
matter. Scientific Results of the Katmai Ex-

plants sent up new shoots, so that in a
few fortunate spots flowers were bloom-
ing in their pristine profusion.

Where the ash remained to a depth of
a foot or more, however, the ground
under the dead trees was absolutely bare.
No vegetation had come through cracks,
as at Kodiak, and indeed such cracks
may not have been formed, because the
deposit here is much coarser grained.

Under the mountain sides, where a
few remnants of the forest remained
alive, various species had suffered in
different ways.

The only large trees were the balsam
poplars (Populus balsamifera) , All of
the growing parts and the ordinary buds
of these had been killed, but deep down
in the bark of old shoots dormant or
adventitious buds had pushed out short,
bushy branches which gave the trees a
most bizarre appearance.

peditions of the National Geographic Society,
No. 4. The Character of the Eruption as Re-
vealed by its Effect on Near-by Vegetation, by
R. F. Griggs. Ohio Journal of Science, vol. 19,
pp. 173-209, 1919.

Digitized by

Google

DEVASTATION IN KATMAI VALLEY

One would not have believed that a
growth of leaves so slight as some we
saw could possibly suffice to keep the
bark and roots of a large tree alive. Nor
did it always do so, for we found many
poplars that had perished from starva-
tion, after a vain attempt to put out
leaves enough to keep the system alive.

The alder (Alnus sinuata), which is
the most characteristic Alaskan bush
everywhere, was simply exterminated.
For our purpose this was somewhat
fortunate, for it was easy to break our
way through the dead thickets which
otherwise would have made traveling
difficult. Not a single live twig of alder
was seen until after we had explored a
wide stretch of country, and then we
found only two or three very small
sprouts coming up from the roots.

The birch {Betula kenaika)^ the Alas-
kan representative of our canoe birch,
which was formerly abundant, suffered
only slightly less than the alder. With-
out exception, this species perished in
the main valley, but a few individuals
with living sprouts were found in Soluka
Valley.

WILLOWS SUFFERED LESS THAN
OTHER TREES

The willows suffered less than the
other trees, but the pussy willows (Salix
scouleriana) were killed down to the
ground. Curiously enough the new
growth did not assume the erect form
characteristic of the species, but instead
produced a rosette of prostrate branches
close to the ground, like the alpine wil-
lows of exposed places.

Of all the trees the big Alaska willow
{SaMx alaxensis) suffered the least or, at
any rate, had made the best recovery by
the time of our visit. This was probably
due to its very great capacity for forming
adventitious roots on burial.

Where shifting streams had disinterred
some of the buried trees, we found that
they had developed a very strong growth
of new roots, just beneath the surface of
the ash. It was interesting to observe
that these new roots were confined to

Photograph by R. F. Griggs
A TALL POPLAR ON WHICH ONLY A FEW
BUDS SURVIVED

It seems incredible that so slight a leaf surface
should have sufficed to keep the extensive trunk
and roots alive for four years.

the area immediately beneath the surface
of the ash, leaving the intermediate sec-
tion of the trunk down to the original
surface of the ground as clean as before
burial (see page 133).

On the mainland as at Kodiak, the
formation of such two-storied root sys-
tems is the characteristic reaction of the
buried plants when their underground
parts show any response at all to burial.

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by D. B. Church

WRITING IN THE TENT DURING THE SAND-STORM

While outside a piece of pumice lodged in my eye with such force that
it had to be dug out with forceps, causing considerable discomfort for a
couple of days.

permanent injury, for
particles of volcanic
dust are sharper even
than bits of broken
glass. After a time,
however, the irrita-
tion stimulated an in-
creased flow from the
tear glands, which
helped to keep the
eyes washed out.

During this day of
dust storm we ex-
plored the valley as
far as Soluka Creek.
The dust heightened
the already weird
character of the land-
scape, giving it an in-
describably uncanny
appearance. The ef-
fect was much like
that of a heavy snow-
storm. This was in-
creased by the out-
lines of the bare dead
trees. Indeed, so
strong were the visual
sensations of snow
that every little while
I would realize with
a start of surprise

On the first day we were able to de-
vote to exploring the valley we awoke
to find that a westerly gale, which had
started during the night, had picked up
the fine dust from the mountains until it
had changed the haze of previous days
into a terrific dust storm. The dust was
so thick that it obliterated all objects
beyond the immediate vicinity.

A DUST STORM

It permeated everything about our
camp, and we were extremely worried
lest it should get into our cameras and
ruin the shutters and films. It matted
our hair so that we could not comb it
for days. It caused acute discomfort in
our eyes.

At first we were afraid it might do us

that it was not cold.

About noon we fell to speculating on
the state of the weather above the dust
cloud and were surprised on searching
the sky to find the sun, whose disc was
just visible, pale white, something like
the moon in daytime, but fainter.

It is impossible to describe our feelings
on this day, as we groped our way for-
ward into the devastated country, ut-
terly different from anything we had
ever seen. Fortunately, the loose sandy
surface of the ash everywhere held our
tracks, so that even without our com-
pass we could hardly have become lost.

We followed all the way a well-worn
bear trail, which skirted the foot of the
mountain, finding that the bears had
selected the easiest going to be had. It
was very noticeable that these trails,

Digitized by

Google

DEVASTATION IN KATMAI VALLEY

Photo and copyright by M. A. Horn«r

MOUNT MARTIN IN I913

except for an occasional side branch
into the mountains, all ran lengthwise,
up and down the valley. The bears
had made no attempt to cross the river.
Apparently they had learned by experi-
ence not to try that.

After four days spent in such prelimi-
nary observations we packed up our
outfit and as much food as we could carry
and started up the valley for the vol-
canoes. The remaining provisions, to-
gether with everything not essential to
our work, were left in the base camp.

Although things were made as snug as
possible, it was not without a good deal of
trepidation that we turned our backs on
our supplies; for in such a desert country
we were absolutely dependent on our
provisions, and if a bear or wolverine
had taken it into his head to wreck the
camp in our absence, we should have
been in a bad way.

Three or four miles up the valley we
came out into the open, where we could
see the distant mountains of the main
range. Standing square across the head
of the valley was Mount Mageik, its
magnificent three-peaked snow-cap bril-
liant in the sunshine.

From a small crater east of the central
peak issued a column of steam which,

although plainly visible on a clear day
for 50 miles out to sea, appeared diminu-
tive in comparison with the bulk of the
mountain. Only rarely does it rise
straight up into the sky, as shown in the
photograph on page 78.

Mount Katmai itself was concealed
beyond the bend of the valley, so that we
were to have no glimpse of it until we
camped at its foot.

DISCOVERY OF MOUNT MARTIN

To the west of Mageik, in a position
where no volcano is indicated on the
chart, was a comparatively low moun-
tain from which rose a column of steam
a thousand feet in diameter and more
than a mile high.

Comparison with Horner's picture
showed at once that this was the moun-
tain he had photographed in July 1913 as
** Mount Katmai.** It was clear enough
from its location west of Mageik that
it could not be Katmai, for that is on
the other side of Mageik. Even from
our position it was evident, however,
that it was at that time the most active
volcano of the district.

The discovery of this volcano was a
complete surprise, for we had had no
intimation of the existence of any such

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

ASH SLIDES IN UPPER KATMAI VALLEY

Wherever the mountains were too steep for the ash to stick, it slid down into the valley, covering the

lower slop>es with great fans of sand.

vent before our landing. There appears
to be neither record nor tradition of any
volcano in this quarter. No map shows
a volcano near its position nor even
names the mountain.

We agreed that no more fitting name
could be given to the new volcano than
that of Dr. George C. Martin, whose
account in the National Geographic
Magazine will always stand as the first
authoritative report of the eruption.

Accordingly, from the first it has been
known to the members of the expeditions
as Mount Martin. Its position is ap-
proximately Lat. 58® 10' N., Long. 155®
21" w., and its altitude about 5,500
feet. Its great column of steam was a
continual challenge to our efforts, but it
was not for us to approach it that year
nor the next, nor did we ever succeed in
studying it as we desired (see chapter xiv) .

The activity of Mount Martin, like
that of all the volcanoes of this district

is remarkably constant. When first
seen by Hesse and Homer in 1913, it was
quietly steaming away, exactly as in
19 19 (compare pages 180 and 182).

In the absence of witnesses it is, of
course, impossible to ascertain exactly
when this quiet activity began, but pre-
sumably its initiation was coincident
with the reawakening of Katmai.

One can hardly image the opening up
of a crater of such dimensions except by
an explosion violent enough to rend the
mountain walls asunder. Great as is the
power necessary to tear open a conduit
like the crater of Martin, however, it is
so slight in comparison with the other
forces operative at the time of the great
eruption, that no mark of its action has
yet been discovered in the surrounding
country. We have diligently searched
for ejecta from both Martin and Mageik,
but no evidence that they threw out
ash or pumice at the time of the recent

Digitized by

Google

DEVASTATION IN KATMAI VALLEY

Photograph by B. B. Fulton

THE AUTHOR STRUGGLING THROUGH THE QUICKSAND OF KATMAI RIVER

After working our way to within a hundred yards of the trees in the distance we had to turn back.
The swift water running over the ash and pumice packs the surface, giving it a crust which sometimes
holds a man and sometimes breaks under nie weight. Crossing these flats is somewhat like traveling
in snow with a weak crust. One will go alon^ easily ankle deep for a few steps and then suddenly
drop down to hb waist. The labor involved m such travel cannot be described, but must be experi-
enced to be appreciated.

eruption has been detected (see page
182).

When we reached Soluka Creek we
found it more formidable than our rec-
onnoiters in the dust storm had indi-
cated. Leaving the others on the bank,
I dropped my pack and waded out
through the dead forest for half a mile in
the icy water. From that distance it
looked wider, deeper and swifter than
from the starting point. I decided,
therefore, that it was impracticable to
attempt to cross it under our heavy
packs, so we camped that night in the
dead forest on the flat near by.

ASH SLIDES MORE THAN 1,500 FEET HIGH

Next morning, starting to hunt for a
practicable ford, we climbed up on to the
shoulder of a mountain where we could

get a bird's-eye view of the creek below
and select the likeliest place.

Here we found a new experience in
climbing the great ash slides with which
the lower slopes were covered. Wher-
ever the mountains were precipitous, too
steep for the ash to stick, it slid down
into the valleys, covering the lower
slopes with great fans of sand, standing
as steep as it will lie. Some of these ash
slopes are more than 1,500 feet high.

When we first entered the country in
191 5 their surface still consisted of loose,
rolling sand, ready to slide down at the
least disturbance (see pages 82 and 86).
With the passage of time they gradually
became more stable, as the angular grains
settled closer together. But, during our
first season, it required all the power we
could muster to climb them.

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

jfnotograpn oy k.. t*. i^nggs

STRAINING DRINKING WATER OUT OF PUMICE

The water we had to drink was so muddy that Folsom refused to wash for three days because he

"would not dirty his face with it."

One sank to his ankles at every step,
while new sand poured down upon him,
either burying his feet or carrying him
bodily down the hill. It was a tread-
mill, in which one must either keep
moving or be shifted to the bottom.
The effort of such climbing soon became
so wearisome that we avoided it when-
ever we could, but occasionally it was
necessary for us to ascend a thousand
feet over such slopes.

In climbing these slides, much the
easiest method of progression was to go
on all fours. But the sharp pumice
soon ground away our finger tips. We
were seriously embarrassed from this
cause the first year. Before we were
aware of the danger, we had worn our
nails down to the quick and so ground
the skin as to make it painful to use our
fingers in any way.

But the labor of climbing the ash slopes
was somewhat compensated by the sport
of sliding down. Since the ash stood at
the critical angle, the effect was practi-
cally to eliminate gravity. One could
cut loose and almost fly down the slopes,

bouncing down the steep slides in great
bounds without the least consideration of
his footing or of where he would bring up.
When we descended to the ford we
had picked out, we found that it was
little better than the place where we
had turned back the night before. But
it was the best that was to be had, so we
decided to try it. It was not the low
temperature nor the swiftness of the
water that alarmed me, but the character
of the bottom, which was a continuous
quicksand clear across.

FORDING A MILE OF QUICKSAND

Sometimes the surface would hold like
the crust of a snowdrift; but we were in
constant fear of going down, for on
sounding with our alpenstocks we found
that the whole length of the stick went
down six feet into the sand anywhere
without striking bottom. Often our foot-
ing gave way and we found ourselves
floundering up to the middle in quick-
sand. But as none of us went deeper
than that, we returned and started across
with our packs (see page 87).

Digitized by

Google

DEVASTATION IN KATMAI VALLEY

Photograph by R. F. Griggs

AN ASH BRIDGE

The caverns formed beneath the ash by melting snows gave us much concern at first,
learned to trust ourselves freely on such bridges.

Later we

^. With all our crossings, in the expedi-
tions which followed, no one ever got in
so deep that he could not get out alone.
But I must confess that, often as we
crossed Soluka Creek, I never got used
to it. There was the ever-present
knowledge that we never touched bottom
and the fear that next time it would
'*get" one of us.

Besides this, the labor of carrying a
pack through such quicksand exceeds
anything else I know. Every step takes
all a man*s strength, and soon his weary
muscles ache from the strain. But once
in, he has no chance to rest until he
reaches the farther shore, for he can
neither lie down nor sit down ; even if he
merely stands still he begins immediately
to sink.

STRAINING DRINKING WATER OUT OF THE
PUMICE

It was very much of a problem to
secure water from the overloaded brooks
beside which we camped. Never more
than an inch deep, the water always
carried such quantities of large angular
chunks of pumice, not to speak of sand

and mud, that it was out of the question
to use it as it was.

If we so much as dipped our hands into
It in an effort to wash, sharp pieces of
pumice stuck thickly to our fingers.
When we tried to rub them off again they
ground into our flesh painfully, prevent-
ing any further attempts at cleanliness.

The density of the ash was so near
that of the water that it would neither
float on the surface nor settle to the
bottom. Fortunately, we had a spare
bag through which we could strain it.
Often we had to strain out a quart of
pumice to get a pint of water. The
stream changed so rapidly that it some-
times ran dry before we could fill our
bucket (see page 88). And even by
the best straining we could take out
only the coarsest grit, leaving the water
still full of mud.

At one of our camps Folsom refused
to wash for three days, because he
'* would not dirty his face with the water
we had to drink.'* Washing is a matter
of choice, but one must drink whether
he will or no.

After crossing Soluka Creek we camped

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by D. B. Church

THE GLOOMY STRETCHES OF SOLUKA CREEK

I must confess that even after many crossings of this sinister stream without mishap I could never
plunge in without a shudder of dread. So wide that from the middle we could see neither shore, its
swift current everywhere churning the quicksand, it presents a formidable obstacle to a man carry-
ing a pack. I was in constant fear lest some member of the party would be mired in its depths,
for, although we seldom sank below our knees, we could plunge the full length of our alpenstocks
into the quicksand anywhere without finding bottom (see text, page 88).

at the foot of the range that had shut
off the view of Mount Katmai all the
way up the valley. We reasoned that
from the summit of these mountains,
which we named the Barrier Range, we
could most quickly obtain that sight
of the volcano for which we were so
anxious.

Haste, we felt, w£is imperative, be-
cause we feared lest the fine weather,
which favored us almost miraculously
the first season, should fail us. I well
knew from previous experience that fair
weather such as we were having was most
exceptional in this district, for one may
often camp here for days without a
glimpse of the high peaks — as we did in
191 7 and 1919, when we waited for
nearly two weeks before we could climb
Katmai. But on our first trip our

provisions were so short that we could
not hope to stay among the volcanoes
for more than a few days.

The day after crossing Soluka Creek
we attempted, therefore, to get a sight
of Katmai from the Barrier Range. In
this we were disappointed, for on reach-
ing the top of the mountain selected we
found a higher peak ahead, cutting off
the view of Mount Katmai.

FEAR OF CAVERNS FORMED BY MELTING
SNOW BENEATH THE ASH

When we tried to cross over to this
higher summit we encountered for the
first time one of the characteristic
features of the ash-covered mountains
that for a long time gave us much con-
cern.

Part way up the slope, one of us, stick-

Digitized by

Google

DEVASTATION IN KATMIA VALLEY

Photograph by D. B. Church

BIRCHES WITH THE BARK CUT OFF BY SAND BLAST

Over large areas such abraded trees give clear testimony of the direction and severity of the prevailing

wind.

ing his staff into the ground harder than
usual, found that it went through into a
cavern beneath. Examination showed
that we were supported on an arch of ash
a foot thick, spanning a deep hole.

We found that the mountains every-
where were deeply covered with snow,
which was concealed by a mantle of ash
and pumice. The snow beneath was
rapidly melting out in the warm weather,
leaving the ash surface standing as
smooth as ever above the cavity.

Such small holes as the one into which
we had accidentally broken were, of
course, of no consequence; but as we
looked down one of the side valleys, we
could see great cave-ins in an apparently
smooth ash field, where a stream, bur-
rowing through the snowdrifts beneath,
had undermined the surface.

For half a mile or so the tunnel thus
made had caved in, and then for another
half mile it was still intact, giving no

indication of its presence to an unwary
traveler.

Alive to the new danger, we carefully
chose a path free from all appearance of
buried snowdrifts. But we had not
gone a hundred yards when, happening
to stamp my foot, I was astonished to
hear the ** ground" ring hollow beneath
me. We quickly retreated, spread out,
and tried another place.

Again we had not gone far when all
three of us at once, though 50 feet apart,
detected a cavern beneath us. We had
absolutely no means of judging whether
the hole was 5 feet deep or 50, nor of
estimating the strength of the roof.
The danger of such a situation seemed
altogether too great, so we reluctantly
turned back, with as yet no view of the
volcano.

In later expeditions we found stretches
thus undermined by melting snow char-
acteristic of much of the area of deeper

Digitized by

Google

92 THE VALLEY OF TEN THOUSAND SMOKES

Photograph by Frank I. Jones
A JOB FOR A TIGHT-ROPE WALKER

THE BRIDGE OVER MARTIN CREEK

All the materials for color photography were in this pack; if Jones had slipped there would have
been no illustrations in color. Some of the boys hesitated a long while before trusting themselves
on those wet, slippery logs.

Digitized by

Google

DEVASTATION IN KATMAI VALLEY

Photograph by K. K. Crriggs

EVENLY BEDDED STRATA OF KATMAI ASH LYING ON TOP OF A SNOWDRIFT

The smooth contact and the absence of the ice, which would have formed on refreezing, indicate
that the falling ash was not hot enough to melt the snow. The ash blanket prevented melting for
five years. On Observation Mountain, about seven miles from the crater.

ashfall and, as we were enabled gradually
to experiment with such caverns, we
learned that the danger from cave-ins is
very slight. Before the exploration was
completed we came to strike out boldly
across places where we dared not venture
in the beginning.

The following day we circled the
mountains into Upper Katmai Valley.
As we proceeded the country became
progressively more desert. Small birds,
which were common in the lower valley,
were absent here.

The stillness of the dead forest was

Digitized by

Google

^ a

Digitized by

Google

DEVASTATION IN KATMAI VALLEY

oppressive. One could travel all day
without hearing a sound but his own
footfalls and the plunge of rushing water.
Bear trails persisted until we turned the
corner into the upper valley, but there
they disappeared.

Every vestige of animal life must
have been destroyed in the eruption,
unless perhaps some wood-boring insects
sheltered beneath the bark of the trees
had escaped. But even of these, few, if
any, remained at the time of our visit.
The bark had dropped off the dead
trees, leaving gaunt white skeletons
standing up out of the deep ash de-
posits.

We saw no signs of animed life around
our camp except a pair of bald eagles,
which flew over at a great elevation the
first night, a few mosquitoes, and a sin-
gle hiunming-bird moth, which seemed
strangely out of place in such a valley of
death.

AWE-INSPIRING DESOLATION OF UPPER
KATMAI VALLEY

Clouds hung low, obscuring every-
thing above a thousand feet. As we
pushed up into the valley a feeling of
tremendous awe possessed us. We had
altogether exhausted our vocabularies in
the lower valley and found ourselves
altogether without means of expressing
the feelings that arose in us or of de-
scribing the scene before us.

Indeed, we did not ourselves fully
realize the terrible desolation of the upper
valley until our return to the base camp.
Coming back into the lower valley after
the total desolation of the country in
the shadow of the volcano was like re-
gaining the earth after an exile in the
inferno.

When we came back across Soluka
Creek we were astonished at our feelings
of exhilaration, as we relaxed from the
unconscious tension of the upper valley.
The broken remnants of the former trees
appeared the most luxuriant verdure.
The branches seemed alive with singing
birds. The mountains seemed covered
with herbage from base to summit. And

this was the country that a few days
before had seemed utterly devastated
in comparison with Kodiak!

Despite the complete devastation, one
of the expected agents of destruction was
conspicuously absent. Nowhere did we
detect the slightest evidence of fire.

We would not have been surprised to
find that the explosion was followed by
flows of molten lava, and we supposed
that the ash falling close around the
vent was hot enough to set the trees
afire. Nowhere in Katmai Valley did we
detect the slightest evidence of high
temperature.

ASH COOL WHEN IT FELL

On the contrary, the deposits give the
clearest testimony that the ash was cool
as it fell. So far as we could see, it did
not melt the snow where it had fallen on
the drifts on the mountains. We found
places where even coarse lapilli had sifted
in among the stems of the dry grass
without visibly affecting them.

The ash did not even cook the plants
on which it fell. Later we found places
well up on the slopes of the volcano,
from which the ash had slid off, where
roots of the old plants were sending up
new shoots. It is difficult to see how
these could have survived, if the ash
had had a temperature even approaching
the boiling point.

The valley had become gradually nar-
rower as we ascended, and was now seen
to terminate in a precipitous canyon,
which, because of our inexperience with
the country, seemed close at hand. Nor
did we realize at first the great height
of the cliffs, which are very impressive
on nearer view (see page Ii8).

Later we found that the mouth of the
canyon was nearly two miles beyond us,
instead of the short half mile it appeared.
Fortunately for our purpose we had
already reached a position which we
judged must be opposite Mount Katmai,
though we could have no assurance of
our location because of the low-hanging
clouds.

Across the valley from our camp we

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

could see what we supposed to be the
base of Mount Katmai. Here we beheld
a formation quite different from anything
we had ever seen.

A ravine that branched off from the
main valley behind a spur of the moun-
tain was filled by what looked like a great
glacier, except that its color was a terra-
cotta red. In every detail of its form,
except for the crevasses, it was exactly
like a glacier. Beginning at a consider-
able elevation, where the ravine was
narrow, it sloped evenly down to the
valley level, widening as it descended, so
as to assume a triangular form.

THE KATMAI MUD FLOW

If the color had not been so different
from everything else in the landscape we
would assuredly have taken it for a
glacier covered with dirt. But so close
to the volcano no glacier could have
escaped a thick covering of the omni-
present ash. We concluded, therefore,
that it must be a mass of mud which
had rolled down the volcano after the
ashfall.

Later, when we visited it, its structure
confirmed this theory. Sections cut by
a marginal stream showed that it lay
on top of the strata of ash, proving that it
had been formed after the ash had fallen.
Although it was hard and firm, affording
easy walking, both its structure and its
form showed clearly that it had reached
its position in a semi-fluid condition.
Like a glacier, it had a relatively steep
front and was convex, highest in the
middle, turning the drainage off to the
edges, along each of which a deep canyon
had been cut.

But despite the indications that it
had once been fluid, we saw no mud
cracks or other evidence of shrinkage
upon drying out, such as one would have
expected to find in a mud flow. Its
length was two and a half miles. Its
highest part was nearly i ,000 feet above
sea-level, from which point it sloped
gradually to about 300 feet at the base.

We were not able to estimate its thick-
ness so well, but along the edge where it

was cut into by the stream a section
about 50 feet thick was exposed. In
the middle it may have been much
thicker, both on account of the convexity
of the surface and the greater depth of
the valley floor.

FIRST CROSSING OF KATMAI RIVER

The river, as it debouched from its
canyon into the wide valley, spread out
in a dozen streams over the broad fan of
gravel that had been carried out of the
canyon. These leaping, ever-changing
streams were the most serious obstacle
we encountered in all our exploration of
the district.

The quicksands may be more danger-
ous, as their passage is certainly more
fatiguing, but the force of these torrents
so overpowers a man that his strength is
useless against their rush. We were
naturally anxious to cross over and
explore the mudflow, whose outlandish
appearance excited our keenest curiosity.
Procuring strong poles we started out.
From a little distance the streams looked
easy enough, but when we waded in we
found the current more than we could
stand against.

It- was extremely hard to lodge our
poles against the bottom or to maintain
our balance while we shifted our props.
We could hear the ominous roar of great
bowlders, rolled along with the current.
Smaller stones were continually bumping
against our shins. Wherever we set our
feet the bottom had a most disconcerting
way of washing out from under us. In
the first attempt both Fulton and I went
down helpless. Luckily, the packing of
our cameras kept them dry, but we
came so near to disaster that we dared
not try it again.

LATER EXPERIENCES IN CROSSING THE
RIVER

During the winter of preparation for
the next expedition, I racked my brains
and sought counsel from all my friends
for some device by which this river could
be mastered. The streams were too
wide and too changeable to permit

Digitized by

Google

ALPINE GLOW ON MOUNT MARTLN
Under the last rays of a setting sun a few small fumaroles are still discernible in the shadowy Valley.

Ill

Digitized by

Google

Natural Color Photograph by Frank I. Jones

FUMAROLES AT THE FOOT OF FALLING MOUNTAIN
Seen from Greased Hill between Baked and Broken Mountains.

Digitized by

Google

DEVASTATION IN KATMAI VALLEY

bridging, even if the expedition could
have summoned force enough to bring
the necessary logs into position. The
best expedient we could devise was a
rope, which was used throughout the
summer of 191 6.

One man took a long rope upstream,
while a second tied the other end around
his shoulders, and, bracing himself with
the rope held on the bank, slowly worked
his way across. After the first man was
over, it was comparatively easy for the
others to slip along the rope, held taut
on both banks. The last man had the
advantage of using the rope as a radius,
swinging downstream.

Even with this assistance, the ford
had to be chosen with extreme care. If
we tried a place where the water was too
strong, the man on the rope went down
and had to be hauled ashore like a fish.
Such an accident is great sport for those
on the bank, but one of the most horrible
of experiences for the man in the water.
As he goes down, struggling his hardest,
a sickening realization of utter impotence
comes over him. Despite the strong
rope, he is sure he is lost and feels all the
terrors of a drowning man.

We gave the problem renewed con-
sideration in planning the expedition of
191 7, and after watching motion pic-
tures of boats in the Grand Canyon,
decided that a boat might be used here,
although the problem was very different,
namely that of holding a boat against
the current, rather than of merely drift-
ing with it. As an experiment, therefore,
we procured a folding canvas boat.

It was with great interest that we
awaited an opportunity to try it at
Martin Creek, which is so swift that
neither oars nor poles could be used.
The procedure was to fasten a line to the
bow, with which the boat could be held
against the current. Then, by means of
a "gee string" attached to the line a few
feet from the bow, the man in the boat
could turn it into the current, forcing the
water to push it across the stream, just
as a kite rises when the wind strikes it at
an angle. The thing worked perfectly,

except that it was hard to land. The first
time I tried, the boat steered out across
the channel close to the opposite shore,
but there it stopped and could not be
made to go farther. I had forgotten
that the current slackens against the
bank, and I suddenly found the power cut
off almost within arm's length of the goal.
Here was a predicament, and not until
one of the men had foundered the boat
in an effort to jump ashore was a solution
found. Even close in, the water was too
swift for oars or poles. So we decided to
try grappling. Throwing an improvised
anchor ashore, we could catch hold in
the gravel and pull ourselves in.

WAITING FOR A SIGHT OF THE VOLCANOES
THROUGH THE CLOUDS

Katmai River so nearly overpowered
us on our first crossing in 19 15 that we
dared not risk another attempt, and so
gave up all expectation of exploring the
volcano that year. But we still hoped
at least to get pictures of it. Up to that
time it had never been photographed nor
even described.

The clouds which shut us in as we
first came up the valley continued to
hang low without any sign of breaking
throughout the next day. All that night
and the next I peeked out under the side
of the tent at frequent intervals, to make
sure that we should not miss a sight of
the mountain if it cleared for a few
moments. But there was no change
and we waited on all through the third
day. As the hours passed, anxiety grew
into fear that we might be altogether
cheated. In this country I had often
seen the clouds hang unbroken for a week
around much lower mountains than these,
and I knew it was not at all unlikely that
our slender food would be exhausted
before the sky cleared.

But finally the clouds lifted, slowly at
first, and then with a sweep. At the
head of the valley stood the 1,000-foot
cliffs which guard the entrance to the
inner canyon of Katmai River, while
behind us, towering aloft over inaccessi-
ble precipices, the summits of Slide and

Digitized by

Google

5
a

B
o

so o
o^ o

6 2

O C0

u c

2 6

3
O
3
O

'S.

Digitized by

Google

DEVASTATION IN KATMAI VALLEY

Avalanche Mountains themselves pre-
sented spectacles fine enough to command
attention in any other setting. But
here they were eclipsed, for on the other
side lay the whole chain of glacier-covered
volcanoes in a continuous series, broken
only by Katmai Pass, whose 2,500 feet
looked low indeed by comparison.

A SECOND NEW VOLCANO, TRIDENT

Farthest to the westward stood Martin
(see page 1 80) , shooting a column of white
vapor a mile into the air; next the three-
peaked summit of Mageik (see page 80),
covered with a mantle of freshly fallen
snow; then, beyond Katmai Pass, an-
other volcano which we named Trident;
and at the end of the line, Katmai.

Trident was a lofty mountain (altitude
6,790 feet) which from our position
showed three peaks, suggesting the name.
The whole mountain was clearly vol-
canic, but it was much eroded, leaving
little indication of the crater from which
the lavas composing it had come.

Evidently it had not been active for a
long time, until, in the general shake-up
accompanying the recent eruption, a small
fissure had opened on the flank toward
us. From this issued a thin column of
smoke that often rose 3,000 feet or more
before dissolving (see page 98).

Between Trident and Mageik lay Kat-
mai Pass, which, though in - no wise
interesting as a piece of scenery, excited
our curiosity because of its historic
interest as the route of the gold seekers
at the time of the Nome stampede.

Could we have known what lay beyond
it, awaiting our discovery, we would have
studied its contours with even greater
interest. Although reputed difficult and
dangerous, it looked very easy from our
position.

FIRST SIGHT OF MOUNT KATMAI

Finally, east of Trident, stood Katmai
Itself at the end of the visible portion of
the range. But how different from the
Katmai of former days!

Before the eruption it must have
dominated the country round about as

Mageik does now, for although much less
favorably located than Mageik, it was
given the name of the town, as were
Katmai River and Katmai Pass. It
must have been the dominant mountain
of the district, but the first glance showed
how completely it had been eviscerated
in the eruption.

Thie three-peaked summit, formerly
towering 7,500 feet high, was entirely
blown away, leaving a great crescentic
rim at the top of the stub. We could
not then suspect the stupendous magni-
tude of the crater that lay within, but it
was evident at once that it was of enor-
mous proportions (see pages 100 and 168).

It must be confessed that at first sight
the appearance of this broken cone was
somewhat disappointing. Its great bulk
so reduced its apparent height that our
first estimates of its altitude had to be
increased by half, as we compared it
more carefully with Trident and Mageik,
whose elevation we knew. No snow was
visible anywhere on its slopes when we
first beheld it, and the dirty ash-covered
surface suffered in comparison with the
glistening snow fields of Mageik.

The slopes were gentle and the ascent
looked easy. But as we studied them
with our glasses we saw that the absence
of snow and ice was only apparent.
Even from our station at a distance of
five miles we could see that a large part
of the mountain was covered with gla-
ciers, whose yawning crevasses we could
make out, despite the heavy mantle of
ash which concealed the ice.

Remembering our experience with
melting snowdrifts beneath the ash, whose
treacherous caverns we still feared greatly,
we decided that the risk of climbing
without an alpine rope was too great to
undertake. We made use of our op-
portunities for observation, however, by
studying the slopes, both to pick out for
future use the best route to the rim and
to ascertain something of the structure
of the mountain.

The first glimpse of the volcanoes
sufficed to reveal something of their
character and to suggest a number of

Digitized by

Google

'-S

100

i 4) 9

Digitized by

Google

DEVASTATION IN KATMAI VALLEY

lOI

considerations which subsequent study
enabled us to confirm. From their out-
line alone it was evident that previous to
the recent outburst none of them had
ever suffered a major eruption of the
explosive type.

VOLCANOES ALL OBVIOUSLY PILES OF
LAVA RATHER THAN CINDER CONES

Nowhere in the district are there any
decapitated mountains or old craters
such as would have been left by a
great explosion. These mountains, more-
over, showed little resemblance to the
typical steep-sided cones formed by alter-
nating deposits of fragmental cinders and
fluid lava, like Shishaldin or Fujiyama.

On the contrary, their round domes
and gentle slopes were built up almost
entirely by successive lava flows, with-
out the intermixture of any considerable
amount of pumiceous material. Some
of the other cones, like Cerberus, not
visible from our position, were indeed
steep enough to remind one of cinder
cones, but an examination proved that
they, too, were built of compact lava,
probably squeezed out in a nearly solid
condition like the well-known spine of
Pel6e.

* This statement refers only to the lavas of the
present volcanic cycle. The numerous lavas of

The lavas were poured out on a plat-
form of sedimentary rocks, Jurassic
sandstones and shales, which suffered re-
markably little change in the formation
of the volcanoes. In many places their
strata are bent up against the flanks of
the volcanoes, as though blistered up by
the injection of a mass of magma beneath,
but otherwise they still lie in nearly
horizontal layers, almost as they were
laid down on the former sea bottom.

Not only the type of eruption but the
composition of the lavas of the district
has been remarkably constant from the
beginning of their activities up to the
present eruption.* All the specimens
examined are basic andesite approaching
basalt in composition.

In view of the uniformity of the ancient
lavas, the strikingly differeni; character
of the ejecta of the recent eruption, which
is an acid rhyolite with a high silica con-
tent, is interesting.

This sharp contrast between old and
new lavas has made possible some very
instructive petrological studies, which bid
fair to carry us farther in an understand-
ing of what really happened in the erup-
tion than any other line of evidence, as
detailed in Chapter xxii.

old Tertiary volcanoes round about have not
been studied extensively.

Digitized by

Google

§ -d
§ °-

° Is
^ J

Digitized by

Google

THE KATMAI FLOOD

When we first landed on the shore of
Katmai Bay, we found the countryside
ravaged by a great flood whose waters
were just subsiding. The whole flat six
miles wide had been covered with water.
Considering the enormous volume of
water necessary to inundate a valley
opening freely to the sea, we might
have realized that such a flood was a
very unusual occurrence.

But we would not have been surprised
at anything we might have encountered
in that weird country where so many
physiographic relations were turned top-
sy-turvy. We did not therefore recognize
at once that the flood was an altogether
unprecedented event.

Even before our landing we began to
observe the evidences of this flood, for
we found Katmai Bay full of uprooted
trees, bushes, and fields of floating pum-
ice, which had but recently been brought
down by the flood waters.

The soft sticky mud previously men-
tioned made it difficult for us to find a
suitable place for our tent. The bed of
pumice that we finally selected was in-
deed only freshly floated into place by
an eddy in the waters. It was possible
as a camp site only because its loose
porous character readily permitted the
water to drain through it. We were in
constant fear, however, lest the water
should suddenly rise again in the night
and drive us out.

EXCEPTIONAL HEIGHT OF FLOOD EVI-
DENCED BY CONDITION OF KATMAI
VILLAGE

When we came to explore the remains
of Katmai Village, to which, fortunately,
the people had not been permitted to
return after the eruption, the magnitude
of the flood was impressed on us as it
could not have been on the brush-covered
flats.

Here we found that the little Russian
church in which the people had wor-
shipped for years without interruption
had been floated loose from its foundations
and left standing in a sea of mud. The
high-water mark could be plainly seen
across the front, about nine feet above
the normal river level (see page 104).

Some of the native houses were solidly
packed with pumice, full to the eaves.
Some had been completely submerged,
as might be seen by the stranded pumice
which had washed on to their roofs. The
roof of one had floated away from the
body of the house and lay at a little
distance.

It was evident enough to us, as we
looked about the village, that the build-
ers of the houses had not contemplated
the danger of any such flood, for there
had been no effort to select high ground.
On the contrary, convenience to the
landing place and shelter from the winds
had apparently been the most important
factors in the choice of building sites.
This was the more significant because of
the manner of construction of the native
"barabaras." These are semi-dugouts
with their rooms deeply excavated in
the ground and, therefore, require a
well-drained site far more urgently than
an ordinary house, built above the
ground.

Directly in front of the houses had
been the landing place. Before the
eruption, boats drawing five feet of
water had freely crossed the bar and
anchored at the village; and even after
the eruption, Martin had succeeded in
entering the river in a 35-foot schooner.
But at the time of our visit it was so
choked with volcanic ash that even the
main channel was impassable to a row-
boat.

Where deep water had been before
there was now only a maze of shifting

103

Digitized by

Google

104

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by B. B. Fulton
KATMAI CHURCH STANDING BLEAK IN THE DESOLATE MUD FLATS LEFT BY THE FLOOD

The high water mark even with the bottom of the bell may still be made out. Some conception
of the volume of the water necessary to fill the valley to such a level may be gained from the fact that
the flat valley is six miles wide at this point. Moreover, it opens freely out to sea so that there was
no barrier to retain the water.

quicksands, spreading out for five miles
all the way to the foot of the mountains
across the valley. Over this vast stretch
the water was on the whole hardly five
inches deep, yet the current was so
strong and the quicksands so treacherous
that it was equally difficult to ford from
side to side or to follow the channel. We
tried once to "track" a light canvas
canoe up the river, but found it out of
the question. We also tried to cross
with skis, which helped considerably
with the quicksands, but became un-
manageable in the currents.

SPECULATING AS TO THE CAUSE OF THE
FLOOD

Conditions at the village greatly in-
creased our respect for the magnitude of
the flood, but failed to enlighten us as
to its cause. The volume of water had
been tremendous, considering the size

of the watershed, for although the main
stream is less than 40 miles long and has
a steep gradient through much of its
course, the water had filled the whole
valley, six miles wide, reaching a maxi-
mum depth of about 10 feet. We knew
that there had been no general storm
which could have caused any unusual
quantity of rain.

As tide-water towns are rarely affected
by any floods, except inundations caused
by incursions of the sea, our first idea
was that the spring tides which had just
passed had overwhelmed the land.

Katmai Valley in particular opens so
freely out to sea that it seemed impossible
that anything less than the ocean itself
could furnish water enough to cover the
whole of it. A little examination showed
also that the water had risen many feet
above any former tide mark. An unusu-
ally high tide is, however, not a local

Digitized by

Google

THE KATMAI FLOOD

105

Photograph by R. F. Griggs

KATMAI RIVER AFTER THE FLOOD

Before the eruption boats drawing as much as five feet of water used to enter the river and anchor
at this point. A good-sized gas-boat entered the river some months after the eruption, but it is now
impossible to navigate a rowboat even in the main channels.

aflfair, and would have affected Kodiak
as well as Katmai.

We saw at once, therefore, that the
flood could not have been tidal, unless,
indeed, the whole valley had been sunk
as a result of the eruption. As the con-
dition of the beach at once negatived
this possibility, it was evident that the
water must have come down from the
mountains, rather than up from the sea.

We then suspected volcanic rains up
the valley, for at that time we had no
knowledge of the condition of the vol-
canoes. As a matter of fact this hy-
pothesis was incorrect, and our explora-
tion of the village left us none the wiser
as to the real nature of the situation we
had to face.

Nevertheless, our examination of the
village was very reassuring, for in giving
a better conception of the enormous
volume of water that had poured down

the valley, it enabled us to rest a little
easier in the thought that such a deluge
would not be likely to come upon us
unawares.

This, as it turned out, was not alto-
gether correct, for the flood must have
pushed down the valley at a high veloc-
ity. Everything at the village, how-
ever, indicated that the flood was an al-
together unprecedented event, not likely
to be repeated. This we later found
more nearly in accord with the facts.

FLOOD CONFINED TO KATMAI RIVER

As we explored our way upstream, we
were surprised to find that the tributaries
gave no evidence of having participated
in the flood. When we thus observed
that, like the smaller brooks, neither
Soluka Creek on the north nor Martin
Creek on the south had contributed in
any way to its waters, our suspicions

Digitized by

Google

io6

THE VALLEY OF TEN THOUSAND SMOKES

r-noiOKrapn oy c. v^. ivoid

A TREE LODGED HIGH UP IN THE BRANCHES BY THE KATMAI FLOOD
Occurring far down near the mouth of the river this is significant evidence of the volume of water.

as to its exceptional character were
strengthened, for all of the water had
evidently come down from under Mount
Katmai itself.

Evidences of damage rapidly increased
upstream, particularly along the south
bank of the river. Indeed, one of the
most striking features of the flood was
the difference in its action on the opposite
banks of the river.

All along the north side of the valley,
which is the concave side of the great
curve formed by the river's course, the
wreckage had evidently been deposited
only from slack water backed up over
the flat. But along the convex south
bank, signs of a strong current were in
evidence far down toward the mouth of
the river. Strange as it may seem, the
high water marks along that bank were
many feet higher than on the opposite
bank. Near the south bank, for ex-
ample, we found a large tree torn up by
the roots that had floated down through
the forest and almost out to sea before
it finally lodged lo feet above the ground
in the forks of another tree. But on

the north side, the wreckage consisted
entirely of logs and sticks, such as could
be carried by shallow water.

Upstream, the high water marks gave
increasingly impressive testimony of the
enormous volume of water. Just above
the mouth of Martin Creek it had reached
25 feet on the slack water (north) bank.

FOREST UPROOTED AND HEELED OVER
BEFORE THE WATERS

The different degrees of violence of the
flood were well manifested by the vary-
ing effect on the forest in its path. In
the outermost zone the trees were simply
piled high with wreckage, as happens in
any ordinary freshet.

In places where the water, though not
approaching the violence of the center of
the stream, yet surpassed the power of
ordinary floods, the trees were laid over
before the current almost flat to the
ground. These trees (see page 107)
were not merely bent over before the
waters as in an ordinary freshet; they
were uprooted or broken and partially
torn loose from the ground.

Digitized by

Google

THE KATMAI FLOOD

107

Photograph by J. W. Shipley
WORK OF THE FLOOD IN A MORE SHELTERED PORTION OF THE VALLEY

Where the flood waters were not so violent as to completely uproot or break down the forest, the
trees were laid over before the current for miles at a stretch. The ash in which the trees are buried
must all have been deposited by the flood itself for it would be impossible to tilt the trunks in this
way while imbedded in the ground.

In the later stages of the flood, ash
and sand were dumped deep around them,
leaving only their stiff straight tops stick-
ing out of the smooth plain at an acute
angle, like the bayonets of an army that
would deny passage to the invader.

The exposed tops of these trees bear
impressive testimony to the abrasive
action of the flood waters. They were
deeply scarred by the fragments of pum-
ice hurled against them by the raging
current. Often the tips of the branches
were thus ground into sharp points,
dangerous to a man traveling upstream.
The bleached whiteness of these sticks,
so small as to be inconspicuous at best,
rendered them almost invisible against
the background of white ash.

TREES SHEARED OFF FOR MILES AT A
STRETCH

But where the forest lay directly across
the path of the flood, the violence of the

waters was too great to admit of the trees
being uprooted. Here for several miles
the sudden violent impact of the water
snapped them off at the level of the
ground (see pages 100 and 102). One
would never suspect that these places
had ever been forested, except where the
stumps had been exposed by subsequent
erosion.

Some of the trees apparently were
sheared off by bowlders carried along
by the waters. Others were bent and
splintered indicating that they were
twisted off by the direct power of the
rushing water itself, without the assist-
ance of rolling bowlders. Occasionally
later erosion has uncovered and washed
out the stumps of such trees, but for the
most part they remain fast in the ground
where they grew (see page no).

There is another zone, where annihila-
tion was still more complete — the area
where the trees were completely extir-

Digitized by

Google

io8

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

THE QUICKSANDS OF KATMAI RIVER

After our experience in Soluka Creek none of us was willing to try crossing the river, but it probably

could be done.

pated, root and branch. But, since no
one knows the country as it was before
the eruption, it is impossible to guess
how extensive this zone of complete
destruction may have been. There is,
however, good reason for the belief that
trees once occupied considerable areas
which now show no evidence of ever
having supported vegetation.

A 250,000-POUND BOWLDER PICKED UP
BY THE FLOOD

The head of the flat, where the river
debouches from Katmai Canyon, is
covered with a great train of bowlders,
carried by the waters and spread for a
distance of two or three miles down the
broad valley. The general distribution
of these bowlders is similar to that of the
bowlder trains to be found along any
mountain stream.

What is peculiar here is the great size
of the rocks. Many are 10 feet in di-
ameter, and a few considerably larger.
The largest of all measures 56 feet (17
meters) in circumference. Although only
7 feet (2 meters) of this bowlder is visi-
ble above the present surface of the

ground, it is probable that there is as
much more below ground. The sand
plain in which it lies is also a product of
the flood, having probably been built up
several feet above the original surface
by deposits from the falling waters, as
may be seen by the half-buried trees in
the picture (see page no).

It is altogether likely, therefore, that
the big bowlder rests on the ground in
which the trees were rooted. It is,
moreover, composed of compact heavy
lava (specific gravity 2.5), and so nearly
spherical as to present a minimum sur-
face for the waters to take hold of.
A little computation shows that its
weight is between 200,000 and 300,000
pounds.

That such enormous rocks had actu-
ally been transported long distances by
a current of water, however, was not
easily to be credited. Only after we had
gone over the situation carefully did we
permit ourselves to accept such a con-
clusion.

At first sight, one would be inclined to
assign a glacial origin to these bowlders,
for to a glacier their size would offer no

Digitized by

Google

THE KATMAl FLOOD

109

r-noLugrayn oy iv. r. vrriKgd

FREAK OF THE FLOOD

Bowlders washed up on top of a bed of ash which was later sectioned by a stream. Under ordinary
conditions the soft sandy ash bank would have melted away at the first touch of the water, but in
the rush of the flood it stood undisturbed by waters capable of carrying large stones. One rock so de-
posited measured 5 by 3 by 2 feet.

such obstacle as to running water. But
their distribution does not conform to
that of a glacial terrane ; they occur exclu-
sively over a fan-shaped area radiating
from the mouth of the canyon, and are
altogether absent from the sides of the
valley, where they would inevitably have
been strewn by a glacier.

Moreover, we later found places where
a relatively insignificant flood from Ma-
geik Creek had rolled large bowlders up
against the trees in such a way as to
leave positive proof of what had hap-
pened. The Great Flood was so much
more violent that it had destroyed those
evidences of its action which are left
behind in an ordinary freshet.

A TEMPORARY NIAGARA OVER AN
OBSTRUCTING RIDGE

Farther upstream, where the valley is
about half a mile wide, it is thwarted by
a transverse ridge of rock stretching al-
most across it like a baflle. At ordinary
stages the river winds from one side of
the valley to the other, around the ridge.

But the marks left behind show that the
swift rush of the flood abandoned the
ordinary channel, for over most of the
ridge, the ash and soil had been washed
away, exposing the bedrock beneath and
showing that the flood waters rose clear
over the obstruction.

At each side were clear-cut high-water
marks, where the ash layers had been
left untouched by the water. As it
poured over this ridge, the flood formed
a tremendous waterfall, a thousand feet
wide and a hundred feet high (see pages
112 and 114).

Observing that these dimensions are
inferior to Niagara, the reader may have
concluded that my caption, ** A Tempo-
rary Niagara,*' is not justified. If I may
anticipate a little, however, I may state
that when our survey was carried up the
canyon we found that the ** flood-fall,'"
while it lasted, had greatly surpassed
Niagara in the volume of its discharge.

At one side of the flood-fall a deep hole,
which still retained a pool of clear water^
had been excavated, offering us a most

Digitized by

Google

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by D. B. Church
A FLOOD-BORNE BOWLDER WEIGHING OVER A HUNDRED TONS

This lava bowlder was carried a mile and a half out of the canyon on to the flat: In the back-
ground the broken forest sheared off at the level of the ground when the flood struck.

welcome opportunity for a plunge in a
country where bathing facilities are few
and far between. It was here that we
came upon the first decisive indications
of the nature of the flood.

It had evidently consisted of a single
quick terrible wave, rather than any
long-continued inundation. So sudden
had been the rush of the waters and so
quickly had they subsided, that in many
places they had not removed all the soil
from the rock on the ridge, but had left
behind considerable masses of earth
and hardpan, which must soon have
crumbled away if the action of the water
had been long continued.

A RIVER WHOSE BOTTOM WAS HIGHER
THAN ITS BANKS

The crest of the rock ridge is not level,
but shows an irregular profile. At the
right is a low notch directly above the
pool, which, by drawing an increased

flow of water, was doubtless responsible
for the excavation of a hole at that point
(see page m).

Beyond this is a high place which
stood up out of the water, as shown by
its cap of ash undisturbed by the current.
At the left the ridge gradually sinks to the
level of the river bed. We were aston-
ished to find that this, the lowest part of
the ridge, marked (A) in the picture, also
had a cap of undisturbed ash, showing
that the water had never risen over it.
All around it is a clear high-water mark,
indicating absolutely the edge of the
stream. This high-water mark is no less
than 40 feet lower than the hill to the
right, from which the water washed all
the soil (see page 112).

That is to say, the water rushed over a
high hill in the middle of the ridge but
did not overflow the lower ground at the
side. In other words, the bottom of the
stream was 40 feet higher than its bank!

Digitized by

Google

THE KATMAI FLOOD

III

Photograph by D. B. Church
A NEARER VIEW OF THE RIGHT HALF OF THE FLOOD FALL, SHOWING THE POOL EX-
CAVATED BY THE FALLING WATER

A man is posted at the high water mark to the left.

The bottom of the stream, as shown in
the photograph, stands at an angle 15
degrees above the horizontal, as seen
from the shore. To this must be added
a further elevation to allow for the depth
of the water over the rock, which must
have been considerable, amounting by
the least calculation to several degrees.
I have sketched in by dotted lines on the
photograph the approximate ''level*'
held by the surface of the water. The
figures are the elevations of the several
parts of the flood-fall in feet above sea-
level, as determined by triangulation by
our topographer, C. F. Maynard.

If one had been so fortunate as to be
standing on the end of the ridge, at the
point marked A in the photograph
(page 114), he would have seen literally
a mountain of water rising far above
him, yet never ''seeking its level" nor
overflowing the lower ground on which
he stood, and could have remained in
safety till it was past! But if on hearing

the warning roar of the flood water up
the canyon, and realizing the danger, he
had attempted to escape by climbing up
the ridge on to the higher ground, he
would have been swept to destruction !

Incredible as this may seem, even in
the face of photographic evidence, yet
it is to be explained by perfectly well-
known principles of hydraulics, whose
operation on a small scale is familiar to
everyone.

Whenever the stopper is pulled out of
a wash bowl, the water swirling around
the vortex is driven altogether out of the
center of the whirlpool. The centrifugal
force becomes greater than the weight
of the water, which is displaced accord-
ingly.

It is seldom that this familiar behavior
of water may be observed on a large
scale, because neither high velocities nor
sharp turns are met with in ordinary
streams. But the flood rushed down the
canyon at such speed that it could not

Digitized by

Google

0) .

•So

8 s^

b
O

Si;

l-H

St!

•c

112

Digitized by

Google

THE KATMAI FLOOD

make the turn around the ridge, and so
ran over it, swelling up especially on the
outer side of the curve.

This is, however, only part of the
story, for it was observed that the high-
water mark was 15 feet higher against
the rock in midstream than against
the outside bank (see page 114). The
stream must, therefore, have bowed up
in the middle. This also is a very well-
known phenomenon to students of hy-
draulics, but is less familiar to most of
us because of our limited experience with
swiftly running streams.

The fact is, however, that any very
swift stream in a clear channel bulges up
in the middle. The reason is that fric-
tion retards the flow so much at the edges
that the water piles up in the middle,
where the path is open. One rarely has
the opportunity of observing such be-
havior in a great river, even in flood,
but in the rapid Alaskan streams it is
by no means uncommon on a smaller
scale.

In the swift channels of Katmai River,
which gave us so much trouble in fording,
the water frequently stood a foot higher
in the middle than at the edges, so that
we often saw the profile of the current of
rushing water against the skyline before
the restraining banks came into view.

In a stream with such tremendous
velocity as the Katmai Flood, this bulg-
ing up in the middle would, of course,
be much greater than in the channels of
a small river. The evidence that it
occurred should not, therefore, excite
particular surprise.

A FREAK OF THE FLOOD

Below the fall, and especially around
the edge of the pool, the water had
washed up a considerable quantity of
coarse shingle. This was like the gravel
brought down by any freshet, except
that the stones were larger.

The remarkable thing about it was its
position. It had been laid up on top of
ash that still lay undisturbed as it fell.
Since the ash here has the character of
coarse, light, loose sand, it melts away

at the first touch of a current of water
under any ordinary conditions.

That a flood capable of carrying good-
sized stones should have left untouched
a sand bank seems impossible. But
here, laid neatly up on top of such a
sand bank, was not only a pile of gravel,
but also stones of large size. Lava cob-
bles, a foot in diameter, were numerous
and there was one block of sandstone
measuring 5x3x2 feet.

The largest of these bowlders were not
so situated that their relations could be
shown by a photograph, but in another
place far downstream we found a bed of
cobbles similarly deposited on top of the
stratified ash in a position that could
be shown clearly in a picture (see page
109).

Folsom and I, who were alone at the
time of our first visit to the place, found
ourselves in complete disagreement as
to the origin of this shingle. From the
crest of the fall, where the origin of the
bowlders was clear, I had concluded
that the stones must be flood-borne.
Folsom had remained below, on one of
the largest bowlders, where the difficul-
ties of such an interpretation were more
manifest. When I told him that the
stone on which he sat had been left on
the sand bank by the flood, he burst into
violent opposition. "Nothing of the
sort could be possible."

"Yes,'' I told him, "I know perfectly
well that it looks impossible; but, never-
theless, there is unquestionable proof
that it actually occurred."

He could not believe it; so I asked him
to explain how else the bowlders could
have reached their present position, and
we began to go over the ground step by
step. The only possible alternative was
that the stones had rolled down from the
mountain side.

But a little examination showed this
hypothesis untenable. They were far
from the base of the cliff. They were
located in one definite hollow, surrounded
by a clear high-water mark, and separated
from the talus of the cliff by a long inter-
val of bare ash. Many of the cobbles

Digitized by

Google

to
O

»
H

P
O

e ^

•§

- 8

bA.M

||
22

-I

c u

> 2

114

Digitized by

Google

THE KATMAI FLOOD

were of dark lava, whereas the cliflf was
composed of light-colored rock. Most
of the cobbles were clearly water-worn,
and bore a very different appearance
from the angular blocks of the talus.

If the bowlders had lain on an ordinary
bed of sand, we might have supposed
that the sand had been deposited by the
receding waters afterward, but here they
lay on top of the undisturbed ash, still
regularly stratified as it had fallen during
the eruption three years before the flood.

It was, therefore, impossible to inter-
pret the bowlders otherwise than as
flood debris; and in the end, though
neither Folsom nor I could comprehend
how the thing had been accomplished,
we came into complete agreement as to
what had happened.

As we came gradually to a fuller
comprehension of the magnitude and
power of the flood, we were more and
more thankful for the good luck that had
delayed our landing. If we had landed
a week earlier, we would almost inevi-
tably have been overwhelmed, for our
work kept us on the flats and there would
have been no time to escape. Even if
by a lucky chance we had been on the
hills out of reach of the waters, our stores
would have been washed away and the
expedition would have been so crippled
that we should probably never have
been able to go on with the exploration,
which was finally to lead us into the
Valley of Ten Thousand Smokes.

FLOOD CAUSED BY THE ESCAPE OF WATER
IMPOUNDED BY A LANDSLIDE

While we studied the ravages of the
flood, our minds were busy with specula-
tions as to its cause, and as we traveled
up the valley and saw increasing evi-
dence of the localization of its source,
one after another of our hypotheses had
to be discarded.

When the evidence began not only to
point to a very local source, but to indi-
cate that it had lasted but a very short
time, there seemed to remain only one
possible explanation . Apparently a lake
somewhere up the canyon, dammed in

perhaps by material from the eruption,
had burst its banks and deluged the
valley. This was, however, merely a
speculative hypothesis, and so remained
for two years.

It was not until 19 17 that we were able
to penetrate the canyon and examine the
Second Valley and the Katmai Lakes,
which, meanwhile, we had discovered
from the mountain top.

When we finally made our way up
through the canyon, we found, high up
on the slopes of Katmai, a thousand-foot
cliff of bright, new, unweathered red
rock, where the mountain side was
gouged out by a great slide. This, fall-
ing into the canyon, had formed an im-
' mense dam, which, after holding for
three years, had yielded just before we
landed, causing the tremendous flood.

The material that broke loose was a
great mass of ancient lava, which, hav-
ing originated in the crater above, was
flowing down the slopes of the volcano
when it solidified. It thus probably
occupied an unstable position which may
have been rendered still more unstable
by undercutting from the river, so that
it was easily shaken down by the earth-
quakes preliminary to the eruption.

That it broke loose just before the
eruption was indicated by the fact that
its surface was everywhere covered by
the layers of ash as they fell after the
great explosions. In this case, there was
apparently only one major slide, for
there were no indications of the contin-
uous activity so prominent in Falling
Mountain and Noisy Mountain.

Although all the material came from
what looks from below like a relatively
small area of the cliff^, it piled up in the
valley to a depth of over 400 feet, and
spread up and down the canyon for
about a mile, making a flat, low dam
capable of resisting much more pressure
from the water than a less massive barrier.

THE SLIDE OCCURRED AT A NATURAL DAM
SITE, IMPOUNDING A LAKE OF 95O ACRES

By a curious coincidence this landslide
occurred at precisely that point which an

Digitized by

Google

Ii6

Digitized by

Google

THE KATMAI FLOOD

117

engineer would have selected for a dam
site, if he had desired to construct a
reservoir of maximum capacity with a
dam of minimum size.

The configuration pf the valley is most
peculiar, for just above the slide the
narrow canyon gives way to a second
broad flat valley in which a very large
quantity of water could be held by a
relatively small dam in the canyon.
This very remarkable combination of
circumstances greatly augmented the
magnitude of the flood.

If the valley had been a normal drain-
age system, gradually narrowing to the
head-waters, the volume of water im-
pounded would have been small and
capable of doing relatively little damage
when the dam went out. Indeed, Ma-
geik Creek furnished an example of just
this condition, for a slide also dammed
up its canyon. But, in this case the
lake, although probably 500 feet deep,
was so small that almost all of its energy
was consumed in destroying the dam,
and as a consequence the flood of Mageik
Creek was comparatively gentle.

In the broad expansion of Katmai
Valley above the dam there is abundant
evidence of the former presence of the
flood lake which it impounded. Across
the base of the mountain is clearly
drawn the line marking the beach left
by the waves of the former lake. This
shore line is 210 feet above the level of
the present Katmai Lake. Our survey
showed that at this level the lake covered
no less than 950 acres. Its maximum
depth was about 400 feet, and the quan-
tity of water impounded more than 8,000
millions of cubic feet (see page 120).

THE PROBABLE CONDITION OF THE DAM

Conditions in the Second Katmai
Valley, where the upper lake is retained
behind the slide from Noisy Mountain
(see page 127), presumably give a fair
idea of the character of the dam that
held back the flood. The slide debris
was probably loose and porous, allowing
a great amount of water to seep through
and appear below as large springs.

The probability is that the dam held
for some time after the reservoir was full,
for the break coincided with a period of
unusually hot weather, when the heavy
snow-melt had raised the waters of the
lake to an abnormal level, giving them
an increased head which enabled them
to cut into their bed and remove the
slide.

No evidence remains to indicate ex-
actly how the dam broke, but when the
waters had done their work the central
portion had been cut away. In its place
remained a trench as straight and dean-
cut as though excavated by man to
carry some great engineering work, like
the Panama Canal (see page 116).

Indeed, I can find nothing with which
to compare it except the Culebra Cut.
It is, to be sure, only about half as large,
involving the excavation of some 33
million cubic yards of material, as com-
pared with the 53 millions called for by
the original estimate at Culebra.

KATMAI CANYON FILLED WITH COBBLES
MANY FEET DEEP

The power of the waters was not suffi-
cient to cut down through the dam to
the original level of the canyon floor.
Not only did the **head" of the rushing
flood decrease as the water was drawn
off from the reservoir, but an immense
amount of debris from the dam itself ac-
cumulated in the canyon below, so that
before the lake had been emptied a bar-
rier was formed sufficiently strong to re-
tain the remainder of the water, which
forms the present lower Katmai Lake.

The present flat gravel floor of the
canyon was formed from the remains
of the dam. Probably nothing at all
similar existed before the catastrophe.

A knowledge of conditions here before
the flood would be of great service, not
only because of our direct interest in the
effect of the flood itself, but because the
canyon is the key to the unraveling of
much of the geological history of the
volcano and surrounding country. But,
as there seems to be little hope of finding
a record of antecedent conditions, we are

Digitized by

Google

2 «*

-st:

c
o

.S
•g

ii8

Digitized by

Google

THE KATMAI FLOOD

119

reduced in this, as in so many other
matters, to speculation.

The Katmai Flood was of an order of
magnitude so far beyond the floods fa-
miliar to residents of the Mississippi Val-
ley, for example, that there is need of
comparing it with other catastrophes of
the same sort, to gain some perspective
by which its rank may be judged.

THIRTY-FIVE TIMES AS GREAT AS THE
JOHNSTOWN FLOOD

For everyone who, like the writer, grew
up in the northeastern United States, the
Johnstown Flood will always remain
the standard of comparison by which all
similar events are to be measured. This
is generally considered to have been the
greatest flood disaster of history.

A reservoir impounding 640 million
cubic feet of water, forming a lake 400
acres in extent, was held by a dam about
400 feet above the city of Johnstown,
which was 12 miles distant in a narrow
but winding valley. On May 31, 1889
the dam broke. Everything within the
path of the waters was dashed to de-
struction. The exact number of victims
could never be ascertained, for few of the
bodies were recovered ; but it is certain
that between 10,000 and 15,000 people
lost their lives.

Every tree within reach of the water
was destroyed root and branch. Di-
mension stones of bridge abutments
were not only carried away, but so
ground up that they could no longer be
distinguished from the cobbles of the
stream. Credible witnesses reported
that of the bricks from the destroyed
buildings not one-twentieth part could
be found after the flood, the rest having
been ground to powder in the waters.
The locomotives in a roundhouse struck
by the waters were scattered far and
wide. Some were found at variable dis-
tances up to 1,500 feet away, but several
were carried off bodily and were never
found. One man, standing in a three
story window, saw one of these "come
dancing along the top of the flood 30
feet from the ground."

Such accounts give but a faint idea
of what the Katmai Flood would have
done in a populous valley. For the
Katmai Flood was much greater than
that of Johnstown. The volume of the
water was twelve and one-half times as
great, and its fall nearly three times as
much, giving the waters a total energy
(which is the measure of the magnitude
of the flood), thirty-five times as great!

THE GREAT GOHNA FLOOD

Although the Johnstown Flood was
the most disastrous in the annals of
engineering, both it and the Katmai
Flood were surpassed in magnitude by
the Gohna Flood of 1895 in India. Like
the Katmai Flood, this was caused by a
gigantic landslide, which formed a dam
900 feet high, impounding more than
16,000 million cubic feet of water.

When the dam broke 10,000 million
cubic feet of this water was discharged in
the space of four hours, producing a flood
260 feet high. Thirteen miles down-
stream the flood was 160 feet high (as
compared with 30 feet at Johnstown 12
miles below the dam) ; 50 miles below it
rose 113 feet; 72 miles downstream it
attained 42 feet; at Hardwar, 150 miles
away, its height was still ii feet.

In the upper part of its course the
Gohna Flood filled the valley with gravel
to a depth of no less than 234 feet, while
13 miles below, the river bed was raised
50 feet. As at Katmai, the basin was
not completely drained, a permanent
lake, two miles long, one mile wide, and
300 feet deep, being left.

But, although the Gohna Flood was
much greater than the Johnstown, and,
like it, occurred in a populous valley, it
stands in great contrast in that not a
single life was lost, except that of a
Hindu fakir who persisted in returning to
the danger zone after being forcibly
removed twice.

Profiting, perhaps, by the fresh mem-
oriesof the Johnstown disaster, the British
engineers began the task of dealing with
the anticipated flood almost as soon as
the landslide occurred.

Digitized by

Google

I§

n PS
B g

£ o
■J

O
O

u
n

120

Digitized by

Google

THE KATMAI FLOOD

121

By means of a careful survey they
were enabled two years before to predict
to within ten days the time when the
flood would occur. They constructed a
telegraph line down the valley to spread
the warning and erected a series of monu-
ments, above which the people could
rest in the assurance of safety. By
these means they converted an otherwise
inevitable disaster into a triumph for
civilization and engineering.

AX ESTIMATE OF THE VIOLENCE OF THE
KATMAI FLOOD

Without going further, these compari-
sons will enable us to fix with sufficient
accuracy the rank of the Katmai Flood.
While not the greatest of recorded floods,
it nmst evidently stand well up in the
front rank among such catastrophes.

In one respect, indeed, it should be
given first place; that is, in the violence
of its waters as they issued from Katmai
Canyon. Conditions here are almost
unique, for the narrow, clean-cut, nearly
straight rock canyon offered a minimum
of resistance (see page Ii8). The vol-
ume of the water, as compared with the
quantity of material that had to be moved
in opening the passage through the dam,
was also very large, so that little of its
energy was expended in clearing a path.

The velocity of the waters as they
issued from the canyon must, therefore,
have been terrific. I have tried in some
way to get a numerical valuation of this
velocity, but find it difficult because
engineers have never had experience
with such enormous torrents. After the
Johnstown Flood, the editor of the
"Engineering News** estimated on theo-
retical grounds that half of the energy of
the rushing waters was consumed in
friction, leaving the remaining energy
free to accelerate the speed of the water.
This estimate he found in close agree-
ment with the actual rate of flow as
recorded by observers on the ground.

If half the energy of the water. was

consumed by friction at Johnstown, prob-
ably a smaller amount was so expended
at Katmai because of the smoother,
straighter, deeper channel and the much
greater volume of water. It will, there-
fore, give us a conservative value of the
velocity of the flood if we assume that
friction consumed as much energy here as
at Johnstown. On this basis the waters
as they issued from the canyon must
have been running about lOO feet per
second, or nearly 70 miles per hour.

In these days of automobiles and air-
planes, 60 or 70 miles an hour may not
appear a very high velocity. As a matter
of fact, however, a river flowing 60 miles
an hour is as extraordinary as would be
an automobile running at three times
that rate, for flood waters never attain
such velocities as have become familiar
to us in other objects. The tremendous
power manifest in a great river in flood
is such an impressive thing that almost
irresistibly one overestimates the speed
of the current.

DESTRUCTIVE POWER A MILLION TIMES
THAT OF A MISSISSIPPI FLOOD

Although floods are often described as
running at ''express train speed," an
examination of the records shows that
even the destructive floods to which the
Mississippi and its tributaries are sub-
ject rarely attain current velocities ex-
ceeding seven miles per hour (10 feet
per second).

The difference in energy between a
stream running 100 feet a second and
one running 10 feet a second is so great
as to be utterly inconceivable. For the
power of water to carry off obstructions
in its bed — its destructive capacity —
varies as the sixth power of its velocity.
That is to say, if the velocity be doybled,
the transporting power is increased sixty-
four times. If, therefore, the Katmai
Flood ran ten times as fast as a Mis-
sissippi River flood, its destructive power
was increased a million times!

Digitized by

Google

Photograph by D. B. Church
CLIMBING OUT OF KATMAI CANYON
The wall opposite is as high as parts of the Grand Canyon, as beautifully colored, and as precipitous.

Digitized by

Google

THE CANYON AND SECOND VALLEY OF THE KATMAI

During the exploration of the country
we found ourselves many times in the
position of the man who on first seeing
Pike's Peak, from Colorado Springs,
started out to walk over before breakfast.

We underestimated the size of almost
every feature we encountered, and sub-
sequently had to enlarge our ideas as
familiarity gave opportunity for better
judgments or the survey supplied ac-
curate measurements.

In no case was this more conspicuously
true than in our estimate of Katmai
Canyon. We paid little attention to it
in the beginning, but when we came into
a position where we could observe it to
advantage we realized that it was one of
the most stupendous features of the
whole district.

Our first approach to the canyon was
by the natural path up the banks of the
stream. We were soon stopped, how-
ever, by the meandering river, which
turned against first one cliff and then the
other. We tried climbing along the
precipitous walls, but after some very
slow and difficult work gave it up and
sought another route. This we found by
way of the Katmai mudflow, which made
an easy path around the obstructions.

BEAUTIFUL WATERFALLS NEAR THE FOOT
OF THE CANYON

Near the head of the mudflow are
four beautiful waterfalls which my wife,
when she camped there in 1919, named
for our four children. Halfway up the
mudflow a clear brook comes cascading
over the rocks, a laughing, babbling
"baby" falls appropriately named, for
the youngest child, Rosamond's Falls.

A little farther on a good-sized stream
comes down from Katmai, leaping the
cliff at a single bound — 255 feet, we
found it, by triangulation. Originating
in the glaciers above, the water is always

full of silt. Such a muddy fall is quite
appropriately named for a harum-scarum
small boy, David's Falls. Rising at one
side of the notch cut in the mountain side
by David's Falls is an impressive domed
cliff of black lava, 1,610 feet high, re-
calling in its form El Capitan in the
Yosemite.

Across, on the south wall of the can-
yon, is a third, a beautiful misty, maid-
enly wisp of a fall, where a small stream
coming out from a hanging glacier above
tumbles over the cliff in a series of cas-
cades 750 feet high. This . is Ruth's
Falls (see page 130).

Up in the middle of the canyon an-
other strong stream comes tumbling over
the edges of the lava flows, dropping al-
together nearly 2,000 feet in a series of
cascades. Two of these are graceful falls
upwards of a hundred feet high. Al-
though too far away to show up to
advantage in the pictures, they may be
made out on page 126. These she called
Julian's Falls.

KATMAI CANYON ALMOST AS DEEP
AS THE GRAND CANYON

Few views in any country can be com-
pared with that from the head of the
mudflow across the canyon. The wall
opposite is higher than any single cliff in
the Grand Canyon, for Katmai Canyon
though only a few miles long is almost as
deep (4,000 feet) as the Canyon of the
Colorado. Its colors — delicate greens,
lit up by pale yellows, blues and soft
pinks — though totally unlike the brilliant
reds and yellows of the Arizona Canyon,
are perhaps quite as beautiful.

At the farther side of a tract of ir-
regular, rounded hills which lie between
the head of the mudflow and the river is
a sharp knob that stands right in the jaws
of the canyon. This hill has been
bisected by the river. When one has

123

Digitized by

Google

124

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

GLACIATED ROCKS AT THE MOUTH OF KATMAI CANYON

These and the similarly planed-off rocks in the bottom of the canyon, shown in section on the opposite
page, are clear evidence that Katmai Canyon is pre-glacial.

climbed up its nearer slopes and reached
the top he finds himself on the brink of a
perpendicular cliff, 500 feet directly
above the roaring river that is under-
mining its foot. From this hill — Prospect
Point, we named it — the finest views of
all are to be obtained (see page 126).

KATMAI CANYON CUT BEFORE THE
GLACIAL PERIOD

The opposite walls of the canyon show
very diverse rock structures. On the
south the wall is made of stratified rocks
dominantly gray-green in color. But
on the north the river is hemmed in by
massive lava flows varying from ma-
hogany color to dead black.

From Prospect Point the flat floor of
the canyon may be reached by a rather
difficult descent, almost straight down to
the swirling waters of the river.

Katmai Canyon differs very markedly
from the other canyons of the region in
having a broad flat bottom, despite its
steep walls; it is U-shaped, not narrowly
V-shaped as are the other canyons of the
district, for example, the Canyon of
Mageik Creek. This flatness could be

accounted for only in part by the debris
left by the flood . Consequently we were
not surprised to find clear evidence of
glacial work in the bottom, proving that
it was already in existence during the
glacial period.

Mageik Canyon on the other hand, al-
though 500 feet deep, is post-glacial. Its
rim is lined with glacial gravels evidently
laid down before the canyon was cut.
A study of the drainage area shows,
moreover, that before the glacial period
Mageik Creek drained into Martin Creek.
It was apparently diverted to its present
course by the glacier that once filled the
depression between Mount Mageik and
Observation Mountain.

PROOF THAT THE LAST LAVA FLOWS FROM
KATMAI WERE POST-GLACIAL

The exposures of Katmai Canyon fur-
nished us also with our first clue to the
date of the previous activity of Katmai.
We discovered that one of the last lava
flows had reached into the very bottom
of the canyon, where its base rests on the
glaciated surface of the sandstone. Evi-
dently then the last eruption, though ante-

Digitized by

Google

THE CANYON AND SECOND VALLEY OF THE KATMAI

125

Photograph by R. F. Griggs
AN OLD LAVA FLOW FROM MOUNT KATMAI RESTING ON GLACIATED ROCK IN KATMAI

CANYON

At the base of the cliff shown on page 126, the bed rock is Naknek sandstone (Jurassic), which is
beautifully planed and striated by glacial action. Here and there a small remnant of till, indicated
by the hammer, still remains below the lava.

dating the historic record, had occurred in
geologically recent times.

GRAVEL BANKS SO UNSTABLE AS TO BE
SET ROLLING BY A SLIGHT BREEZE

At the remains of the dam, which had
held back the waters of the Katmai
Flood, we found that the impetus of the
torrent had been exhausted before the
dam was entirely cut away. There re-
mained a considerable bank of unstable
material into which the river was busily
gnawing at the time of our visit, lowering
its level perhaps ten feet a year.

Near the position of the former crest
of the dam the stream was constricted
into tumultuous rapids and was under-
mining its banks so rapidly that they
stood at the critical angle, sometimes
misnamed **the angle of repose.**

So unstable was the equilibrium of the
banks at this point that even a good
breeze was sufficient to start them tum-
bling into the river. During calm weather
we could clamber over the bowlders up

and down the gorge with safety, but I
have seen the wind tear up yard upon
yard of loose material and send it rolling
into the river.

Because of my ignorance of this con-
dition I had a narrow escape the first time
I traversed the canyon. I was standing
on a rock, photographing the rapids,
when a sudden gust started the stones
above me rolling, and before I realized
what had happened a big one went whiz-
zing into the water, just missing my head.
It was only the first fragment of a small
avalanche that was coming my way.

THE SECOND VALLEY OF THE KATMAI

Above the remains of the dam we
emerged into the Second Valley of Kat-
mai River. Across the base of the
mountain opposite was the conspicuous
horizontal line made by the beaches of
the old flood lake, 210 feet above the
level of the lower Katmai Lake, which
still persisted in the lowest portion of
the basin (see page 120).

Digitized by

Google

Photograph by R. F. Grigga
STEAM RISING FROM MOUNT KATMAI : SEEN FROM PROSPECT POINT

The ash slides of the recent eruption contrast with the massive ancient lava flows. At the right are
two fine waterfalls. The summit stands about a mile above the observer.

126

Digitized by

Google

THE CANYON AND SECOND VALLEY OF THE KATMAI

127

Before the last lava flow the canyon
evidently lay closer to the volcano than
the present gorge. By filling the river
bed the lava forced the stream against the
opposite bank, compelling it to cut into
the south wall and form the present cliflf.

From the general appearance of the
volcano we were inclined to suspect that
such a process had occurred repeatedly,
but we could not find evidence enough to
reach a definite conclusion as to the
original course of the drainage.

CURIOUS PHYSIOGRAPHY

The Second Valley lies in such a posi-
tion, however, as to suggest that it might
once have formed a continuation of Knife
Valley (see map, back cover). If Kat-
mai Volcano were removed bodily the
Second Valley would drain straight
across through the north arm of the
Valley of Ten Thousand Smokes to the
lower end of Iliuk Arm of Naknek Lake.

The evidence is not sufficient to war-
rant the assertion that Knife Valley was
an antecedent feature, beheaded by the
formation of Katmai Volcano, yet it is
difficult to understand how its broad
basin could have been eroded by its
present drainage.

As we passed up the Valley we were
much perplexed by the conditions we en-
countered. In contrast to the peaceful
beauty which marked the lower lake, the
rough piles of broken rocks that cover the
isthmus between it and the upper lake
had a certain rawness that obtruded the
desolation of the place on our senses.

Apart from its appearance of total dev-
astation, the conspicuous features of the
landscape are a great number of very
r^ular conical mounds, which are strewn
over the isthmus and rise from the waters
of the upper lake (see page 128).

The irregular disposition of these hill-
ocks recalled a glacial moraine, but we
were puzzled by the complete absence of
vegetation, for the moraines of an old
glacier should long since have been
occupied by bushes. Moreover, they
lacked the characteristic disposition of
morainic deposits. Some of them were

covered with the ash, lying in regularly
bedded layers; others were entirely free
from it and many of these were so situ-
ated that it was difficult to see how the
ash could have been removed by erosion.

The absence of wood gave me much
concern the first time we came into the
place. We were tired and anxious to
camp, but I could find no suitable site.
Finally I turned to the others for counsel.

'*Why, here is plenty of wood right
here," said Folsom; **here is a good
place.** I looked incredulously about me.
Not the smallest bush was to be seen.

'*Well, if you will provide the wood,
we'll camp," I replied.

So, dropping his pack, he proceeded to
one side a few yards and pulled a stick
out of the ground. My eyes had not
been sharp enough. When we began
searching we pulled out a considerable
quantity of sticks from the rock debris.
We even managed to find a couple that
would answer for tent poles. So we set
up our camp on the bank of the river,
between the two lakes.

These upper reaches of Katmai River
are curious, indeed. Although it is a tu-
multuous mountain torrent, the bowlders
of its bed are sharp and angular, entirely
different from the rounded cobbles usu-
ally found in such streams. The river
occupies no proper bed of its own, but
simply finds its way at random over the
irregular surface, now cascading over a
group of bowlders, now slackening to fill
a depression with a small pond.

The rock-covered ground is so loose
and porous that much of the water
strikes in and travels along beneath the
surface. There are several hollows filled
with clear, deep, green water, forming
little ponds. These have neither inlet
nor outlet, but are kept sweet by the
rapid underground seepage. At the foot
of the belt of hills the underground water
reappears in several large streams which
come from the ground as springs.

NOISY MOUNTAIN

Immediately behind our camp rose a
tall mountain whose base was screened

Digitized by

Google

128

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by D. B. Church

A PORTION OF THE ROCK SLIDE FROM THE SLOPES OF NOISY MOUNTAIN

The cones of detritus in the middle distance are everywhere the characteristic feature of the surface
of the slides. In the foreground is Katmai River, which spreads out irregularly over the angular
bowlders of the slide.

by an enormous mass of talus. Ac-
customed to the avalanches, which were
abundant everywhere in the volcanic
country, we paid no attention to the
frequent rock falls from the precipices of
this peak until after we had settled our-
selves for the night.

But all night long our sleep was dis-
turbed by fresh avalanches. Never
during our sojourn was there a lull of five
minutes. '* Noisy Mountain," for so we
named it, was evidently a counterpart of
'* Falling Mountain,*' which we discov-
ered in the Valley of Ten Thousand
Smokes (see pages 241-243). Its con-
stant avalanches gave us a new clue to
the anomalous character of the country.

Next morning, when we could climb
up and get a general view of our sur-
roundings, the explanation of the curious
physiography of the valley became
evident enough. We could see where
the whole side of the mountain had
broken loose and slid out into the valley,

burying several square miles under a
mantle of debris.

As in the other big slides, the bright,
fresh rock cliffs, newly exposed, con-
trasted strikingly with the old, weathered
slopes of the mountain. We could form
no intelligent opinion as to the quantity
of material that had broken loose, but
some of the piles of debris appeared to be
300 feet high.

CONICAL MOUNDS CONSPICUOUS

Like the other great slides of the district,
the rockfall from Noisy Mountain shows
little of the hummocky surface charac-
teristic of the ordinary slow-moving land-
slide; instead are the above mentioned
conical mounds (see also Chapter xi).

Another feature not to be found in the
ordinary landslide is the frequent pres-
ence of shattered rocks. These evi-
dently started from the top of the cliff as
solid masses and held together during the
avalanche. But they were subjected to

Digitized by

Google

THE CANYON AND SECOND VALLEY OF THE KATMAI

such strains that, although they still
retain their original form and appearance,
they fall to pieces at a touch.

NOISY MOUNTAIN IN ACTION

Directly in front of the mountain, and
circling its base, was a mass of detritus
which, having been discharged with less
violence than the rest, was not shoved
out into the valley like the bulk of the
material. The rim of the ridge thus
formed stood 600 to 800 feet above the
valley floor, completely screening it
from further slides by forming a retaining
wall behind which all the new debris
from the mountain was caught.

The rounding top of the ridge like the
rest of the slide area was covered with
cones, but around its base the surface
was thrown into something resembling
the hummocky ridges of the ordinary
landslide, indicating, perhaps, a slow
settling into position rather than a sud-
den drop into the Valley.

This crescentic barrier largely cut off
the view from below, hiding the face of
the mountain whence came the incessant
noises that had disturbed us. But its
crest was a fine vantage point from
which to watch the activity of the
mountain.

The noises were of two sorts: a low
ominous rumble, continued for a con-
siderable period, and the familiar boom,
rattle, and bang of falling rocks loosened
from their hold.

Some of the falls consisted of large
masses of rock, weighing several tons,
which would let go and come down all at
once, breaking into fragments and rais-
ing a cloud of dust, with a terrific roar
which lasted several minutes. More
commonly they were single bowlders
which came rolling and bounding down
the slopes. This noise continued night
and day.

But the other type of disturbance, the
long-continued rumble, was confined to
the hours when the snow was melting
rapidly. It occurred at short intervals on
bright sunny days and less frequently in
cloudy weather. The first of these
10

129

rumbling noises that came when we were
in a position to see the face of the
mountain furnished the explanation.

The continuous small falls from the
cliff periodically choked the precipitous
channels which carried the drainage of
the steep slope, filling them with detritus.
As water from the melting snow trickled
down the channel, the debris clogging it
was gradually changed to mud. Eventu-
ally the point of flowage was suddenly
reached and the whole mass would then
start moving downstream, gathering
momentum and picking up additional
material as it went roaring down the
steep slope.

This moving mass of rock, mud, and
water had the consistency of a batch of
coarse concrete, as it comes fresh from
the mixer. The rock fragments which
it carried varied all the way from mud
and sand to bowlders three feet in di-
ameter, which were rolled, spattering,
end over end, down the channel with
the rest.

Starting high up on the mountain,
these '* water slides'* flowed for the first
few hundred feet down a precipitous
slope approaching the perpendicular,
rapidly gathering momentum which
carried them across the gentler slopes
below, until they brought up in the de-
pression behind the barrier-slide.

Considering the consistency of the
flows, it was rather surprising how far
they ran out into the comparatively level
basin. For several hundred feet at the
terminus the gradient of their courses
was hardly more than ten per cent.
The total drop was about 1,500 feet.

The quantity of material brought
down in a single flow often amounted to
more than a hundred cubic yards. The
slides occurred several times a day all
summer and kept up at least for several
years, accomplishing in the aggregate a
considerable amount of erosion. Adding
to these the rock falls, which certainly
move as much material as the bowlder
flows, one gains a new conception of the
forces here at work. Yet the whole of the
material brought down since the great

Digitized by

Google

Fhotograpb by U. B. Churcb
ACROSS KATMAI CANYON FROM THE LOWER SLOPES OF MOUNT KATMAI

The scale may be judged by the man, who may barely be made out on the trail near the center of the
picture. The waterfall, Ruth's Falls, is 750 feet high.

130

Digitized by

Google

THE CANYON AND SECOND VALLEY OF THE KATMAI

131

landslide is insignificant in comparison
with the enormous mass of broken rock
that fills the valley.

AN INTERMEDIATE BETWEEN LANDSLIDES
AND RIVERS

The bowlder flows or "water slides"
stand as curious intermediates, con-
necting two agents of erosion not
ordinarily thought of as intergrading,
namely, landslides and rivers.

They have many characteristics of
both, which make it possible to consider
them either as slips or as overloaded
streams. Being largely made up of rocks,
which grate and grind and jostle each
other on the way down, they are like
slides. Moreover, at their termini they
often assume the crumpled, hummocky
surface of landslides.

They are closer yet in their relations
to streams loaded with more detritus
than they can carry. They follow a
single definite course in their upper
channels and spread out into divergent
distributaries like a delta, when they ap-
proach their termini. At the point where
they debouch from the single channel,
they build up regular convex *' alluvial
fans," highest in the middle, like those of
lateral streams in a main valley.

Like all overloaded streams from the
great Mississippi to the diminutive
brooklet filled with pumice from the
eruption, they build their banks higher
than the level over which they flow,
forming dikes on each side of the chan-
nel.

The automatic flushing to which the
channels of Noisy Mountain are peri-
odically subjected furnished the key to
a number of unusual features in the ab-
normal physiography of the devastated
district.

Comparison of the moving flows with
the forms assumed by the detritus after
motion had ceased permitted the in-
terpretation of the larger and more
important terrane that forms the subject
of the next chapter, which otherwise
would have been hard to understand.

The ability of the thick, heavy mud

of the flows to transport large bowlders
made it clear how the Mageik Flood had
rolled great rocks through the forest
without breaking down the trees. That
also must have consisted of mud so thick
as to move rather slowly, and so heavy
as to have a power of flotation much
greater than ordinary water.

Moreover, it was evident that if these
bowlder flows had had a higher propor-
tion of water their velocity would have
been greatly increased and the stones
they carried would have become formi-
dable missiles capable of doing great
damage, like the rocks carried in the
swift Katmai Flood.

On the other hand, seeing in action
a mass so largely composed of solid
material aided greatly in a comprehen-
sion of the amazing fluid character evi-
dently possessed by the great landslides,
such as that described in the following
chapter.

THE UPPER PORTION OF THE SECOND
VALLEY

One of the primary objects of the ex-
ploration of the Second Valley was to
cross a low pass which we had descried
at its head from the slopes of Katmai.
Even in 191 7 we were anxious to find
some better entrance to the country than
Katmai Bay.

As the Second Valley trended toward
Kinak Bay, which we knew to be a good
harbor, we hoped that the pass at its
head might provide an easy avenue for
passage thither.

The morning after our arrival, there-
fore, we set out skirting the west shore of
the upper lake. Beyond the corner of
Noisy Mountain we came upon a large
glacier, a valley tongue coming down
from the n€v€ that lies between Katmai
and Snowy Mountain (see map, back
cover).

This n^v6 is much more extensive than
any other snowfield of the region, and
the glacier — Princess Glacier, Church
called it — though not yet surveyed, is
clearly the largest thus far discovered in
the district. It is not improbable, how-

Digitized by

Google

132

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by £. C. Kolb

THE LAST LEAP IN THE DESCENT INTO
KATMAI CANYON

The descent is not so difficult or hazardous as
might be supposed from this picture. Before taking
off, I took good care to see that there was a way
back.

ever, that other glaciers of equal or
greater size will be found in the unex-
plored sections of the area. Princess
Glacier is the chief source of Katmai
River, for it furnishes much more water

than the brook which comes down
from the end of the valley.

When we came near enough to ex-
amine the head of the valley in detail
we found two passes rather than one,
both comparatively easy. The ques-
tion was which of the two would lead
us to Kinak Bay, which we knew lay a
dozen miles to the east. Consulting
my compass I decided to try the more
southerly. This turned out to be the
wrong choice.

Our mistake was of no consequence,
however , for theentrance to thecountry
discovered in 19 19 is far superior to
either of these passes, which are above
2,000 feet high and always deeply
covered with snow.

The valley that we descended on
the opposite side was wild and rugged
in the extreme — a narrow V, con-
stricted and difficult of passage. Some-
times the cliffs forced us to wade
in the stream and sometimes cascades
compelled us to climb a thousand feet
up the mountain side.

Before we had gone far we realized
that we had taken the wrong pass, for
the valley, trending at first south,
turned more and more westerly in-
stead of easterly, as we had hoped.

Had it Of)ened to the east there
would have been time to reach the
sea and return again the same day.
But it was evident that we were not
making much progress seaward by
traveling southwesterly, parallel with
the coast. Yet hoping for a bend in
the valley, we kept on till five or six
o'clock. But there was no indication
of a turn, so at last we gave up and
turned back.

A HOT SPRING DISCOVERED

On our way down we had passed a
fine hot spring close to a sightly water-
fall, which appeared at a place where
a fault seemed to cross the valley. We
decided to return and sf)end the night
there. Though we had neither supper
nor blankets, we could at least enjoy the
luxury of a hot bath.

Digitized by

Google

THE CANYON AND SECOND VALLEY OF THE KATMAI

133

Photograph by D. B. Church
ROOTS PUT OUT BY A AMLLOW AT THE SURFACE OF THE ASH

Later washed out by a stream. The new roots are 8 feet above the original surface of the ground.
All the new roots were put out immediately below the surface.

This was much desired by the natives,
who have great faith in the curative
properties of hot springs for the rheu-
matic complaints with which they are
affected. The temperature was just
right for the best effect. We could bear
our hands in it, though it was too hot for
our feet.

The drizzle which had been falling all
day continued most of the night. All the
wood was pretty thoroughly wet, but by
whittling off the outside we procured
some dry shavings and soon had a roar-
ing fire, partly compensating for the lack
of blankets, and enabling us to get a few
hours of sleep. Having no breakfast to
delay our departure, we were off at day-
light and reached camp about noon.

When we laid out our compass trav-
erse of the new valley, the most probable

outlet appeared to be Soluka Creek. We
were loth to admit that with all our
trouble we had merely walked around
the Barrier Range, and tried to believe
that our stream flowed at least into
Takkala Bay. The survey made in 1919
showed, however, that we were wrong.
Had we continued half an hour longer,
we should have been able to look out into
Katmai Bay.

If we had only known it, we might
have attained our quest that day. Ha-
gelbarger*s Pass, by which the route from
Geographic Harbor crosses into Soluka
Valley, lies just below our hot spring.
When the surveyors came across from the
coast in 19 19 the first thing they saw
from the pass was our waterfall and hot
spring, which they recognized from the
description.

Digitized by

Google

3 S

•s-1

H
O
U

a> o

5?;

§5
o d

s e
« §

C 4-.

•11

.s-s

134

C
3

Digitized by

Google

THE MAGEIK LANDSLIDE

When our explorations carried us into
the Valley of Martin Creek, early in 1917,
we came upon a remarkable terrane, as
noteworthy, in its own class, as Katmai
among eruptions.

For several miles the valley is strewn
with enormous bowlders thrown promis-
cuously about. M. A. Horner, who saw
them in 1 91 3, was rather more impressed
by these bowlders than by the volcano
itself. Realizing that they were con-
temporaneous with the eruption, he con-
cluded that they must have been pro-
jected from the crater. Accordingly he
labeled his pictures of them ** Rocks
thrown out by the Volcano." Postcards
bearing this legend are probably still for
sale in Seward.

BOWLDERS AS BIG AS A HOUSE STREWN
ACROSS THE VALLEY

Bowlders 10 feet long are abundant,
and much larger rocks are frequently
encountered, some of them as large as a
good-sized house. Many measure 30 or
40 feet. The largest single stone ob-
served was 75 feet long and 35 feet wide.
Its height could hardly have been less
than 20 feet, though this could not be
measured, for it lay half buried in the
surrounding debris.

There could be no question but that
these bowlders had in some way been
carried far from their original position.
The valley is cut, through the greater
part of its course, in sedimentary rock,
but the waste that covers its floor is
mostly composed of lava fragments.

Yet it was evident that these rocks
could not have been discharged from the
crater of any of the volcanoes. Mixed
with the piles of lava bowlders was much
material of an entirely different nature.
Considerable masses of old soil, great
chunks of peat, remnants of vegetation,
fragments of sandstone, and bits of tuff

from an ancient eruption were all jum-
bled together in great confusion.

In addition to the heterogeneous
character of the materials, their position
and attitude furnished positive proof
that they had not been thrown out from
one of the volcanoes. If the bowlders
had been hurled through the air, the
marks of their impact upon the ground
as they struck would have been conspic-
uous, but none were to be found. If
they had been thrown out of one of the
craters, they would have been strewn
indiscriminately on all sides of the vent,
but they are confined to a single sharply
limited sector.

The terrane is restricted to the bottom
of the valley. Beginning high up at the
head of the stream, the debris is spread
for several miles down the floor of the
broad U-shaped valley, occupying the
same position as a glacier. From a dis-
tance, indeed, it looks like a glacier com-
pletely covered with rock fragments.

On nearer approach, however, it is
seen to contain no ice. Yet it resembles
glacier work in so many ways that we
supposed at first it must be some unusual
form of glacial deposit.

In places its surface is pitted with
characteristic circular puddles, exact
miniatures of the '* kettle-hole ponds" so
familiar to students of glacial geology
(see page 143). Since these are most
readily accounted for by the melting
away of an embedded block of ice or
snow after the mass reached its present
position, their occurrence gave color for
a time to the glacial theory.

But the terrane has several important
features which show at once that it was
not the work of any glacier. It has no
moraines, no ridges of any kind at its end
or edges, nor is there any sign of the push
and shove always characteristic of ice
work.

135

Digitized by

Google

8
J?

'Ai

O
<

o I

s I
^ !

136

Digitized by

Google

THE MAGEIK LANDSLIDE

137

Yet the terminal edge is perfectly
clear-cut and distinct; its barrenness con-
trasts sharply with the green grass be-
yond. In some places the margin is a
steep convex bank, several feet high
(see page 141). In others, the dfebris is
spread only a few inches deep over the
ground (see page 144).

BROKEN BUSHES INTERMINGLED
WITH THE DEBRIS

The effect on the bushes that lay in its
path likewise affords clear and positive
evidence that this is no glacial terrane.

An advancing glacier slowly over-
whelms the trees that stand in its way,
gradually pushing them over by piling
debris against their trunks. But here
the willows and alders were snapped off
clean, as though struck by flying rocks
(see page 141). The bushes were never
bent over before the advancing mass
nor were they uprooted by its force; the
line of destruction is as sharp as the edge
of the debris itself.

There were many places where one
could observe the clean-broken stumps
protruding through their covering, with-
in a few feet of similar bushes standing
untouched beyond the margin of the
drift.

The terminal portion of the mass is com-
posed more largely of fine materials with
fewer rocks. Here also are great chunks
of black peat soil, sometimes scattered
irregularly through it, sometimes thrown
into rude windrows.

At the lateral edges, several miles
back from the terminus, the debris in-
stead of piling up sank away from the
highest level reached, leaving a hollow
rather than a ridge.

This is particularly well shown along
a spur of one of the neighboring
mountains, so located as to have stuck
out squarely into the path of the flow.
Here the rocks had evidently ground up
against the obstruction, clearing the
vegetation and soil from the hillside to a
height of nearly 100 feet above the pres-
ent level of the detritus.

But instead of lodging against the

barrier, as glacial bowlders would have
done, they slumped away again, leaving
the hillside almost bare, and forming a
deep hollow around its base in which has
gathered a considerable body of clear
green water in a horseshoe pond (see
page 142).

Another very curious feature of this
bowlder train is the fact that half-way
down it forked, sending separate tongues
down different portions of the flat val-
ley. Its margin likewise showed a de-
cided tendency to splay out into minor
branches.

A forking glacier is, to be sure, no im-
possibility, but a mass of ice can be di-
vided only when it meets an insuperable
obstacle, capable of withstanding the
enormous power of its advance. The di-
vide on which this terrane split is merely
the inconsequential rise between the two
streams which come together farther
down the valley.

It is evident, therefore, that whatever
the explanation of this mass of dfebris
may be, its movement was delicately
adjusted to slight variations in the gra-
dient of its bed. Notwithstanding its
high percentage of rocks it gives every
indication of having sought its level
almost like water. This amazing evi-
dence of fluidity is strengthened rather
than weakened by detailed study.

ORIGINAL PLANTS STILL GROWING ON THE
BOWLDERS

This bowlder flow, if we may call it
such, shows unmistakable evidence of
having moved much more rapidly than a
glacier. It bears numerous fragments of
the original vegetation which were not
destroyed by the catastrophe. Some of
these, indeed, still remain alive and have
begun to grow again, as though never
disturbed.

No such thing would have been pos-
sible in a glacier. It would require
many years for a glacier to carry a
bowlder like that shown on page 1 38, from
the mountain top down into the valley,
and all vegetation would almost in-
evitably disapp)ear in the process. These

Digitized by

Google

138

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs
ONE OF THE LARGER ROCKS OF THE SLIDE, HALF BURIED IN SMALLER DEBRIS

The largest bowlder observed, which was 75 by 35 by 20 feet, was not favorably situated for a

photograph.

plants were, however, evidently carried
** right side up with care'* all the way
down from the mountain top to their
present resting place.

INDICATIONS AS TO DATE OF OCCURRENCE

Along with the indications of sudden-
ness of formation is other evidence
that fixes rather closely the date at
which this curious mass of d6bris reached

its present position. By all the criteria
of physiography it is very young — indeed,
infantile. The two streams which trav-
erse its surface have scarcely begun to
erode their beds. They spread out in
most irregular fashion over the uneven
surface of the debris, without having so
much as scooped out a channel for them-
selves (see page 136).

The bowlders in their beds are still

Digitized by

Google

THE MAGEIK LANDSLIDE

139

angular, like the rest of the material, in
sharp- contrast to the rounded cobbles of
typical streams. The surface is so fresh
that, despite the large content of soil in
places, new vegetation had just begun
to take hold in 1919.

Many pieces of wood, fragments of the
overwhelmed bushes, lie embedded in
the debris in such a way as to leave no
doubt that they were contemporaneous
with it. They will, of course, rot away
in a few years, but as yet they show little
sign of decay. These circumstances com-
bine to make it certain that the bowlders
were thrown into position very recently.

On the other hand, it is certain that
the bowlder flow goes back to the time of
the eruption, for its surface is covered in
many places with the layers of ash from
Katmai. Because of local irregularities
in deposition, I have not been able to
satisfy myself whether all layers of ash
are present, or only the upper strata.
But in other spots near by the ash is
either absent or thoroughly mixed with
the detritus. This is taken to indicate
that while the mass must, in the main,
have reached its position before the end
of the eruption, there were minor move-
ments after the ashfall. All the evi-
dence taken together seems to show defi-
nitely that the bowlder flow originated
at the time of the eruption.

EVIDENCE THAT IT WAS SOME SORT OF
LANDSLIDE

The manifest suddenness of its forma-
tion, its occurrence at the time of the
eruption, and other features suggest that
this bowlder flow was some sort of land-
slide. This idea is strengthened by the
presence of tufts of antecedent vegeta-
tion here and there.

Fragments of the original ground
were similarly transported in the great
landslide at Frank, Alberta, as described
by McConnell and Brock,^who report,
**One large bowlder retains on its upper
surface a coating of moss and a fragment
of rotten log lying quite undisturbed.*'

Yet it is altogether different from the
familiar type of landslide. The ordinary

landslip carries the evidence of its mode
of origin so plainly marked in its every
feature that no one could mistake it.
While still fresh the tree trunks, pointing
every which way out of the broken
ground, advertise what has hapf)ened as
plainly as though it were posted on a
billboard. Even after the vegetation
has adjusted itself to the new conditions
and the broken ground has settled to-
gether, the hummocky, billowed surface
and the crescentic ridges identify it in-
dubitably to the student.

None of the recognized features of
landslides occur here. In place of the
irregular broken hummocks and billows,
this terrane has a peculiar topography
quite different from anything I have ever
seen or read of in regions of ordinary
physiography.

SURFACE STUDDED WITH REMARKABLE
CONES

Like the Noisy Mountain Slide de-
scribed on page 127, the Mageik Slide is
covered with remarkable conical piles of
detritus which present many unusual
features that make detailed discussion
advisable (see pages 128 and 134).

The general textbooks on physiog-
raphy make no mention of the occurrence
of such cones in landslide debris. Never-
theless they are mentioned incidentally
and even figured in the accounts of a
number of notable landslides in other
parts of the world. Observation of the
slides of the Katmai district leads me to
believe that these cones are character-
istic features occurring regularly in land-
slides of the most violent tyf)e.

While there is nothing to indicate their
mode of origin, I am inclined to suspect
that each represents a separate mass of
material which held together and trav-
eled intact throughout the movement of
the slide, and settled down in a single
spot when motion ceased.

While most of them are made up of
lava fragments, occasionally one is found

1 McConnell, R. G. and Brock. R. W. Report
on the Great Landslide at Frank, Alberta, 1903.
Ann. Rept. Dept. Interior, Canada, 1903, p. 10

Digitized by

Google

I40

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

ORIGINAL VEGETATION PERSISTING ON A ROCK IN MAGEIK SLIDE

By some freak this grass with the soil in which it grew was transported, "right side up with care,"
for two miles down the valley from its original position at the top of the cliff.

composed entirely of sandstone and
shale, as though derived from a single
block of sedimentary rock. Very likely
these piles were at first irregular in shape,
and assumed their conical form because
the original block was too thoroughly
shattered to hold together, its fragments
rattling down symmetrically on all sides.

GEOLOGICAL STRUCTURE FAVORABLE FOR
LANDSLIDES

An examination of the geological
structure of the area where the slide
occurred reveals conditions favorable for
the development of landslides. Massive
lava flows several hundred feet thick
have been poured out on a platform of
sandstone dipping away from the vents.
The strata are bent up against the vol-
canoes, as though blistered up by an in-
jection beneath.

The lava flowing down the inclined
surface of the strata often congealed be-
fore reaching the bottom of the slope.
As it hardened, joint planes perpendicu-
lar to the sloping surface developed, more
or less changing the solid flow into a
series of slanting columns standing in

unstable equilibrium like a **row of
soldiers*' ready to be knocked over.

Closely adjacent to the part of the cliflf
that fell away, colossal columns several
hundred feet high are still hanging, ap-
parently so ready to topple over that one
wonders how they survived the shock
that set oflf their neighbors (see page

134).

Through the joint cracks, water doubt-
less reached the original surface and,
working along, may have facilitated
movement by lubricating a bed for the
unstable mass above.

Of the immediate cause which set off
the fall we have no knowledge. Un-
doubtedly there were earthquakes in con-
nection with the eruption, but they seem
not to have been particularly violent.
It is not impossible that there may have
been a further uptilting of the strata
by the injection of new magma below.
There is little if any independent evi-
dence of recent uplift. Yet an uplift
of this sort would be the simplest ex-
planation of the continual avalanches
of Falling apd Noisy Mountains (see
page 241).

Digitized by

Google

THE MAGEIK LANDSLIDE

141

Photograph by R. F. Griggs

TERMINAL EDGE OF MAGEIK SLIDE

Although obliterating the alder thickets wherever it touched them, it showed no tendency to form
ridges resembling moraines. At this edge the slide covered the ground only four or five feet deep.
The black chunks are masses of peat.

It is interesting further to note that
while the cliff at the head of the Mageik
Slide was quiet during our observations
in 1917, it was subject to frequent heavy
rock falls in August, 1919.

DIMENSIONS OF THE SLIDE

The top of the cliff from which the
rock started stands at an elevation of
about 3,000 feet. The area from which
the fall occurred is about 2,000 feet wide
and 750 feet high. It is a great gouged-
out cirque, whose bright unweathered
rock contrasts strongly with the duller
surface of the old cliffs. We have no
means of estimating how far in front of
the present cliff the original face may
have stood.

The head of the slide is situated on a
branch to one side of the main valley.
From this point the rocks swept down
half a mile of steep slope, falling about
1,000 feet; then across the flat valley
floor for nearly two miles to the opposite
mountain side, 400 to 500 feet lower.
Then, deflected partly by the obstruct-
ing spur and partly by the mountain wall

beyond, they turned 60® from their
course and continued down the mile-
wide valley three miles farther, descend-
ing 600 feet more (see map, back cover).

The area covered by the debris may be
estimated as two and one-quarter square
miles. Of the quantity of material that
fell away we are not in a position to make
even a very intelligent guess, since we
have no knowledge either of the original
contours of the cliff or of the depth of the
debris covering the ground.

From the reports on other slides of the
same class it appears conservative, how-
ever, to estimate the average thickness
of the debris at ten yards. On this
assumption the cubage of the whole
mass would be in the neighborhood of
70,000,000 cubic yards.

The upper portion of the slide, just
below the cliff, has somewhat the charac-
ter of an ordinary talus slope; but its
bowlders are of very great size, many of
them 30 and 40 feet in diameter, and its
general contours are concave, not convex
as in an ordinary talus fan. Along with
the large size of its bowlders goes a cer-

Digitized by

Google

142

THE VALLEY OF TEN THOUSAND SMOKES

A HILL SCOURED OFF BY THE BOWLDER FLOW

The flying rocks scoured off the soil nearly a hundred feet up this hillside, yet before movement
ceased the mass slumped away again, leaving a deep depression now filled with water, forming the
Horseshoe Pond.

tain roughness and irregularity of surface
not found in the typical talus.

The largest and most perfectly de-
veloped cones occur in the second por-
tion, which occupies the flat valley above
the bend. Here, also, the percentage of
large rocks is much higher than in the
lower portion . Cones and large bowlders
gradually become smaller and less numer-
ous toward the tip, although some cones
and a few large bowlders occur near the
end.

SURPRISING LIQUIDITY MANIFESTED BY
THE SLIDE

Of all the curious features of this re-
markable bowlder train none is more sur-
prising to the observer than the clear evi-
dence it presents of having behaved like
a liquid during the period of its motion.

In default of a better term, I have
called it a landslide, but it cannot be con-
sidered as having slid down the valley in
any accepted meaning of that term. It
must have resembled rather the bowlder
flows observed on Noisy Mountain
(see page 129). It acted more like a

flow of thin mud than a mass of dry
bowlders.

The grade over which the debris
moved in the lower part of its course
slopes less than 200 feet to the mile (4
per cent). Even if a bowlder were
highly polished and its bed smooth, it
would not slide on a grade of i to 25.
The attitude of many of the rocks shows
plainly that they did not slide into their
present place but were rather floated into
position, buoyed up by the mass of finer
debris around them.

It is incomprehensible that a mass of
rough angular rocks could by any means
be made to flow like a liquid. Neverthe-
less, the evidence convinces everyone
who studies the terrane that this is ex-
actly what happened.

DESCRIPTION OF A SIMILAR SLIDE IN
MOTION

When opportunity came to look up
the accounts of violent landslides else-
where, my conclusions as to the remark-
able fluidity of the mass were confirmed
and strengthened. A similar ** slide'*

Digitized by

Google

THE MAGEIK LANDSLIDE

143

Photograph by R. F. Griggs
A DIMINUTIVE ** KETTLE HOLE** POND NEAR THE TIP OF THE BOWLDER FLOW
This may have been caused by the melting away of a chunk of ice or snow.

at Elm, Switzerland, was witnessed by
a number of persons. From their stories
Sir William Conway ^ has written a de-
scription of the event which closely
parallels what must have occurred in the
Mageik slide.

" The last phase of the catastrophe is
the hardest to imagine and was the most
difficult to foresee. The actual facts
are these. Ten million cubic meters of
rock fell down a depth (on an average) of
about 450 meters, shot across the valley
and up the opposite (Diiniberg) slope to
a height of 100 meters, where they were
bent 25° out of their first direction and
poured, almost like a liquid, over a hori-
zontal plain, covering it uniformly
throughout a distance of 1,500 meters
and over an area of about 900,000
square meters to a depth of from 10 to
20 meters. The internal friction of the
mass and the friction between it and the
ground were insignificant forces com-
pared with the tremendous momentum
that was generated by the fall. The
stuff flowed like a liquid.*'

* Conway, Sir William Martin, The Alps from
End to End, pp. 176-184, 1900.

Conway*s narrative goes on to describe
the death and destruction wrought by
the catastrophe in the village of Elm.
Fortunately, we have no need to follow
him here. We may rather be thankful
again for the good luck that the Mageik
Slide occurred in an uninhabited valley,
where no harm was done.

MAGEIK SLIDE COMPARED WITH OTHERS

The Elm Slide, although one of the
best known examples of its class, is so far
inferior in size as to make it interesting
to inquire how the Mageik Slide should
be ranked among similar phenomena.

The largest American slide of which I
find record occurred some years ago in
the Canadian Rockies, when a portion of
Turtle Mountain slid down on the town
of Frank, entailing great destruction.
This was about three times as great as
the Elm Slide, involving an amount of
material estimated at 36,000,000 cubic
yards, spread out over i .03 square miles to
an estimated average depth of 15 yards.*

» McConnell, R. G. and Brock, R. W. Report
on the Great Landslide at Frank, Alberta, 1903.
Annual Report, Dept. Interior, Canada.

Digitized by

Google

144

THE VALLEY OF TEN THOUSAND SMOKES

Photosraph by R. F. Griggs
A THIN EDGE OF THE SLIDE WHERE IT RAN UP THE SIDE OF A HILLOCK

The grass land at the left was not touched. The veneer of debris is so thin that it scarcely alters the

contour of the hill.

A slide at Rossberg, in the Alps, was
still larger. This was 4,300 yards long,
349 yards wide, and totalled 51,000,000
cubic yards.*

Of the landslides in the Katmai dis-
trict three require mention here. The
slide which dammed Katmai Canyon,
causing the Great Flood when it broke,
has been estimated by Maynard from his
survey to contain 88,000,000 cubic yards.
The Noisy Mountain Slide appears far
greater than the one in Katmai Canyon;
but the survey has not yet been carried
into that area, so that it is impossible to
give any estimate of its dimensions.
Yet both of these seem to an average
observer on the ground far inferior to
the Mageik Slide, for which perhaps we
have assumed too small an average
thickness.

The slide from Falling Mountain,
which was the opening event in the
eruption (see Chapter xxi), may have
exceeded all others in the Katmai dis-
trict. Certainly the cirque on the
broken face of the mountain looks as

though fully as much rock had been
gouged out as in any of the others. But
there is no way of estimating the amount
of the fall because all the debris is
covered up by the deposits of the erup-
tion.

As the slides described in the papers
cited above were considered very re-
markable events, it is evident that those
of the Katmai district are to be ranked
among the most notable examples on
record. They are, however, by no means
the greatest landslides known.

Aside from records of slow-moving
landslips of the ordinary type, with
which we have no concern, there are ac-
counts of other catastrophes belonging
to the same violent type, which so far
exceed those we have described as to
dwarf them by comparison.

THE GREATEST RECORDED SLIDES

The Gohna slide in the Himalayas in
1893 was much greater. An enormous
mass of rock fell 4,000 feet from Mount
Maithana into the Bhira Gunga at its

* Zay, Karl, Goldau and Seine Gegend, page 318, Zurich, 1807.

Digitized by

Google

THE MAGEIK LANDSLIDE

145

foot. The violence of the fall was at-
tested by the formation of great clouds
of dust which darkened the neighbor-
hood for miles around, and on settling
covered the ground like snow.

The falls lasted for three days and con-
tinued in rainy weather for many
months. The detritus formed a great
dam, stretching along the river for two
miles, filling the channel to a depth of
900 feet; the length across the gorge was
600 feet in the bottom and 3,000 feet at
the top; the thickness of the dam up and
down stream 11,000 feet at the bottom
and 2,000 feet at the top.

Above the dam a great lake with a
maximum depth of 777 feet accumulated,
and was later released in one of the
greatest floods of record (see Chapter
ix). Although the surface covered by
the dfebris was only 423 acres, it is
evident that the quantity of material
involved far exceeds those mentioned
above, being estimated at 800,000,000
tons, perhaps 5,000,000,000 cubic yards.*

Greater still was a slide at Bandai
San,* Japan, in connection with the
eruption of 1888. In this case a stu-
pendous mass of rock, set in motion by a
volcanic disturbance, came flowing down
from the mountain into the lowland,
wreaking tremendous havoc on the in-
habitants.

This event has gone into the annals of
geology as an eruption, but it was clearly
recognized that " the main feature of this
eruption was the deluge of rock and
earth." Consequently it has quite as
much interest to students of landslides
as to volcanologists. Its similarity to
the Mageik Slide will be apparent from
the following description:

* National Geographic Magazine, vol. 31, p.
277, April, 1910.

• Sekiya, S. and Kikuchi, Y., The Eruption of
Bandai San, Journal, College of Science, Impe-
rial University, Tokyo, vol. 3: pp. 91-172, 1889.

"Among the various phenomena that
constantly bewilder the eyes of visitors
to the scene of the eruption, not the
least striking are the numerous big bowl-
ders, some of them measuring from 5 to
10 meters each way, that are to be seen
resting on the surface of the d6bris far
away from the crater. These have evi-
dently been carried along as part of the
mud current, and not hurled through the
air. Not less curious are the quantities
of small cones, varying from a few meters
up to 15 meters in height, which are
scattered here and there over the sur-
face, standing out of the d6bris like so
many miniature Fujiyamas."^

The material thus precipitated down
the mountain side covered no less than
27 square miles (70 sq. km.). Its bulk
was estimated by the Japanese writers
as 1,587,000,000 cubic yards, more than
twenty times as great as the Mageik
Slide, putting it in a class altogether by
itself.

The velocity of the Bandai San Slide
was tremendous. Our authors estimate
it as 48 miles, or 77 kilometers per hour.
It was accompanied by terrific wind
blasts (minimum velocity 90 miles an
hour), which overthrew houses and tore
up trees by the roots.

With the perspective furnished by
these comparisons, we are enabled to
rank the Mageik Slide among similar
phenomena that have occurred else-
where. Although a pigmy in compari-
son with Bandai San, it is so much
larger than most of the better known
slides and shows so many features of
exceptional interest, that it must be con-
sidered one of the most remarkable in
history.

^ This account is accompanied by drawings
showing the innumerable cones studding the
surface of the flow, in perhaps even more striking
fashion than the photographs of the Mageik
Slide reproduced herewith.

Digitized by

Google

jraoLograpn oy d. d. ruuuu
AN ASH ACCUMULATION ON A TRIBUTARY OF SOLUKA CREEK

The streams covered their beds with many feet of ash after the eruption. Later they began to re-
move the ash, sometimes cutting deep canyons. All of the material from the water up to the highest
bank in front of the bushes is ash. The barometer indicated a thickness of 70 feet.

146

Digitized by

Google

XII

KATMAI VALLEY IN LATER YEARS— PHYSIOGRAPHIC INSTABILITY
—ASHFALL— RETURN OF PLANTS AND ANIMALS

When we landed on Katmai Beach
again in 191 6 we found the general ap-
pearance of the country much the same
as the year before, though the mountains
were a little greener and even on the flat
a few seedlings were beginning to start;
but when we began to examine the old
landmarks we observed great changes in
. detail.

The site of our base camp of the pre-
vious year, for example, was found buried
under twenty inches of fresh pumice
washed off the mountain side, and a
brook had cut its bed across the place
where our tent had stood. The year be-
fore this stream had been 50 yards away.
Indeed, it was by reason of just such
changes thatthecountry retained thesame
general appearance from year to year.

If it once settled down so that the
ground ** stayed put,*' vegetation would
begin to creep in and the whole face of
the country would soon change.^

INSTABILITY THE KEY TO THE ABNORMAL
PHYSIOGRAPHY

Instability is, in fact, the key to the
abnormal physiography of the ash-
covered region. Whether one observes
the mountain tops, the hillsides, or the
valleys of the streams, the one most
striking feature of the landscape is its
continually shifting surface.

On returning in 19 18 to the camp-site
at the base of Observation Mountain
used in 191 7, Say re and Hagelbarger
found that the ground on which the tents
had stood was covered by about two feet
of fresh pumice washed in during the
interval. By 1919 the process had been
reversed. The stream had not only re-
moved the new accumulation but had
cut away all the ash, clear down to the

* More detailed discussions of the problems of
the revegctation of the country can be found in

original soil. On page 161 is shown the,
stump of a tree that was cut at a con-
venient height in 1917. In 1919 the cut
end stood nine feet above the ground.
Some of the treadmill ash slides al-
luded to in Chapter viii were in continual
motion. Near the base camp of 1915-
16 was a cliff from which there was an
almost constant stream of fine ash slid-
ing down a little channel and dropping
off the edge, like the sand in an hour-
glass. But the "hour" of this glass
never ended, for week after week the
same trickle kept steadily sliding down,
varying only with the wind, increasing
as it rose and decreasing as it fell.

FREQUENT AVALANCHES FROM THE OVER-
LOADED MOUNTAINS

The general unsettlement of the
mountains produced by the great loads
of ash dumped upon them was not al-
together relieved when the surface ash
had slid off. Once started, the moun-
tain sides continued to send down ava-
lanches for a long time.

Some of the ash-cones were thus deeply
covered with earth and rock, which fell
after the slides of ash had ceased. On
page 154 is shown a section of a slide
produced under such conditions.

In 1915 rockfalls on a small scale were
of constant occurrence; never an hour
passed, night or day, without at least one
being heard. Often we were awakened
out of a sound sleep by their noise.
While most of the rocks that fell were of
comparatively small size, some were
large bowlders capable of doing much
damage (see page 148).

Entirely unused to such happenings,
we could not at first believe that these

the Scientific Results of the Katmai Expeditions
of the National Geographic Society, Ohio Journal
of Science, vol. 19, pp. 318 et seq.

147

Digitized by

Google

148

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

A ROCK THAT ROLLED OFF THE MOUNTAIN SIDE NEAR
OUR BASE CAMP, I916

frequent falls were due to natural causes,
and for a long time supposed they must
have been started by bears digging roots
on the mountain side. But, after re-
peated failures to find Bruin with our
glctsses, it gradually dawned on us that
they were merely part of the readjust-
ment of the physiography after the erup-
tion. They were, however, a passing
phase, much less frequent in 1916 than
the year before.

BURIED SNOWDRIFTS PERSIST FOR YEARS

Everywhere throughout the region one
is made to feel the importance of wind

action on the land-
scape. Almost every
part of the surface of
the country shows the
effect of wind, being
either eroded away by
the powerful blasts or
built up by deposits of
wind-borne sand.

There are, to be
sure, few such dunes
of ash as have been
formed near Kodiak,
for the ash on the
mainland is so much
coarser and heavier
that it seldom shows
the characteristics of
drifting snow. The
deposits here are
smoother, forming
more regular veneers
over the hills, and so
fitting their contours
as to be much less
noticeable than the
smaller deposits
among the plants at
Kodiak.

Allusion has already
been made to the cav-
erns formed by the
melting of ash-covered
drifts. We were as-
tonished to find that
in some cases an ac-
cumulation of ash two
feet deep had been piled on the drifts
since the last snowfall.

Ash is so full of air spaces that a layer
of it acts like the sawdust of an ice house,
protecting the snow from the sun's rays
and greatly retarding melting.

At higher altitudes, where the ash
covers the snow only in spots, its insulat-
ing property is very strikingly brought
out by the differences in melting between
the bare spots and those covered with
ash. Even where the covering is very
thin, melting is retarded and the pro-
tected spots soon stand up above the
general surface. In this way the snow-

Digitized by

Google

KATMAI VALLEY IN LATER YEARS

149

Photograph by D. B. Church

A SNOWDRIFT COVERED BY TWO FEET OF WIND-BLOWN ASH, NEAR KATMAI VILLAGE,

AT SEA-LEVEL, JULY 1 5

Thus protected from the sun, melting of the snow is so retarded that in many places formerly un-
covered early in the season the snow now fails to melt away and is accumulating year by year.

drifts come to be covered with the most
curious, regular, ash-covered cones of ice
(see page 156).

Where the ash-cover is deeper the
melting of the snow is almost prevented,
making permanent many drifts which
formerly vanished early in the season.
Thus, a secondary effect of the eruption
may be a lowering of the snow line
and an increase in the glaciers of the
mountains.

In this connection it is interesting
to observe that snowdrifts covered by
the original falls of ash at the time of
the eruption have melted away very
slowly.

Five years afterward we found cave-
ins, where such drifts were just be-
ginning to melt from under the regu-
lar strata of ash which covered them.
As previously stated, the ash was not

hot enough as it fell to melt the snow
appreciably.

WIND EROSION OF ASH SLOPES

In more exposed places the wind has
greatly eroded the deposits, smoothing
the rough places and rounding off the
whole surface. In planing down irregu-
larities the wind has often cut through
the various layers of ash, leaving the hill-
sides marked with bands and circles
where deposits of different colors are
alternately laid bare (see page 130).

Where wind-swept situations were
occupied by birches, their dead trunks
bear evidence of the poWer of the gales.
On the northwest side the bark has all
been cut away and in many cases the
wood as well has been deeply abraded by
pieces of ash and pumice flying before
the wind (see page 91).

Digitized by

Google

ISO

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by D. B. Church

A CAMP SITE OF I915

On the first expedition our camp stood on the bank of Fickle Creek, whose channel, six feet deep,
occupied, the foreground of the present picture. During the year the channel was filled up and oblit-
erated, while a new channel was dug a thousand feet away. Compare the picture on the opposite page.

One of the most astonishing instances
of the instability of the country was
noted near the mouth of Soluka Creek
in 1916. Here, as elsewhere, everything
looked perfectly familiar, but when we
tried to find our last year's camp our
memories seemed to have failed. No
one could locate it. How we could have
missed it was a mystery, for it was conven-
iently placed on the bank of a tumultuous
torrent which supplied us with water.

FICKLE CREEK SHIFTS ITS COURSE I,000
FEET IN A YEAR

Curious to check up such an unusual
lapse of memory, we hunted and hunted
through the dead forest in search of the
old camp. Finally we found the tent
pins and the coals of the fire, just as we
had left them, but the creek was no-
where in the vicinity. It had moved a
thousand feet away.

Not only was the stream gone; its very
bed was missing as well. The year be-
fore it had flowed in a steep-sided trench,
six feet below the general level, but now
the ground was smoothed oflf so perfectly
that we could not, after the most careful
search, detect the position of the former
banks.

We were not so much astonished that
the stream had filled up its bed and
moved away, for that meant simply a
fresh supply of pumice from a slide some-
where upstream; but that it could have
done it so perfectly as to leave no trace
was incredible.

When we studied such streams at
work, however, we were enabled to un-
derstand the process. We observed that
as they filled their channels the most
tumultuous torrents gradually became
subdued, more and more of their volume
disappearing in the interstices of their

Digitized by

Google

KATMAI VALLEY IN LATER YEARS

151

Photograph by R. F. Griggs

THE BED OF FICKLE CREEK IN I916

It had shifted a thousand feet in the year. We would not have believed that such a channel could be
filled up so smoothly, without touching the coals of our old camp fire.

loose porous beds. Thus, their power
was gradually dissipated, until at the end
they became mere idle trickles spread
out over broad flat plains of smooth-laid
pumice.

In the case of Fickle Creek, a break in
the banks upstream had stopped this
slow process of aggrading just after the
channel had filled up, turning the water
in a new direction and giving it a steep
gradient which enabled it to cut out
another steep-sided trench.

ANTICS OF PUMICE-CHOKED STREAMS

Such streams, choked with ash and
pumice, were extremely interesting
things to watch. They spread out over
the whole width of their flood plains,
which had been filled with pumice several
feet above the original level, wandering
this way and that through the dead
forest in a most fantastic way, changing
their courses very rapidly, never the
same for half an hour at a time.

As the supply of ash and pumice was
practically unlimited, the water was al-
ways choked with more than it could

carry. The pumice was so full of cavities
that much of it floated, and the rest was
but slightly heavier than the water.

The stream thus carried a much
greater load of this sort of material than
it could of ordinary sand or gravel, and
this gave it a most peculiar appearance.
The whole bottom was rapidly traveling
downstream, its continuous steady mo-
tion reminding one of those moving plat-
forms on endless belts sometimes used to
transport passengers.

A brook near one of our camps had cut
entirely through the accumulated mass
of ash just below a fall, forming a bluff
about 70 feet high (see page 146).

A hundred yards downstream, how-
ever, the slope flattened out somewhat
and the brook had been completely over-
come by the enormous quantity of
pumice in its way. It was ludicrous to
watch the struggles of this stream as it
wrestled with the pumice in its bed.

Dammed up in the failure of a previous
attempt, it would gradually accumulate
energy enough for a new effort. Then,
suddenly breaking its bonds, it would

Digitized by

Google

152

THE VALLEY OF TEN THOUSAND SMOKES

Photosraph by R. F. Griggs

RAISED CHANNELS BUILT UP BY OVERLOADED STREAMS

By watching such streams one can see how the mighty Mississippi builds dikes in its channel, raising it
above the level of the surrounding bottom lands.

rush forward down the slope, pushing a
pile of grating pumice before it as though
to engulf the onlooker, writhing this way
and that like a live thing, picking up
pieces of pumice and floating them along
as it came.

Before it had gone far, however, its
new load would literally choke it, and it
would give up the struggle in a hiss of
rattling pumice stones, only to repeat
the process at another point in a few
minutes.

CARICATURES OF AGGRADING RIVERS
BUILDING FANS, DIKES, AND DELTAS

Watching these overloaded streams
was instructive as well as amusing, for in
their activities they furnished perfect
miniatures — caricatures made diagram-
matically simple — of the processes by
which mighty rivers build their beds.

Here one could see in the course of a
few minutes changes in channel which
rivers require centuries to accomplish.

The lower Mississippi, for example,
has built its banks high above the ad-
jacent bottom lands, so high that when a

break occurs the water spreads out far
and wide, inundating everything except
the immediate bank. The fact is evi-
dent, but I had never quite understood
how a river could build such dikes for
itself. It seemed too much like "lifting
oneself by one*s own boot straps." These
streams furnish the demonstration.

Every one of their channels is built up
high above the general surface, hemmed
in by steep-sided dikes as regular as man
himself could build. In an ordinary
photograph it is possible to show only
the finished channel standing as the
stream left it, but in our motion pictures
the whole process from beginning to end
is beautifully portrayed.

Here one may see why it is that the
deltas at mouths of rivers always assume
a finger-form shape. For the process of
building up the enclosing dikes soon
reaches a limit. The stream bed is built
so high above the general level that the
banks can no longer restrain the waters,
and they break out at the weakest place.

The ensuing flood bursts over the
lower ground, at first spreading over the

Digitized by

Google

KATMAI VALLEY IN LATER YEARS

153

Photograph by D. B. Church
PATCHES OF RESURRECTED HERBAGE IN CLEARED AREAS

Contrast the bare ash-covered ground round about. The roots from which these plants grew were
buried for three years, but when uncovered by the great flood of July, 19 15, quickly put up new
growth.

whole surface, but soon cutting a new
trench where the current is swiftest.
The stream is thus entrenched in a new
channel, which in turn is diked in and
built up like an aqueduct.

Whether the slackening of current
that causes the stream to drop its load is
due to entering a larger body of water or
to a change in grade, the land form pro-
duced is much the same. The repetition
of the process of grading up, over and
over again at different points, results in
the formation of a smooth fan, spread
out radially from the point where the
current is slackened.

All the while the little brook, like the
great river, is moving its load of pumice
downstream. Before long there comes a
time when it approaches the completion
of its task. The supply of pumice at
the headwaters begins to fail. As soon
as this happens the energy of the water
is released in part, and with renewed
power the stream begins to cut away
the earlier accumulation.

In this process, as in building up, the
channels shift rapidly from side to side,
cutting first here, then there, along their
banks. This results in the formation of
terraces, like those often found in large
river valleys.

Sometimes one may count as many as
a dozen of them, differing by no more
than a foot in level, beautifully carved
out by the shifting channels of one
of these diminutive rivulets (see page
248).

OBSERVATIONS ON THE DEPTH AND
CHARACTER OF THE ASH DEPOSITS

As we pushed our exploration into dif-
ferent parts of the ash-covered country,
we kept records of the amount and
character of the ashfall at various lo-
calities. The results of our study, com-
bined with the earlier observations of
Martin, are shown graphically by the
map facing page one.

The most striking feature of the maps
is the great irregularity in the courses of

Digitized by

Google

i 1

-c

■*-»

T3
B

:z;

o
a.

ft)

•4J

•f

4-»

•w

•M

C/3

-g

C/)

•g

c
a

«j

C/2

154

Digitized by

Google

KATMAI VALLEY IN LATER YEARS

155

the contours. It is probable, moreover,
that if we were in possession of full data
from the unexplored areas north and
east of the volcano, the lines would fol-
low even more sinuous courses than those
shown.

The occurrence of such irregularities,
some of them so glaring £ts to appear
almost like errors in the record, at first
seems surprising.

It must be remembered, however, that
the distribution of ashfall is largely con-
trolled by the caprice of the variable
winds. Whea one recalls the well-known
irregularity of air currents around high
mountains, the sinuous contours of the
ashfall find, perhaps, a sufficient explana-
tion.

In the deeper deposits of the mainland
there are naturally many more strata
than in the thinner falls at greater dis-
tance. Nevertheless, the major strata,
a middle brown layer between two of a
gray color, observable at Kodiak, are
dominant everywhere in the sector south-
east of the crater.

Down Naknek Lake and on the Be-
ring Sea side of the range generally the
stratification seems not to be correlated
readily with that of the Pacific slope.

When we came up Katmai Valley
under the volcano, we found a fourth
layer of yellowish ash on top of the three
primary strata. Close to the volcano
this had a thickness comparable with
the others, but it thinned out rapidly and
disappeared within a few miles.

This layer we took to be due to a
subsequent explosion of lesser violence,
which had not thrown its ejecta high
enough in the air to effect a wide dis-
tribution.

The very top of the deposits on the
mainland is formed by a curious layer of
dark, very fine, brown material so re-
tentive of moisture that, under the cli-
matic conditions obtaining, it always ap-
pears as mud rather than dust. On the
slopes of the volcano this top mud layer
is several inches in thickness. At greater
distances it becomes extremely thin, but
it is definitely recognizable as far away as

Katmai Village, where it gives a slight
brown tinge to the top of the deposit.
This layer differs from all the others in
that it shows no tendency to drift about
befoi:e the winds. Its presence and its
thickness are so uniform £ts to suggest
that, even as it fell, it was not subject to
wind dispersal, i, e, that it may have been
precipitated with the rainfall rather than
by settling out of the air as dry dust.

PLANTS LIE DORMANT, UNTIL UNCOVERED
BY FLOOD WASH

The most surprising change in con-
ditions in 1916 was a resurrection of
plants in places washed bare of ash by
the great flood. The evidence that
plants could survive burial for three
years was here much more positive than
that which we found later at Kodiak
(see page 153).

At the ** Flood Fall" and in numerous
other places that had been deeply
covered with ash until cleared by the
flood waters, numerous plants had ap-
peared. Most of them were survivals,
coming up from old roots, which had re-
mained deeply buried ever since the
eruption.

Examination of the roots showed that
there had been no attempt at growth
during the three years intervening be-
tween the eruption and the flood, for the
fresh shoots of the season were evidently
the first to start since the old stems
had been crushed down beneath the
ash.

Resurrection was not restricted to any
particular species, but appeared to in-
clude practically all kinds of plants that
had originally occupied the ground laid
bare.

Among them were several of the most
characteristic species of the region, e.g.,
blue-top grass {CalamagrosUs scabra),
horsetail {Equisetum arvense)^ a sedge
{Carex sp.), salmon-berry (Rubus spec-
tabilis), a willow {Salix arctica)^ stone-
crop {Khodiola rosea), bumet {Sanguis
sorba sitchensis), wormwood {Artemisia
tilesii), twisted-stalk {Streptopus amplexi-
folius) , bitter cress {Cardamine umbellata) ,

Digitized by

Google

•c
o

2 "2

^
M

5S 2

< a

o S

^ .s

en 2
2 8

-J
Si
z ^

156

Digitized by

Google

KATMAI VALLEY IN LATER YEARS

157

Photograph by R. F. Griggs

DETAIL FROM ASH-COVERED SNOW FIELD

When sectioned the cones are found to consist of ice which, blanketed by the ash, has melted more

slowly than the bare snow round about.

chickweed {Cerastium sp.), a grass {Poa
sp.), and a rush {Juncus sp.).

Together with these root-survivals
were numerous seedlings, sprung from
seeds that had lain dormant beneath the
ash. Sedges {Carex spp.) and elder-
berry {Sambucus pubens) were abundant
in all such stations; Geranium (G.
erianthum), and alum root {Heuchera gla-
bra) were found less commonly, as was
Polemonium (P. coeruleum), which may,
however, have been a root survival
rather than a seedling.

If these plants had not been found in
connection with the high-water marks of
the great flood, the date of which was
known, we should not have believed that
they could endure such a long period of
enforced dormancy.

But these survivals, together with
what we found around Kodiak, led us to
suspect that the resurrection of old plants
usually plays an important r61e in re-
covery of vegetation after an eruption.

THE BEGINNING OF REVEGETATION

The re^stablishment of a plant cover
over an area deeply covered by ash is.

however, quite distinct from the recovery
of scattered survivors, for the isolated
spots where the ash is removed comprise
only an insignificant fraction of the
whole surface.

Real revegetation, the colonization of
deep deposits by new plants, is obviously
much slower and much more difficult
than the rejuvenation of old roots.

New plants not only must start from
tiny seeds, as contrasted with the ex-
tensive rootstocks of survivors, but they
must draw their nutrition from the sterile
ash rather than from the fertile soil be-
neath.

At the time of our first visit it could
not be said that revegetation proper had
begun, but in 1916 and in subsequent
years definite though very slight begin-
nings in the formation of a new plant
cover had started.

The handicaps that must be overcome
by new plants in getting a foothold on
the shifting ash are very great. When
one considers the physiographic insta-
bility of the country, he marvels that
seeds can find any situations stable
enough to permit their starting.

Digitized by

Google

I.S8

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

FRUITING LUPINES GROWING IN DEEP ASH DEPOSITS

These plants are the most successful pioneers in the process of revegetation, because, on account of
their root tubercles, they obtain nitrogen from the air, while most other plants are dependent on

Xnic compounds. The vi^or of the plants and the abundance of the fruit in the Katmai region
d ample evidence that soil conditions are not unfavorable in areas where plants are not killed by
sand blast.

One would suppose that the extreme
poverty of the ash in all those salts neces-
sary for plant nutrition, especially in
nitrogenous compounds (as shown by
analysis, page 46), would be the crit-
ical factor in retarding revegetation.
Observation, however, shows that in
reality it is the mechanical rather
than the chemical factors which limit
the process.

The instability of the soil discussed
above in its physiographic bearings is, in
fact, one of the greatest obstacles to re-
vegetation.

The effect of sand blast on the coloni-
zation of the soil is not easily observed
directly, because at best the seedlings are
few and far between. It was clear, how-
ever, that revegetation was restricted
to areas relatively free from both sand
blast and accumulating drifts.

Seedlings nevertheless often revealed

an astonishing degree of tolerance to
drifting ash. Lupines, for example, had
become thoroughly established in an area
where the sand blast was severe enough
to decorticate woody stems.

SEEDS, CARRIED BEFORE THE WIND, FAIL
TO LODGE IN PUMICE FI-ATS

In addition to affecting revegetation
by keeping the surface soil always drift-
ing about, the wind controls to a large
extent seed dispersal.

Regardless of the various adaptations
of the different species for dissemination,
the distribution of seed in this district is
effected largely by the high winds which
prevail. They are so powerful that they
can carry even heavy seed for long dis-
tances, thus assuring that adequate sup-
ply of seed throughout the devastated
area without which revegetation would
be impossible.

Digitized by

Google

KATMAI VALLEY IN LATER YEARS

159

SOCKEYE SALMON SPAWNING IN A SMALL BROOK
This was the only stream in the Valley in which they were to be found (see text page 161).

I found by experiment that a breeze of
only 15 miles an hour was able to pick up
and carry off lupine beans, one of the
heaviest of native seeds. It is evident
that the frequent gales must carry seeds
beyond into every corner of the ash-
covered district.

Under the circumstances the rel-
ative distances of different areas from
sources of seed supply are of little conse-
quence.

Yet the wind interferes greatly with
the seeding down of any particular
situation. The picture of the bear
trail, reproduced on page 162, gives the
clearest testimony as to the condi-
tions. The seeds, carried before the
wind, found no lodgment on the smooth
flat, but were caught in the depres-
sions, perpetuating the transitory foot-
prints.

Likewise, we found the old dry heaps
of pumice deposited by the shifting
streams the abode of numerous seedlings,

in contrast to the bare ash round about.
Evidently the seeds had been carried
down by the water and planted in the
pumice out of reach of the wind.

LUPINES THE MOST IMPORTANT PIONEER
PLANTS

In wet places on the flats, seedlings of
many sorts were springing up sparingly,
but in well-drained situations the lupine
was much the most effective pioneer.
It was not until 1919, however, that
new plants were found coming up in
deep deposits on the hillsides, and then
they had gained a foothold only on
slopes sheltered from the most violent
winds.

Yet in some localities they were so
abundant and so thrifty as to seem to
give a clear indication of the course that
revegetation was to take.

There is every reason to believe that
with the years they will gradually spread
into more and more exposed places, for it

Digitized by

Google

i6o

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by E. C. Kolb
MOUNT MAGEIK FROM THE ASH FLATS
The elderberries and bunches of grass have come up through a heavy blanket of ash and pumice.

is evident that the extreme instability
of the country is gradually subsiding and
that every new plant serves as a soil
binder, helping to prevent further shift-
ing in the soil which it occupies (see
pages 46 and 47).

The lupine is in several ways peculiarly
well adapted as a pioneer. Its heavy
seeds can lodge where the lighter seeds
of grass would be blown away. The
large quantity of food stored in the seed
enables the young plant to grow with
great rapidity and establish itself in a
much shorter time than species with
smaller seeds.

It is, moreover, a leguminous plant,
and like all the members of the bean
family is provided with bacterial root
tubercles, which enable it to utilize the
free nitrogen of the air.

The lupines growing in the ash de-
posits all had very abundant and thrifty
tubercles. Their ability to obtain the
nitrogen necessary for their sustenance
from the air must be an advantage of
prime importance in a soil from which

organic nitrogen compounds are practi-
cally absent.

Nature has here carried out on an ex-
tensive scale a field repetition of Hel-
riegeFs famous pot experiments by which
he proved that legumes could live on
nitrogen drawn from the air in soils
where other plants perished for lack of
that ability (see page 158).

Since nitrogenous compounds are essen-
tial to plant life, the conditions here would
seem to prohibit the growth of any plants
except those provided with some means
of rendering the nitrogen of the air avail-
able.

With great interest, therefore, we ob-
served that a few individuals of other
plants had obtained a foothold in the
same deep ash. They had somehow
managed to obtain enough nitrogen to
reach maturity and produce seed. These
included blue- top grass {Calamagroslis
scabra), a sedge {Carex sp.), and a rush
(Juncoides sp.). Just how these plants
managed to subsist in a soil so nearly
devoid of all nitrogenous compounds is

Digitized by

Google

KATMAI VALLEY IN LATER YEARS

I6l

Photograph by W. L. Henning

EROSION IN VOLCANIC ASH

We cut oflF the top of this tree in 1917 at a then convenient height. So much ash washed out in the
ensuing years that in 1919 the cut end stood nine feet above the ground.

a mystery, but that a few individuals
did so is beyond question.

A FEW FISH

One might suppose that no fish of any
kind could have survived in the pumice-
choked streams.

We were thus surprised on our first
trip up the river to find a few small fish,
like minnows, playing over the quick-
sands. These were, however, unusual.
We saw no more fish of any sort until
191 7, when we discovered some Dolly
Varden trout in a brook on the south side
of the Valley, where the ashfall had been
about 18 inches on the level.

As such a quantity of ash seemed suffi-
cient to choke the stream, we supposed at
first these must be migrants from out-
side the devastated area. But an ex-
amination of the scales of one of them,
by Professor Charles H. Gilbert of Stan-
ford University, indicated that it had
been spawned in 1915, probably in the
stream where found, so it seems more

likely that its ancestors had survived in a
particularly deep pool which happened
to occur in the course of the stream. No
trout were found in neighboring brooks.

A PERVERTED RUN OF SALMON

Everywhere we kept a sharp lookout
for salmon, but found none until 1917,
when we discovered one little brook in
which red salmon were running in con-
siderable numbers. This stream did not
appear different in any way from a score
of others which coursed down the moun-
tain side to join Katmai River. But
although we could find no fish in any of
the other streams, we caught them here
almost every time we came by, taking
altogether about 50 in the course of a
month (see page 159).

Readers unfamiliar with the spawning
habits of the salmon may be surprised
to learn that fish two feet long could
be found in a stream only a few inches
deep; but the initiated will recognize
that there is nothing unusual in such an

Digitized by

Google

«-o

162

Q
52;

C/5

5 ^

Digitized by

Google

KATMAI VALLEY IN LATER YEARS

163

occurrence. It is well known that some
species of salmon work their way up into
brooks so shallow that their backs stick
out before they deposit their eggs.

The remarkable thing about this run
was that all of the fish taken were red
(sockeye) salmon, a species which spawns
in lakes and normally seeks only streams
which have accessible lakes in their
courses.

When we observed that our fish were
sockeyes, one of the party examined the
whole length of the stream to see whether
there could be a lake on its course, but
he reported that there was no possibility
of there ever having been a pond in the
stream.

SALMON PROBABLY HATCHED WHERE
• FOUND

Thinking, therefore, that the fish
might be strays, diverted to this stream
by the abnormal character of the
country, we sent some of their scales to
Professor Gilbert, who, after examining
them, reported:

'*It seems impossible to accept a sug-
gestion that these fish were strays from
some other region. . . . Each of the
salmon from which you collected material
had had an identical history. Each was
five years old (that is, was in its fifth year
when captured, having been spawned the
year of the eruption) and each had lived
two years in fresh water before going to
sea. ... I think it decidedly prob-
able that they were hatched in the very
tributary of the Katmai River in which
alone they were to be found.*'

We were anxious to see whether similar
runs would continue in subsequent years.
The stream was visited again in 1918,
but it was on the 4th of July, which is
rather too early for the fish. In 191 9 we
searched diligently for them a number of
times during the spawning season, but
none were found, either in this stream or
any of the others.

We have little basis for theorizing on
the significance of this anomalous run,
but it is clear that here is a biological
problem of great interest and importance.

If we understood fully why these fish
ran into that particular stream at that
particular time, we might gain an insight
into the causes which lead to the well-
known peculiarities of their runs: the
great variation in their numbers from
year to year, their great abundance in
some streams and their scarcity in others.

Beginning at Soluka Creek, a marked
change in the animal life of the upper
valley was evident when we returned in
19 16. The year before we had seen in
this area only a solitary duck and a few
small birds, all evidently waifs, though
insects had been fairly numerous — in-
cluding mosquitoes, black flies, and
abundant dipterous larvae under the
bark of the decaying trees.

In 19 16 there were numerous flocks of
chickadees and sparrows, and the insect
life showed a marked increase, both in
abundance and variety.

In upper Katmai Valley the change
was much greater, for where there had
been practically no life the year before,
insects of many species were fairly swarm-
ing. They were everywhere, not only in
the valley but high up on the mountains
as well.

FLIES DYING OF STARVATION

The insects were all ravenously hungry.
When we ate our noonday lunches they
swarmed over our food until they be-
came a nuisance.

They were, moreover, tame — so tame
that we had no difficulty in catching
specimens of all of them with our hands.
This may have been merely the result of
hunger, with its consequent weakness.
However the larvae of these insects
might have subsisted, there could be no
doubt that the adults which we en-
countered all faced the same end — star-
vation.

At one place in the valley we found a
little bunch of grass standing by itself,
far away from any other herbage. We
were astonished to observe, on examin-
ing it, that the ground beneath was
covered with dead flies — half a teacupful,
of all species.

Digitized by

Google

164

THE VALLEY OF TEN THOl SAND SMOKES

The blades had t>een gnawed and
chewed by the scores of living insects
which were clinging to them in the vain
hope of sustenance from the lone bit of
green 'see page 162). That the grass
was not the proper food for them was
evident from the heaps of dead on the
ground. But it was not until the speci-
mens could be determined at home that
we fully realized the unusual character
of this assemblage of insects.

LVSECTS Mf>STLY PREDACEOUS OR
PARASITIC

When the collection was examined by
Professor Hine it was found that less
than 30 per cent were vegetable feeders,
the remaining 70 per cent being composed
of parasitic predaceous, blood-sucking,
and dung- feeding forms.*

We were very much puzzled to account
for these insects in a region from which
they had been absent the year before.
It was hardly supposable that they had
bred in the vicinity. But if they had been
blown in by the wind, why had they been
absent the year before, when the gales
had been more numerous? And if they
had come in this way, why should there
be such a curious assemblage of parasitic
and predaceous forms, rather than a
representative collection of the whole
insect fauna?

Such study as we could give the sub-
ject did not enable us to solve these
problems. After a severe gale which
might have been expected to bring in a
new supply, if indeed they were wind
borne, we could detect no change in the
insect population around us. Moreover,
when we returned to the Valley in 191 7,
we found it inhabited by the same species
of insects that we had observed the year
before.

Although we were unable to account
satisfactorily for the peculiar assemblage

* There were 35 species altogether, distributed
as follows: 10 parasitic Hymenoptera (parasitic
on other insects), 2 moths (larvae plant feeders),
2 mosquitoes, i lepid fly (blood-sucking), 10
syrphid flies (larvae mostly predaceous on plant
lice), 6 anthomyid flies (mostly plant feeders),
I scatophagid fly (dung feeder), i bibonid fly

of insects which occupied the devastated
count r\-, we could draw a certain parallel
to it in the character of the mammals
encountered. These were all beasts of
prcA'. Signs of bears, foxes, wolverines,
and wolves were common. To these we
added subsequently a weasel; but in the
whole course of the work not a single
herbivorous animal, nor the track of one,
was obserxed until 1919.

Not a sign of caribou, which formerly
were abundant, nor of moose, was to be
seen. Neither did we find any of the
rodents which had swarmed over the
country, including marmots, ground
squirrels, rabbits, porcupines, and numer-
ous mice. The only exception was one
small ** mouse" seen by one of the men,
but this was quite as likely a shrew as a
rodent.

NOTABLE INCREASE OF HERBIVOROUS
ANIMALS IN 19 19

The progress in the return of animal
life beti^-een 191 7 and 191 9 ^"as striking.
There was not only a notable increase in
the number of animals, but herbivorous
species began to come in.

The upper valley of Martin Creek, for
example, where we spent a number of
days in 191 7, was then almost without
larger animals. I particularly remember
being startled, while wandering alone one
day, upon finding a solitary loon floating
about in Horseshoe Pond. My mind,
which had been absorbed in studying the
geological problems round about, was
suddenly brought back to the utter deso-
lation and solitude of the country by the
sight of this lonesome bird, paddling about
close by with no fear of the intruder.

But when we went up this same valley
two years later, animals of many sorts

(usually a feeder on vegetable matter), i lace-
wing (larva predaceous).

We were interested to observe that the sca-
tophagid fly, which is a common dung feeder all
through Alaska, altered its habits under the
stimulus of hunger to such an extent as to become
predaceous, catching and devouring mosquitoes
in the tent, giving a further indication of the
straits to which these creatures were reduced by
the lack of food.

Digitized by

Google

KATMAI VALLEY IN LATER YEARS

I6S

were abundant. Our footsteps were
dogged all the way by the* monotonous
note of the golden crowned sparrow, " Oh
dear me! Oh, dear me!", in endless
repetition.

On one of the little ponds we found not
only several loons, but two or three pairs
of golden-eye ducks, geese, and even a
swan, these contributing an element to
our larder that had been sadly lacking in
former years. On the hillside round
about ptarmigan were nearly always to
be found, and there were a number of
colonies of ground squirrels, each with
several hundred individuals, like a
'*town" of prairie dogs. To complete
the fauna there were many short-tailed
mice, busy about their affairs, as they
worked along the ground with little heed
to approaching footsteps.

Strangely enough, the return of all
these animals did not appear to be cor-
related with a corresponding increase in
their food supply, for the vegetation in
this particular area had been abundant
from the first and had not increased ap-
preciably in the interval.

Similar conditions were observed
among the birds about Kodiak, where
with little doubt the food supply was
sufficient after the first year. The well
known varied thrush, or ** Oregon robin,**
is a good example, since it is a noisy
bird whose presence can hardly be over-
looked. The year after the eruption I
saw not one in my rambles through the
woods. Two years later one or two
were found and hailed as great rarities.

They became gradually more common
in 1916 and 1917, and in 1919, after
seven years, were everywhere in great
abundance.

The explanation lies, I suspect, in the
homing habits of animals, which seek to
breed in the very spots where they them-
selves were reared. If the birds that
came to the forest at Kodiak every sum-
mer were merely chance visitors, likely
to settle in any favorable situation at the
end of their migration, they should have
reappeared in full force the year after the
eruption.

But, if the bird life of any locality con-
sists mainly of individuals that were
hatched in the vicinity, the long period
of scarcity followed by a sudden increase
in numbers is just what would be ex-
pected. The varied thrush was so
nearly exterminated at Kodiak by the
eruption that if repopulation were re-
stricted to the progeny of survivors a
number of years must elapse before it
would reappear in force. Once fairly
started, however, the repopulation would
tend to proceed in a geometrical ratio
at a rapidly increasing rate.

The delay in the reappearance of
ground squirrels on the mainland is
readily explained in the same way.
These animals were almost exterminated
by the eruption. Although abundant
only a few miles away, none seem to have
migrated into the devastated district
for some time. After the first pairs
were established, however, large colonies
quickly sprang up.

Digitized by

Google

C/2

•s

«*•

o
<

S§

— *

••1

i66

Digitized by

Google

XIII

KATMAI CRATER

As we plodded up Katmai Valley the
second year (191 6), our objective was the
volcano. The day after our arrival,
July 19, dawned clear and fair, so we
were up and away toward the mountain
at an early hour.

When we started we had little idea of
making the ascent, expecting to content
ourselves with reconnoitering the lower
slopes. Leaving the mudflow at about
800 feet, we proceeded by an easy grade
over good hard ground up the long ridge
that runs out parallel with the river, at
the end of which we came into a magnifi-
cent view of the canyon and .the lakes
of the upper valley — the first sight we
had had of either. Here the real climb
begins.

Should we turn back or should we go
on? The weather was perfect, or so it
seemed. The climb thus far had been
so easy that as we proceeded we had be-
come more and more anxious to try for
the rim.

CLIMBING THE MUD-PLASTERED SLOPES

The slopes were all plastered with mud
of varied colors — gray, yellow, chocolate,
red, black, and blue — thrown out by the
last spasms of the eruption.

At the lower levels the mud was dry
and hard, making easy going; but it soon
became slippery as we ascended and, a
little higher, soft and sticky. Most of
the way it was about ankle deep, but in
spots it let us down nearly to our knees.
At times it required all our strength to
pull ourselves out (see page 169).

For many hundred feet we went up on
the crest of the lateral moraine of a
tumbling glacier which stretches from
the lowest notch in the crater rim, near
the sharp rock shown in the pictures,
down to about 2,000 feet. The eroded
badlands were grotesque enough, but the

picture of confusion presented by the
parti-colored, ash-clothed seracs of this
glacier was bizarre beyond description.

At about 3,500 feet patches of snow
began to alternate with mud and these,
when covered with wind-blown mud and
pumice, made much better traveling.
But above 4,000 feet the way was mostly
through fresh soft snow, with only oc-
casional mud patches, and the slopes
became steeper as we advanced (see
page 170).

From the higher levels the scenery was
superb. Over the tops of the nearer
mountains, which presented a magnifi-
cent mass of sharp peaks with interven-
ing snowfields and glaciers, we could see
the blue waters of Shelikof Strait and
the distant snowcaps of Kodiak Island,
50 miles away.

Notwithstanding the exceeding sharp-
ness of the individual peaks the general
concordance of their elevations suggested
that they might be the remnants of an
ancient land surface — a peneplain — as
believed by Atwood.^

When we approached the to p it began
to cloud up, and we were afraid we should
not be able to see anything if we did
reach the rim. All the other summits
for miles around remained clear, but
Katmai became densely covered with
black, heavy clouds which permitted
only occasional glimpses of the crest.

Furthermore, we were on the 1 ee side of
the crater instead of to windward, as we
should have been. We knew the clouds
must be due, in part at least, to t he activ-
ity of the crater, because of the strong
sulphurous odor which filled the air, but
we could not tell how much was to be at-
tributed to this cause and how much
simply to the greater altitude of the vol-
cano.

The clouds were in rapid motion.

^ United States Geological Survey, Bulletin 467, page 83, paragraph 2.

167

Digitized by

Google

168

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

MOUNT KATMAI FROM NEAR THE BASE CAMP, I917

From a distance of over 20 miles the volcano loomed up much bigger than the nearer mountains.
The great jagged edges gave us all a new conception of the immensity of the crater within.

coming straight up out of the crater.
What if we should reach the rim only to
poke our noses into a steam jet, through
which we could see nothing!

Yet we were unwilling to give up now,
without at least a try, and so we pressed
on. The slope increased in pitch toward
the top, making the climb progressively
more wearisome. Church expressed the
feelings of all when he complained that
the rim was farther away than when we
started. No matter how far we went he
insisted, " It's still a thousand feet to the
top." And so it appeared up to the
very crest. Finally at 5,500 feet we
reached the rim.

THE KNIFE-EDGE CRATER RIM

We came to the rim so suddenly that
we might almost have walked right over
and tumbled down the precipice inside.
The edge was unbelievably sharp — so
sharp that a single step lifted the interior

into full view. We stopped in our tracks,
sank to our knees, then stretched out at
full length, and, propped up on our
elbows, peered over.

On this knife edge we lay, trying to
fathom the cloud of steam that rose in
our faces. For our fears had been real-
ized : before we gained the rim the clouds
had shut down, enveloping us in mist.
Then there came a little rift, and for a
second we could make out dimly some-
thing blue, far, far below. Then the
steam cut us off again, and again we
waited.

We could see only a few feet in any
direction. To left and right a little of
the precipitous wall could be made out,
standing almost vertical. Yet the crest
was made of loose fragmental stuff — ash
from the eruption. Its knife-edge sharp-
ness was maintained by frequent slump-
ing away into the depths.

Back from the crest was a series of

Digitized by

Google

KATMAI CRATER

169

* Photograph by R. F. Griggs

STUCK IN THE MUD ON THE WAY UP TO THE CRATER — FIRST ASCENT
The slopes of the volcano were covered with soft, sticky mud and slush (see text, page 167).

fissures running parallel with the edge,
where the rim was cracked off, ready to
drop over the cliff. These had com-
pelled us to detour till we could find a
place at least without visible signs of
slumping off.

The ash lay in the same even bedded
layers as below. Although it was here
40 or 50 feet deep, three times as much as
at the foot of the mountain, the frag-
ments of which it was composed were
little coarser than there (see page 171).

Our suspense as we waited, perched on
that eery crumbling edge, cut off by the
encircling mist from everything else in
the universe, would be hard to describe.
Were we to be cheated, after all our
effort?

Finally, the steam blew away, and we
were struck speechless by the sight below.

FIRST VIEW OF THE CRATER

We found ourselves hanging over the
brink of an abyss of such immensity that,
as the event proved, we were powerless
even to guess its size. Down, down,

down, we looked until the cliff shelved
off and we could follow it no further.

In the bottom lay a wonderful lake, of
a weird vitriolic robin's-egg blue, milky,
like one of the glacial lakes of Syritzer-
land. The middle of this was set with a
horseshoe island, the remnant of a cinder
cone evidently thrown up in the last
spasms of the eruption. Around the
margin hissed columns of steam, issuing
from every crevice. On the opposite
side of the crater were spots of yellow
sulphur, and in two corners where the
slopes were less precipitous were exten-
sive snowfields (see pages 172 and 178).

All this we took in almost at a glance,
for we were given no time to study de-
tails. Before we could get our tripod
set up the cloud closed in again, and we
waited amid a roar of escaping steam.
What if we should fail to get the coveted
pictures for lack of a clear moment!
Finally there came a little rift and fran-
tically we made our exposures.

I had planned to take bearings and
measurements which would afford some

Digitized by

Google

I70

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by L. G. Folaom

ON THE FIRST ASCENT WE ROPED UP FOR FEAR OF CAVE-INS
IN THE MUD-COVERED SNOW

The climbers are within a few hundred feet of the crater rim.

sort of determination of the depth and
size of the crater, but we were vouch-
safed so few clear moments that it was
impossible to do so.

We had reached the rim at 5.05 p.m.
We lingered on the edge for 50 minutes,
hoping for better views. But as the
clear intervals became less and less fre-
quent we had to give it up and start
home.

None of us fully realized, I think, how
far we had come till we found how long it
took us to get back. Down hill as it
was, it seemed as though we should never

finish with that inter-
minable succession of
gullies and mud holes.
It was a weary crowd
that finally dragged
into camp, but the
awe inspired by the
sight we had beheld
counterbalanced the
exhaustion of our
bodies and kept our
minds busy with
speculations on the
significance of what we
had seen.

Next day, on writ-
ing up our observa-
tions, Folsom and I fell
into argument as to
the size of the crater.
He maintained that it
was about 500 feet
deep, while I insisted
that it must be more
nearly 1,500 feet.

Later we were to
find how ridiculously
small both estimates
were, for the abyss is
so big that it is liter-
ally beyond the power
of the imagination to
grasp its dimensions,
even when standing on
the rim itself. The
uncertainty of this and
other details, which we
had not been able to
observe satisfactorily in our few brief
glimpses of the abyss, proved to us that
we must make another ascent to clear
up the discrepancies in our observations.

THE SECOND ASCENT

But the opportunity for a second as-
cent did not come until July 30, when
for the first time since our arrival in the
Valley, the steam from Mageik rose
straight up into a cloudless sky. This
time we chose a route up on to the lava
plateau from near the mouth of Mageik
Creek.

Digitized by

Google

KATMAI CRATER

171

The ground did not seem especially
favorable from below and we were by no
means sure of reaching the rim when we
started ; but we were anxious to examine
Trident at closer range, and especially to
see what might lie behind the isthmus
connecting this volcano with Katmai,
because of our suspicions of activity in
that direction.

THE CRATER OF TRIDENT

This course gave a fine view of Trident,
whose crater proved to be a simple
fissure, out of which steam was con-
tinually issuing in comparatively small
volume.

But we were disappointed in our hopes
of seeing anything over the divide be-
tween Trident and Katmai. Although
we traversed the whole length of the
nearly level n6v6 at an altitude of about
4,200 feet, we could see no indications of
volcanic activity beyond. There were
several jagged minor summits, but as we
saw no large mountain and no clouds we
quite dismissed the idea of a volcano in
that quarter. How greatly in error we
were in this conclusion we were to find
out only the next day.

On our first ascent we had followed
straight up a single ridge, and so avoided
the necessity of crossing the gullies. But
this time, for a good share of the way be-
yond 2,000 feet, our path lay across the
lines of drainage, which had gashed the
level surface of the ash with innumerable
seams anywhere from two to ten feet
deep. We soon found that continued
jumping and scrambling across these
gullies was very fatiguing, and we were
thoroughly tired of the job long before
we got through them.

AN ASH-COVERED GLACIER

For the last 1,500 feet our way led
across crevassed snowfields and ash-
covered glaciers. The appearance of
these ice fields was curious in the extreme.

Where a glacier was moving rapidly,
the variously colored layers of ash were
dumped about promiscuously among the
glacial seracs, making a bizarre and

Photograph by R. F. Griggs
THE CRATER RIM
The extreme sharpness of the edge is evident.

highly colored picture, the reds, browns,
blacks, and yellows in vivid contrast to
the pure blue ice.

Where the glacial motion was slower
and steadier the ash deposit still lay as it
fell. In such places drainage systems
had developed which followed the slope
exactly as on ordinary hillsides, regard-
less of the transverse crevasses beneath.

These lines of drainage were some-
times able to maintain their courses in
competition with the crevasses, giving
rise to a curious double system of topog-
raphy. The square, straight-sided blocks
of ash looked like so many huge slices of
layer cake laid side by side.

Glaciers of this peculiar type extended
clear up to the very rim of the crater,
above whose depths the loose blocks of
ice hung with such a precarious hold that
we dared not approach.

Turning back, we made our way around
the ice, descending somewhat until we
finally reached the rim in the lowest

Digitized by

Google

Photograph by J. W. Shipley
THE NORTHERN WALL OF THE CRATER

The steam from the numerous fissures evaporates about 1,500 feet above the lake, but condenses again
into a great cumulus cloud on meeting the outside atmosphere.

172

Digitized by

Google

KATMAI CRATER

173

notch (altitude 5,200 feet), beside the
sharp tooth of rock which projects at
that point (see page 100).

From the foot of the mountain this
tooth appeared as a small rock. Near
at hand it proved to be a great column
of jointed lava, two or three hundred
feet high. Its position and structure
suggested that it may be the solidified
core of an old vent (see page 175).

From our position directly under the
column, its perpendicular cliflfs, though
insignificant from the Valley, appeared
immeasurably high! Frequently, in this
land of stupendous proportions, we
had occasion to realize how little con-
ception we could really form of the true
dimensions of the scenery around us.

INABILITY TO JUDGE HEIGHT OR DISTANCE

When one stands directly beneath a
cliff, or looks down from the brink, 200
feet appears an immeasurably great
height. Ten times as much appears
little greater, unless there are trees,
houses, or some such familiar objects by
which the distance may be judged inde-
pendently.

In a desert country without any such
criteria we were frequently unable to
form any estimate at all of the various
features that met our view. We had
an amusing instance of this when, send-
ing a man up on a great ash slide to
serve as a scale for a picture I found to
my amazement he was hardly visible to
the naked eye, and utterly lost on the
ground glass.

We nearly always found our estimates
too small rather than too large, .and
generally had to revise our statements
upwards when accurate data were sup-
plied by the survey.

Standing on the edge of the crater, we
recognized our total inability to form
any judgment of its depth. Using every
standard of comparison possible, we con-
cluded our former guesses were too small
and that it must be at least 2, 000 feet deep.
Thus by slow degrees did we approach the
real figure. But even yet our conjecture
was barely more than half big enough.

Estimation of the width of the crater
was somewhat easier because we could
see the rim from below. We knew that
its great jagged arc looked far bigger
twenty miles away than from the foot
of the mountain, and by laying it off
on the chart reckoned that it must be
about three miles in diameter. Never-
theless, as we stood there on the crest it
was impossible to believe that the dis-
tance across to the high point opposite
was the equivalent of forty city blocks.

THE SECOND VIEW OF THE CRATER

Both the weather conditions and our
position were more favorable for observa-
tion of the crater than on our first ascent.
The sun shone brightly, and it became
evident why steam had given us so much
trouble the first time. We found that
the point which we had reached before
stood directly above a prominent fissure
stretching out from the edge of the
lake.

The hot lake was now covered with
wisps of steam, curling up everywhere
from its surface. The vapor thus given
off condensed into a hazy cloud, which
hung in the mouth of the crater, conceal-
ing the rim opposite.

At the northeast angle we could see
another low notch in the rim at about
the same altitude as the one where we
stood. But this one was occupied by a
wall of ice which rose vertically, flush
with the crater walls, as though sheared
off by the explosion.

It was surprising at first sight to find
that a glacier could have persisted on the
crater rim, unmelted by the heat of the
eruption. It might be supposed that, as
the eruption approached, the tempera-
ture of the mountain would be raised
enough to melt away its glaciers, or, at
any rate, that the heat of the molten
magma laid bare by the explosions would
promptly liquefy all the ice near by.

It is clear, however, that no such thing
happened. Ice cliffs, the exposed ends
of beheaded glaciers, stretch for several
miles along the crater rim. The glaciers
bear little if any indication of having

Digitized by

Google

174

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by L. G. Folaom
A LONE WOMAN VISITOR AT THE CRATER's EDGE

Mrs. Griggs, spurred by an unconquerable optimism, climbed through the fog all the way, while a
party of the men lingered behind, waiting in vain for a break in the clouds. She ^'as rewarded by
commg into clear air just before reaching the rim, but some of the men went home without having
ever beheld this wonder of wonders.

dwindled during the eruption. The
parts preserved on the outer slopes were
outside the blast of the explosions, which
must have been directed skyward.

The ice cliff, however, probably did
not stand so close to the brink during the
eruption. The crater wall has caved in
considerably since the eruption and is
still slumping away at short intervals.
It is in such an unstable condition that it
is no infrequent experience, while stand-
ing on the crater rim, to be startled by a
roar as something lets go, and on search-
ing for the cause to find a mass of ice and
rock bouncing down the half-mile preci-
pices into the abyss.

On our first ascent we had not fully
appreciated the coloring of the walls,
which are made up of variously tinted
lavas and tuffs. The prevailing tones
are soft warm browns, harmoniously
framing the turquoise lake. But in
many places these give place to brilliant

reds and yellows, in strong rich contrast
to lake and sky.

ASCENTS IN 191 7

It was with very different feelings that
we undertook the climb when we returned
to the scene in 191 7, for then Katmai was
a secondary matter in comparison with
the Valley of Ten Thousand Smokes.
Our purpose was to survey the crater to
ascertain its dimensions.

But for me personally there was an
added incentive: I wanted to see how the
new members of the expedition would
act when they first popped their heads
over the rim. I was not disappointed,
for sated with wonders though they were,
after a month's sojourn in the Valley of
Ten Thousand Smokes, they were, if
possible, more impressed with the crater
than with the Valley.

They accused me of not having given
them any true description of it. Yet I

Digitized by

Google

KATMAI CRATER

175

Photograph by J. W. Shipley

TOPOGRAPHERS WORKING ON THE CRATER RIM

The peak at the left is the inconspicuous tooth seen in the notch of the crater rim from below (see page
100). A portion of the beheaded glacier lining the rim is shown beyond.

had done my best and they had studied
our photographs just before starting.
Sayre epitomized the situation. He
told the photographers: '*Go ahead and

take your d pictures. But you

might aswell tear them up now. They *11
never help anybody to imagine what this
place is like."

Recalling our previous difficulties with
mud, we had decided to leave the ascent
until late in the season, when the winter's
snow was gone and the mud had had
time to dry up somewhat.

Days for climbing the high mountains
were very few in 191 7, and we had finally
to wait a full week before there came a
chance to try. Even then we were
cheated. As we ascended, the clouds
began to gather; and just as we reached
the crest, they completely shut down.
We had one fleeting glimpse of the crater,
but before we could so much as set up a
camera it was gone. We had to wait
three days longer for another chance.

Our decision to defer the ascent till
late in the season was well justified. In
places where in 191 6 we had floundered

up the slopes in slippery mud, ankle
deep, we found the ground hard and firm,
so that the climb, which before had taxed
our strength and endurance, was now
made easily in four hours, even under
30-pound packs.

Those who had not gone through the
previous experience had difficulty in be-
lieving the first climb could have been so
much harder, until we came upon our old
trail. It was broken so deeply into the
mud that it had persisted in spite of
freezing and thawing through the winter,
and was still deep and clear a good share
of the way!

In spite of my experience of the pre-
vious year I found myself overawed by
the grandeur of the abyss. Its im-
mensity little comported with my recol-
lection, huge as I had conceived it to be.

The proportions are so perfect that no
one dimension appears exaggerated at
the expense of the rest. If the crater
were not so deep, its area would be more
evident; if its walls were not so precipi-
tous, one could better measure with his
eye the distance to the bottom.

Digitized by

Google

176

THE VALLEY OF TEN THOUSAND SMOKES

v*J

(/)

'•t>^

•v3

-4J

&«

52;

•s

t-H

-5

•0

tfi

"bb

As one looks into the abyss, the
dominant impression is one of
enormous depth, but this is hard
to render in the photographs,
most of which give rather the
impression of great width. The
magnitude of the crater is literally
inconceivable.

Yet as one stands on the rim
and projects in his mind's eye the
outer slopes up to the original
summit, which stood 4,200 feet
above the lake within, he does
gain a conception, perhaps not to
be duplicated elsewhere, of the
tremendous forces that destroyed
the majestic mountain that only a
little while ago was Katmai.

There have been no • evident
changes in the crater during the
four years covered by our obser-
vations. Some of the talus slopes
have extended a little farther into
the water, but there is no demon-
strable change in the height of the
lake.

Unless there is a subterranean
outlet, which is very unlikely, its
level must be controlled by the
temperature of the crater floor.
If the quantity of heat given off
should increase, the water would
be vaporized more rapidly, lower-
ing the lake. If the rock should
cool off, water from the abundant
precipitation would accumulate
and the level of the lake rise.
The height of the lake is thus a
good indicator of the quantity of
heat supplied from the conduit
below.

We wanted to estimate the
depth of. the lake, but could find
no means of doing so and had to
be content with such surmises as
we could make from the charac-
ter of the banks, which vary con-
siderably.

Beneath the northeast cliff a
conspicuous talus slope nearly
1,000 feet high has accumulated.
But on the north and northwest

Digitized by

Google

KATMAI CRATER

177

the cliffs dip off into deep water, with
no sign of talus.

When the extremely unstable charac-
ter of the cliffs and the frequency of ava-
lanches are recalled, it will be seen that
the amount of detrftus dumped into the
lake in the years since the eruption must
have been prodigious. That this rock
waste could have entirely disappeared
beneath the waters suggests that the lake
is at least several hundred feet deep close
up to the shore.

A WEIRD CLOUD-ENCIRCLED CRATER

The clouds hung off long enough for
the topographers to make the observa-
tions necessary to triangulate the crater,
but even after waiting so many days we
were permitted only a brief view. Be-
fore long the clouds began to gather
again. The struggle between the in-
creasing mists and the heat rising from
the abyss created one of the weirdest
spectacles imaginable.

Long after the gathering clouds had
blotted out all else, the crater cup re-
mained perfectly clear. Above the cliffs
on all sides stood impenetrable walls of
mist, 2,000 feet high. The sun shone
down on the lake out of a blue sky, just
as it had before. The mist which kept
filtering into the crater through the low
notches in the rim was quickly dissipated
by the heat rising from below.

For half an hour the elements main-
tained their balance in this uncanny
fashion, but finally the increasing cloud-
iness gained the upper hand, and the lake
slowly faded from view.

DIMENSIONS OF CRATER

From his survey Maynard has com-
puted the dimensions of the crater as
follows: The width of the crater rim, as
seen from below, which includes all of
the concavity on top of the mountain, is
3 miles (4.8 km.). The circumference,
measured along the highest point of the
rim, is 8.4 miles (13.5 km.). The area is
4.6 square miles (12 sq, km.).

* The capacity of the crater is 4,500 million
cubic yards (3,400 million cubic meters). It
13

The precipitous abyss, which does not
extend to the rim on the southwest side,
is somewhat shorter, measuring 2.6 miles
(4.1 km.) in length, 7.6 miles (12.5 km.)
in circumference, and 4.2 square miles
(10.8 sq. km.) in area. The milky blue
lake in the bottom is 1.4 miles (2.2 km.)
long and 0.9 mile (1.4 km.) wide, with an
area of 1. 1 square miles (2.8 sq. km.).
The little crescent-shaped island in the
lake measures 400 feet (120 meters)
from point to point. The precipice
from the lake to the highest point of the
rim is 3,700 feet (1,130 meters) high.

THE CRATER WOULD HOLD WATER ENOUGH
TO SUPPLY NEW YORK OVER FOUR YEARS

Statistics of dimensions of such mag-
nitude, however, convey little meaning.
Suppose we wished to fill the lake to the
crater rim and had available the pump-
ing stations which are employed to keep
New York City supplied with water.
If we should turn all the streams that
supply the six million people of that great
city into Katmai crater the pumps would
have to continue working steadily for
I »635 days — four and one half years —
before the water would reach the lowest
notch in the crater rim.^

This does not begin to tell the story,
however. If the hole were filled with
water the cliffs still standing up around
the lake would form a spectacle almost
as impressive as the abyss we see today.
The great cliff on the eastern side would
stand 1,700 feet above the water's edge
and nowhere, except at the low notch
forming the outlet, could we approach
within hundreds of feet of the water level.

Because of the great irregularity of the
crater rim, computation of the quantity
of water necessary to fill it does not give
an idea of its true size. The real magni-
tude of the hole is indicated by the
quantity of material blown off the
mountain during the eruption.

A comparison of the present stub with
the original mountain shows a difference
of more than 1 1 ,000 million cubic yards,

would hold 900,000 million gallons of water.
New York City uses 550 million gallons daily.

Digitized by

Google

o "Z

178

5 .§S

P'S

a;
bo

3
cS

O
■M

■M

c
(d
«*■
o

Digitized by

Google

KATMAI CRATER

179

approximately 2 cubic miles (8.5 cubic
kilometers). This is more than 40 times
the amount of material dug in the great-
est excavation ever attempted by man —
the Panama Canal.

THE SKYSCRAPERS OF NEW YORK WOULD
NOT FILL THE CRATER

In seeking more familiar comparisons
to make vivid the proportions of the
crater, I recalled my bewilderment when
I first tried to see New York City. The
endless rows of closely built blocks,
series on series, supply an apparently in-
exhaustible number of objects large
enough for units to measure cubic ca-
pacity of almost any dimensions.

If one could drop the blocks of New
York into the crater of Katmai how
many would it take to fill it? The truth
is that if an average six-story tenement
house were set against the crater wall, it
would be hard to find in the picture.

But the skyscrapers impress the visitor
to New York more than the tenement
houses. How would they appear in the
crater? The probability is that all of the
skyscrapers of Lower Manhattan could
be submerged in the lake and never a
tower project. Before the bottom of the
crater were well covered, all the build-
ings from the Battery far up town would
have to be set in. The more put in, the
more impressive the hole would become.
And if every structure on Manhattan
Island were thrown into the crater
hardly a good beginning would have been
made. As a matter of fact, all the build-
ings of all the boroughs of Greater New
York — Brooklyn, Queens, the Bronx,
and Richmond, as well as Manhattan —
would not fill the crater.

Indeed, according to the best estimate
I can secure * the buildings of fifteen
cities like Greater New York would be
less bulky than the material blown off
the mountain!

KATMAI LARGER THAN KILAUEA

Kilauea, in the Hawaiian Islands, has
always been accounted the largest active
crater in the world, but it is clear that it

must now yield the palm to Katmai.
Kilauea's greatest diameter is 2.93 miles,
its circumference 7.85 miles, and its area
4.14 square miles. Each of these di-
mensions is somewhat less than the coi#
responding figure for Katmai.

But the most significant difference is in
the depth, which is less than one-seventh
as great as that of Katmai. The highest
point in the walls of Kilauea is only 500
feet above the floor of the crater, as com-
pared with 3,700 feet in Katmai.

There is a good automobile road lead-
ing through a low notch in the rim into
the caldera of Kilauea. But no one who
has ever looked into the abyss of Katmai
has been tempted to try the descent.
For, interesting as a close-up examina-
tion of the bottom of the crater would
certainly prove, no one has been able to
suggest a means by which those sheer
precipices might be ascended again, if by
good luck one managed to reach the bot-
tom with no broken bones.

As a matter of fact, there is little justi-
fication for comparing Katmai with Ki-
lauea. Each is unique and unrivalled in
its own way. Visitors to Kilauea are not
so much impressed by the magnitude of
the crater as by its incomparable fiery
lake of molten lava, whichj however, is a
scant thousand feet in diameter. On
the other hand, there is nothing at Kil-
auea similar in the least to the stu-
pendous abyss of Katmai, its turquoise
lake set round about by multi-colored
cliffs.

Katmai and Kilauea stand, in fact, at
opposite extremes in the series of the
world's volcanoes. While Katmai, dur-
ing the brief period of explosive activity,
furnished one of the most terrible
examples of the titanic forces of vol-
canism, Kilauea has been called "the
world's safest volcano." Its continual
quiet welling-up of highly liquid lava is
the antithesis of the explosive expansion
of Katmai, wherein all the magma was
blown to fragments.

' From data furnished by Herbert S. Swan,
Secretary of the Zoning Committee, New York
City.

Digitized by

Google

o
c

'uvJ
*j o

c
B

o
c

i8o

Digitized by

Google

XIV

MOUNT MARTIN

After the attainment of Katmai
Crater, Mount Martin, which was far
more active than Katmai itself, became
our prime objective. But for several
reasons it is the most difficult of the
volcanoes to study.

To get within striking distance is of it-
self somewhat of a problem, since it is
located in the most inaccessible part of
the range. Despite the definiteness of
our plans for its ascent, pressure of other
work prevented us from establishing a
camp at its foot in 1917 until late in the
season. During the week we watched,
the mountain was continuously shrouded
in a cloud, and we had to give up the
attempt.

Unfortunately, Mount Martin is so
placed that it is the first peak to cloud up
with an approaching storm, and the last
to clear again. Worse yet, the strong
updraft created by its column of steam
sucks up from the lowlands a body of
warm air, whose moisture promptly con-
denses into heavy clouds when it reaches
the upper levels.

While frequently clear for a time at
morning and evening, it is often swathed
in clouds during the middle of the day.
It is, therefore, hard to choose a day
suitable for the climb, for it is almost
impossible to tell in the morning whether
it will remain clear or cloud up before
noon.

Knowing something of these difficulties
in advance, we planned in 1919 to make
the ascent early in June, when the
weather is most favorable. But here
again we were foiled, for we found that
snow persists longer at the head of Mar-
tin Creek than anywhere else in the dis-
trict, concealing the features we most
wished to study.

Accordingly, it was not until August
that we found ourselves encamped in the
valley at its foot, waiting our chance.

The crater was clear on the morning of
the first day after our arrival. So up I
started without waiting for Hagelbarger
and Henning, who were to meet me.

But before I had gone far a hard rain
commenced, and although I could still see
the volcano it was of no use to try for
pictures. Nevertheless, the day was not
wasted, for I was enabled to pick out the
best route to the summit and to study the
glaciers and lava flows round about.

MOUNT MARTIN A SUBORPINATE CONE ON
THE FLANKS OF MAGEIK

From the level of the lava platform on
which the cone stands, it is evident that
Mount Martin, although located di*
rectly on a projection of the Katmai-
Mageik axis, and the most active vent
in the whole region, has somewhat the
character of a parasitic cone on the flank
of Mageik.

From the Valley some of the lava
flows comprising the plateau look as
though they might have come from
Martin, but from the higher levels it is
evident that all the more extensive bodies
of lava were poured out by Mageik.

Martin stands superposed on the slop-
ing surface of these lavas. Its cone is
composed of short, massive flows of the
same basic andesite that characterize
the whole region. But apparently the
lava was so viscous when extruded that
it solidified close around the orifice. As
a consequence, a steep-sided pile was
built up, which from the valley looks
like a cinder cone, rather than a mass of
solid lava.

The crater, which is about 1,000 feet
in diameter, appears from below as a
funnel blown out of the top of the
mountain. The lowest point in the rim,
which lies on the east side, is several
hundred feet lower than the high point
opposite.

i8t

Digitized by

Google

I82

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

MOUNT MARTIN FROM THE PACIFIC SIDE

Three times we climbed this volcano and once descended inside, but we were baffled in our attempt
to study its crater t)y the clouds which shut down before we could reach the rim.

Often the whole cavity is filled with a
dense cloud of steam, which may rise
high in the air in a solid column. But
on a very dry day it can be seen that
most of the steam comes out in two
vents at opposite sides of the crater.

MARTIN SEEMS NOT TO HAVE THROWN OUT
ASH OR PUMICE IN THE ERUPTION

We had hoped to be able to determine
the character of the convulsions by which
Martin was opened up, and the r61e it
played during the eruption. The fact
that its later activity has been greater
than that of any other vent in the dis-
trict suggests that it may have con-
tributed a considerable share to the
general disturbance. But we saw no
evidence on the flanks of the volcano
that it had ejected any ash or pumice.

The only deposit of unusual character
observed in the vicinity was a consider-
able amount of coarse sand and fine
gravel, made up of fragments of old
crystalline rocks, which lay among the

crevices of the lava flows, as though
showered onto them by an explosion
from the top of the volcano.

The summit, before the eruption,
doubtless bore a quantity of glacial till
that may very well have been the source
of such a deposit. If the crater of
Martin was produced by explosion, and
this sand is the only debris left behind,
the outburst must indeed have been
mild, compared with the terrible blasts
which wrecked Katmai.

ALL VEGETATION IN THE VICINITY KEPT
DOWN BY ACID RAIN

Upon examining the neighboring
country, I was astonished to see that for
a distance of five miles or more in every
direction the naked brown ground,
though fertile and free from heavy ash
deposits, remained bare of vegetation.
It was quite different in this respect
from any other area in the district.

Even if all vegetation had been de-
stroyed seven years before, it should

Digitized by

Google

MOUNT MARTIN

183

have returned by this time; everywhere
else plants were growing vigorously where-
ever their roots could reach the old soil.

I could not imagine what might be
preventing revegetation until I chanced
to experience a local rain, in which the
drops had been so strongly acidulated by
the fumes that they were painful to the
eyes and other sensitive parts. Such acid
rains probably destroy the plant life as
fast as it comes back within their range.

This country, indeed, resembles
strongly the smelter regions of the West,
where vegetation has been destroyed by
noxious fumes.

On the second attempt to climb, clouds
settled down on us at the 3,500-foot level,
but, thanks to bearings secured the first
time, we were able to keep on through
the obscurity for another hour in hope
of a break that would permit us to scale
the summit.

Finally, having reached the main divide
of the range, we sat down and ate our
lunch, hoping for a rift that would show
us where we were, for we had gone as far
as we dared without new observations.
Later, we found that we were right under
the last steep pitch of the cone and could
have reached the rim in a few minutes
more, could we have seen our way.

But it would have been a useless ex-
ploit to reach the top under such condi-
tions, since we should have seen nothing
when we got there. It was a mournful
party that descended the mountain that
evening, for the demands of other work
were imperative, and camp had to be
broken next day with the crater yet un-
seen after two weeks of waiting for a
chance to climb.

THE THIRD CLIMB

Five days later, as I lay awake at
Baked Mountain Camp, I crawled out
into the night to look at the Ten Thou-
sand Smokes and the volcanoes in the
spectral light of a wonderful full moon.
There was Martin puffing away, beauti-
fully clear, its unconquered steam col-
umn rising majestically over all. Why
not do it tomorrow?

We were now on the wrong side of the
range, and far from a favorable starting
point; but it was the only chance, for
there were unmistakable signs that the
good weather that had favored us for
two days was about to change. It
meant covering a mile in altitude and 30
miles in distance, but I was not to be
turned back if there was any possible
way of getting there, and I knew Charlie
Yori was as anxious as I to try it.

Indeed, I should never have dared at-
tempt to cross the glaciers that guard
Mount Martin on the Valley side with-
out the benefit of his experience in ice
work. So I waited until a seasonable
time to rouse the camp, and finding that
we could leave without interfering with
the chemists* work, Fenner, Yori, and I
got ready in a hurry and started off
double quick, in our eagerness to get to
the top before anything should happen.

The whole of our course, after we had
climbed out of the Valley, lay across
glaciers and snowfields. When we sur-
mounted the first slope, we found to our
sorrow that between us and our goal lay
a considerable hollow, into which it was
necessary to descend several hundred
feet before starting up the slopes that
led to the cone.

The glacier that occupied this hollow
was more extensive than any of the
others which have developed on the
flanks of Mageik. From its crest our
view stretched out for many miles
across the interior basin back of the low
coast range.

A few miles beyond us the central
range, which bears the volcanoes, gave
way to a broad open lowland, occupied
by a luxuriant growth of tall grass. Be-
yond this lay Becharof Lake, from whose
shores these fine lands are readily ac-
cessible in every direction. Farther
still, from the opposite shore, the sym-
metrical white cone of Peulik, the next
volcano of the series, rose a mile high
directly out of the lowland.

For the most part going was not
difficult, except that we had to waste
much time winding in and out around

Digitized by

Google

1 84

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by Frank I. Jones
MOUNT MARTIN FROM BAKED MOUNTAIN CAMP

The crater is thirteen miles distant in a straight line. After leaving the first slopes, five miles away,
our path led entirely over ice and snow. Some of the glacier-work was extremely hazardous.

the irregularities of the ash-covered
glacier, here following a drainage gully,
there cutting across the ridges.

A "real GLACIER*'

But between the hollow into which
we had dipped and the high snowfields
surrounding the cone lay an area badly
cut up by close-set crevasses. As we
entered this, Yori remarked, *'This is a
real glacier, all right enough.**

With his customary hardihood, he pro-
fessed to scorn a rope, and took a sort
of fiendish glee in trying me out in the
most '* ticklish'* places he could find.
I will not deny that I was somewhat
timid, for my hobnailed boots were worn
out and I had been compelled to come in
rubber-soled shoe-packs, which could

get no grip on the slippery surface. This
bothered me greatly, though the ice-axe
which I carried probably counter-bal-
anced the disadvantage.

Our way lay, as Fenner expressed it,
"along the ridge pole," following narrow
crests of glaring blue ice, themselves
sloping both ways, between bottomless
crevasses on each side. A slip would
have meant certain death. But we
crept along, using all the care we could,
and finally reached the n6v6 above with-
out mishap.

Without stopping for the much-de-
sired pictures we pressed forward fever-
ishly, in the hope of reaching the summit
while it was yet clear. A breeze had
sprung up out of the northeast and we
knew it was a question of only a few

Digitized by

Google

MOUNT MARTIN

185

hours before our mountain would be
hidden in clouds. Indeed, they had
already begun to gather.

We made rapid progress now, travers-
ing the smooth snowfields in a wide
semicircle around the base of the cone
toward the low point in the crater rim
opposite. We were in the clouds here,
but had a glimpse out on the other side
of the range to the Pacific.

When we came around under the low
notch in the rim we sat down and waited
for a break in the clouds, to get our bear-
ings before taking the last steep slope,
and ate lunch as we waited.

Hardly more than a hundred yards
away was the hillock where we had
stopped on our attempt from Martin
Creek. This time, however, we never
doubted but that we should soon have
our chance, for the clouds were only
beginning to gather. Indeed, just as we
started to eat they blew off a little.

There was the crater rim, seemingly
only a few steps directly above us.
Great masses of steam came rolling up
close against it, but as we were on the
windward side they were quickly carried
off in the opposite direction. From a
distance the smoke of Martin always ap-
pears snow white, but from our position
it took on a weird lemon-yellow color,
which Fenner suggested was probably
due to reflection from a large body of
sulphur within the crater.

CLIMBING A 60-DEGREE DIRT SLOPE

The rim seemed so close that, dropping
my cracker, I started for it. Before I
could go a dozen steps the clouds closed
in again, continuing until we had
finished our lunch. But we had our
bearings now, and started up, so as to be
on the rim when the next break came.

That last pitch, 250 feet it proved
when we measured it, was the steepest
slope I have ever seen. If it had been
rock it would have been easy climbing;
but it was bowlder-clay, left there by a
glacier that had capped the mountain
during its dormant period.

The slope was 60 degrees by the cli-

nometer (as compared with about 30 de-
grees in a steep railroad embankment).
The round bowlders on which we de-
pended for foot- and hand-holds were so
loosely held in the uncemented clay that
it was extremely difficult to hang on.

Finally we reached the top, at 5,300
feet, but were unable to see anything in
the cloud and steam that surrounded us.
Inside the first sharp edge we found a
slight depression, and then a second
similar sharp inner rim. The original
rim had evidently broken loose and
slumped down into the hole a little.

In the depression between the two rims
was a little pool, over which we bent to
secure a drink, for we had had only snow
with our lunch, but, ugh! it was strong
acid.

The fumes at the rim were very dis-
agreeably acid, and I was glad to retreat
into a little hollow where I could take
notes in comparative comfort.

DESCENDING INTO THE CRATER IN THICK
STEAM

After a little, Fenner came out of the
cloud and reported that it was possible
to go down inside the crater, using
handkerchiefs as respirators. So we all
put our handkerchiefs to our noses and
plunged over the edge.

On the rim we could distinguish the
outlines of objects 50 feet away; but
once inside the steam thickened rapidly.
A few feet down we could hardly see each
other, though standing only four or five
feet apart. Whenever a gust of wind
swept the smoke back a few feet, we
leaped on farther down, until the ob-
scurity closed over us again and we were
compelled to halt for fear of stepping off
the edge of the precipice into the roaring
vent beneath.

There we stood huddled together, like
ninnies, panting through our handker-
chiefs and pulling down our hats in futile
efforts to protect our smarting eyes. If
we loosened the handkerchiefs a little to
get a freer breath, we met a suffocating
draft that at once compelled us to clamp
the protection back again.

Digitized by

Google

H
O

H
<

>
OH

u
(/)

QQ
O

u
<

o
<

i86

Digitized by

Google

MOUNT MARTIN

187

Silent as ghosts we stood, until one of
us, caught by the ridiculous attitudes of
his companions, burst out laughing, set-
ting us all a-snickering behind our hand-
kerchiefs like school boys who fear the
teacher's wrath, for we durst not lift our
mufflers.

Thus we waited for nearly half an
hour, but there came no further op-
portunity to proceed. We could see
nothing around us. So we decided to
give it up, and after picking up a few
specimens of the rocks immediately
around us (ordinary andesites) we
scrambled back to the rim.

RETURNING THROUGH THE LOWERING
CLOUD

Here again we sat down and waited
for an hour. But there was not the
slightest sign of a let-up; so, reluctantly,
we started for camp. The momentary
break that had come while we were
eating was the only near-by glimpse of
the crater we were to have.

Climbing up a dirt slope at an angle of
60 degrees is as nothing compared to
wriggling down again. There was im-
minent danger of dislodging a loose
, bowlder to fall on one's companions, be-
low, and there were places where we
simply had to "let go** and fall off, trust-
ing to luck not to hit anything hard be-
fore stopping.

Regaining the snowfields, we found
that the cloud cap had settled much

farther down the slopes, and we had to
descend several hundred feet before we
came out into clear day again.

We had not turned back a minute too
soon. Of that we needed no second
hint. We all knew that to be caught in
a fog on that crevassed glacier, where we
could neither follow the route by which
we had come nor choose a new path,
might mean disaster.

As we worked our way back and forth
among the crevasses, we got into an
area of snow bridges which made even
Charlie, with his long experience on
glaciers, hesitate. He kept protesting
that "he didn't like that ice,'* and we
went forward with great caution.

But no better way was to be found,
and the certain menace of the thickening
clouds compelled us to take the uncertain
risks of rotten snow bridges and go
ahead regardless of danger. "All's well
that ends well," however, and finally we
crossed the last crevasse, reaching the
solid floor of the Valley again without
accident.

There was still a long four-mile drag
across to camp, but we made it before
half past eight, tired and disappointed,
but yet with the knowledge that we had
succeeded in a difficult feat, even though
we had been denied the fruits of our labors.

And this, the sixteenth of August, had
been the best opportunity to climb
Mount Martin since the twenty-fifth of
July!

Digitized by

Google

■n
O

C/3

S "
O E

O o
§ .2

_ ^

.c
a
o

l88

Digitized by

Google

DISCOVERY OF THE TEN THOUSAND SMOKES

July 31, 19 1 6, dawned clear and
bright as had the day before, when we
had made the second ascent of the vol-
cano. But this time the steam from
Mageik drifted off to the northwest,
and small clouds were beginning to
gather on the west side of the valley, so
we knew it was to be the last day of
good weather.

We had originally planned a two-
days' trip across the Pass, to see if we
could find the source of the clouds that
had aroused our suspicions. But, re-
membering the bad name given Katmai
Pass by Spurr (see page 268) , who states
that it was impossible to cross it except
in good weather, and fearing to be
caught short of provisions on the wrong
side, we gave up the projected trip and
decided merely to reconnoiter.

Planning to make an easy day of it —
for we were tired after our climb to the
rim the day before — we climbed around
the shoulder of Observation Mountain
and descended into the upper valley of
Mageik Creek.

Here we found the most striking ac-
cumulation of ejecta we had yet encoun-
tered — a quantity of fine material which
resembled the Katmai Mud Flow in
its massive non-stratified character, al-
though it was much more extensive and
thicker, having weathered out into bad
lands with bluffs nearly 100 feet high.

ASCENT TO KATMAI PASS

After stopping a little while to study
the flow and to eat our lunch, we made
our way across the bad lands toward
the Pass, following now the ridges of
the flow, now the bottom of the canyon,
which rose at a gentle grade.

As we ascended the valley past the
highest peak of Trident we came within
view of the hollow between it and the
next peak, from which we had thought

several times we saw clear, indications
of rising steam. The sun illumined it
brightly, and there was not the smallest
puff of steam anywhere.

We were now up 2,500 feet, only a
little lower than the Pass, and could
see a long way through it. No steam
was visible there, either (see map, page
68.)

So again we concluded, as we had the
day before, that we had seen nothing
more than the ordinary clouds which
gather so easily around the summits of
all high mountains.

Church, incapacitated by too many
flapjacks at breakfast, gave out, so
Folsom and I left him with the packs,
much against his wishes, while we went
forward a little farther to see what we
could discover.

Traveling transversely across the gul-
lied bad lands of the mudflow, a route
made necessary by the condition of the
canyon below, was very laborious, and
before long I was ready to turn back,
satisfied with having seen through the
Pass and, as I believed, having laid an-
other ghost.

THE FIRST FUMAROLE

But just as I was about to suggest
returning, I caught sight of a tiny puff
of vapor in the floor of the Pass. I
rubbed my eyes and looked again. Yes,
there it was, a miniature volcano send-
ing up a little jet of steam right in the
Pass.

The very smallness of this steam jet
made it of as much interest as a large
volcano (see page 192), for we had been
continually surprised at the absence of
manifestations of secondary activity —
parasitic cones, fumaroles, mud craters,
hot springs, and the like — features al-
most invariably associated with so great
an eruption.

189

Digitized by

Google

igo

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by L. G. Folsom

WARMING MY HANDS AT THE FIRST LITTLE FUMAROLE IN THE PASS

These little sentinels of the Pass continued unchanged- through 191 8, but had gone out in 1 919. It was
the sight of these which led to the discovery of the Valley.

Thus far we had seen absolutely no
indication of signs of recent volcanism
outside Katmai Crater, except in the old
volcanoes Mageik, Trident, and Martin.
These, however, were singularly far re-
moved from the main crater, and we
had been at a loss to understand why
activity should appear at such widely
separated points, with no intermediate
vents.

Earlier in the day we had found the
streams from the hot springs near the
Pass, which continued as described by
Spurr before the eruption, except that
the water now was not nearly so hot as
stated in his account (see page 269).
Aside from that, this fumarole was the
first thing of its sort to be observed.

When we reached the Pass we found
its floor all shot through with cracks
and small fissures from which issued
half a dozen good-sized jets of steam
and perhaps a hundred small ones.

With some trepidation we approached
over the fissured surface. Most of the
steam escaped from small openings a few
inches in diameter, whence it came with

considerable velocity, giving forth a low,
roaring sound. We could go quite close
and warm our hands in the steam, which,
though very hot as it emerged, soon
cooled, like the moist vapor from a tea-
kettle.

Coming off with the steam were vari-
ous substances which gave rise to curi-
ous evil odors and precipitated a highly
colored crust on the ground. Prom-
inent among these was hydrogen sul-
phide, with its characteristic **rotten
egg** smell, and sulphur, which gave a
yellow tinge to the parti-colored subli-
mates of the crust.

I was anxious to return to Church,
for we had alreajdy been gone much
longer than we expected, and had
reached a little eminence, when, turn-
ing round to urge Folsom to hasten
back, I saw far down the valley, over
the top of some rising ground beyond
us, a puff of steam.

This had not been there when we
came over the Pass, and was evidently a
good deal larger than the jets we had
been examining. As the obstructing

Digitized by

Google

DISCOVERY OF THE TEN THOUSAND SMOKES

191

Photograph by D. B. Church

THE TRAIL UP TO KATMAI PASS

The way led over permanent snowdrifts, for the snow had been covered with ash during the erup-
tion of Mt. Katmai, and this coating protects the drifts from the sun's rays. All the supplies for the
camp had to be carried on the backs of the men through this pass into the valley.

hill was not far away I decided, late as
it was, to go forward to "have a look."

DISCOVERY OF THE STEAMING VALLEY

The sight that flashed into view as
we surmounted the hillock was one of
the most amazing visions ever beheld by
mortal eye. The whole valley as far as
the eye could reach was full of hundreds,
no thousands — literally, tens of thou-
sands — of smokes curling up from its
fissured floor.

From our position they looked as
small as the little fumaroles near by, but
realizing something of their distance we
knew many of them must be gigantic.
Some were sending up columns of steam
which rose a thousand feet before dis-
solving.

After a careful estimate, we judged
there must be a thousand whose columns
exceeded 500 feet. A dozen miles away
the valley turned behind a blue moun-

tain in the distance. Plainly the smokes
extended that far. How much farther
we could not tell.

The first glance was enough to assure
us that we had stumbled into another
Yellowstone Park — unseen and un-
suspected by white man and native
alike until this hour.

I tried to "keep my head** and ob-
serve carefully, yet I exposed two films
from my one precious roll in trying for
pictures that I should have known were
impossible. For a few moments we
stood gaping at the awe-inspiring vision
before us, then plunged down to get a
nearer view.

It was as though all the steam en-
gines in the world, assembled together,
had popped their safety valves at once
and were letting off surplus steam in
concert.

Some of the fumaroles were seen to
be closely grouped in lines along common

Digitized by

Google

192

THE VALLEY OF TEN THOUSAND SMOKES

fissures; others stood apart. The big-
gest of all, the one whose steam had first
caught my eye, stood well up on a hill-
side, in a nest of fissures, reminding me
of the crevasses of a glacier, and was
big enough to be plainly visible though
three miles away.

Fortunately a strong wind was blow-
ing through the Pass, carrying all the
fumes down the Valley and away from
us, or we might not have dared to go
on. In addition to the active fissures
there were thousands more, inactive at
the time of our visit, which had en-
crusted the ground round about with
colored deposits like the others.

Even from the hasty examination we
were able to make on the day of dis-
covery, we could see that the vents
must be counted by the tens of thou-
sands. And from this first impression
came the name, The Valley of Ten
Thousand Smokes.

CHARACTER OF THE VENTS

In some cases the steam issued from
a large deep hole; in others there was
no opening at all — the vapors simply
escaped through the interstices of the
soil particles.

There was no relation between the
size of the vent and its output. Some
of the largest jets had very narrow
throats, while from some cavernous
holes there issued only faint breaths of
steam. In many cases steam poured
from the sides of the drainage gullies,
where it did not break through the more
compact surface layer of sand.

The ground was hot in places where
no visible exhalation was being given
off, as we found to our dismay when we
sat down to rest on a bank seemingly
safe enough.

Fearing to ruin our shoe leather we
dared not enter any areas except those
which appeared moderately cool. But
even with such precautions the heat in
many places penetrated our thick soles.

I wish my vocabulary were adequate
to describe the curious mixture of foul

odors which the "Smokes" gave forth.
Mixed with the omnipresent sulphurous
gases were others which had a strangely
organic smell, recalling at once burning
wool, the musky smell of a fox den, and
the odors of decay.

We could not tell to what extent, if
any, odorless asphyxiating gases, such
as carbon monoxide, might be present
in the complex. We did not experience
any ill effects from the fumes, but we
took good care to keep to windward
most of the time.

ENTRANCE TO THE MAIN VALLEY

A mile and a half from the Pass we
came out from between the two lava
cones ^ that guard the main entrance,
into full view of the Valley. It was
clear that the fumaroles stretched along
the range all the way from Katmai to
Martin, linking into one huge area of con-
tinuous activity the widely separated
volcanoes, which had hitherto appeared
as isolated vents.

We did not take time to examine con-
ditions to the west, for we beheld in the
opposite direction a prodigious steam
column from a volcano which we later
named Novarupta.

As we drew nearer we saw that the
main body of this steam was rising from
a central mass of rock (stiffened lava),
surrounded by a comparatively low
crater ring, evidently composed of ma-
terials thrown out of the vent. The
surface of the cooling lava plug was
covered in most fantastic fashion with
sharp angular fragments, which on
closer inspection proved to be rocks sev-
eral yards in diameter. These had been
broken off by the too sudden cooling of
the molten magma, just as a piece of
hot glass is shattered if suddenly plunged
into cold water (see page 276).

AN INTERPRETATION OF THE FUMAROLES

Even the hurried observations we
were able to make were sufficient to
bring out clearly, in its larger outlines,
the significance of the Valley. With
^ Named later Cerberus and Falling Mountain.

Digitized by

Google

Natural Color Photograph by Frank I. Jones
MOUNT MAGEIK, WITH CERBERUS CROUCHING AT ITS FOOT, STANDS GUARD AT THE

HEAD OF THE VALLEY
The sloping "high sand mark" comes down from the Pass across the base of Cerberus.

Digitized by

Google

Natural Color Photograph by Frank I. Jones

THE OVEN WHERE CORN BREAD WAS BAKED BY NATURE'S FIRES
If left too long it would quickly burn to a cinder over the greatly superheated steam issuing here.

Digitized by

Google

DISCOVERY OF THE TEN THOUSAND SMOKES

193

the very first glance it was evident that
although little fumaroles in the Pass
might be secondary manifestations, the
great steamers of the Valley must be
true volcanic vents. They were alto-
gether too big and too hot to be anything
else.

The long fissures everywhere so much
in evidence were apparently the outlets
of fractures in the bedrock through which
the emanations from a mass of intruded
magma were escaping to the surface.

The gases rising through these funda-
mental fissures and seeping through the
mass of accumulations from recent
eruptions evidently found their vent in
the myriad fumaroles, very much as
many small leaks will break out on the
surface of an old bicycle tire from a
single puncture of the inner layer of
rubber.

We were surprised to observe that the
main trend of the activity was at right
angles to the chain of volcanoes. We
would have expected exactly the reverse,
for in any general rise of the range, such
as might have accompanied the eruption,
the fracture lines would run parallel with
the axis.

Another remarkable feature was the
fact that so far as we could see the
fumaroles were strictly confined to the
low ground of the valleys and were not
breaking out on the mountains as well.
This indicated that whatever the nature
of the intrusion below, it must have come
so close to the surface that the position of
the fractures was determined by rela-
tively small differences in surface relief,
i,e. in the thickness of the roof above
the magma.

The great distance to which the
fumaroles extended from the axis of the
old volcanoes — 15 miles, as we found by
later surveys — further indicated that
the subterranean magma must have a
considerable extent.

It was late, and as we had now seen
all that could be observed without ex-
tended exploration, we turned our steps
homeward and hurried to rejoin Church,
who had shivered in the cold wind for
I*

five hours, even with all our extra
clothes. Once across the gullies, which
were more than ever a terror to us now
that we were nearly exhausted, we made
good speed back to camp, which we
reached a little after ten o 'clock.

We had expected, of course, to outfit
and return to our Valley of Ten Thou-
sand Smokes for further exploration,
and especially for pictures which would
prove our story. For, on the day of
discovery, so little did we anticipate
what was before us that we had left the
photographic outfit behind with Church,
and had only a kodak with a single roll
of film.

But the river had risen so fast under
the influence of the warm weather that
we dared not remain longer on the far
side. We decided it was better to save
what pictures we had than to try for
more at the risk of not getting any of
them out.

REFLECTIONS ON THE NIGHT
OF THE DISCOVERY

Sleep that night was impossible. I
had as yet only a very inadequate con-
ception of the place we had discovered,
but I had seen enough to know that we
had accidentally discovered one of the
great wonders of the world.

The responsibilities and the oppor-
tunities of our position came over me in
a flood of problems that I could not drop.
Here were we, alone of all the world,
possessed of the knowledge of the exist-
ence of this marvelous place.

I kept wondering how the discoverer of
the Yellowstone felt when he first beheld
the spout of Old Faithful. I remem-
bered Jim Bridger, and his stories — so
incredible that no one would believe
them, yet so phrased as to prove in-
dubitably to posterity his familiarity
with phenomena which by that time had
become well known.

I remembered how the editor of a
prominent Western paper had refused
to publish any of "Jim Bridger's lies."
It was sweet to recall that finally, years
afterward though it was, that editor had

Digitized by

Google

.2
2

< X

9 **
Q g

« 1

O "^
te §
« o

u .

CO ^

< 3

e >

194

Digitized by

Google

DISCOVERY OF THE TEN THOUSAND SMOKES

195

publicly apologized for the injustice he
had done the great pioneer.

And when the reality of the marvels of
the Yellowstone was definitely estab-
lished everyone believed them hopelessly
inaccessible. Yet what has happened?
Anyone may visit the Yellowstone,
surrounded by all the luxuries of modern
civilization and within the lifetime of
Jim Bridger too!

Could I hope for as much? Would
our story be believed when we got back?
Or would we be regarded as modern
Mimchausens? I did not underestimate
the difficulty of the task that had been
set me by that day's discovery.

I recognized at once that the Katmai
district must be made a great national
park accessible to all the people, like the
Yellowstone. To make it known, to
have it set aside as a National Park, and
to secure the means necessary for its
development would, I foresaw, require
a tremendous amount of effort.

In the* years that have followed I
have had occasion many, many times to
be thankful for the good fortune which
had sent me out under the auspicies of
the National Geographic Society. For
in all the world there is no other organ-
ization so well fitted to bring such a dis-
covery to the knowledge of the public.

The task of making the place known
and of getting it set aside as a National
Monument proved much easier than I
dared hope, as I lay awake that night.

The President of the National Geo-

graphic Society did not refuse credence
to our account and deny us publicity.
On the contrary he was quick to grasp
the significance of our discovery, and
did all in his power to assist further ex-
ploration and to stimulate interest in the
reservation of the district as a part of
our great National Park System.

Many times also have I been thankful
for the tools which modern invention
had placed in our hands. With kodaks
and halftones and motion pictures to
record our discoveries, our advantage
over the old-time explorer, who could
supplement his account with nothing
better than sketches, was enormous.

Along with the difficulties, real and
imaginary, that plagued me there was
one practical problem which from that
night assumed a large place in my mind.

How was the public ever to reach this
region? Where was a landing-place to
be found on that dangerous and in-
hospitable coast? For it went without
saying that no great number of people
could land through the surf, as we had
done; that before it could become the
great National Park it was to be, a har-
bor must be found somewhere along that
difficult shore.

A study of the chart next morning
held out little hope. But fortunately
for our quest, the coast was very im*
perfectly charted and we discovered
afterward a fine harbor within easy dis-
tance, lying in an area shown on^the
charts as dry land.

Digitized by

Google

•Si

£ 4>

a -5

2.S
g.S

v- c
.2 3

n >

cs O

2 88

O
O
U

OJ3
O «

H
H

o a>

3^
O «
«^

•S p

o *-

§*

o ♦-

«5

IS o«

196

Digitized by

Google

XVI
EXPLORATIONS IN THE TEN THOUSAND SMOKES IN 1917

Anxious as we were to see more of our
wonderful Valley of Volcanoes, condi-
tions compelled us to leave with only the
scantiest evidence to substantiate the
story of our discovery.

Fortunately the pictures secured from
our one roll of film were enough to prove
that our tales were not altogether men-
dacious, and the Board of Managers of
The Society proposed to send us back in
191 7 equipped to explore the place in
greater detail.

It was with great impatience that we
awaited the slow course of the interven-
ing months. After we had landed, the
days required to pack our supplies up
Katmai Valley to our camp at the foot of
Observation Mountain (see map, page
68) seemed weeks of waiting. But we
thought it necessary, before attempting
to cross, to accumulate sufficient stores
at the foot of the Pass to insure the
maintenance of the party in the Valley.

Finally my impatience would brook no
further delay, and I decided to celebrate
the Fourth of July by a preliminary ex-
cursion across the Pass, to have one look
at the place. So, luring Walter along
with the prospect of a bath in the hot
springs below Trident, I started out,
without announcing my intentions.

SMOKES UNDIMINISHED AFTER

A year's interval

When we reached the Pass I saw at
once that everything was just as it had
been the previous year. There were the
two little fumaroles which we had first
discovered, steaming away exactly as
they had the year before.

This was decidedly reassuring, for I
had been tormented with the fear that
we had seen only a passing'stage in the
declining activity, and that after all the
time and effort spent in preparation for
the' expedition, we might find the Valley

dead when we returned, and all its vol-
canoes things of the past.

When I got back to camp and reported
the conditions, I found that some other
members of the party had secretly been
entertaining the opposite fear — that
the whole Valley might blow up sud-
denly while we were in it!

On the contrary, all that we have seen
indicates that the activity of this dis-
trict, like that of the Yellowstone Park,
has reached a stable stage which will
continue without much change for a rela-
tively long period.

Wherever we went, conditions were
the same. All the vents, big and little,
are remarkably constant in their ac-
tivity. And now that the district has
been made a National Monument, this
constancy assumes a twofold interest and
importance. In the first place, its won-
derful sights are likely to continue for
the admiration of tourists for years to
come. Secondly, there is little likeli-
hood of a sudden return to a condition
of explosive activity.

EFFECT OF THE VALLEY ON AN
UNEDUCATED NATIVE

Walter, who had been keeping camp
on the day of the discovery the year be-
fore, had listened to our accounts of its
wonders with polite incredulity. I was
anxious to watch the effect of the Valley
on an uneducated native, with no scien-
tific interest to give significance to the
sight.

When we had examined the little fu-
maroles in the Pass and had looked at
the dozen or so of others round about,
he turned to me with an air of, "Well, I
thought so all the time,'* and asked,
"And is this the Valley of Ten Thousand
Smokes?"

"Oh! No!*' I answered, "that is over
the rise yonder.*'

197

Digitized by

Google

4^
V

o
S

si
s^

&S
Sl

-I

CQ V

6:5
*- («

s is

Si*

198

Digitized by

Google

EXPLORATIONS IN THE TEN THOUSAND SMOKES IN 1917

199

When we reached a point from which
we could see on down the Valley, his
amazement was unbounded. *'Why," he
exclaimed, **a whole big valley all full of
smoke!"

I had planned merely to look in and
turn back, for we had come a long way —
so far that on the way up he had fre-
quently remarked how far we had come
and how fast I was walking, and had
even asked to rest.

But once he caught sight of the Val-
ley, he must go on. It was my turn to
call a halt now, for I was thinking of the
long way back to camp. But before I
could stop him, he had gone a couple of
miles beyond the Pass.

He came home with shining eyes, very
much excited, and though very tired he
kept talking to Andrean about the won-
der he had seen until late into the night.
How I wished I could understand his
Russian!

CROSSING THE MOUNTAINS AGAINST A
HAIL OF PUMICE

It was four days later before all was in
readiness for the whole party to go over.
The wind, which had been blowing un-
comfortably hard for several days, fresh-
ened during the night until it began to
carry away our dishes. The gauge in the
sheltered nook we had selected for our
camp showed a velocity of 25 miles an
hour.

Out on the mountain it was blowing
twice as hard, directly in our faces, lift-
ing us off our feet at times. But worse
than the wind itself was the hail of sharp
pumice which it raised. This cut like a
knife whenever it struck our flesh. The
others protected their eyes with close-
fitting goggles; but I could not avail my-
self of that relief because of the necessity
of finding the way.

Fortunately these conditions lasted
only round the shoulder of Observation
Mountain, and from there on the going
was comparatively easy. This was by
no means an extreme wind, however.

On another occasion the men bringing
up the supplies were unable to make it

and had to turn back. The wind gauge
at the sheltered camp that day regis-
tered 60 miles an hour steadily, and
much higher during the gusts.

MANY MORE THAN TEN THOUSAND VENTS

When the party reached the Valley,
they were even more impressed with the
sublimity of the spectacle than Walter
had been. None had imagined anything
so wonderful. All agreed that no de-
scription could convey any conception
of its immensity and grandeur.

I found that my matter-of-fact chem-
ist was counting the smokes, to see
whether I had been justified in asserting
that there were 10,000 of them ! He soon
announced that I was quite well within
the truth.

There were certainly several times ten
thousand to be seen, even on a clear day;
and when the weather was moist myr-
iads more appeared, for then the smoke
from the millions of little holes, whose
gases were ordinarily invisible, con-
densed until there were a thousand times
ten thousand.

One member of the party, who had
traveled considerably and found many
of the sights of the world overdrawn, had
been somewhat skeptical in advance
about the Ten Thousand Smokes.

When once he felt its thrall, however,
he repeated over and over again, **Why,
you couldn't exaggerate it!" This state-
ment is perfectly true. Although the
statistics of length, area, and the like
could be falsified, the enlarged figures
would no more convey any idea of the
stupendous number and size of the
Smokes than do the real ones.

It can not be described. Only after
one has spent many days exploring the
Valley and realizes how many active
areas still remain unseen, does one begin
to get an idea of its magnitude.

No amount of experience seemed suffi-
cient to enable us to grasp the propor-
tions of the place, for as long as we re-
mained we habitually underestimated, in
spite of ourselves, its distances and the
size of its fumaroles. As we moved to

Digitized by

Google

o
c

-o

'5.

O
H

pes

•s

IS!
•S "

!•-

^ s

C
3

200

Digitized by

Google

EXPLORATIONS IN THE TEN THOUSAND SMOKES IN 1917

20 1

any corner of the Valley and came up to
what we had supposed from a distance to
be little fumaroles, we found instead
monster vents, each group a more won-
derful spectacle than the whole seen in
panorama, so much larger was the area
of activity than our powers of compre-
hension.

OVERAWED BY THE WONDERFUL
VALLEY

No later visitors can ever experience
the sensations that came over us as we
prepared to camp for the first time within
its confines. They will know in advance
that it is safe, but we could not tell what
dangers might lurk in its fumaroles.

Our feeling of admiration soon gave
way to one of stupefaction. We were
overawed. For a while we could neither
think nor act in normal fashion.

It was a situation calculated to instill
dread. Would we wake again, or would
that insidious poison, carbon monoxide
— colorless, tasteless, odorless — attack
us and render us incapable of the exer-
tion of climbing out, even if we did
waken? We knew that it must be ema-
nating from the fumaroles in considera-
ble quantities and that it was deadly in
concentrations as low as one- twentieth of
one per cent. It was one thing to go into
this *' Death Valley" with all our facul-
ties alert to ward off danger, but quite
another to lie down and allow ourselves
to lapse into unconscious helplessness in
such a place.

I do not know whether the others were
as worried as I. If so they, like myself,
kept silent. But I must confess that
when it came to lying down I chose for
my bed the place next the window,
where I would get the best air possible.

How I was paid for that selfishness!
It was the windward side, and through
all the fierce storms that followed I lay
and took the brunt of it. The tent al-
ways leaked first on my side. My bed
was always the first wet, my position
the coldest and most uncomfortable.

No one slept much the first night; but
as nothing happened to justify our fears.

and dawn found us all hale and in good
spirits, one great bugaboo had been laid.
It was a long while, however, before we
could regain anything like a normal men-
tal attitude.

At night, as I lay in my blankets with
the sight of the Valley cut off, I would
upbraid myself for not having accom-
plished what I had planned, and make
out a list of things I wanted to do the
next day; but when the morning came I
could not move myself to action. I
could only look and gape.

Shipley, our chemist, was easily the
most self-possessed member of the party.
But for him we might have turned around
and come home without the scientific
material we had gone to collect.

After all, the whole Valley is very
much of a gigantic chemical laboratory,
and perhaps that acccounts for his
greater command of himself. Yet, on
the third day, even he remarked that he
"did not feel like monkeying with his
little bottles of chemicals."

One of the boys was frankly scared to
death. He did what I told him, but
when not at work sat in a dull-eyed
stupor, like one at the funeral of his
sweetheart, and from this state no efforts
of ours could rouse him.

I can only guess the effort it must have
cost him to go up to the fumaroles and
get pictures of them. He said himself
that he expected to go crazy before he got
out again. When relieved of his duties
and sent down to the lower camp he
quickly regained his nerve, and in the
end got as good control of himself as any
of us.

I was utterly unprepared for the feel-
ings which thus overcame us. At the
time of the discovery I had not been able
to stay long enough to get beyond the
first sensations of wonder and admira-
tion. I had by no means grasped the
situation sufficiently to report it ac-
curately.

This region should have been named
"The Valley of a Million Smokes," for
there are certainly not merely thousands,
but millions of them all told.

Digitized by

Google

ca *"*

'C o

« e
a «

l«

o s:

Digitized by

Google

EXPLORATIONS IN THE TEN THOUSAND SMOKES IN 1917

203

A large factor in my feelings was plain
fear; perhaps I ought in honesty to say
cowardice. The Smokes were so much
bigger, so much hotter, so much more
numerous than I remembered them that
I was badly scared by the job I had
undertaken.

THE FEAR OF CAVE-INS AND FUMES

As we explored the margin of the
Valley (the worst place, as we after-
ward found), we could plainly hear the
ground ring hollow beneath the ''tunk'*
of our staffs, and more than once we felt
it shake under our blows.

What if it should suddenly give way
and precipitate us into a steaming cal-
dron? Occasionally a breath of the
steam blown around us by a chance
breeze gave an uncomfortable burn. We
knew that if once a man fell into such a
place he would be instantly parboiled.

At first we roped up as for mountain
climbing, and spread out, so that if one
man went through the others could pull
him out. When we came better to un-
derstand conditions, however, we dis-
carded the ropes, having decided that if a
man once broke through it would be
more merciful to leave him than attempt
to pull him out.

We had been assured by the best au-
thority that there could be no danger
from the fumes. Nevertheless, I brought
along a chemist to keep us out of the
dangerous places, for I knew this Valley
to be unlike any other place in the world
and reasoned that there could be no real
basis for the assurances given.

Our apprehension was, to be sure,
greatly relieved after passing the first
night safely, and our fears diminished
with each day. There was no telling,
however, what we might encounter as we
extended our observations into new areas.

Had we known, for example, that there
was enough hydrofluoric acid in some of
the fumes to etch the inside of the glass
tubes in which they were collected, we
should have been still more worried, for
even a small dose of hydrofluoric acid is
dangerous.

But, like most bugaboos, these dan-
gers proved much less real than our im-
agination pictured. Experience showed
that there was always plenty of air to
breathe, and we found no insidious gases
strong enough to strike us down without
warning. Our noses always gave ample
notice of dangerous places.

We suffered no injury beyond slight
headaches and irritation of the mu-
cous membranes caused by the corrosive
gases.

LEARNING TO TRAVEL SAFELY

So also with cave-ins. As we grew
familiar with conditions we soon learned
to pick our way about safely. The sub-
limations brought up by the fumaroles
themselves so encrusted their throats
and the ground round about that a thin
roof over a cavern would support a man
with safety.

The most dangerous places were those
where fissures had been bridged over by
ash and mud, so as to leave nothing to
indicate their presence. After we had
been in the Valley several days we had
some experiences with such "crevasses"
that probably would have turned us
back, had they occurred when we first
arrived.

Several times when we accidentally
put a foot through a thin place in the
crust, steam would come spouting out
of the hole, forming a new fumarole.
But it was always one foot only, and
the owner did not take long to back
out.

Once, while walking across a place
that looked perfectly solid, I noticed a
new hole midway between two old fis-
sures and, on investigating, found that a
steaming fissure two feet wide and ten
feet deep was roofed over for 50 feet by a
layer of mud so thin that I could per-
forate it anywhere by a slight thrust of
my ice ax.

But such experiences rapidly led to a
sort of technique like that of the moun-
tain climber, by which we could judge
the ground. It was, moreover, some-
thing like one's boyhood "banters" on

Digitized by

Google

204

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by L. G. Folaom

THE MOUTH OF A VENT IN THE SmE OF A GULLY

The entomologist with his bug net seems incongruous in such a place, but around some of the vents
are thousands of dead insects, killed by flying into the hot steam.

thin ice. After the first fellow had made
it anybody was willing to follow.

If, like the actuary, we may judge the
risk by the mortality, the Valley is far
safer than Broadway, for never in the
course of the four expeditions which en-
tered it was there the slightest accident
to mar our success.

To be sure, one could easily walk into
a volcano if he were careless, but the very
character of the Smokes engenders a

respectful attitude on the part of the
spectator, so that there is little if any
real likelihood of accident.

There is no doubt that, with ordinary
precautions, the Valley can be made as
safe for tourist travel as the Yellowstone
Park.

In many places the ground around the
vents is covered with a peculiar blue
mud, thinly coated with a chestnut-
brown crust, which sometimes supports

Digitized by

Google

EXPLORATIONS IN THE TEN THOUSAND SMOKES IN 1917

205

one and sometimes gives way suddenly,
letting one down to his shoetops in the
soft scalding mud beneath. At such
times one is apt to feel that in all verity
his "feet are taking hold on hell," par-
ticularly if the place happens to look
"ticklish" otherwise.

EFFECT OF THE FUMAROLES ON SHOES

Sinking deep into scalding mud would
appear to be very hard on shoe leather.
Many have asked me what sort of foot
gear we used. We wore our ordinary
shoes. The ash is such a poor conductor
of heat that even where it is hot enough
to melt zinc a few inches underground
the surface is little above the air tem-
perature.

It is only when the surface is covered
up, confining the heat, that the tempera-
ture rises. Consequently, one may walk
— so long as he keeps moving — over
ground hot enough to burn up his shoes
in short order if he stood still, without
any marked sensation even of unusual
warmth.

Once, when we were careless enough
to leave a packsack lying near a hot
fumarole, we found its straps burned up
in a few minutes. But our shoes had
suffered no injury.

As for the patches of scalding mud, one
naturally gets out of them with as little
delay as possible. I fully expected, how-
ever, that the soil would be so impreg-
nated with acids as to eat up our shoe
leather with great rapidity, but fortu-
nately there is here little or none of the
free sulphuric acid prevalent in some
volcanic districts.

Our shoes lasted far better in the Val-
ley than on the pumice plains outside.
Indeed, the pair in which I worked in
191 7, with a new pair of soles, went back
into the Valley in 1919, and, again re-
soled, are still good for a trip or two more.

COOKING AT A FUMAROLE

We chose for our camp a site 250 feet
above the Valley floor close beside a
melting snowdrift. Here, although we
were denied the pleasure of a camp fire

— for not a stick of wood remains any-
where in the upper Valley — we had
** all the comforts of home." Fifty yards
to the rear was our refrigerator, where
we could keep everything freezing cold
until needed (see page 198).

Just in front was our cookstove — a
mild-mannered fumarole — into which
we hung pots to cook our food. We were
somewhat dubious beforehand as to the
feasibility of this method of cooking, be-
cause of the noxious gases that came off
along with the steam; but the results
were more than satisfactory. We never
detected the faintest taint in any of our
food.

Everything was always done exactly
right. Since the pots were surrounded
by an atmosphere of live steam just at
the point of condensing, nothing ever
boiled away, cooked to pieces, or burned,
no matter how long neglected or for-
gotten.

There was only one drawback; while
we were in the Valley we had to do with-
out our old stand-bys, bacon and flap-
jacks, for the *' stove** we used that year
would not fry. There were, however,
many vents in the Valley quite hot
enough to fry bacon, as we found to our
satisfaction in 1919 (see page 294). But
in 191 7 we were too much afraid of the
big vents to try cooking in them.

We could not even measure their tem-
peratures, for supposing in advance all
the fumaroles to be merely '* steamers*'
at the boiling point we had come en-
tirely unprepared for high temperatures.
Our thermometers did not read high
enough, so we had to content ourselves
with guesses until the following year,
when Sayre and Hagelbarger, provided
with pyrometers supplied by the Geo-
physical Laboratory of the Carnegie
Institution, made a special study of tem-
peratures. Thus equipped, they meas-
ured temperatures ranging all the way
from boiling up to 432° C. (810° F.),
hot enough to melt zinc^ (see page 221).

^ For a detailed account of the temperature
observations in 191 8, see Ohio Journal of Science,
vol. 19, pp. 249-278.

Digitized by

Google

a
£

J! .
ft. "^ o

< Ml

^ g s
- S^

n c V

c
o

3
u
0)

S
I

■4-»

.s
&

206

Digitized by

Google

EXPLORATIONS IN THE TEN THOUSAND SMOKES IN 1917

207

When we turned in the first night we
were astonished to find that the ground
under our tent was decidedly warm.
On examination we found that a ther-
mometer thrust six inches into the
ground promptly rose to the boiling
point. This was indeed a surprise, for
the place had been vacated only recently
by the retreating snowbank behind us.

A STEAM-HEATED TENT

We put most of our bedding under us
to keep us cool !

Before long our blankets were as hot
as the ground. Close to the snowdrift
as we were, and at an altitude of about
2,300 feet, the air was at times decidedly
cold ; so while we steamed on one side we
froze on the other and had to keep turn-
ing over and over in an effort to strike a
balance. Everybody expected to *' catch
his death of cold.*'

After a little we discovered that our
beds were not only hot, but wet from in-
visible vafKjrs which were everywhere
seeping up through the soil. The con-
densation of the steam soon saturated
our blankets, so that by morning we were
in a most curious case. The sensations
that greeted us on awakening in these
warm, wet beds can in justice be com-
pared only with certain distressing mem-
ories of one's childhood days.

This state of affairs worried us very
much indeed, for such conditions were
the worst possible for the films on which
we were depending to vouch for our
story.

By building a sort of crib with the
walking sticks we had brought from the
lower camp, we managed to keep them
off the floor, and thus reasonabl>^ cool.

Our instruments also took up water
and swelled, so that we feared we should
lose everything. A tripod that had suc-
cessfully stood the climate of a tropical
rain forest jammed so hard that it could
not be hammered loose. The cameras
swelled until their focal points shifted.

A panoramic outfit uix)n which high
hopes had been built refused to work and
was altogether useless for the rest of the

summer. It is because of this failure
that the frontispiece of this book shows
only the main arm of the Valley. After
many trials and much jogging, we got it
to rotate as far as Baked Mountain, but
there it stopped dead.

As I saw everything thus rapidly soak-
ing up moisture, I was very much dis-
turbed lest we should be caught in the
rain; a fumarole might be an ideal cook-
stove but it was no good to dry clothes
by. With the tent steamy also there
would be absolutely no way of drying
out when once we got thoroughly wet,
as was sure to happen before many days.
Transportation was so difficult that we
had brought no change of clothing.

But in these fears I was most happily
disappointed. We found that though
everything in the tent was soon suffused
with moisture, nevertheless anything
brought in dripping wet from outside,
being supersaturated, quickly became
merely damp.

When we came in soaked through and
chilled after a ducking, therefore, we
merely crawled into our blankets. After
a while both clothes and bedding would
become as *'dry" as when we started.

THERAPEUTIC POSSIBILITIES
OF THE VALLEY

In spite of the expxjsure to which we
were daily subjected, there was not a
symptom of *' colds" or other illness in
the party.

On the contrary, the constant steam-
ing seemed a good treatment for the
rheumatic pains which usually develop
on such expeditions. During our stay in
the Valley, and for some time thereafter,
we were as free from such aches as if we
had taken the '*cure" at a hot spring.

We had no physician along, nor did we
have time for any experimenting, but
one does not have to be very well versed
in medical lore to see the therapeutic
possibilities of such a place. The healing
powers of thermal waters are a matter of
common knowledge; the thousands of
patients who annually journey to such
places sufficiently attest their efficacy.

Digitized by

Google

208

THE VALLEY OF TEN THOUSAND SMOKES

The hot mud, which everywhere
abounds, may possibly excel more than
one hot spring in its healing properties.
The virtues of a clay poultice are well
known to us all, beginning with the mud
plasters we used to employ when as boys
we were stung in digging up a bumble-
bees' nest, and ending with the high-
sounding '*antiphlogistene" poultice we
now purchase at the drug stores to relieve
bruises and inflammation.

One's imagination can readily picture
the benefits that might accrue from
the scientific employment of the hot
muds of the Valley in the treatment
of all disorders that yield to external
applications.

KATMAI PASS A NOTORIOUS
WEATHER-BREEDER

We came at length to appreciate
greatly our steam-heated tent, for we
found it always warm and comfortable,
and there were times when the driving
wind and rain outside were so bitter that
we could scarcely have endured the place
otherwise.

It would be a mistake, however, to
suppose that with all our conveniences
life in the Valley was wholly ideal. The
Alaska Peninsula is notorious as a storm-
breeder, and before the eruption Katmai
Pass had a reputation for bad weather
not to be matched elsewhere on the
American Continent.

Now, with such enormous quantities
of hot steam rushing into the air close
beside the extensive glaciers and snow
fields of the mountains, the weather is
even worse than before. From the head
of the Valley, where conditions made it
necessary for us to carnp, we could often
look out of our door through a storm that
threatened every moment to tear the
tent from the ground, and see bright sun-
shine and good weather a short five miles
down the Valley.

During the season of 191 7 there was
rain almost every day of our stay — not
the gentle mist familiar to dwellers of
southeastern Alaska, but real rain in big
drops, driven before the gusty winds

that penetrated everything, until our
tent roof leaked like a basket.

How we wished to study the Valley
from the shelter of a house with a real
roof, where we could keep things dry and
contemplate the wonders of nature with
some degree of personal comfort!

But in the intervals between rains, the
sunshine made up for all the hardships
we endured. The weather here reminds
one of the little girl with the curl. When
it is bad it is undeniably horrid, but
when it is good it is so very, very good
that one straightway forgives its evil
moods. Whenever the skies cleared we in-
stantly forgot the discomforts and one and
all gave ourselves up to admiration of the
surpassing beauty of our surroundings.

Having established ourselves in the
Valley, we prepared to study the many
scientific problems presented by thb
unique place.

A FEW SIGNS OF ANIMAL LIFE

One of our first observations was the
number of dead insects around many of
the vents, where they had been killed by
flying into the live steam. Hine, there-
fore, came up for a few days to study the
insects with the object of ascertaining
how they got into the Valley and where
they bred (see page 204).

The larger animals were practically
absent, but we found occasional tracks
of bears, wolves, and wolverines, which
had crossed the Valley from one range
to the other. Most of these were old,
but one day I found the tracks of a bear
that had crossed during the night.

I wished I might have watched him
when his feet sank into the patches of
soft scalding mud that lay in his way.
He must have been treated to the sur-
prise of his life! But however he felt, he
kept right on, straight across the Valley,
without making the slightest deviation
to avoid the bad places, often sinking
deep into the hot mud.

SURVEYING THE VALLEY

Maynard, with one of the others for
assistant, toiled up to the summits day

Digitized by

Google

BRILLIANT INCRrSTATIOXS FROM THE FUMAROLES
The yellow of sulphur is largely lost in reproduction, materially subduing the effect.

VII

Digitized by

Google

Natural Color Photograph by Frank I. Jones
A ROW OF GORGEOUS "STOVE HOLES"
Here the rice and beans could be left to simmer indefinitely, held exactly at the boiling point.

VIII

Digitized by

Google

EXPLORATIONS IN THE TEN THOUSAND SMOKES IN 1917

209

after day to secure the topx^graphic map
which is the necessary basis of all our
statements of areas and sizes.

His was arduous work and the effort
was often wasted, for perfectly clear
days, which are necessary for this work,
are rare around mountain passes any-
where, and here especially so.

Sometimes the weather seemed to take
an almost fiendish delight in opposing
his plans. Several times as we watched
from the Valley we could see a thin
cloud hanging all day to the very summit
on which he stood shivering, while the
mountains all around were clear.

The first time he tried to work Baked
Mountain the wind happened to be just
right to carry the steam from the biggest
fumarole in the vicinity, No. 46, square
across the summit, preventing observa-
tions as effectually as though the whole
region were enveloped in fog.

More than once it seemed as if there
would not be enough clear days to com-
plete the observations, but in the end he
succeeded in getting the data for an
excellent map.

COLLECTING GASES FOR ANALYSIS

The most disagreeable, as well as one
of the most difficult tasks, fell to Shipley,
who collected samples of gas from the
vents for analysis (see page 246).

It looks easy, as the work is laid out in
advance, to poke a glass tube into a vent
and pump the gas into a container; but
in the field all sorts of difficulties crop up,
and to surmount these requires great
patience and resourcefulness. Appara-
tus will not do what is expected of it;
tubes clog prematurely or snap in the
heat.

Moreover, a fumarole is not an easy
thing to deal with at close range. When,
after some trouble, one gets in a position
where he can take his sample, and a sud-
den shift of wind brings a cloud of hot,
acid vapor around him, he is placed in a
difficult, not to say dangerous, situation.

More than once our gas collectors be-
came lost in the blinding vapors and re-
gained a place of safety only after some

minutes of anxious groping through the
steam. But fortunately the precious
samples were all secured without mishap,
and a considerable amount of other val-
uable chemical work done.

PRACTICALLY ALL PLANT LIFE DESTROYED

Only the botanists were without em-
ployment, for in the formation of the
Valley all life was annihilated. Plants
were practically absent in 1917. Not
quite so, however, for around some of
the mild vents moss and algae were be-
ginning to start in places bathed by the
warm breath of the fumarole.

From the vapors they derived, besides
constant moisture, their supply of nitro-
gen in the form of ammonia, which is
given off in considerable quantity by the
vents. Beans dropped on the ''kitchen
floor" near our fumarole sprouted and
grew rapidly on the warm ground, soon
making a bright spot of green. But they
were short-lived; the roots were killed
wherever they touched the acid soil.

The absence of vegetation gave me op-
portunity to study the unique exhibition
of geological forces here presented. Ob-
servation of the different types of fuma-
roles, of their relations to each other,
to the old rocks and to the new deposits
that filled the valley, endeavors to cor-
relate temperatures and incrustations,
and similar problems of a merely de-
scriptive character were more than
enough to occupy my time.

But beyond these were the more im-
portant questions of the relations of the
fumaroles to the old volcanoes, to the
subterranean magma whose rise brought
on the cataclysm, and to the great forces
of the inner earth responsible for the
crumpling and fissuring of the outer
shell, the building of mountain systems
and the formation of volcanic chains —
processes in which the greatest of erup-
tions are but incidental details.

TESTIMONY OF MY ASSOCLA.TES

Consideration of these questions will
best be deferred, however, until the com-
pletion of the narrative of the expedi-

Digitized by

Google

> a.

o s

O <«

O rt

2 §

P =3

a
oT

H X

u
<

210

Digitized by

Google

EXPLORATIONS IN THE TEN THOUSAND SMOKES IN 1917

211

tions. My account of our experiences is
necessarily influenced by the limitations
of my own interests and point of view.
The impressions which my associates
have set down at my request, reflecting a
wide diversity of temperament and ex-
perience, will give a broader view:

L. G. Folsom: '*You ask me to write
of my impressions upx^n seeing for the
first time the 'Valley of Ten Thousand
Smokes/ I do not recall your ever hav-
ing given me yours, and I suspect that
you feel as helpless as I do in attempting
to put impressions into words and for the
same reason — that it would require a
special vocabulary not found in any dic-
tionary of present publication. It can-
not be denied that such impressions are
determined by point of view, and ix)int
of view by temperament, so I must be
sincere or my statements are of no real
value.

*' You, and only you, whose companion
I was in the discovery, can have an
adequate conception of the emotions of
one who knows he is the first of his
fellows to look ufKjn such a marvel of
nature. In fact, the emotions of a dis-
coverer must bear some similitude to the
emotions of a creator.

''To think again of what we had al-
ready seen, to feel again the explorer's
urge, to suffer again the pains of pro-
tracted exertion, to experience again the
thrill and exaltation of achievement, pro-
duces a sort of exquisite languor, which,
were it not that there is still so much to
do, one might wish to be lasting.

"The thousands of little volcanoes
were almost a more magnificent sight
than the wonderful Katmai Crater we
had visited the day before. We were in
danger of being surfeited with 'world
wonders.'

"There comes in upx^n my conscious-
ness yet, occasionally, a sort of startling
memory of the experience. It almost
hurts. Then I fall to dreaming of the
causes and probable progress of such
phenomena; of the unusual opportuni-
ties for study; the wonderful scenic dis-
play of natural forces.

" Upon first walking among the hissing
fissures there was such a sensation as one
might imagine if sitting upon the roof
of a powder magazine — sputtering fuse
— Kingdom-come. And yet, I cannot
say I ever feared the Valley, but was
awed by it. I could not decide whether
it was Valhalla, 'World of the Gods,' or a
rendezvous of devils; an Elysium or a
Tartarus.

"There was always a very conscious
realization that the ground upon which
we were walking was holey ground, also
that it was not wholly ground, but in
large part, cavity, filled with odorous
and malodorous gases, chiefly the latter.

"In its bigness it is truly Alaskan; in
its scenic effect it is marvelous; in its
display of mighty force it is magnificent.

"You perhaps recall Walter Metro-
kin's story of his dogs, who had once
caught a porcupine, with 'needles all
chuck full o' mouth.* While this first
wonder of the world is not a thing to in-
spire humor, Walter's twisted vernacular
would fittingly characterize it as 'smokes
all chuck full o' valley.'"

EACH INDIVIDUAL VENT MORE BEAUTIFUL
THAN VALLEY AS A WHOLE

Paul R. Hagelbarger, Assistant Botan-
ist: "Bright sunshine bathed the Val-
ley when I first saw it. Even though
several miles away, I was awe-struck by
the surprisingly large size and striking
beauty of the spectacle. There were so
many more steam jets than I had even
hoped to see that I could only gaze in
silent admiration.

"After living in the Valley and work-
ing among the fumaroles, my impres-
sions began to change. My amazement
at the great area was intensified by the
knowledge gained on many trips across
the Valley floor. The beauty of each
individual vent was even more than that
of the Valley as a whole.

"The thing that stupefied me, how-
ever, was the ever-present proof that
some terrific energy or force had only re-
cently exerted itself. Everything seemed
on such a huge scale. Our tents looked

Digitized by

Google

212

§.^•6 8

Digitized by

Google

EXPLORATIONS IN THE TEN THOUSAND SMOKES IN 1917

213

insignificant amid the roaring volcanic
vents.

"As I came daily to know the area
better, I was more and more impressed
by the titanic forces that had been
at work here. Human endeavor and
achievement seemed dwarfed to insig-
nificance by comparison. I felt out of
place and like an intruder in this Land of
the Gods. This Valley appeared to be
on another planet that was in the proc-
ess of formation.

** I spent 16 days in the Valley and was
glad to leave, as will be seen in my diary
for August 2 : * Came out of the steam-
ing Valley for good. Lucky to get out.
Glad to see trees and grass again. Feel
like I am just awakening after a two
weeks' nightmare. Valley is wonderful,
but no place to camp. Walter says,
" Lots of steam. Hell of a place." Heart-
ily agree.*"

SURPASSED HIS WILDEST DREAMS

J. D. Sayre, Assistant Botanist: ** My
sensation on first seeing the Valley of
Ten Thousand Smokes was one of won-
der and astonishment. I was astonished
at the great dimensions of the Valley and
at the countless numbers of fumaroles
and fissures out of which the steam is-
sued, to say nothing of the many other
gorgeous and magnificent displays of na-
ture. Never in my wildest dreams had
I imagined anything to compare with
these.

*' Greatest of all was my surprise that
so much energy could be released in such
an e2isy and quiet manner without ap-
parent injury or danger to any one or
any thing. I experienced no sensation of
fear while staying in the Valley, perhaps
because my mind was so filled with as-
tonishment and admiration at this great
marvel of nature, or because I was fool-
hardy and did not realize the grave dan-
ger of falling into one of those hot places.

*'I had no hatred of the place during
my short stay there, although we were
surrounded by many discomforts, and I
said, soon after we left, that I would like
to come back some time and see the

place again. I am very proud to say
that I was a member of the expedition
which overcame the difficulties and
hardships and first explored such a won-
derful place.**

THE COMPLAINT OF A TOPOGRAPHER

Clarence F. Maynard, Topographer:
"To me the Valley of Ten Thousand
Smokes is a stretch of country that offers
all the usual difficulties of topographic
surveying in Alaska, with a few rather
unusual ones thrown in for good meas-
ure. It is hardly a country to make the
heart of a topographer glad.

"The smokes did not impress me with
their grandeur or with their wonder as a
natural phenomenon. Their ability to
make surveying next to impossible did,
however, make a very decided impres-
sion on me. On the occasional clear days
when the sun was shining down the Val-
ley they seemed to be always at their
best, as Griggs would put it, but to my
mind at their worst.

"On these, the few rare days when it
was not raining and the wind was not
doing its best to move our camp (rather
good judgment on the part of the wind, I
should say), they would shoot forth jets
of steam which soon took the form of
clouds and obscured the country we
were trying to work.

"I finally began to believe that the
smokes were out to buck me, and be-
came convinced of it when, on one of the
rare fine days, I ascended to a peak
which immediately became enveloped in
fog. This was not unusual, but I was
impressed on returning to camp to hear
from the more fortunate members of the
party that the whole Valley had been
clear with the exception of the peak 1
occupied.

"I am not a v^etarian; furthermore,
tea cooked in a steam pit is not tea. A
tent that never sheds a drop of water is
not a tent. A wool comfort placed on the
ground which is iio'^ F. in the above
tent will steam beautifully. It is a nat-
ural phenomenon, but it is not a good
bed.

Digitized by

Google

Photograph by J. W. Shipley
HINE INSPECTING THE CAVERNOUS MAW OF A GAS-EMITTING VENT

The gases from these openings are transparent until they begin to condense in the atmosphere.
Consequently it is frequently possible to look far into the depths of the earth.

214

Digitized by

Google

EXPLORATIONS IN THE TEN THOUSAND SMOKES IN 1917

215

" I believe I mentioned that I am not
a vegetarian. I like bacon in the morn-
ing; I like it fried. A steam jet, in spite
of its being a glorious and natural phe-
nomenon, will not do this. I am from
New England and have decided ideas on
baked beans. Again the steam jet fell
down. It needs New England training.
Baked beans are beyond the limit of its
capabilities.

*' I should say the coming of the smokes
ruined what might otherwise have been
a perfectly good country. My opinion,
however, is probably valueless, as being
out of tobacco always colors my views."

Fortunately Maynard was by no
means the confirmed grouch that he
makes himself out, but on the contrary
was one of the best fellows in the crowd,
his good-natured raillery contributing
greatly to the pleasure of the summer's
work.

THE MODERN INFERNO

James S. Hine, Zoologist: "A hike of
miles over devastation wrought by nat-
ural disturbances in the Katmai country
naturally puts one into a peculiar state
of mind. He is deeply impressed with
the enormity of the whole affair, and
everything seems beyond comprehen-
sion. The unusual circumstance of sum-
mer with no plant life and no animal life
surely is a strange realization.

** Having reached the summit of Kat-
mai Pass, the Valley of Ten Thousand
Smokes spreads out before one with no
part of the view obstructed. My first
thought was: We have reached the
modem inferno. I was horrified, and yet,
curiosity to see all at close range capti-
vated me. Although sure that at almost
every step I would sink beneath the
earth's crust into a chasm intensely hot,
I pushed on as soon as I found my-
self safely over a particularly dangerous-
appearing area. I didn't like it, and yet
I did.

*'I felt like a boy at a circus, for I
couldn't take time to study the attrac-
tion before me because I suspected some-
thing more captivating farther on. Nor

was I ever disappointed, for nothing was
exactly like anything else.

''The broken hills, the falling moun-
tains, the magnificent glaciers, the steam-
ing fumaroles, and the rolling steams
can all be described, but their wonderful
profusion and the manner in which they
encroached upon one another must re-
main largely in possession of him who is
fortunate enough to make a visit to the
locality where these things abound in
extraordinary splendor."

LIKE A HUGE CHEMICAL MANUFACTURING
PLANT

J. W. Shipley, Chemist: "On first
entering the Valley from between the two
guardian volcanic cones, I experienced
the same sensation as the man who on
seeing a giraffe for the first time ex-
claimed, 'There ain't no such animal.'
The quiet evolution of myriads of col-
umns of vapor from the floor of a wide,
desolate valley, the encompassing moun-
tain ridges, the sequestered isolation,
the avalanches of rocks, all vividly re-
called Sindbad's adventures in the 'Ara-
bian Nights.* It is so unreal!

"Hot streams issue from beneath
banks of snow; extensive glaciers hob-
nob with steaming fumaroles; icebergs
and hot water are found in the same
little lake (see page 244). Enormous
mud flows appear to have run uphill.
A stick chars when thrust into a jet of
steam. It is uncannily unreal.

"But the unreality suddenly vanishes
when one's foot breaks through the crust
and hot volcanic gases rush out. It is
also sufficiently realistic to have ava-
lanches of rocks galloping down the
2,000-foot face of Falling Mountain
while we are collecting gases in the
bottom of a 20-foot hole at its base.

"The familiar fumes of hydrogen sul-
phide, sulphur dioxide, and hydrochloric
acid transform the Valley into a huge
chemical manufacturing plant, roofed
over by a permanent cloud of vapx^rs.
And when one is cold and wet, it is
rather comfortably lonesome to lean
against the hot walls of a sheltered crev-

Digitized by

Google

2l6

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

ONE OF THE GAPING FISSURES ALONG THE MARGIN OF THE VALLEY

The two sides were evidently pulled apart by contraction of the Valley floor; compare the diagram-
matic section of the Valley given on page 263.

ice and meditate on the dead bodies of
hundreds of flies lying around the orifice
of the subterranean chimneys."

A SPECTACLE OF AWESOME MAGNITUDE

D. B. Church, Photographer: *' Re-
gardless of our packs, we hurried through
the Pass leaving behind the few faint,
wispy steam jets that mounted from its
floor just over the divide, craning to
glimpse the first steam cloud to rise from
the Valley beyond. There floated over
the spur of the ridge to the north a bil-
lowy cloud that marked the largest
steamer.

''Reaching the higher ground that
had hidden my view, I gazed at the
panorama before me. Flanked by Cer-
berus and Falling Mountain, spread the
Valley — a maze of pearly columns that
billowed skyward and bent before the
strong westerly wind. Down a narrow
canyon we trudged and climbed out over
its painted ash walls on to the Valley floor.

, "The meager pictures of the previous
year, and even the graphic descriptions
of Griggs and Folsom, had not prepared
me to face such a spectacle of awesome
magnitude. I had pictured the Valley
as large; the actual view dwarfed my
wildest imagery to insignificance.

"I started for the nearest fumarole; it
seemed a few hundred yards distant. I
found it half a mile away. It was a
small fumarole and I crept cautiously
up to its edge. From its red-painted
throat, which vanished in deep black-
ness, the sulphur-reeking steam roared
forth in a smothering blast.

"Passing back, I found a crack in the
rocklike crust of the mud flow, through
which sizzled the scorching steam and
gas. A few prods with my staff opened a
hole into the underground conduit, from
which the steam hissed forth. The
fragility of the crust and knowledge of
the result of a misstep startled me. My
fears began to awaken — fears that for

Digitized by

Google

EXPLORATIONS IN THE TEN THOUSAND SMOKES IN 1917

217

several. days made me tiptoe over spots
where the earth rang hollow beneath my
feet. Familiarity gave me greater con-
fidence, but I never ceased to tread care-
fully the color-daubed regions of sub-
surface activity.

WORK DROVE AWAY FEAR

"The next day I b^an my work in
the Valley. This day activity and in-
terest drove fear from me. The one
conception that pervaded me was: How
like this place to Dante's conception of
the 'Inferno.' It seemed to me, as we
stood on the edge of Novarupta, that
this was the Devil's own private comer
in hell itself.

*' It seemed, as I gazed at the seething
steam clouds that rushed from the cool-
ing lava plug, and at the shattered,
steam-smothered furnace that filled the
rising vale beyond, that there was some
vague fantastic form, a horrid dream, a
hideous, potent 'thing,' which was not
for human eyes to see nor human ears to
hear.

''Then an endless night on the hot,
moisture-teeming ground; an endless
rolling from side to side to escape the
torment of the penetrating heat that
seeped up from the hot, sodden ground;
and always, as I looked down the Valley
through the open tent door, shone the
marble-like steam columns, which, like
tall, writhing specters, swayed in the
dim twilight.

"There was always a certain awe-
someness about the Valley which clung
to me throughout my stay. I looked for-
ward with relief to the time when I
could put from my sight the curling
steamy billows that rose from f umaroles
and mounted ever skyward.

"Pictures cannot bring back the Val-
ley of the Smokes. They have lost the
awesomeness that lies in the setting.
You may build in memory, but never
reproduce the scenes which lie beyond
the Katmai Pass. They seem too big to
be a part of the rest of the world. They
do not connect up with the little things
which are built into our lives.

"Outstanding in my memory is the
Valley as I left it. It was a brilliant day,
with puffy silver clouds that floated on
a sky of deepest blue, and sunlight that
glinted on opalescent steam jets and
sparkled on peaks fresh-capped with
snow.

"As, homeward bound, we skirted
Cerberus, the steamers turned in the
dying sunlight to shimmering gold and
the snowy crests of distant mountains
glinted yellow. I forgot the heavy pack
which bowed my shoulders as I glanced
backward at the growing beauty which
filled the Valley. Through its giant
gateway the Valley of Ten Thousand
Smokes sank from sight as we dropped
over the Pass, and the sky above red-
dened to a crimson halo in the fading
rays of the sinking sun."

Digitized by

Google

a
a

o
o

:^
o

pes
c/)

O
en

en
;:>
O

H
55

b
O

>^
.J

H
b
O

pes
O
U

2 >»

c.ti
g.iS

3 w

£ c

••SI

^ ^'^

♦- o

•5^

2l8

CO
(0

Digitized by

Google

XVII
THE TEN THOUSAND SMOKES IN 1919

When, after months of preparation,
we approached the Valley again on the
first day of June, 1919, we were intensely
curious not only to see what changes
time had wrought but also to observe
the fumaroles at that early season.

It was thirty-four days earlier than
anyone had entered the Valley before —
so early that the snow had only begun
to disappear from the mountains and
the great torrents that pour from the
glaciers in July and August had not yet
started . Many feet of snow still covered
the south approach to the Pass, and the
area of little fumaroles between Cerberus
and Mageik was blanketed ten feet
deep.

But the Valley itself looked just as in
midsummer except for two or three
diminutive white patches in quiet cor-
ners away from fumaroles. We had
been curious to know how the fumaroles
behave through the winter, and this view
in early spring was next best to over-
wintering in the region.

FUMAROLES NOT VISIBLY AFFECTED BY
SEASONAL VARIATIONS IN STREAMS

Aside from a somewhat idle curiosity
to see the struggle between the elements
when the winter's cold should intensify
the temperature contrasts, there are
scientific deductions of much importance
to be made from a comparison of the
winter activity of the Valley with its
summer condition.

Practically all the drainage from the
watershed tributary to the Valley is
vaporized by the heat and returned to
the air, either through the fumaroles or
as diffuse invisible vapor. The only
way to estimate how large a proportion
of the steam from the fumaroles is com-
posed of surface water is to compare the
condition of the Valley in winter, when
practically no water runs into it, with

that in summer, when the accumulated
run-off of the whole year is largely con-
centrated in two months. If the fuma-
roles are due to surface waters they
should dwindle in winter and grow with
the increasing melt in summer. If on the
other hand surface water plays but a
small part, they should remain nearly
constant at all seasons.

So far as our observations go, they
point definitely to the latter suppo-
sition. The Valley looked exactly the
same on June i, 1919, as it had in Au-
gust, 191 7, and we could not detect any
differences beyond the daily fluctuations
due to changing weather conditions as
the season progressed. Yet in early
June the slight streams from the gla-
ciers quickly dried up, leaving dry beds
through the main portion of the Valley,
where in August ran foaming mountain
torrents.

SOME INDICATIONS OF SLACKENED
ACTIVITY

While it was evident at once that the
general appearance of the Valley was
the same as before, after we had had
time to examine the ground in detail we
could observe some slight indications of
a slackening of activity around the edges.

The little fumaroles between Mount
Cerberus and Mount Mageik at the
head of the Valley, of which there were
about a hundred in 1916, 1917, and 1918,
were reduced to two or three in 191 9.
At the foot of the Valley a large number
of springs of boiling water had started
up in areas where there had been noth-
ing but steam previously. The temper-
atures of some of the fumaroles in the
same area were markedly lower than the
year before.

Falling Mountain, likewise, though
quite active enough to satisfy the new-
comers, did not shoot forth its avalanches

219

Digitized by

Google

220

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

FALLING MOUNTAIN FROM ACROSS THE VALLEY

One of the most remarkable phenomena of the region is the constancy of the avalanches down the

face of Falling Mountain.

with quite the same frequency as in for-
mer years. The old volcanoes, Katmai,
Trident, Mageik, and Martin, also were
less vigorous than in 1916 and 191 7.

This is, however, by no means a cer-
tain indication that the activity of the
region is dying out. The big volcanoes,
which alone were known at that time,
had a similar period of slackening in
1915. We could not assure ourselves
then that Katmai was really active, and
steam from Trident could be seen only
at certain times.

But the two following years the out-
put of both these vents increased measur-
ably, so that no one could question the
reality of the steam clouds issuing from
them. Whether the slackening of the
activity of the Valley vents observed in
1919 was simply such a temporary
fluctuation, or whether it represented a
permanent quieting down can only be
judged by the future.

In either case there is no reason to
expect any sudden extinction of the
Smokes, for the changes observed were

so slight that we could not be certain
that there was a real slowing-down until
after many days of observation and com-
parison.

MELTING LEAD AND ZINC IN
THE FUMAROLES

Despite the signs of decreasing ac-
tivity that were detected around the
edges of the Valley, the geophysicists
found fumaroles very much hotter than
any that had been measured before — hot-
ter, indeed, than any of us had supposed
occurred in the Valley.

It was late in the season before any of
these extremely high temperatures were
discovered. We all felt that more and
perhaps hotter places would have come
to light if only there had been time to
continue the work. But increasingly
bad weather prevented further search.

There was great excitement one night
when Dr. Allen, coming in late, an-
nounced, "Three times have we melted
zinc this day." His party had found
three widely separated fumaroles with

Digitized by

Google

THE TEN THOUSAND SMOKES IN 1919

221

Photograph by Jasper D. Sayre
MEASURING THE TEMPERATURE OF A FUMAROLE

The temperature of this insignificant little hole was 300° C. (572° F.). Others in the same line near
by ran above 400** C. (752° F.). The difference in temperature between the hot and cold junctions
at opposite ends of the thermocouple sets up an electrical current whose intensity, recorded by the
meter in the foreground, is a measure of the temperature (compare Plate XV, facing page 304).

temperatures above the melting point of
zinc (420*^ C).

The hottest of these was practically
500° (496^ C, 868° F.). This was found
in a fumarole where the steam broke
through a myriad of small holes in the
roof of a bridged-over fissure.

The volume of gas coming from any
one of these holes was so small that we
could approach as closely as desired. We
could easily reach into the hole with a foot
rule. Within five seconds the end of a
stick held in the steam began to smoke
violently and when drawn out was a glow-
ing coal. In less time than that a bar
of lead slumped down and melted away.

KINDLING A FIRE BY PLUNGING
A STICK INTO WATER

But even this performance was
eclipsed next day, when the chemists

came in reporting several temperatures
above 500*^. The highest, 645*^ C. (nearly
1,200° F.), was obtained in a small orifice
not more than two inches in diameter,
located at the bottom of a crater-like
pit eight or ten feet across.

This was an inconspicuous little fuma-
role which had not excited our curiosity,
although it was but little removed from
the trail that we used constantly in trav-
eling up and down the Valley. The gas
rising in the vicinity was, however, so
blue as to arouse Dr. Zies' suspicions,
and lead him to examine the place in
detail.

When we put an aluminum cup into
the steam it was quickly softened, so
that it could be cut with a knife like pew-
ter. Yet it showed no signs of melting.
The tinned handle of the cup was vigor-
ously attacked by the gas, but the alu-

Digitized by

Google

222

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by E. C. Kolb
OUR DRINKING POOL AT THE MOUTH OF A SNOW CAVE

By moving the tents a few feet back or forward, we could obtain any desired floor temperature. Only
a few rods beyond the tents was our cookstove.

minum body was neither corroded nor
tarnished in the least degree.

Then we cut the end of a walking
stick into a brush of shavings, such as one
uses to light a fire, and thrust it into the
fumarole. Instantly it began to smoke
and char, and when jerked quickly into
the air burst into flame, from which we
kindled a bonfire by breaking up what
was left of our canes. Since the vapor
from this fumarole was almost pure
steam — that is to say, water — what we
really did was to kindle a fire by plung-
ing a stick into water (see page 296).

''a red-hot fumarole"

All of the very high temperatures
measured were found in such relatively
small and inconspicuous fumaroles as

1 A number of people whose knowledge of steam
at higher temperatures was derived from experi-
ence with boilers have asked me why the Valley
was not in great danger of blowing up from the
tremendous pressures accompanying steam tem-
perature of 1,200° F. It may be well to remark
that while steam heated in a confined boiler to
such a temperature would develop an enormous

these, rather than in the big vents which
one would suppose to be hottest.

The temperature of the gas as it
first emerges is probably quite as high in
the big vents as in the little ones, but
the wide-open throats, which the force of
the escaping gases has blasted out, per-
mit the emanations to cool down consid-
erably before reaching the orifice. On
this account the largest and most im-
pressive vents, those which are actually
delivering by far the greatest quantities
of heat, are seldom more than two or
three times as hot as ordinary steam,
that is, 200° to 300° C. This tempera-
ture, although far beyond that ever
found in a steam boiler, is yet much
below the kindling point of wood.*

Even at 300°, however, the steam is
perfectly dry as it emerges, and some-
pressure, in an open receptacle permitting free
expansion, it could be raised to a high tempera-
ture, as readily as so much air, without dex'elop-
ing increased pressure.

The gases from the fumaroles all come off at
relatively low pressures, but little above the
atmosphere.

Digitized by

Google

THE TEN THOUSAND SMOKES IN 1919

223

times does not condense until 20 feet
from the vent. The dancing, swirling,
feathery wisps which first appear in such
a condensing column hold a peculiar
fascination for the onlooker, as he
watches them thicken and coalesce into
a heavy cloud.

In the hottest vent the steam is not
only so perfectly dry as to show no sign
of condensing for a long distance, but
so hot that in the dark the throat must
glow with a faint redness. All of us
would have liked to see a red-hot fuma-
role, and there was much talk of going
down to the vent to observe it at night;
but when it came to the test no one was
quite ready to undertake the trip, for
none cared to try picking his way a-
round among the fumaroles in the dark-
ness.

VALLEY BECOMING GREEN IN SPOTS
WITH MOSS AND ALGAE

The most conspicuous change in the
appearance of the Valley was due to
quite another cause — the beginning of
an invasion of its hot soil by plant life.
In the two years since 191 7 the few
minute beginnings of vegetation had in-
creased until they formed in places con-
spicuous spots of bright green, visible
sometimes for a mile.

The new growth was composed of
"blue-green" algae and moss, which
made a dense carpet around some of the
fumaroles. The algae consisted of sev-
eral genera that are common in hot
springs; the moss was uniformly sterile.

These plants here show a very decided
preference for hot ground. In fact, they
are strictly confined to spots moistened
by the steam. Sometimes a small fuma-
role will be surrounded by a mat of moss
which approaches the orifice on all sides
until its advance is stopped by the high
temperature. Besides the living plants,
one often finds brown seared remains of
others that have been killed by the fuma-
roles, after having managed in some way
to make a good growth inside the dan-
ger zone.

In addition to these visible plants.

Sayre found by incubating samples of the
soil that the hot earth is inhabited by
bacteria.

The scientific problem raised by the
invasion of the Valley by these lowly
plants is one of the greatest fascina-
tion. If one could work out the whole
sequence of organisms from the lowest to
the highest as vegetation is reestablished
he would perhaps be enabled to under-
stand a great deal more than we now
know concerning the peopling of this
planet and its preparation for habitation
by the higher animals and man.

BEARS INVESTIGATE THE VALLEY

In the seven years since the formation
of the Smokes, the surrounding country
had gradually been occupied by a new
population of bears which, having prob-
ably grown up in the vicinity of the
Valley, had come to regard it as one of
the normal phenomena of their world.

When first discovered, the active area
was as absolutely devoid of evidence of
living creatures as can be imagined. The
next year, as previously noted, a single
bear ventured to cross the Valley dur-
ing our stay.

But in 1919 bears Were frequent visi-
tors, comprising indeed the only tourists
that had yet visited its confines. Not
merely here and there, but in many
places, their tracks were to be found all
through the Valley. These were not the
work of any single bear, nor made at any
one time, but were of many sizes, show-
ing that most of the bears of the region
probably enter the place from time to
time.

They were not satisfied with merely
crossing through the steaming areas, but
were apparently attracted in some de-
gree by the Smokes. Their tracks were
often to be found close around the largest
vents, even far up toward the head of the
Valley, many miles from any possible
source of food.

From their behavior, indeed, it seems
not at all improbable that they may
have enjoyed lying on the warm ground,
seeking good places to bask in the heat,

Digitized by

Google

224

THE VALLEY OF TEN THOUSAND SMOKES

just as a dog picks for his bed the hottest
place behind the kitchen stove. We
could not assure ourselves in this matter,
for we could never catch sight of them in
the Valley, and the ground around the
big fumaroles is baked so hard that only
claw marks remained to show where
they had walked. If they lay down
their shaggy coats left no mark on the
hard crust, so we were little the wiser
concerning their real habits.

It was not unusual to find the tracks
of a bear leading straight up to one of
the large vents, where evidently he had
stopped to peer into the mysterious hot
hole. In one of the steaming areas
Hagelbarger found a place where the
hot ground had apparently excited the
bear's curiosity, for he had dug into it
until he started a small fumarole of his
own. The appearance of a cloud of
steam under his claws as he broke into
the hot crust must have given him a great
surprise. It did not scare him away,
however, for not satisfied with a single
experiment, he tried again in several
places, each time digging down till he
started the steam before turning away.

BAKED MOUNTAIN CAMP

When we first camped in the Valley
in 191 7 we were so afraid of its fumaroles
that our guiding maxim was "safety
first." But when we returned in 191 9
we were willing to take a chance and try
experiments we had not dared attempt
before.

At the suggestion of the photogra-
phers, the camp was placed at a corner
of Baked Mountain (see map, page 68),
close beside some of the biggest vents
in the Valley, a situation that command-
ed magnificent views of Mount Mageik
and Mount Martin and was at the same
time convenient of access from all direc-
tions.

Although this position proved unten-
able, as was afterwards demonstrated by
a series of disastrous experiences, it per-
mitted us to secure some photographs
which could not otherwise have been
obtained; for pictorial conditions in the

Valley are so dependent on fleeting
light effects that one cannot hope to
obtain the pictures he wants except by
camping on the spot and rushing out
with his camera when the light is right.
So far as conveniences for the camper
were concerned, this location was all
that could be desired. The big fuma-
roles in front furnished all degrees of
heat for cooking, while the snowdrift
directly behind the tents supplied an
abundance of good water and served as
refrigerator.

ACID FROM THE COOKSTOVE
EATS THE POTS

The new location called for new
methods of cooking. We soon found
that it was impossible to hang a pot
down into the fumarole, as we had done
before. The steam from this group of
vents was highly charged with acid
(either hydrochloric or hydrofluoric),
which in the course of a few moments
attacked the rope to such an extent that
it fell to pieces when we tried to pull
out the pot. More embarrassing still
was its effect on the pots themselves, for
it was not long before they developed
leaks where the fumes had eaten away
the aluminum.

Instead of hanging the pots in a fuma-
role, therefore, we dug holes in the hot
ground and banked up the steaming
earth around them. The different holes
stood at different temperatures, so that
we were able to vary our procedure
according to the results desired. If, on
leaving in the morning, we wished to
provide supper against our return, or if
we wished simply to keep a pot warm,
we would select a "slow** hole. But if
we were in a hurry for something, we put
it into a hot hole, where it would boil
away vigorously.

The hottest places would burn up the
food if we left it too long, just as a fire
does. But a pot could be left indefinitely
in the slow holes without coming to
any harm. Once, when driven out by a
storm, we found the oatmeal that had
been put on for breakfast in prime con-

Digitized by

Google

Natural Color Photograph by Frank I. Jones
THE CAVERNOUS MOUTH OF ONE OF THE FUMAROLES
A long passage full of superheated steam could be followed with the eye for sixty or a hundred feet

into the depths.

Digitized by

Google

Natural Color Photograph by Frank I. Jones
A RED-BURNED FISSURE ON THE WEST SIDE OF THE VALLEY, LOOKING TOWARD THE

GLACIERS OF MAGEIK

Digitized by

Google

THE TEN THOUSAND SMOKES IN 1919

225

dition on our return two days later. The
wind had entirely buried the pot in drift-
ing pumice, so that it had to be dug out
with a spade; but it was none the worse
for having simmered away ten times as
long as intended.

CORN PONE A LA FUMAROLE

With no fumaroles hotter than ordi-
nary steam at hand we had been com-
pelled in 191 7 to limit our menu to boiled
dishes. But the range of temperature
at Baked Mountain was great enough to
permit any sort of cookery we might
desire.

The situation of the "cookstove"
which made such varied temperatures
readily available in a small compass was
characteristic of many areas. It was
located on the roof of one of the great
bridged-over fissures that encircle the
margin of the Valley. The cavern be-
low was full of superheated steam under
pressure (see plate ix, facing page 224).

Every little crevice in the arching
bridge was the path for a small seepage
of steam, the temperature of which, as it
emerged, depended on the volume. The
slow cookers were dug in places with
little seepage; the hot spots had close
connection with the steam below. We
could not tell in advance, however, how
effective any hole would prove as a
cooker, for there was little visible emana-
tion from any of those we used.

In the hottest of our ''stove holes"
the temperature was high enough to fry
bacon or bake bread with ease. For
some reason the tastes of the party cen-
tered on johnny cake rather than on
white bread or biscuits, although the
latter were occasionally made with
equal success. The procedure was to
put the pan of batter in a covered
aluminum cooking pot which was then
simply set on the ground in the proper
place.

Baking over the f umarole required the
same sort of watchful care that it does
anywhere else. It would not do to go off
and forget the bread, as we did the oat-
meal, for if it was left longer than the

allotted hour it was burned up, as it
would have been in any other oven.
When taken off at just the right time,
however, it came out with beautiful crispy
brown crust, as fine corn bread as was
ever seen (see plate vi, facing page 193).

FRYING BACON IN SUPER-HEATED STEAM

For frying bacon we found it most con-
venient to repair to **Fumarole No. 5,"
at a little distance from the camp. Here
a column of very hot steam emerged
under pressure from a round hole about
a foot and a half in diameter in such a
way that it could be readily approached.
The gas at this place was so hot and
came out with such a rush that if we
tried to pour a cup of water into the hole
it never reached the ground, but was
caught up by the outrushing steam and
carried away before our eyes, vaporizing
within a few inches. When we threw
our hats into it, they would go sailing
away 30 feet into the air before coming
down.

The frying pan had to be held down
against the steam, for even the weight
of the pan and the five-foot stick we em-
ployed as a handle was not sufficient to
balance the pressure of the f umarole. It
would keep wobbling around, up and
down, in and out of the rushing steam
unless firmly held. Needless to say, the
bacon began to sizzle promptly and was
soon as well crisped as when cooked over
the best of camp fires.

While experimenting to find the best
place to hold the pan, we tried pushing
it down into the cavern below the orifice,
but immediately it was caught by the
back draft and — piff! the bacon was
whisked out of the pan, flying through
the air in every direction, to be eagerly
caught and devoured by the spectators,
who howled with delight at this sudden
turn of events. Discovered accidentally,
this trick was repeated again and again
till we tired of chasing the flying slices.

With such facilities at our command
and a full stock of dehydrated fruits and
vegetables, there was little in the way of
**grub" that could not be concocted at

Digitized by

Google

226

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Grigss
THE ''defenses" OF BAKED MOUNTAIN CAMP

Profiting by several sorry experiences, Yori finally shored up the tents so thoroughly as to bid de-
fiance to the winds but it was no use. The condition of the camp when the next storm cleared away
is shown on the opposite page.

Baked Mountain. The staples were
oatn-eal, rice, beans, combread, dehy-
drated potatoes with abundant butter,
cheese, and pilot bread. These were
varied with an occasional mess of corn-
beef hash, or of trout when someone
brought them up from the lake.

For green vegetables there were string
beans, spinach, and ** boiled dinner vege-
tables," the latter soon becoming a joke
because of the superabundance of carrot,
of which everyone quickly tired. Our
fruits included apples, raisins, peaches,
pears, apricots, loganberries, cranberries,
and cherries, the last being such a uni-
versal favorite that they were soon used
up. Soluble coffee was varied with co-
coa and an occasional cup of tea. One
hot day we even indulged in iced tea,
but for the most part we preferred hot
dishes.

THE ''WILLIWAWS"

All went well at Baked Mountain
Camp as long as the weather remained

good, but when the storms struck we
encountered a fury that no tent could
withstand. The conformation of the
mountains is such that Katmai Pass is
a double-ended funnel, through which
the wind sucks with terrific violence
whichever way it blows (see map, p>age
68).

Curiously enough it is not on the
windward, but on the lee side of the Pass
where the heavy ''williwaws" or '*wool-
ies," as the boys called them, are ex-
perienced. The winds we had learned
to dread on the Pacific slope were the
westerly gales, but on the Valley side it
is the northeasters that make the trouble.

We were blown out four times before
we finally learned our lesson and moved
to a safer, if less convenient, location.
Fortunately, these storms came in a
series of increasing violence, so that each
time we were better prepared for trouble
than before. Otherwise, they would have
been even more disastrous than they
were.

Digitized by

Google

THE TEN THOUSAND SMOKES IN 1919

227

Photograph by R. F. Griggs

THE WRECK OF BAKED MOUNTAIN CAMP

We could never have believed that the boasted strength of our camp would be so completely
demolished. Nearly a thousand feet of cordage had been used in lashing the poles together to resist
the weather. The wreck was so heavily drifted over by flying pumice that it took the combined
strength of two of us to peel back the fallen roof and get at the duffel on the floor.

As we came across the Valley on our
way back from the ascent of Mount
Martin (described in Chapter xiv), the
northeast wind freshened in a way which
I knew meant that the '*woolies*' were
to visit us that night. We were too
tired from the climb to care much, how-
ever, for we believed we could sleep
through anything. And the tents had
been so shored up as a result of pre-
vious experiences that, like the skipper of
the Hesperus, we believed they "could
weather the roughest gale that ever wind
did blow."

The frame of the grub tent had been
strengthened by a multiplicity of poles
and braces sunk deep in the ground un-
til it formed a veritable cage, inside of
which we had patched together parts
of four tents. Fully a thousand feet of
cord had been used in lashing the struc-
ture together.

All the guys were anchored to bowl-
ders as big as a man could roll, deeply
buried in the ground. On every side

except the front there were at least two
thicknesses of cloth to protect us from
flying pumice. A heavy duck tarpaulin
had been thrown over the second tent
and buried in the ground on the wind-
ward side, to reduce its resistance and
prevent the wind from getting in under
the eaves.

THE FINAL WRECK OF BAKED
MOUNTAIN CAMP

After the first few warning blasts I
called to Yori, questioning whether we
had not best take additional measures
to protect the camp, but he let out a
whoop of defiance, '*0h, let it blow; we
can stand it."

I was too tired to get up and go out
looking for trouble, so I lay down again
to get the rest I so much wanted. If I
had gone out then I could have saved the
apparatus tent, which was afterwards
sorely needed. But weariness is a great
foe to will power, and I missed my
chance.

Digitized by

Google

228

THE VALLEY OF TEN THOUSAND SMOKES

Photograph from E. G. Zies
DR. ZIES LEAVING THE WRECK OF BAKED
MOUNTAIN CAMP

The expression of grim determination caught
by the camera was not accidental. It was due
to the experiences recently passed through.

By the time it was fairly dark the
storm had increased to considerable
proportions. It came in great inter-
mittent gusts, with intervals of quiet
between. We could hear them coming
over the mountains long before they
reached us, their terrible roar as they
tore their way down through the Pass
for a while more frightful than the blow
when it struck. Quaking with dread,
we lay huddled together beneath the
tent awaiting the crash.

APPARATUS AND SLEEPING TENTS
GO DOWN

The apparatus tent went first, torn
to shreds and exposing all the chemists'
instruments, as well as cameras and
other valuable equipment, to the soaking
rain and driving pumice which flew be-
fore the wind in sheets. By ten o'clock
the poles of our tent snapped, the broken
ends tearing great rents in the fabric as

the>' went down. Fortunately they
went one at a time, gi\'ing us a chance
to move our beds and duffel into the
grub tent, which now alone remained
standing.

After the last blanket in-as removed
we managed to lay hold of the flapping
wreck and weight it douTi with rocks,
to prevent its being torn to pieces. It
was all we could do to accomplish that,
for by this time the fur>' of the u-ind
was such that one could hardly manage
even so small a thing as a prostrate tent,
and the hail of flying pumice was im-
possible to face even with goggles for
protection. We were well repaid for our
efforts, however, for the saving of that
torn tent was all that made a resump-
tion of work possible after the storm.

EXTRA TARP TEARS AWAY

Crawling into the remaining tent,
I stretched out in my sleeping bag to
get a little rest, if I could. But it was
not for long that such relaxation was
permitted. It soon became a question
whether even so strongly built a tent
as ours could withstand the pressure.

Before long the spare tent, with which
the lee side had been reinforced, gave
way and, still holding to the frame by
the corners, went flopping around in the
wind to increase the general confusion.
Two days later, when we returned, we
found it, a new tent that had never been
set up, torn clean in two!

No sort of rest was longer possible.
So I got up, packed my bed with a little
clothing inside my packsack to keep it
dry, and put on all my extra sweaters
and coats to keep warm, for the driving
rain was intensely cold. Some of the
others who continued to use their bed-
ding as robes were less fortunate, for it
was extremely difficult to gather things
together and to keep them dry in the
final break-up.

About one o'clock the extra tarp
lashed to the windward side tore away,
leaving only one thin sheet between
us and the gale. It seemed impossible
that any light cotton fabric could en-

Digitized by

Google

THE TEN THOUSAND SMOKES IN 1919

229

dure the strain. It soon broke all our
boasted guying and then was held only
by the frame.

The pumice came beating against it
with the noise of a hailstorm on a tin
roof; but these stones were sharp at
every corner, not round and smooth
like hail. The impact was so heavy
that our flesh would not tolerate the
pain if, trying to keep the tent in place,
we pressed our bare hands or arms against
the wall. Why the flying pumice did
not instantly cut the tent to shreds was
more than we could understand. But
it held for half an hour or more, and then
gave way all at once with the crack of
a cannon, as a great rent opened from
roof to floor.

HUDDLED TOGETHER IN THE LAST
CORNER OF THE TENT

The hail of pumice that greeted us as
we crawled out to see if anything could
be done cannot be imagined. It could
not be endured on face or hands for an
instant. It hurt clear through our
clothing. It drove in around our ''dust
proof" goggles, a constant menace to our
eyes. Many of the pieces were as large
as hickory nuts and all were armed with
sharp corners that made them terrible
missiles.

Once the tent ripped, the wet pumice
began to collect all over us, down our
necks, in our pockets, in our ears, in our
noses, in our mouths — everything was*
permeated with it. Soon it was four
inches deep on the floor, though more
blew out through the other side of the
tent than lodged within.

The five of us — Allen, Zies, Fenner,
Yori, and I — huddled together against
the windward corner that gave the last
little shelter remaining. It was sug-
gested that we crawl into a gully; but
we remembered that such places were
collecting grounds for the pumice which
would hail down on us in greater quan-
tities than ever. So we agreed to stick
it out as long as possible, praying fer-
vently for the dawn.

The nearest relief was at Ukak, ten

Photograph by E. G. Zies
CHARLIE YORI
Leaving the wreck of Baked Mountain Camp.

miles down the Valley. The violence
of the wind left us little control over our
bodily movements, and we dared not
attempt to flee down through that maze
of fumaroles in the darkness of the night.
How we watched the diminishing
shreds of our tent, and counted time
against the wasting fabric! Could it
shelter us until it was light enough to
leave? Time and again the wind would
pick us up as we braced our backs uni-
tedly against the wall and roll us into a
pile in the middle of the floor. I sat
on a flat valise weighing 50 pounds,
on which I had placed for protection a
big rock weighing as much more, yet the
wind repeatedly picked up the whole
bundle — man, suitcase, rock, and all —
and rolled us into the middle of the tent.

Digitized by

Google

230

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by L. G. Fobom
SORTING OUT THE WRECKAGE AFTER THE STORM AT BAKED MOUNTAIN

We looked and felt very much like rag-pickers on the dumps, but among the wreckage were many of

our most cherished possessions.

Yori, with characteristic self-sacri-
fice, chose the most exposed end of the
line. He drew a tarpaulin over his head
and shoulders in a vain effort to miti-
gate the blast. Soon it was the only
protection he had, and for some time
he sat practically outside while the wet
pumice was driven down his neck. He
must have been bitterly cold, but never
a complaint escaped him.

THE WIND PICKS ME CLEAR OFF THE
GROUND

Finally, at four o'clock, the gray
shapes of the fumaroles began to be
distinguishable in the general blackness.
I crawled outside and decided it was
little worse than within, for there one
could at least move around. So, though
all were fearful of being driven bodily
into a fumarole, it was agreed that the
time had come when we must try it
down the Valley.

Zies had a little flashlight which made
it possible to gather up the things we

most needed. But before they were
collected Allen came out and was almost
literally carried off before the wind.

Feeling that it was of the utmost im-
portance for safety's sake that we keep
together, I made frantic efforts to de-
tain him, but he was gone like a ghost in
the night, beyond recall, out of earshot
almost at once. Then I tried to hurry
the others. Zies and Fenner were ready
and started ahead but Charlie was busy
trying to save some extra conveniences
for the geophysicists, and I turned back
to urge him to come along.

To face the gale was impossible; so
I started backing toward the tent, brac-
ing with all my might against the wind.
Suddenly I found myself flying through
the air — I felt I was being carried off
down the Valley, God knew where.
My whole life flashed before me! The
gratitude I experienced when my face
landed in the mud only two gullies
away was inexpressible. My feet kept
on and nearly turned me another somer-

Digitized by

Google

THE TEN THOUSAND SMOKES IN 1919

231

sault, but I was safe on the ground
again!

FLIGHT TO UKAK

After that I decided it was useless to
try to go back» so I started on in an
attempt to catch Allen and slow him
down. Down the hill I sped, carried
to windward with little chance of choos-
ing my course. In the distance I spied
the gray smoke of one of the big fu-
maroles, dead ahead. I knew that I
must avoid the yawning chasm by which
it issued, if I were to live.

By desperate scratching I managed
to deflect my course a little, but slid
by so close that had there come just then
another gust such as had picked me up
I could not have avoided it. It was
some time before I could find Fenner
and Zies, who were waiting for Charlie
and me in a deep gully. They had seen
Allen go sailing by. We soon caught
him and proceeded down the Valley in
somewhat more orderly, if none the less
precipitate, fashion.

The wind came obliquely over Baked
Mountain, making it extremely diflicult
for us to avoid being carried out into the
middle of the Valley; whereas our safety
depended on keeping to the east side,
in order to make the ford of Knife Creek
above the point where it plunges into an
impassable canyon. For if we missed
the ford our flight would have ended in
a cul-de-sac among the canyons, from
which there would have been no escape.

Our legs fairly ran away with us.
Despite the accumulated weariness of the
day before and the strain of the night,
we made the trip to Ukak in record
time, covering the distance — nearly
ten miles — in less than an hour and a
half. Everyone jumped up to let us
crawl into the warm sleeping bags.

and we were soon sound asleep. The
gale continued all that day, but in the
night it calmed down enough to permit
us to visit the ruins on the following
day.

The site of our former happy camp
presented a scene of complete deso-
lation. Everything was covered with
pumice blown into the fallen heaps of
wreckage until no more could stick on.
Where we had rolled a big stone over
the sleeping tent to hold it down the
pumice had drifted a foot deep. It lay
so heavy on the tent that it took the
united strength of Yori and myself to
roll back the fallen roof, to expose the
things that lay on the floor.

The friction-top tin cans in which we
stored most of our supplies were scat-
tered all over that section of the Valley.
Fortunately for us the gullies had inter-
cepted most of them and their lids had
stuck tight. When we came to take
account of stock, we found the real
damage less than seemed possible. The
single cot which had served primarily
as a bench for keeping the photographic
materials off the ground was removed
from the fallen tent by ripping a hole
through the wall, and the scattered
effects gathered together and stowed on
it in a secure cache.

Meanwhile we found that another
party, camped at the old place under
Cerberus, had experienced no difficulty
with the gale. None of us who had
been through that experience would
have cared to try camping at Baked
Mountain again. But here was another
chance. Enough remained in the wreck-
age to permit a resumption of work on a
restricted scale. No one was slow to
grasp the opportunity, for the season
was far advanced and we knew it was
now or never.

Digitized by

Google

<
>

X
H

o
o

O
<
P
<

o
z,

232

Digitized by

Google

XVIII

A DESCRIPTIVE ACCOUNT OF THE VALLEY OF
TEN THOUSAND SMOKES

If one will turn upside down the map
reproduced on page 68 he will note that
the Valley occupied by the Ten Thou-
sand Smokes has the form of a gigan-
tic heavy-faced Y, stretching from the
mountain axis of the Alaska Peninsula
toward Naknek Lake. The height of
the Y, perpendicular to the trend of the
main range of volcanoes, is 13 miles
(20.9 km.). Its width across the top is
9 miles (14.95 km.). The width of the
stem and of the right arm of the Y is
about 2 miles; the left arm is approxi-
mately half as wide.

Each of the three extremities of the

Y is expanded into a heavy ceriph,
transforming the simple crotched Y,
such as might be described in the basin
of any forked stream, into a heavy
Roman Y of very different form.

The expansion at the bottom of the

Y is due to the fact that the Valley di-
vides at the base into two broad arms
which reach Naknek Lake independently
some ten miles apart. If, therefore, the
volcanic activity had continued beyond
the present terminus, the Smokes would
have assumed the form of an X still
more gigantic.

So broad and open, indeed, is each of
these lower valleys that travelers who
formerly traversed the Valley in crossing
Katmai Pass used to go sometimes by
one, sometimes by the other arm, de-
pending on weather conditions.

It is at the constricted area between
the two upper ceriphs of the Y that the
climax of activity occurs, for here are
located Novarupta and Falling Moun-
tain, as well as the most bewildering
mazes of smokes to be found in the
whole Valley.

The area enclosed between the two
arms of the Y is occupied by a highland
of which Baked Mountain and Broken

Mountain are the principal summits.
This is bisected by a small valley, which
extends up from the stem of the Y.
Like the main Valley, this area is active.
Its mouth is the seat of some of the
most vigorous vents in the region, and
smaller fumaroles occur farther up (see
page 294).

The main entrance to the Valley is
through a narrow branch which extends
from Katmai Pass down between Cer-
berus and Falling Mountain. There
is abundant evidence that this Valley
also at one time gave vent to numerous
fumaroles, though it is no longer active.

Near the summit of the Pass another
branch runs out between Cerberus and
Mageik, in which a considerable number
of small fumaroles were still active in
1917 and 1918. The nearest of these
were the first little smokes that lured us
on to the Pass on that memorable 31st
of July, when we discovered the Valley
(see page 190).

VALLEY UNDERLAIN BY SEDIMENTARY,
NOT BY VOLCANIC ROCKS

It is interesting and important to note
that the rocks of the Valley are not
volcanic, nor even igneous, but ordin-
ary fossiliferous sandstones and shales.
Where the lavas of the old volcanoes,
Mageik, Trident, and Knife Peak have
been poured out, the rock is, of course,
igneous. But elsewhere the walls and
floor of the Valley consist of sedimen-
tary rocks lying in nearly horizontal,
evenly bedded strata, almost as they
were laid down on the bottom of the
Jurassic sea which once occupied the
region.

To attempt a catalogue of the vents
or any description of the individual fu-
maroles would be futile. They vary all
the way from microscopic jets of steam

233

Digitized by

Google

g g

CO S

w s

r c
o a

c/)

234

Digitized by

Google

A DESCRIPTIVE ACCOUNT OF THE VALLEY

235

Up to mighty columns that overtop the
mountains. To explore the Valley
thoroughly and become acquainted with
the character of the various vents would
require much more time than has ever
been spent within its confines.

There are several square miles within
the active area which no one as yet has
so much as entered, much less examined
carefully. The north arm of the Valley,
under Knife Peak, which was supposed
from the cursory inspection of 191 7 to
be of only subordinate interest, was
found, for example, to be extremely ac-
tive, containing the most remarkable
craters of the whole Valley.

There can be little doubt, moreover,
that a more thorouj^h examination of
areas whose general features have be-
come familiar will yield further results
of importance. We were continually
finding new and interesting fumaroles
in places which we thought we had ex-
plored carefully. The extremely high
temperature previously mentioned was
discovered close to the trail, where we
had all passed scores of times without
noting anything unusual.

THE FISSURE FUMAROLES

By far the greater number of the fu-
maroles find vent through the fissures of
the Valley. These readily fall into
two classes: bands of marginal fissures,
several together, running around the
edge of the Valley in parallel lines; and
single central fissures which criss-cross
the floor in all directions.

While the marginal fissures are often
evident for a distance of half a mile or
more, the central fissures are usually
marked only by the lines of incrustations
or of small fumaroles stretched along
their courses. The fissures in the center
of the Valley are mostly closed, but
one may frequently peer far down into
those along the margin. They are
great clefts in the tuff which forms the
floor of the Valley. These narrow
chasms are sometimes 10 feet wide at
the surface, and extend to bottomless
depths, like the crevasses of a glacier.

For the most of their courses, however,
the fissures do not gape open, but are
closed over by curious bridges of strat-
ified ash which have been stretched
across in some manner. These arched
spans usually sag a little, revealing both
the width and direction of the tunnel
beneath. But they are remarkably
strong, considering how much they re-
semble the proverbial rope of sand, and
even when only a few inches in thickness
will support the weight of a large party.

We soon learned to cross and recross
such places indifferently, notwithstand-
ing the proximity of superheated steam
in the cavern beneath. It was indeed
the steam below that made the bridge
safe, for its incrustations partially ce-
mented the grains of ash together. After
a soaking rain, however, they sometimes
broke down.

One such cave-in occurred on the
bridge which we used continually in
going down to our cookstove. When
we came out the morning after a heavy
storm we found our path broken by a
yawning chasm 20 feet long and half as
wide, from which issued a new and im-
pressive fumarole.

Most of the great fumaroles that
fringe the edge of the Valley emerge
from just such openings. Since the
whole fissure for a long distance is full
of steam under pressure, it is evident
that any hole that is broken through
will become the seat of a new fumarole.
Variation in the number and appearance
of these fumaroles does not imply,
therefore, any change more fundamental
than the mere opening of new orifices
at the surface.

From some of the less active of these
fissures the vapor lazily rises in diffuse
clouds. From others it comes pouring
forth in great rolling, expanding columns,
reaching many hundred feet aloft or
drifting off before the wind for a mile
or more.

In other places, where a narrow crevice
is the only opening from a large cavern,
the steam may come out under pressure,
emitting a low hissing or roaring sound.

Digitized by

Google

PQ
Q

b«

•n

•s

§0

«•«

(2.

»«

•c

e
s

236

Digitized by

Google

A DESCRIPTIVE ACCOUNT OF THE VALLEY

237

If one tosses pebbles into such vents,
they are buoyed up by the rising steam
and are either immediately spewed out,
or sink slowly down, like feathers settling
to earth. It was at such a vent that we
fried our bacon, as described in Chapter

XVIL

DOMED FUMAROLES

Some of the largest and most con-
spicuous of the fumaroles come, not
from fissures, but from isolated eleva-
tions in the flat Valley, which may per-
haps be best described as low domes.
These appear to have been upheaved
after the general level of the Valley floor
had been fixed. They are confined to
the lower half of the Valley, and are
especially well developed at the junction
of the two arms of the Y.

The height of these domes is perhaps
50 feet and their diameter several hun-
dred yards, but the slopes are so gentle
that no definite boundary line can be
drawn around them. The principal fu-
marole comes from a hole in the summit,
whence cracks radiate in difi^erent direc-
tions. Surrounding the main vent are
numerous satellites of lower temperature,
which are interpreted as radial leaks
from the central conduit.

The domes give ofi^ tremendous quan-
tities of steam, making some of them
the most conspicuous objects in the
Valley. Sometimes the gigantic column
of "No. 32,*' the biggest of all, rises ver-
tically high in air, when all its lesser
neighbors are held by the breeze close
to earth.

CRATERS

The fissure fumaroles give no evidence
of ever having been the seat of any sort
of activity other than the present quiet
evolution of vapor, but there is another
class of craters which, though now steam-
ing quietly or extinct, were evidently at
one time explosive.

These vents have blown great funnel-
form pits in the floor of the Valley
(see page 292). Around the edge the
ground is commonly raised several feet

above the general level by debris thrown
out of the hole. The ejected material,
for the most part, consists of small frag-
ments, but occasionally includes lava
blocks up to three feet in diameter.

Some of these craters stand apart
without evident relation to other vents,
but by far the larger numbers occur in
two regions: there are a score or more
of them, mostly extinct, near the out-
side edge of the west arm of the Y under
Buttress Mountain ; in the north arm of
the Valley, under Knife Peak, they are
easily the dominant feature.

In these tracts they often stand close
together with intersecting edges, forming
compound craters. In a number of in-
stances they are so numerous and thick-
set as to have coalesced into veritable
nests of craters.

Two such nests in the north arm, now
extinct and filled with smooth floors of
inwashed pumice, are perhaps half a mile
long and half as wide. Throughout each
of these areas the floor of the Valley
was blown out completely, giving some-
what the impression of a single, gigantic
explosion crater. But the crenate mar-
gin, made up of the arcs of the individual
craters, testifies to its composite char-
acter.

In other cases the craters stretch out
in long lines, their position evidently
determined by the presence of a deep-
seated fissure in the Valley floor. In a
few instances where explosive action
has built up mounds around points of
special activity, the chain of miniature
volcanoes stretching along the fissure
furnishes a diagrammatic model of the
conditions believed to underlie the great
lines of volcanoes that girdle the earth
(see page 292).

STEAMING MUD AREAS

In most of the fumaroles the vapors
find exit through some evident break in
the surface of the Valley floor; yet there
is a large number where the escaping
gases have no manifest outlet, but seep
through the surface in innumerable
minute hissing jets. The diff^use activ-

Digitized by

Google

238

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Grins
THE TOP OF THE LAVA PLUG OF NOVARUPTA

Where the rock is in place in the center of the picture, the parallel concentric banded structure is

evident.

ity of these steaming mud areas often
covers an acre or more.

Throughout the area covered by the
"mud hissers** the ground is everywhere
full of steam, sometimes under pressure,
ready to escape through any new orifice
which may be broken through the sur-
face. The surface itself is generally
covered with a crust of some strength,
stained a rusty brown by iron oxide.

But beneath, the vapors (which are
never of a high temperature) commonly
keep the ground in the condition of soft
wet mud, having the appearance and
consistency of steaming hot asphalt, as
one sees it applied to a street in process
of being paved. The illusion is height-
ened by the curious "organic" odors
that are often liberated when the mud is
stirred up, but the chemists tell us that
the blackness is here due to iron sulphide.

STEAM RISING THROUGH UNDISTURBED
SANDSTONE STRATA

One of the most conspicuous features
of the Valley is the strictness with which

the activity is confined to the Valley floor.
All the big fumaroles come forth below
the "high water mark*' which is the
boundary of the Valley, as described in
the next chapter.

Careful examination showed, however,
that there were some exceptions to this
general rule of distribution. Steam was
seen coming from a number of places on
the slopes of the mountains which rise
above the Valley basin.

Upon examination of a number of these
high-level fumaroles, it was found that
the steam rose directly through the
sandstone strata of which the mountains
are composed. In some cases the sur-
face was covered with a mantle of frag-
ments which obscured the relations
somewhat, but in others it could be seen
with the utmost clearness that the
steam came directly out of crevices in
the undisturbed sandstone strata.

By far the greatest number of the
fumaroles coming from the sandstone oc-
curred on Baked and Broken Mountains,
though they were not confined to this

Digitized by

Google

A DESCRIPTIVE ACCOUNT OF THE VALLEY

239

locality, similar small steamers being
found across the Valley at the base of
Buttress Mountain. At the foot of the
Valley, also, in the vicinity of ** Three
Forks" (the junction of Knile Creek,
Lethe, and Windy Creek), we found
places in the underlying sandstone which,
though no longer active, had evidently
once been the seat of fumaroles.

To a thoughtful observer these fuma-
roles, despite their diminutive size, are
extremely interesting, for it is hardly
necessary to state that they could not
originate in the sandstone rocks from
which they emerge. Neither could such
activity be due to the gradual cooling
down of a great mass of material thrown
out by the recent eruption, as, conceiv-
ably, might some of the fumaroles of the
Valley floor, which undeniably rise
through a great mass of ejecta from the
eruption. The only possible source of
these fumaroles is a mass of magma in-
truded beneath the sandstone.

HOT SPRINGS

Hot springs, mud pots, geysers, and
the like are such familiar phenomena
in most regions where volcanic activities
are spread out over wide areas that we
naturally looked for them in the Valley
of Ten Thousand Smokes.

As for hot springs, nothing of the sort
was found in 191 7. The minimum tem-
perature observed was the boiling point,
so that all water appeared as vapor.

In 191 8 Say re and Hagelbarger found
a bathing pool at Three Forks, provided
"with hot and cold water, as a white
man *s bathtub should be. " A considerable
stream of water (95° C.) poured into the
cold creek in such fashion that one could
obtain any temperature desired simply by
moving from one side of the pool to the
other. The next year there were a num-
ber of such hot springs in the lower end
of the Valley, in areas known to have
had only fumaroles in 1917.

Among the most interesting of them
were numerous hot spots which we dis-
covered in 1919 in the bed of Ukak River,
while fording at the foot of the Valley

across ground formerly occupied by the
sand flow which came down during the
eruption (see Chapter xix).

So much hot water was poured into
the river at this point that, although it is
a considerable stream of swift water, it
was distinctly warm to the touch across
a good share of the channel.

In this connection it is worth noting
that, according to trappers, Iliuk arm
of Naknek Lake did not freeze over
until January 20, in the winter of 191 8-
19 19, a circumstance which suggests
that the total amount of heat delivered
by the inflowing streams may be very
large.

The chemists brought back some sam-
ples of the water of these recently formed
hot springs, but it has not yet been de-
termined whether they represent merely
the heating of surface waters by contact
with the hot ground or are of deep-seated
origin, like some long-established hot
springs.

MUD POTS

True mud pots of long-continued
activity, which are so characteristic of
the Yellowstone and similar places, have
not been found in the Valley. But
occasionally a sort of ephemeral mud
volcano develops, which for the time
furnishes a counterpart.

After one of the heavy rains common
in the summer months, streams of water
run freely over the hot surface of the
Valley. Naturally such water often
finds its way into the basins of some of
the little fumaroles, which, choked off^
from their usual method of escape, boil
and bubble up through the water, quick-
ly stirring up a thick soupy mud which
sputters and splashes with miniature ex-
plosions until the rain water is boiled
away, and the erstwhile mud pot be-
comes a steamer again.

Since it is the continuous disturbance
of the sputtering mud that gives them
their character, these mud pots make
but poor **stiir' pictures (see page 284).
In the motion pictures they are very in-
teresting objects.

Digitized by

Google

240

THE VALLEY OF TEN THOUSAND SMOKES

One other type of vent requires men-
tion. These, from their excrescent ap-
pearance, are what the boys have called
** pimples*' (see page 294). They are
confined to the lower half of the Valley,
where dust storms are frequent. Their
formation appears to be due to the
moisture which condenses around small
vents so as to catch and hold the blowing
sand, that is later cemented into place.
The process is repeated again and again
until a mound several feet in diameter is
built up over the vent.

These little fumaroles are so smothered
by their covering that they usually ap-
pear dead, but whenever we dug into
them we found them hot, proving that
the apparent lack of activity was due
largely to the covering of sand which
clogged the orifice and stopped visible
exhalations.

THE NEW VOLCANO, NOVARUPTA

There is only one place in the Valley
where the underlying magma from which
the Smokes take their origin has come
through to the surface in such a way that
its connection with the present activity
can be observed. This is in the throat
of Novarupta, a new volcano of typical
form and character which has arisen in
the floor of the Valley between Broken
Mountain and Trident.
• Here, surrounded by a crater ring 0.8
mile (1.25 km.) in diameter, evidently
formed by ejected materials, is a "dome"
of lava 800 feet in diameter and 200 feet
high that was pushed up through the
throat of the volcano in a stiff, viscous
condition (see pages 280 and 300).

The lava of the plug is a rhyolite whose
predominant color is a light gray. Su-
perficially it appears to be solid massive
rock, but on close examination it is seen
to be all blown full of minute bubbles
which, under a lens, give it a very loose
spongy texture.

In the process of cooling, this viscous
lava naturally developed strains which
shattered the upper portion into a con-
fusion of broken blocks, heaped topsy-
turvy over the surface, concealing for

the most part the bedrock and giving
the lava plug the appearance of a huge
stone pile (see page 238).

These great bowlders, standing in the
most precarious attitudes of unstable
equilibrium, would make the top of the
lava plug one of the most dangerous
places in the whole region, even if there
were no stifling vapors to interfere wit6
one's movements.

Many times, as Fenner and I walked
over it on the day we examined it to-
gether, we found great blocks, weighing
tons, balanced so nicely that they rocked
back and forth under our feet. There
was imminent danger of tumbling them
down on ourselves.

Once, a pile of rocks on which I was
standing suddenly crumbled, and when
it came to rest again I found myself
dropped down several feet. But as I
was on top of the moving bowlders, I
was in small danger of being caught be-
neath them.

In spite of the unstable rocks we man-
aged to examine nearly the whole of the
plug, and that without very great diflft-
culty; but an hour later, as we looked
down on it from above (see page 280), it
seemed altogether beyond the bounds
of possibility for anyone to pick his way
through the thick clouds that rose every-
where from among the loose rocks.

As we looked, we wondered whether
we had been specially favored by atmos-
pheric conditions at the time of our as-
cent, or whether the plug was really
easier than it looked from a distance.

Like most of the ejecta of the recent
eruption, the lava of the dome is not pure
rhyolite, but contains a considerable ad-
mixture of basic andesite, like that com-
posing the old volcanoes, as well as a
small quantity of fragments of the sur-
rounding sedimentary rocks.

These foreign inclusions, which show
clear evidence of having been partially
digested by the new magma, are arranged
in striking dark bands, alternating with
the lighter rhyolite. Where the rock of
the dome can be seen in place, it be-
comes evident that the bands are con-

Digitized by

Google

Natural Color Photograph by Frank I. Jones
MOSAICS OF ALUM DEPOSITED IN AN AREA OF DIFFUSE ACTIVITY
When the steam finds no open vent it permeates the whole ground over a considerable area, escaping in

innumerable minute jets.

Digitized by

Google

Natural Color Photograph by Frank I. Jones
THE GAUDY MUDS OF THE VALLEY MAKE GOOD PIGMENTS FOR A PAINTER
With no other materials, very creditable pictures have been painted.

XII

Digitized by

Google

A DESCRIPTIVE ACCOUNT OF THE VALLEY

241

centric, running tangentially, parallel
with the circumference of the lava plug,
thus affording some indication of the
manner of action of the forces of extru-
sion (see page 238).

The blocks of lava were evidently
broken apart before they had entirely
solidified ; their surfaces frequently show
evidence of flowage after breaking up.
This is especially striking where the
banded structure is well marked, for
there it is evident that the rhyolite was
more fluid than the andesite.

In many places the bands of rhyolite
stand out beyond the alternating inclu-
sions, giving the blocks a fluted appear-
ance. When this is developed at the
corner of a block, a striking cornice struc-
ture results (see page 296).

The greater fluidity of the rhyolite is
both surprising and significant, since its
melting point is undoubtedly higher than
that of the andesitic inclusions. The
explanation of the anomaly lies, I be-
lieve, in the fact that the andesite in the
dark bands was *'dead" lava in process of
digestion by the new '*live" magma — a
matter whose interpretation is discussed
in some detail in Chapter xxii.

FALLING MOUNTAIN

While the attention of the visitor who
comes into that vicinity for the first time
is certain to be absorbed by the sight of
Novarupta, he could not gaze long upon
its steaming lava plug without inter-
ruption.

Within a few minutes he would be
startled by a crash behind him and,
turning, would see an avalanche of fall-
ing stones, which had started high up on
the face of the mountain side, galloping
down the precipitous slope toward him.

Whenever it has been under obser-
vation, ever since its discovery in 1916,
this rock face, which we named Falling
Mountain, has been subject to frequent
avalanches. The constancy of their oc-
currence is one of the most impressive
features of the Valley.

The falls of rock vary in size all the
way from single bowlders up to large

masses weighing many tons, which let
go with a crash distinctly audible for
miles around. During the first two sea-
sons, a five-minute period never passed
without a fall of some sort, but in 1919
the mountain was often quiet for half an
hour at a time.

FALLING MOUNTAIN AS A FOG SIGNAL

The first time we heard it in 1919 was
memorable. **Biir' Henning and I had
started for Katmai Pass in a fog, from
a camp pitched in the valley between
Baked and Broken mountains. Neither
had ever been over the ground before,
and the fog was so thick that we could
not see 20 yards.

We took a compass-bearing from the
camp and started out over '* Greased
Hill,** which connects Baked and Broken
mountains. As usually happens in a fog,
the land took on an unexpected aspect
before we reached the divide, and we
soon became uncertain of our direction
and suspicious of our compass.

When we started down the far slope,
we ran into a series of hot steaming fis-
sures directly athwart our course. In
reality, they were not very deep but in the
dense fog we could not see the bottom.
The slippery, variegated muds reeking
with steam made progress both difficult
and uncertain. It seemed that we were
in danger of sliding straight into '*the
bottomless pit.'*

We had traveled thus, working our
way among the fissures for what seemed
a long time, when suddenly, out of the
distance, a great booming rock fall let go
on the slopes of Falling Mountain, con-
firming us in our course and setting us
at ease once more.

CATACLYSM BEGAN WITH A GREAT SLIDE
FROM FALLING MOUNTAIN

Examination shows that Falling Moun-
tain is a counterpart of Noisy Mountain,
discovered later (see page 127), and of
the cliff from which the Mageik Land-
slide originated (see Chapter xi). It
is obvious from a glance at the great
gouged-out face of the mountain that an

Digitized by

Google

242

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Grisg*

TAKING THE TEMPERATURE OF A HOT ONE

Many of the vents were far beyond the range of the thermometers we carried the first year.

pare the picture on the opposite page.

Com-

enormous mass of rock has fallen away
recently — a mass quite comparable with
the Mageik Slide or with the Noisy
Mountain Slide.

But, although the broken cliffs give
the clearest evidence that such a slide
occurred, one looks about in vain for the
debris. Nothing but a few rock cones
are to be seen on the smooth-floored
Valley. It is as though the whole side
of the mountain had not only fallen from
its place, but had been swallowed up in-
to the bowels of the earth.

In a sense that is exactly what has
happened, for we can only conclude that
the slide debris is so deeply buried under
the ejecta of the eruption that the form-
er irregularities of its surface have been
entirely concealed. In other words, this
slide must have occurred before the out-
pouring of the ejecta. It is thus the
first event in the recent volcanic disturb-
ance of which we have record.

As one watches the frequent rockfalls
which come bounding down the face of
the mountain, he gains, as perhaps is
possible in no other way, some realiza-
tion of the magnitude of the great mass
of material that fell in the early slides.
For while the present avalanches often
involve the fall of an impressive mass of
rock, the accumulated pile of fragments
at the foot of the mountain has not
grown appreciably in the four years
since its discovery. It is now, as then,
utterly insignificant in comparison with
the total volume of the material that
must have fallen away.

FUMAROLE ACTION ON FALLING MOUNTAIN

In addition to its frequent avalanches,
Falling Mountain possesses another fea-
ture of considerable interest. Far up
on its bare rock surfaces are a number of
small fumaroles, the highest vents in the
region (except the craters of the old vol-

Digitized by

Google

A DESCRIPTIVE ACCOUNT OF THE VALLEY

243

Photograph by R. F. Grigss

THE SAME VENT FROM A LITTLE DISTANCE

The steam is so hot that it does not condense until ten feet from the vent. The scale may be judged
from the little steamer enclosed by the circle which was our cookstove (see page 196).

canoes) through which the subterranean
gases reach the surface.

The petrological studies made by Fen-
ner show that the gases from these, com-
paratively small though they are, have
wrought notable changes in the andesitic
rock through which they have come,
apparently reacting with the constituent
minerals in such a way as to form vola-
tile compounds, "so that quantities of
material have actually been removed by
gaseous transfer." ^

Many of the pyroxene crystals have
been replaced by aggregates of hematite
scales, while tridymite was abundantly
formed. As a result of this action, the
rock has lost much of its cohesive
strength, which fact may have some
bearing on the continuance of rock falls.

BAKED AND BROKEN MOUNTAINS

One of the most significant indications
of the character of the subterranean dis-
turbances at the time of the eruption is
the present condition of Baked and

Broken mountains, which occupy the
central space between the arms of the
Y-shaped Valley, entirely surrounded
by active vents. These mountains,
which rise 1,500 to 2,000 feet above the
Valley floor, are, like the Valley itself,
not volcanic, but carved out of horizon-
tal sandstone strata.

Their mass is now criss-crossed by wide
gaping fissures, accompanied by faults
whose individual displacement is as
great as 50 feet. These fault scarps
show the deepest deposits of ash and
pumice found anywhere in the district.
In the sections here exposed, 50 feet of
stratified ash may be seen.

The scarps do not, however, cut
through to the original surface, so there
is no means of telling how much greater
the total thickness of the ashfall may
have been (see page 244).

Where the slopes of these mountains
are so steep that the ash has slid off^ and
exposed the rocks beneath, they show
a curious change wrought by the erup-

* C. N. Fenner, Journal of Geology, vol. 27, p. 586, 1920.

Digitized by

Google

244

THE VALLEY OF TEX THOUSAND SMOKES

Photograph by R. F. Griggs

A FAULT SCARP ON BROKEN MOUNTAIN

The face of the scarp is composed of stratified ash mostly from Novarupta. The sequence of the
ash strata shows that Novarupta burst forth before the explosion of Katmai, but this faulting occurred
after the eruption.

tion, their gray-green sandstone having
been turned a dull brick-red . The chem-
ical change that took place is exactly the
same as that which occurs when clay
blocks are burned to red brick. But it
is not certain that heat was the only
agency at work in the coloration of Baked
Mountain. Some of these cliffs are so
extensive and conspicuous that their
character may be seen from more than a
mile away.

FISSURE LAKE

Near the foot of the glaciers that
wind down from Mageik occurs the
largest fissure to be found anywhere in
the Valley. It is 200 to 400 feet wide,
with perpendicular walls, one of which
stands about 35 feet higher than the
other. The depth could not be ascer-
tained, because it was filled by a beauti-
ful lake of clear, green water.

Standing just at the foot of the glaciers,
Fissure Lake when first seen was one of
the most picturesque spots in the whole

Valley. Along the sides were numerous
snowdrifts, from which miniature bergs
broke off and floated away (see page
212).

Unfortunately, the proximity to the
glacier which gave the place its greatest
beauty will result in its early extinction.
The lake of necessity receives a large
amount of glacial debris, and will be
filled up in the course of a few years. A
large delta had, indeed, been built out
into it in 1919, considerably impairing
its charm.

WARM WATER FROM SNOWDRIFTS

Fed by the glaciers and melting snows.
Fissure Lake would be expected to be
icy cold, but on the contrary it was de-
cidedly tepid in spots, where heat was
evidently received from below. One of
the most amusing incidents of the whole
work occurred when Shipley, the chem-
ist, poking his thermometer into every-
thing, discovered this fact.

I was coming along a little behind and

Digitized by

Google

A DESCRIPTIVE ACCOUNT OF THE VALLEY

245

Photograph by R. F. Griggs

THE RIVER LETHE CROSSING A LINE OF FUMAROLES

The steam in places actually bubbles up through the cold water. The volcanoes and the surface
drainage manifestly have no connection with each other.

he, pretending to need my assistance,
asked me to tell him the temperature of
the water coming out from under the
edge of a snow field. Willing to answer
even a foolish question, I had the words
" ice cold " on the tip of my tongue when
my fingers touched the water. The
speaking expression froze on my face
and I carefully dipped my hand in again.
It was actually warm! How he did
laugh at my discomfiture!

DRAINAGE WATER NEARLY ALL VAPOR-
IZED IN THE HOT VALLEY

The snow fields which surround the
Valley send trickling rills down its slopes,
but these disappear long before the floor
of its basin is reached.

Even the streams from the half dozen
good-sized glaciers which discharge into
the Valley were able to hold their courses
clear across the hot area only during
the warmest months, when the melt is
greatest, and then in such reduced vol-

umes that they were often scarcely more
than muddy trickles at the lower end.

But in early June, when we first en-
tered the Valley in 1919, the bed of the
River Lethe, half way up the Valley, was
one mass of hissing fumaroles, without a
sign of the muddy torrent which coursed
across the hot ground a month later.

It was very striking, and at first sight
surprising, to see how little the direction
of the waters was affected by the fuma-
roles. As in ordinary streams, their
courses were controlled by gravity, re-
gardless of the position of the fumaroles
which they often approached closely or
even overran. When this happened the
waters quickly washed in a thick deposit
of ash, effectually blanketing the fuma-
role, whose vapors probably found a
means of escape through some other
outlet.

In some places, however, we found the
steam still bubbling up through the cold
water. But the waters of the river

Digitized by

Google

246

THE VALLEY OF TEN THOUSAND SMOKES

passed the hot spots quickly and were
only slightly heated by their close con-
tact with very large hot areas. We
found that the River Lethe was still
48** F., although the ash of its banks
stood above the boiling point.

On the other hand there is no clear
evidence that the steam of the fuma-
roles is sensibly increased by the addi-
tion of surface water, for, as observed
on page 219, it does not appear to fluc-
tuate with the seasonal variations in
drainage.

A BEWILDERING COMPLEX OF ODORS

The mingling of the many substances
rising through the vents results in an ex-
tremely curious combination of odors,
which Dr. Shipley, with the trained nose
of a chemist, thus describes:

"As we entered the Valley along a
deep, dry watercourse, we observed from
time to time a peculiar, indefinable and
not unpleasant odor. Passing close to
the active vents, the odors of hydro-
chloric acid and hydrogen sulphide
could be detected easily. From certain
of the active areas a disagreeable smell,
unlike any odor that we had ever en-
countered, arose. It was somewhat sug-
gestive of a pigsty, a horse stable, and
sewer gas, yet we could not relate it defi-
nitely to any previously observed smell.

''Whatever the gases are, that rise
from the vents in the floor of this won-
derful Valley, collectively they offer a
considerable task to the olfactory organs
in differentiating the known from the

' The statements herein are summarized from
the followmg papers by the chemists:

Shipley, J. W. Some Chemical Observations
on the Volcanic Emanations and Incrustations in
the Valley of 10,000 Smokes, Katmai, Alaska.
Am. Jour. Sci., vol. 50, pp. 141-153, 1920.

. The Nitrogen Content of Volcanic

Ash in the Katmai Eruption of 19 12. Ohio Jour-
nal of Science, vol. 19, pp. 213-223, 1919.

. The Water-Soluble Salt Content, the

Ferrous Iron Content and the Acidity of Katmai
Volcanic Ash. Ibid., pp. 224-229.

. Ammonia and Nitrous Nitrogen in

the Rainwater of Southwestern Alaska. Ibid.,
pp. 230-234.

Allen, E. T. Chemical Aspects of Volcanism

unknown. At a distance of 20 miles
from the Valley, one was certain one
moment that the gas was sulphur dioxide
which the wind bore to him, the next
moment it was hydrogen sulphide, and
the next both or neither.

''This same elusive uncertainty clung
throughout the whole period of our stay
in the Valley. It was only in the vicin-
ity of a vent that the individual gases
could be identified with certainty by the
sense of smell."

THE CHEMISTRY OF THE FUMAROLES

A thorough chemical study of the
fumaroles would undoubtedly greatly
assist in the interpretation of the dy-
namics of the volcanoes. The difficulties
in the way of such a study are so great,
however, that as yet only a good begin-
ning has been made.*

Steam is the predominant constituent
in all the gases examined, whatever the
temperature. In the samples collected
by Allen and Zies it varied from 98.4 to
99.99 per cent by volume. The most im-
portant remaining gases are hydrochloric
acid, carbon dioxide, hydrogen sulphide,
nitrogen, hydrofluoric acid, and some-
times marsh gas. Minor constituents
are oxygen, carbon monoxide, argon, and
ammonia.

Although part of the steam is un-
doubtedly derived from the vaporiza-
tion of surface drainage, as stated above,
much of it must be truly volcanic in
origin* (see also Chapter xvii).

The ammonia comes up largely in the

with a Collection of the Analyses of Volcanic
Gases. Journal, Franklin Institute, vol. 193,
pp. 29-80, 1922.

' Allen and Zies have shown that the remaining
volcanic gases by themselves could not possibly
be hot enough to vaporize surface water in the
quantities that issue from the fumaroles.

There is in this region no relation between the
temperature of a vent and the composition of its
gases. Deville's generalization that there is a
progressive increase in the percentage of steam
and of carbon dioxide with declining activity and
falling temperature finds no support here. Nor
is there any simple system of temperature dis-
tribution, except, as would naturally be ex-
pected, that high temperature vents do not

Digitized by

Google

A DESCRIPTIVE ACCOUNT OF THE VALLEY

^47

form of the chloride, sal ammoniac,
which may sometimes be gathered in
masses as pure and white as when pur-
chased in a drug store. (One sample
analyzed was 98 per cent pure.)

AN UNPARALLELED PERCENTAGE OF
FLUORINE

The most surprising chemical discov-
ery was the abundance of fluorine. The
hydrofluoric add in the fumes attacked
the collecting apparatus, etching it until
it looked as though made of ground glass.
Even one of the vacuum tubes was at-
tacked in the same fashion, the relatively
minute quantity of gas it contained hav-
ing enough hydrofluoric acid to destroy
the polished inside surface.

The majority of the incrustations were
found on analysis to contain fluorine.
In one soft deliquescent orange deposit
Shipley found 7 per cent of it. Such high
percentages of fluorine have never been
reported previously.

Sulphur is, of course, abundant. It
occurs most frequently in small veins —
cracks in the mud which have been filled
up. Rarely such veins reach a thickness
of six inches. The throats of many fuma-
roles are lined with the most beautiful
delicate needle-like outgrowths of sul-
phur crystals, appropriately named flow-
ers of sulphur.

In 191 7 we found one large deposit
where sulphur crystals could have been
gathered by the bushel, but by 191 9 it
had disappeared. Sulphuric acid, which
is troublesome in many volcanic areas, is
uncommon in the Valley, but Shipley
found one fumarole in which it reached a
high concentration.

Arsenic was discovered associated with
sulphur as an incrustation over several
of the fumarole cracks. The mass con-
sisted of yellow crystals of orpiment —
^'King's Yellow" (arsenic trisulphide
As2Ss) — with some red crystals which
may have been the other sulphide of

occur around the edges. In other words the
Valley does not conform with the central type of
volcano, where the climax of activity occurs in
the main crater and its intensity decreases stead-

arsenic, realgar (AS2S2). These deposits
analyzed 19 per cent arsenic.

Corundum^ aluminum oxide (AliOs),
which in impure form is used as emery,
was found as small crystals in several of
the incrustations.

Potassium alum was perhaps the most
conspicuous incrustation, forming strik-
ing lichen-like deposits in all the active
areas. These lichenoid growths would be-
come very conspicuous after a few hours
of good weather, only to disappear when
the rain began again (see color plate).

Iron is one of the most abundant ele-
ments in the Valley, and is largely re-
sponsible for its beautiful coloring. Well-
developed crystals of pyrite (FeS2) were
found in several places. In other situa-
tions the mud had the lustre of fool's
gold, presumably imparted by myriads
of tiny crystals of pyrite.

Considerable deposits of 'Venetian
red" (Fe208) were forming in one gully
examined by Shipley. In another fu-
marole discovered by Zies a quantity of
good-sized black crystals of magnetic
iron ore (Fe804) had been deposited, sub-
limed by the hot vapors.

PETROLEUM-LIKE EMANATIONS

Another remarkable and puzzling as-
sociation of products was found by both
Shipley and the geophysicists in the vi-
cinity of Novarupta. The gases here con-
tain considerable amounts of marsh gas
(CH4), amounting in one case to 12 per
cent of the total gases aside from steam.

The interest of this finding is that
marsh gas, which is a principal constitu-
ent of natural gas, is generally assumed
to be of organic origin. Its significance
here is increased by the occurrence of
other products, discussed by Shipley as
follows:

"A most interesting organic deposit
resembling coal tar in odor and appear-
ance was observed saturating the ash
around fumaroles on the western slope of

ily with the distance from that point. The wide
distribution of activity here indicates rather the
close approach of the magma to the surface over
a large area.

Digitized by

Google

248

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by E. G. Zies

STEAM RISING FROM HOT SPRINGS IN THE LOWER END OF THE VALLEY

These streams were running cold in 1917 and 1918. At that time the hot springs of 1919 were
f umaroles, whose steam escaped into the air without warming the surface waters. The terraces in the
background are the exact counterpart of those often found in large river valleys (see text page 153).

Novarupta volcano and in the immediate
neighborhood of the NH4CI [Ammonium
chloride] fumaroles.

" One fumarole in this same area was
emitting nitrous acid. The gases from
fumaroles in this area possessed a similar
tarry odor and emitted much H2S [Hy-
drogen sulphide]. On analysis of the im-
pregnated ash the deposit proved to be
of a very complex character.

" A 20-gram sample was extracted with
the following solvents successively until
no further extract from each was ob-
tained. The solvents in the order used
were: ether, petroleum ether, carbon di-
sulphide, benzene, carbon tetrachloride,
ethyl alcohol and glacial acetic acid.
Each of the solvents extracted organic
matter varying in amount from half a
gram to a centigram, CS2 [Carbon bi-

sulphide] and (C2H6)20 [Ether] being the
most effective. Several of the solvents
extracted considerable quantities of free
sulphur.

'*On evaporation, a brownish-black
resinous mass was left by each solvent.
None of the extracts belonged to the
aliphatic series; none contained nitrogen
and none were unsaturated.

"Steam distillation carried over the or-
ganic matter together with the free sul-
phur, and the first portion of the distillate
gave an unmistakable odor of naphtha-
lene.

"The occurrence of this deposit of
aromatic hydrocarbons may have some
significance in the theory of volcanic
origin for such compounds. The de-
posit lies close to the most active area of
the district and is associated with NH4CI

Digitized by

Google

A DESCRIPTIVE ACCOUNT OF THE VALLEY

249

and sulphur, while H2S is unusually con-
centrated in the escaping gases.

" On the other hand they might equally
well be attributed to steam distillation
from the sedimentary rocks with subse-
quent condensation in the ash on the
surface. The Jurassic sandstones meet
the igneous intrusion in the immediate
neighborhood.'* *

The origin and relationships of these
various products are most interesting
and deserve much further study. It is
not possible at the present time to sug-
gest any satisfactory explanation of
those of the ** organic*' series.

THE COLORS OF THE DEPOSITS

The members of the expedition of 1919,
while agreeing that the published ac-
counts described the fumaroles accu-
rately, were unanimous in the opinion
that the printed accounts had given them
no idea of the colors of the deposits.
The previous articles had, to be sure,
stated that the ground was painted
**with all the colors of the rainbow" and
that the *' fissures were baked bright red
for miles at a stretch,** but somehow it
was imagined that these must be wild
exaggerations rather than careful scien-
tific statements.

No one was more impressed than Kolb,
who, having lived for years on the brink
of the Grand Canyon, was not oversus-
ceptible to striking colors. Along with
most people, he had supposed that the
Canyon represented the climax of na-
ture's coloring, and had never expected
to see anything more brilliant.

The coloring is, however, so altogether
different in character from that of the
Canyon that the two cannot properly be
compared. In the Canyon it is in the dis-
tance that the color is most remarkable.
In the Valley it is in the foreground.

There are no bright and sharply con-
trasting masses of rock in the walls of
the Valley. On the contrary, the Valley
itself, seen under the heavy clouds which
usually cover it, is apt to convey an idea
of sombreness. But, when one comes

* Quoted from Shipley, American Journal

Up to the fumaroles, he is almost over-
whelmed by the indescribable riot of
color. Not so in the Grand Canyon ; its
rocks are ordinary enough when closely
approached. I have before me two vials
— one given me by the keeper of the
Bright Angel Trail, containing, according
to his statement, sands of all the colors of
the Canyon, the other filled with mud
from the Valley. The one is so ordinary
as never to attract a second glance; the
brilliant hues of the other arouse uni-
versal interest.

The colors of the Canyon are due al-
most entirely to the wonderful atmos-
phere and the brilliant light which floods
its recesses.

Another reason why the Valley as a
whole fails to show much color is that
when the hues of the spectrum lie close
together, they blend into neutral grays
or browns. But, while largely losing their
effectiveness at a distance, the colors are,
for this very reason, all the more strik-
ing when seen close up, for then each is
heightened by contrast with the other.

The throats of the fumaroles and the
ground around them are most often
burned into some of the various shades
of red which are familiar in the different
tints of brick in common use. Some-
times it will be a light, pinkish tone;
again a bright scarlet; or, in still hotter
places, a dark, rich crimson, passing into
purple and black in some of the very hot-
test vents. With such deep-burned pur-
ple patches there is frequently associated
a deliquescent incrustation of brilliant
orange. These two colors are most often
found together in fumaroles bearing a
high concentration of acid which has
eaten away all soluble constituents, leav-
ing the general ground snow-white.

In some places considerable areas are
thus leached to a gleaming whiteness by
fhe acid fumes. When to this are added
faint tinges of pink and yellow, there
results a beautiful flesh color of the great-
est delicacy. Again, the ground color
of red, white, or gray may be overlaid
with copious deposits of yellow sulphur,
of Science, vol. 50, pp. 147-148, 1920.

Digitized by

Google

250

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by L. G. Polsom

KOLB IN THE ROLE OF SAMSON

Balancing such a "rock" in the air is not so much of a feat as it appears, for the mass is a chunk of
pumice blown full of bubbles by the gases of the eruption. It might even float on water. Such large

gieces of pumice are confined to the vicinity of Novarupta. All the ejecta of Katmai itself are very
nely divided, because of the greater violence of its explosion.

THE TRAVELER S SHOES DAUBED WITH
BRILLIANT PIGMENTS

Bright colors are by no means limited
to the large and vigorous fumaroles.
Over many hundreds of acres in the
Valley, where steam is everywhere seep-
ing up from beneath, the ground becomes
a multicolored mud with the most bril-
liant colors imaginable.

Sometimes it is black, from iron sul-
phide, with the character and consis-
tency of asphalt. This grades through
various shades of blue into delicate light
pearl grays, or alternates (as the chem-
ical conditions change from reducing to
oxidizing) with the various shades of red,
due to iron oxide. Round about are the
more ordinary ochraceous yellows and
browns.

Digitized by

Google

A DESCRIPTIVE ACCOUNT OF THE VALLEY

251

After a trip across such an area, one *s
shoes, covered with the particolored
muds, take on a resemblance to an art-
ist's palette, daubed with every pos-
sible color in a confused medley.

The fine-grained mud is indeed so sim-
ilar to artist's pigments that it may
readily be used as a substitute for them.
With no other materials than mud from
the Valley, and a piece of canvas from
a ruined tent, Kolb painted several ex-
cellent pictures.

Where the ground is not too hot to
prevent their growth, the patches of
steaming mud are covered with a layer
of bright green algae, adding the last
color needed to complete the spectrum.
One of the most striking color combi-
nations in the whole Valley was produced
by a green carpet of such algae in one
of the great conical craters where the
general rock color was a bright orange-
ochre.

In another place I came across a trail,
made a few days previously, where the
depressed tracks had served to collect
a little water which had favored the
growth of algae, making the tracks stand
out green against the general brown sur-
face. Occasionally a green, perhaps one
of the salts of copper, is found among
the mineral incrustations, but such de-
posits were never seen in quantity.

In other places one may find the most
astonishing combinations of reds and
blues and yellows by digging into the
incrustations in the vicinity of a fuma-
role. The brilliance of the colors of such
freshly exposed blocks is indeed quite
beyond belief. One can simply stand
and marvel at the never-ending variety
of hues he uncovers, for each block is dif-
ferent from all the rest (see color plates).

THE COLOR PHOTOGRAPHS

To portray the gorgeous coloring by
verbal description is impossible. Even
pictures colored on the ground as care-
fully as might be would not serve. The
most gaudy mixture of colors that could
be daubed over the canvas would not

surpass the shrieking effects presented
by the impressionistic artists as their
idea of the coloration of ordinary land-
scapes; wherefore many would doubt
the accuracy of any painted pictures.

Ever since I first beheld the display
of colors in the Valley I had been anxious
to record them by color photographs,
which should present what we had seen
without the possibility of the personal
equation entering in.

The problem was by no means easy of
solution, for the obstacles to successful
color photography, which at best are
difficult to overcome, become greatly
intensified in such a region as the Ten
Thousand Smokes.

The plates are sensitive to adverse
climatic influences and must be guarded
from the hot damp ground with the most
jealous care. The dust clouds fre-
quently stirred up by the wind make it
extremely difficult to keep things clean,
and dust is much more serious in color
photography than in ordinary black and
white work. Films can be changed just
before exposure, but the plates must be
loaded beforehand. On a black and
white picture it is easy to touch out a
spot, but in the Paget process which we
used any imperfections on taking-screen,
plate, or viewing screen, must remain a
permanent blotch on the picture.

Success in ordinary photography is
dependent on one's ability to regulate
two or three mechanical processes —
shutter-speed, diaphragm, and focus —
and at the same time to watch the sub-
ject and make the exposure when con-
ditions are exactly right. But, in color
work, the number of mechanical factors
is increased and the demands of artistic
conception on the part of the operator
are far greater.

It is, therefore, a rare man who can do
such work successfully in the rough-and-
ready conditions under which we were
forced to live. The results are so beau-
tiful and preserve the natural colors
with such fidelity as to reflect great
credit on the artist.

Digitized by

Google

c
o

K
O

<
>

c
'•2 .

'£8

CO OS

O CO

bfi O
3 u

«- «

« E

^ 3

252

Digitized by

Google

XIX

THE GREAT HOT SANDFLOW OF THE VALLEY

The first thing that attracted our
attention as we explored the fumarolic
area — when we could look beyond the
Smokes themselves — was a curious line
that encircled the Valley almost like the
high water mark of a flood. Above
this line were the ordinary gray ash
slopes; below it, covering the Valley
floor, a great mass of compact, firm,
terra-cotta tuff, which has no counter-
part on the surrounding hills.

The interpretation of the character
and origin of this tuff was, for a long
while, our chief concern as we explored
the Valley and its branches, for its
evident peculiarity had at once excited
our curiosity.

Surrounded as it is by high and rugged
mountains, the most striking feature of
the conformation of the Valley of Ten
Thousand Smokes is the flatness of its
floor. One could ride a bicycle for miles
along its smooth surface, and there are
many places between the lines of activity
that would be ideal landing fields for air-
planes.

IMMENSE DEPOSIT OF TUFF ON THE FLOOR
OF THE VALLEY

The traveler peers into the canyons cut
by streams in the hope of finding some
clue to the thickness of the deposit, but in
this he is disappointed. In such places
one finds cuts 40, 60, or even 100 feet
deep, but none of them, except those near
the foot of the Valley, reach its base
and reveal the character of the ground
beneath. Inasmuch as most of the
trenches are located near the edges of
the Valley, where the deposit is relatively
thin, it may be inferred that the total
thickness is very great. An attempt
to reconstruct in the mind's eye the
original contours of the Valley before
the tuff was poured out leaves the

impression that its maximum thickness
can hardly be less than 500 feet (see
page 264).

This is very puzzling and very im-
pressive, for when it is observed that the
tuff maintains a great thickness over
wide areas it becomes evident .that the
volume of the formation is enormous.
The great quantity of this material was
further emphasized as we extended our
exploration through the Valley and found
it spread over the whole of the active
area.

First we discovered that the branch
valley between Baked and Broken
mountains was covered with tuff. Then
we saw the great valley between Broken
Mountain and Knife Peak filled in the
same way, clear back to Novarupta
Volcano, connecting with the main valley
and forming a complete circuit (see map
page 68).

Later we came to recognize that the
massive '*mud flow** we had found as
we climbed up to Katmai Pass was part
of the same formation.

From the point where we first en-
countered it, back of Observation Moun-
tain, to its extremity down Ukak Valley
is a distance of 20 miles (32 km.). From
the divide back of Novarupta to the
tip at the foot of the Valley is 15 miles
(24 km.). The greatest width is about
9 miles (14 km.).

Altogether it occupies an area of 53
square miles (137 sq. km.). At its high-
est points, near Novarupta and in Kat-
mai Pass, it reaches an altitude of nearly
3,000 feet, while at the tip it extends
down to within about 200 feet of sea level.
If its average thickness over this area is
only 100 feet, its volume is a full cubic
mile.

The bulk of the deposit is composed of
fine fragments, many of them dust-like,

253

Digitized by

Google

254

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by Frank I. Jones

KNIFE CREEK CANYON

The streams have cut most curious sinuous canyons into the
stiffened mass of the sandflow. Although in places these canyons
are a hundred feet deep, they do not cut through the flow to the soil
beneath, except in the lower part of the valley.

but there are included numerous lumps
of pumice which in places make up a con-
siderable fraction of the whole. There
is no trace of stratification of the mate-
rials, except where they were obviously
subject to secondary readjustments after
deposition. Where cut into by erosion or
sheared by faulting the tuff breaks with
irregular fractures, running in any direc-
tion according to the stress. From the
character of its cleavage, it is often broken
or weathered into fantastic blocks like the
hoodoos of the bad lands (see page 259).

VALLEY SURROUNDED BY

A CONSPICUOUS "high

WATER mark"

The'* high water mark"
above alluded to, which
separates the chocolate
tuff-covered basin of the
Valley from the gray
ash slop)es of the moun-
tains, impresses one more
and more as he studies
its relations. It follows,
in a general way, a con-
tour line 200 to 300
feet above the floor of
the Valley.

Although it thus re-
minds one of the shore
line of a pond, it is ob-
viously not quite level,
like the shore of a body
of water, but slopes grad-
ually toward the foot of
the valley, as though due
to the movement of a
heavy liquid like mo-
lasses, which had suc-
ceeded only imperfectly
in finding its level.

The most conspicuous
example of this occurs in
the southwest comer of
the Valley, under the
glaciers of Mageik, where
the ''high water mark"
is more than a hundred
feet lower than farther
east along the foot of
Mount Cerberus (see page 262).

All around the margin of the Valley,
just below the ''high water mark,"
runs the series of gaping marginal fissures
described in the last chapter, as though
the surface had been stretched by sub-
sidence after its formation.

The situation reminds one of a tem-
porary puddle which, after having frozen
over heavily during a rise in a river, is
drained again so as to let the ice down on
to the bottom, with consequent stretch-
ing and cracking all along the shore.

Digitized by

Google

THE GREAT HOT SANDFLOW OF THE VALLEY

255

The analogy is carried
further by the criss-
cross fissures which run
in all directions across
the middle, like the
contraction cracks of a
frozen pond. Where
the ridges from the
shoulders of the moun-
tains project, the mar-
ginal fissures coalesce
and run far out into
the Valley, in just such
a Y form as do the ice
cracks at the point of
a peninsula in our
drained pond.

In another place,
where apparently a
considerable detached
hill stood in the floor
of the original valley,
there remains a high
area whose crest is oc-
cupied by a notable
fissure, like that formed
when the ice is let
down over a hump in
the bottom of the pond
(see page 263).

About ten miles
down the Valley there
is a distinct horizontal
line of the same red
tuff plastered high up
on the slopes of But-
tress Mountain, a hun-
dred feet above the
present "high water
mark," as though the
liquid that filled the
Valley had stood for
this higher level and
along the bank before subsiding.

PROOF THAT THE ACTIVITY OF THE

VALLEY BEGAN BEFORE THE

EXPLOSION OF KATMAI

Detailed examination of the surround-
ing mountains confirmed the suggestion
of the ''high water mark," that the tuff
was confined to the floor of the Valley,

Photograph by L. G. Folsom

PROOF THAT THE VALLEY OF TEN THOUSAND SMOKES
OPENED BEFORE KATMAI BLEW UP

Beneath the three layers of ash from Katmai is seen the massive sand-
flow, which must already have run its course before the ash began to faU.
This picture was taken close to the terminus of the sandflow. The char-
coal log, a foot in diameter, is striking evidence of the heat still retained
by the sand 16 miles from the head of the flow at Novarupta.

a short time at
''frozen" a little

for it has absolutely no counterpart on
the slopes above. Like all the rest of the
country the slopes are covered only by
the layers of stratified ash from Katmai.
More significant yet was the obser-
vation that these same layers of strati-
fied ash also extend out over the tuff it-
self, which they everywhere cover (see
also page 258) . This is indubitable proof
that the tuff had reached its present posi-
tion before the ashfall. Since, as will

Digitized by

Google

256

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by P. R. Hagelbarger

THE EDGE OF THE HOT SANDFLOW

The lower ground, exposed in the drainage gully, was originally covered with forest like the bank
on the left. Here and there a burned stump remains to show what happened when it was overflowed
by the hot sand. The destruction shown in the right half of the picture is complete and total.

be seen, the tuff is intimately related to
the fumaroles, this can only mean that
the Valley of Ten Thousand Smokes had
already come into existence before
Katmai blew up.

RELATIONS TO FOREST SHOW THAT THE
TUFF WAS ONCE FLUID

When our work carried us down to the
foot of the Valley, we found clear-cut
and positive evidence of the manner of
deposition of the tuff, making it certain
that the sand of which it was composed
must once have flowed down the Valley
like a viscous liquid.

In the lower Valley, where the remains
of the former forest still persist to tell the
tale, the ''high water mark*' becomes
yet more conspicuous and significant
than it is in the upper Valley.

The remains of the forest embedded
in the tuff show plainly that it flowed
down the valley as a thick liquid would
have done. Right down to the edge of
the flow the trees stand undisturbed, but
below that level they are overridden,

twisted, and bent as before a moving
mass. The remarkable sharpness of
the line separating the uninjured forest
from that covered by the tuff is plainly
shown on the photograph reproduced on
page 260.

At the bend in the Valley, where the
forest begins, the flow encountered a belt
of morainic hills among which it pursued
a most irregular course, overtopping
some, leaving others standing free above
the surface, slopping over into the ravines,
and in every way showing that it ran
rather freely. In this vicinity a creek
was dammed by the flow, forming a
fairly large pond.

Beyond the bend in the Valley it con-
tinued for more than a mile, gradually
thinning out until for some distance
back from the tip it was only ten feet
thick, in striking contrast to its massive
character farther up the Valley. Every-
where it showed a surprising capacity to
adjust itself to variations in the level of
the ground over which it flowed.

At one place, near the middle of the

Digitized by

Google

THE GREAT HOT SANDFLOW OF THE VALLEY

257

Photograph by P. R. Hagelbarger
LOOKING ACROSS THE SANDFLOW NEAR ITS TERMINUS

This illustration is panoramic with the picture on the opposite page. The massive character of the
flow and its relations to the undisturbed forest covering the hills beyond its reach are evident. The
fiery torrent consumed everything it touched.

level flow, I noticed two or three trees
projecting from it. Investigation showed
that they had grown on the summit of
a morainic hill which, after having been
deeply buried, had been all but laid bare
again by readjustment of the mass after
the first wave had passed.

But there was none of that evidence
of violent damage which would have
accompanied the rush of a flood of
water down the Valley at high speed.
Although the bushes were bent and
twisted beneath the weight of the sand,
they were not broken or uprooted as
might have been expected. This seems
to prove that the mass moved rather
slowly.

EVERY STICK BURIED BY THE SANDFLOW
REDUCED TO CHARCOAL

Near the edge of the flow a few trees
remained sticking up out of the sand.
When we took hold of them we found
to our astonishment that they broke off"
a foot beneath the surface. The rest of
the buried portion had turned to char-
coal. Later, in gullies cut into the flow
18

by erosion, we discovered that every
particle of wood buried by the sand had
been completely turned to charcoal.

In some places the mat of old tundra
vegetation had been transformed into a
thin sheet of charcoal lying between the
flow and the old surface of the ground.
Some of the trees engulfed had been a
foot in diameter, but they were as com-
pletely charred as the fine twigs (see
pages 255 and 260).

On returning up the Valley we ob-
served that the degree of destruction
steadily increases toward the supposed
source of the sand. Around the foot
of the Valley the trees had evidently
charred rather slowly and there was
no visible eff'ect above the high sand
mark. A little farther up, evidence of
occasional grass fires appeared. Half
way up, on the flank of Buttress Moun-
tain, the bushes on the mountain side
had been burned to the ground; but
the roots remained, though charred at
the surface.

Finally, around the head of the Valley,
where no trace of vegetation remains,

Digitized by

Google

258

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs
A ** BUTTE*' ERODED OUT OF THE SANDFLOW ON THE MARGIN OF FISSURE LAKE

The angular blocks that have broken oflF show the irregular cleavage. The horizontal strata at the
top are the layers of Katmai ash which manifestly must have fallen after the occurrence of the_^hot
sandflow.

the heat of the flow must have been so
intense as to have completely consumed
all plants, leaving no remains to prove
their former abundance.

When proof of the extreme heat of
the sandflow was added to evidence of
its remarkable fluidity and tremendous
proportions, it became apparent that the
eruption had involved a cataclysm far
more extensive than at first supposed.

It was clear that the explosion of Katmai,
which alone was known to the outside
world, was only one event in a disturb-
ance involving much more than a single
localized crater.

SEARCH FOR THE ORIGIN OF THE SAND-
FLOW

Even before it had become certain
that the sandflow had been intensely

Digitized by

Google

THE GREAT HOT SANDFLOW OF THE VALLEY

259

hot, we began search-
ing the Valley for evi-
dence as to its origin.

Katmai Crater was
naturally the first
place to be considered.
But an examination
made it certain that
the Valley flow could
not have come from
Katmai. The slopes
of the volcano over
which the sand must
have flowed, if it
reached the Valley
from that source, are
covered with great gla-
ciers, and it is hardly
to be imagined that a
, mass of material hot
enough to consume
every vestige of vege-
tation could have
rolled down over a gla-
cier without melting
it.

The high mark left
by the sand, moreover,
sweeps across the foot
of this glacier in a
steady grade, which
shows very plainly that
the sand flowed past
the foot of the moun-
tain, not down from
it.

z'liuiuKiapii uy j. tt • ouiyicy

Similar evidence
proves that the sand
did not come from Mageik; the sloping
" high water mark " separating the rugged
glacier-covered slopes from the sand-filled
Valley could not have been sharper (see
pages 262 and 264).

THE HOT SAND MUST HAVE BROKEN

THROUGH THE FLOOR OF

THE VALLEY ITSELF

The fact that the high sand mark is
practically continuous all around the Val-
ley excludes, in fact, all outside sources.
If the hot sand had flowed into the
Valley from any of the surrounding

A HOODOO IN THE SOLIDIFIED SANDFLOW

These formations show the irregular fracture and the massive unst rat-
ified character of the deposit. They were broken out by faulting.

mountains, the mark would be broken
at the point of influx.

The absence of such a break can only
mean that the material welled up from
within the Valley itself, just as many a
lake is fed by submerged springs. The
sand covers up and conceals the orifices
from which it came.

Under the circumstances, we should
not expect to be able to locate any inlet.
Yet irregularities in the slopes assumed
by the sandflow make it certain that
no one vent could have supplied all the
material. It flowed both ways from the

Digitized by

Google

THE EDGE OF THE CHARCOAL FOREST ABUTTING UNINJURED TREES ABOVE THE
EDGE OF THE ONCE FIERY TORRENT (SEE PAGE 256)

Photographs by R. F. Griggs
STUMPS OF TREES BURNED OFF BY THE HOT SANDFLOW; EXPOSED BY EROSION

Near the terminus of the flow, where this picture was taken, the sand had so far cooled as not to bum

the trees clear to the ground.

260

Digitized by

Google

THE GREAT HOT SANDFLOW OF THE VALLEY

261

summit of Katmai Pass and from the
divide back of Novarupta, proving that
at the very least there must have been
vents near each of these saddles.

Novarupta bears every appearance of
having contributed a large share of the
materiaL But there is no considerable
vent in the vicinity of the pass, nor any
evidence that one existed at the time of
the eruption. Appearances make it more
probable that the tuff in the Pass and in
Mageik Creek came from a number of
small vents rather than a single large one.

So through the Valley as a whole.
The remarkably uniform distribution of
the tuff; the occurrence of high temper-
ature vents at very distant points;
the widely scattered craters, marking
places that were once seats of such
mildly explosive activity as we may im-
agine to be responsible for the outpour
of sand ; the striking coincidence between
the boundaries of the active area and
the edge of the flow — all these suggest
strongly that the incandescent sand came
from a large number of scattered vents.
The most probable hypothesis is, there-
fore, that it was extruded from vents es-
sentially*similar to the present f umaroles.

SANDFLOW AN ERUPTION OF
A NEW TYPE

Whether this theory is correct or not,
it is clear that the present outburst
differed radically from an eruption from
a typical volcano of the central type.

In the past many eruptions are known,
wherein great floods of magma poured
out from fissures scattered over wide
areas; but in all cases previously re-
corded, the ejecta consisted not of ash and
pumice but of molten lava, which hard-
ened into solid rock, basalt in every case.

The incandescent sandflow of the Val-
ley of Ten Thousand Smokes is, there-
fore, an eruption of a type hitherto
unknown. Its nearest analogue is to be
found in the incandescent avalanches
from Mount Pel6e, which overwhelmed
St. Pierre. But our sandflow differs as
much from the Pelean eruption as does the
latter from the ordinary explosive type.

If a relatively quiet upwelling of
magma such as filled the Valley had
followed the explosion of Katmai, it
would not have been so surprising. But
for the quiet extrusion of a great quan-
tity of magma to precede a violent par-
oxysm reverses the usual order of events
and runs contrary to the accepted theo-
ries of volcanism.

The extrusion of the enormous amount
of material entering into the sandflow
would be expected to have relieved the
pressure of the subterranean magma
sufficiently to prevent the explosion of
Katmai. But it did not even delay it. On
the contrary, it might be argued that the
Valley flow set off the great explosions.

On yet a third count, this remarkable
Valley flow gives a new viewpoint of
great importance to an understanding
of volcanism.

Inasmuch as the sand of the Valley is
as finely fragmented as the ash of Kat-
mai, it seems fair to suppose that the
amount of gas given off must have been
comparable in the two cases. Yet the
ash of Katmai was blown so high as to
spread over the whole earth, while the
sand of the Valley was never lifted
beyond the "high water mark." The
reason for this great difference in be-
havior of magma closely similar both in
original composition and in end-products
is not known.

The unique character of the incan-
descent sand thus raises a number of
speculative questions of the greatest
fascination to all interested in finding out
the workings of this old world of ours.
The aim of this chapter is, however,
merely to present the facts as observed in
the field and point out their possible
significance, not to deduce conclusions of
general application.

In Chapter xxii we shall attempt, on
the basis of the observed evidence, to
reconstruct the eruption sufficiently to
give a picture of its events as they might
have appeared to an eyewitness. Mean-
while we shall bring together the evi-
dence of the condition of the Valley and
of the volcanoes before the eruption.

Digitized by

Google

a
2

.2
O ^

en «»
•en r

< -^

262

Digitized by

Google

0) a>

2 ed

^'3

o o
(d «

.S:S3
cd p^

§8*5
2 3 "^

^ ^^-%

2 ^

£ h^

3 cd u

>•*

•3**^ IS

vs a"
gi§

>> s

> •♦-»

Digitized by

Google

t
a

Q B

D 3

o -^^

O §

^ Q.
n .s

^ ft)
Z. =5

« -s

:^ S

< 8
:z; 3

a
W u

c
<d

264

Digitized by

Google

THE CONDITION OF THE COUNTRY BEFORE THE ERUPTION

In his perusal of the preceding chap-
ters the reader must have wondered
many times what sort of country this
may have been before the eruption. It
h2is seemed best, however, to defer con-
sideration of the previous condition of
the region until the description of the
devastation was complete. But we,
like everyone else, had begun to inquire
concerning the previous history of the
volcano as soon as we heard of its recent
eruption.

Had it ever been active before? What
warning had been given? What changes
in the face of the country were wrought
by the disturbance?

And when we discovered the Valley
of Ten Thousand Smokes and realized
that the explosion of Katmai was only
one of several manifestations of the dis-
turbance, our interest in the previous
condition of the country was redoubled.

But, unfortunately, knowledge of
these antecedent conditions is very frag-
mentary. I have been at some pains
to search out information concerning the
former state of the country and espe-
cially to secure photographs of it, but
the results are far from satisfactory.

petrof's description of the
katmai district

From time immemorial Katmai Pass
had been one of the most important
routes of travel across the Alaska Penin-
sula. A very good general picture of
the aboriginal conditions is given by
Ivan Petrof ^ in his report on Alaska for
the Tenth Census, page 24, as follows:

**The country between Bering Sea
and the Alpine chain of mountains ex-
tending along the eastern shore is a grad-
ually ascending plain, dotted with lakes
fed from the glaciers and eternal snows
in the east, and having their outlets in
the west.

265

** In the northern portion of the penin-
sula a belt of timber reaches down in the
center to the vicinity of Lake Becharof,
but beyond this the forest disappears,
and only the deep ravines exhibit a
stunted growth of creeping willow and
alder brush.

*'The reindeer browse in herds all over
this region, retreating during the sum-
mer up to their inaccessible retreats
among the snowy peaks of the mountain
range, where they are often seen by the
traveler below as a moving line of black
dots winding around the summits. Dur-
ing the autumn and winter they seek the
vicinity of the lakes and scatter over
the tundra, where they are hunted with
comparative ease.

** Foxes, land otters, martens, and
minks are plentiful throughout this sec-
tion, and the gigantic brown bear of con-
tinental Alaska rivals the native fisher-
man in the wholesale destruction of the
finny inhabitants of lake and stream.

'*The people of Port MoUer and Ooga-
shik are of the Aleutian tribe, which in
former years made warlike expeditions
along this coast, extending as far to the
northward as the Naknek River and
Lake Walker [Naknek Lake]. At the
village situated on one of the feeders of
the latter lake the present inhabitants
still tell the story of the night attack
made by the * blood-thirsty* Aleuts long
years ago, when every soul in the place
was dispatched without mercy, with the
exception of one man, who hid himself
under a waterfall [Brooks' Falls, close
by the village] and thus survived to tell
the tale.

*'The peculiar formation of this coun-
try led to the discovery at an early
date of several portage routes across the

^ Petrof, Ivan, Report on the Population, In-
dustries and Resources of Alaska in the Tenth
Census, Vol. 2, 1884.

Digitized by

Google

256

THE VALLEY OF TEN THOUSAND SMOKES

A POND IN THE SOUTHWEST CORNER OF THE KATMAI NATIONAL MONUMENT

This valley, adjacent to the devastated country, gives a good idea of the general character of the
country before the eruption, but in Katmai Valley the trees were much larger, forming a heavy forest.

peninsula. The Russian promyshleniks
first made their way to Bristol Bay and
Nushagak across the peninsula from
Kodiak, and found abundant evidence
to show that this route of communication
had been an inter-tribal highway for
ages past.

*' During the early years of the Russian-
American Company's sway in these re-
gions, a large proportion of their traffic
was carried on in this way from Kodiak
to Bristol Bay, and thence to the Kus-
kokwim and Yukon Rivers, and Saint
Michael. This was a tedious and ex-
pensive mode of transporting merchan-
dise, but it was long preferred to the risk
and uncertainty of sending sailing craft
around to Saint Michael through the
shallow and stormy waters of Bering Sea.

"On the eastern side of the peninsula
the mountains rise abruptly from the
sea, a short day's climbing transplant-
ing the traveler from tide- water into
the midst of glaciers and eternal snows
and scenes of alpine grandeur and soli-
tude. ...

"The settlement of Katmai, in this

vicinity, was once the central point of
transit for travel and traffic across the
peninsula. Three different routes con-
verged here and made the station a point
of some importance; now Katmai 's com-
mercial glory has departed, and its pop-
ulation, consisting of less than two hun-
dred Creoles and Innuits, depend upon
the sea-otter alone for existence.

"The men could have reindeer in
plenty by climbing the mountains that
rear their snow-covered summits im-
mediately behind them, but they prefer
to brave the dangers of the deep and to
put up with all the discomfort and in-
convenience connected with sea-otter
hunting, and in case of success purchase
canned meats and fruit from the trading
store, leaving the deer on the mountains
undisturbed.

"The people of the two villages across
the divide, in the vicinity of Lake Walker
[Naknek Lake], come down to Katmai
to do their shopping and to dispose of
their furs, undertaking a very fatiguing
tramp over mountains and glaciers and
across deep and dangerous streams in

Digitized by

Google

THE CONDITION OF THE COUNTRY BEFORE THE ERUPTION

267

Photograph by P. R. Hagelbarger
SPRUCE FOREST NEAR THE SMOKES

Much of the area now covered by the sandflow was occupied by forest like this before the eruption.
The spruce here occurs at altitudes above 1000 feet.

preference to the canoe journey to the
Bristol Bay stations.

"Only at long intervals a small party
will proceed to Nushagak to visit the
Russian •nissionary stationed there, to
whose spiritual care they have been as-
signed without regard to locality or con-
venience.*'

THE IMPORTANCE OF KATMAI REVIVED BY
DISCOVERY OF NOME

The importance of the Katmai route
across the Peninsula was revived by the
discovery of gold at Nome in 1898. In
the eager ** stampede*' to the gold fields,
many availed themselves of the oppor-
tunity to shorten by several hundred
miles the distance to the goal afforded
by the route across Katmai Pass.

This sudden influx of travel must have
utterly bewildered the poor natives, who
were unable to provide accommodation
for the travelers. As a solution of the
difficulty, the village trader built a
'*Bunk House" at a little distance from
the village, where a large number of tran-
sients could be accommodated.

Many of the men who thus crossed the
Pass must be still living, and some doubt-
less retain vivid memories of the char-
acter of the country, but I have not yet
found any photographs of the country
covered by the eruption.

One of these pioneers was ** Charlie"
Carter, who carried the mail by dog
sledge from Nome to Katmai one winter.
He has told me that on the upper part of
Naknek Lake the ice was very treacherous,
sometimes thawing out when the air tem-
perature had not risen above 15° F., which
clearly indicated to him the presence of
hot springs somewhere in the vicinity. As
the country was altogether covered with a
blanket of snow at the time of his trips,
he learned little else of its character.

SPURR 'S ACCOUNT OF KATMAI PASS

The only scientific man who has left
an account of the district is J. E. Spurr,
who finished his expedition by crossing
the Katmai Trail. His account is, there-
fore, almost the only published source
of information concerning its condition
before the eruption.

Digitized by

Google

268

THE VALLEY OF TEN THOUSAND SMOKES

Before quoting from Spurr's narrative
it should be recalled that the reason for
the location of the trail across Katmai
Pass lay in the extraordinary fashion in
which great valleys from both sides of
the Peninsula penetrate the heart of the
mountains at that point. Katmai Val-
ley from the Pacific lines up almost pre-
cisely with the valley now occupied by
the Ten Thousand Smokes on the Bering
Sea side (see map). Spurr's account of
his trip from the head of Naknek Lake
is, therefore, a description of the former
condition of the main Valley of Ten Thou-
sand Smokes.

He left Savonoski on October 14, 1898,
and camped that night in the vicinity
of the present terminus of the sandflow.
The next night he camped in the middle
of what is now the main arm of the Val-
ley of Ten Thousand Smokes.

**On the 1 2 th of October we started
paddling across Naknek Lake, which is
the largest body of fresh water that we
encountered on our trip. The water
was very calm and, on account of the
danger of heavy gales in these mountain
lakes, the natives were willing to push
ahead as fast as possible. At dark, after
havii^ made about 40 miles, a stop was
made for supper, after which about 20
miles more was made at night till the
head of the lake was reached.

'*The next morning we reached the
native village of Savonoski. At Savon-
oski our Nushagak natives went back,
being in continual fear of being frozen in
by the increasing cold weather. In
order to make as quickly as possible the
60 miles or so of mountains which lay
between us and the coast, ten natives
from Savonoski were hired to help us
carry over, in one trip, our outfit and
necessary provisions, everything which
was inessential being left behind.

''The trip from the lake to the sea-
coast occupied three and one-half days,
an average of nearly 20 miles a day, all
of which we made on foot, largely
through swamps and deep moss. On the
1 6th of October we crossed the mountain
pass and descended to the other side.

"This pass lies between two extinct
volcanoes and is high, snowy, and rocky,
and has no definite trail. The wind
is often so cold and violent here, even
in summer, that the natives do not dare
to cross except in calm weather, for
the gusts are so powerful that stones
of considerable size are carried along by
them.

"On the sea side of the pass we came
to a considerable stream of hot water
which emerged from the side of one of
the volcanoes and flowed down, steam-
ing, to reach the cooler water of the
other mountain drainage. [These hot
springs still gush forth from the slopes
of Trident, but now come out as many
streams of warm, rather than hot, water.]

"On the 17th of October we arrived
at the Aleut village of Katmai, where we
found a Russian trader. Having now
reached the coast, we were disappointed
to find that the difficulties in getting
farther were even greater than they had
been in coming so far, for there were no
boats at Katmai except a couple of open
dories, belonging to the trader, and one
skin bidarka.

"We knew that in a few days the last
steamer would leave St. Paul (Kodiak],
on Kodiak Island, 120 miles away, for
Seattle; but the Shelikof Strait which
lay between was so rough and windy
that the trader advised us very strongly
not to attempt the trip in dories.

"Finally, the sole bidarka in the vil-
lage was sent across to St. Paul with
three expert native paddlers, with a
letter to the Agent of the Alaska Com-
mercial Company at that place, asking
that a schooner or some other craft
should be sent to us in order that we
might make connection with the steamer.
For several days the winds were so vio-
lent that the natives would not cross, and
after they had crossed no news was heard
of them for a long time, till considerably
after we knew that the last steamer must
have gone.

"On the 31st of October, however,
we were awakened in our camp, on the
beach, by the whistle of a steamer which

Digitized by

Google

THE CONDITION OF THE COUNTRY BEFORE THE ERUPTION

269

we promptly boarded. We found it to
be the Alaska Commercial Company's
boat Dora? which had been delayed in
Cook Inlet considerably past her usual
time for returning home on her last trip,
and so had got to St. Paul in time to re-
ceive our letter and to come to our as-
sistance, a trip which involved a delay
of two days for the steamer and its
passengers. In the Dora we reached
Seattle on the nth of November.' . . .

** During the first part of the land jour-
ney the ascent is comparatively gradual,
and the traveler passes over extensive
swampy areas with drier knolls, where
walking is made difficult by tufts of
grass, called 'niggerheads' by the pros-
pector or Utes de femme by the Canadian
voyageur.

'*The trail leads along the hillside
above the bed of the small stream which
runs into the lake at Savonoski, and fre-
quently boiling brooks, tributaries to
this stream , must be forded . The moun-
tains on both sides of the small valley
[Le, Valley of Ten Thousand Smokes]
grow higher, and, as one approaches the
summit of the range, it is seen to be com-
posed of a continuous chain of volcanoes,
none of which, however, is at present
active, although the natives informed
us that one of them occasionally smokes.
''From the sides of some of these
highest volcanoes splendid glaciers wind
down into the valley, and in other places
great walls of moraine, damming moun-
tain gorges, mark the former positions
of glaciers which have now somewhat
retreated. So far as observed, these
glaciers seem to be more extensive on
the northwestern than on the south-
eastern side of the range.

**The Katmai Pass, which leads down
to the village of that name, lies between
two volcanoes and is extremely wild and
rugged, being the most difficult moun-
tain pass we crossed during the journey.

' It will be recalled that it was this same old
Dora which was in Shelikof Strait at the time of
the eruption and brought the first news to the
outside world (see pages 10 and 15).

*J. E. Spurr. A Reconnaissance in South-

For several miles on both sides of the
summit there is no trace of vegetation,
the surface being composed of huge an-
gular fragments of rock, piled together
without even a covering of moss.

"Through this debris and the under-
lying lava the mountain streams have
cut deep gorges. On all the upper part of
the pass the snow lay thick at the time
of our crossing, in the middle of October.
Many natives have perished here by be-
ing caught in gales, for during storms,
even in summer, the wind blows with in-
tensity and piercing coldness. At such
times stones of considerable size are
picked up by the wind and carried
through the narrow defiles where the
traveler must walk, and we found many
of these stones lying upon the snow. Ow-
ing to this danger the natives cannot be
induced to cross except in perfectly calm
and clear weather.

''Extensive hot springs emerge from
the Katmai side of the mountains below
the pass, and there are very frequent
earthquakes and other evidences of vol-
canic activity. Our party itself experi-
enced a slight earthquake just after
crossing.

**The descent from the summit in the
direction of Katmai is much more abrupt
than on the northwestern side, and in
about 10 miles one passes from an alti-
tude of nearly 3,000 feet to a broad,
level flat which is at about the level of
the sea.

"In this gravel flat, several miles
wide, the Katmai River flows and on
both sides of it are high and rugged moun-
tains which run quite down to the coast,
and, indeed, extend below the sea, as is
shown by the fact that the water close to
the shore is very deep, as it is nearly
everywhere from here southward and
eastward along the Alaskan coast. The

western Alaska in 1898. U. S. Geological Survey
Annual Report 20, pt. 7, page 59.

Owing to the style of the Geological Survey
Reports, Spurr's narrative is repeated three
times with increasing amplification; hence the
apparent repetition in the triplicate quotation
here given.

Digitized by

Google

2^o

THE VALLEY OF TEN THOUSAND SMOKES

— <\^— --. \

^ — X \

x\\\\\ \
r/ i ? \i

! 1 ' 1 / / ,'

\
\
\

1 A

/ /

/ 1

\
\
\

1
1

:-■// / / V

• / / \ >

\
\
\

\
\
1

/
/
/

I :

E.S
3 B

£11

8"*-

o V ^

Digitized by

Google

THE CONDITION OF THE COUNTRY BEFORE THE ERUPTION

271

gravel flat of Katmai Valley forms on
the seacoast a sand beach several miles
long, which is effectually shut in on both
sides by high mountainous promon-
tories.*

**The first 15 miles of the trail from
Savonoski to Katmai lies through
swampy flats in the broad valley which
forms a continuation of the lake. After-
wards a series of rises over slight benches
brings' one to a larger plateau valley
[Valley of Ten Thousand Smokes],
which, however, is level and generally
swampy.

"The trail lies along this until near
the pass, where the plateau valley and
the deeper valley of the stream come to-
gether. The mountains on both sides
have well marked horizontal benches,
up to 1,000 feet at least.

''At the first rock bluff above Savono-
ski, where the trail leaves the flats, is an
outcrop of dark green granitic arkose,
with nearly horizontal stratification. The
rock of the sharply eroded mountains on
both sides, nearly as far as the pass, is
of the same horizontal bare green strata,
which, by fragments included in the
drift of the valley, are shown to be green
arkoses and limestones. Jurassic fos-
sils are found abundantly in these.

"Near the summit of the range is a
continuous chain of volcanic peaks,
some with the sides more or less furrowed
by drainage, others smooth and perfect.
Against one of these volcanoes the hori-
zontal rocks are turned up, this being
the only folding observed anywhere.

"Along the axis of the mountains are
large and well-developed glaciers, espe-
cially on the sides of a tall volcano which
stands directly at the head of the valley
up which the trail runs [Mageik]. Here
the glaciers are many and distinct, and
have splendid moraines.

"The main valley is filled with hori-
zontally stratified sand, in which are
great angular bowlders, which are also
beautifully arranged in layers. This
deposit has a generally level top, but
there are many sharp hillocks formed of

bunches of huge bowlders. The streams
have cut deep channels through this.

"On each side of the chain of volca-
noes which form the axis of the range,
the stratified rocks dip away very gently
and are slightly undulating. On the
Katmai side of the divide the green fos-
siliferous rock has perfect crystals of
mica in it, giving it an igneous appear-
ance. These have probably developed
as a consequence of metamorphism.

"The pass lies between two volcanoes,
the one on the right [Mageik] having a
cone scarcely modified by erosion, while
that on the left is somewhat furrowed,
but carries no glaciers [Trident]. Below
the pass, on the Katmai side, streams
of very hot water burst out of the ground
in many places and, joining together,
form a considerable rapid stream. This
water contains much sulphurated hydro-
gen and makes copious precipitates of
iron and sulphur, so that the earth
through which it oozes is colored a bril-
liant yellowish red [just as it is today].

"The valleys of the streams which
drain both sides of the pass are deeply
excavated, especially that on the north-
western side, which has cut down at
least 100 feet through lava bowlders.
Owing to the excessive frost action the
whole surface of the volcanoes is covered
thickly with these large bowlders. On
the northwestern side, near the summit,
a lake half a mile long is dammed in by
the d6bris from three volcanoes which
rise above it. [Remains of lake still
present.]

"On the sea side of the pass no gla-
ciers or glacial phenomena on a large
scale were observed. On this side are
broad benches of great height, the mate-
rial of which is rolled bowlders from the
mountains above; on the surface the
frost has cleaved these bowlders into
jagged fragments.

"From the mountain sides below the
pass one looks down upon a broad valley,
a flat several miles wide, mostly covered
with bare gravel, giving the appearance
from a distance of an arm of the sea and
page 91.

Digitized by

Google

272

THE VALLEY OF TEN THOUSAND SMOKES

o
n

a
o

to
O
tfi

9
u

C
ed

<
s

•s

Digitized by

Google

THE CONDITION OF THE COUNTRY BEFORE THE ERUPTION

273

through this flat the Katmai River runs
in many channels."*

KATMAI PASS PROBABLY IN UNSTABLE
CONDITION BEFORE THE ERUPTION

When I first informed the members
of the National Geographic Expedition
of 191 7 that the wind in Katmai Pass
was reported strong enough to pick up
stones and throw them agaijist travelers
to their destruction, the boys very natu-
rally hooted at the idea. But I reminded
them that such a statement was written
into an official publication of the United
States Geological Survey, which is not
given to wild exaggeration.

The heavy deposit of ejecta from the
eruption has filled up the defiles men-
tioned by Spurr, and reduced the danger
from rolling rocks. During 191 7 we had
little difficulty from this source, but the
wind around the Pass is so fierce that
before they had finished with their al-
most daily trips through it, all hands
were quite ready to subscribe to any
statement that might be made concern-
ing Katmai Pciss.

The assertion that the wind used to
start rocks rolling down on travelers
should not be dismissed as a fantastic
invention of the natives. It may be
recalled in this connection that I myself
came near being killed in exactly that
fashion in Katmai Canyon, where many
tons of rocks were started down the
banks by no more than a good breeze.

The importance of the statement lies
in the fact that the wind could only
start such stones as were already in an
unstable condition, ready to topple down
at the least impulse. These stories
should be considered in connection with
the frequent earthquake shocks of which
Spurr makes mention.

It will be recalled also that the incan-
descent sand described in the last chap-
ter was poured out in the Pass and in
both approaches. This feature and the
presence of extinct fumaroles on the
Pacific slope as well as on the Valley
side is clear evidence that Katmai Pass

is traversed by a fissure such as opened
in the Valley.

Slow movement along a fault line may
have caused a continual unsettlement,
keeping a considerable number of bowl-
ders always in the unstable equilibrium
indicated by the frequent rock falls.

To Spurr *s account should be added
the narrative given by Rex Beach in his
"Silver Horde," pages 75-83. In a work
of fiction of this sort the writer was of
course at liberty to draw freely on his
imagination for his descriptions. Nev-
ertheless, his story of the crossing of
Katmai Pass, drawn partly from Spurr *s
account and partly from his own ex-
perience in making the trip, is a very
good description of conditions.

In crossing the Pass repeatedly I be-
came well acquainted with its evil char-
acter and found Beach's account cor-
rect even in numerous minor details
which no one would have expected him
to grasp in a single trip. Our expe-
rience being confined to the summer
season had, of course, no danger of the
tragic ending that befell the party of the
novel, but it required no effort of the
imagination to believe that in winter
temperatures the trip would be fully
as dangerous as Beach pictures it.

DIFFICULT TO JUDGE THE WEATHER
ACROSS THE PASS

It is very difficult from either side to
judge the weather on the other side of
the range. The experience of Beach's
hero, who, after waiting for four days,
started out in calm weather only to find
a raging gale on the opposite side was
duplicated by our parties several times.
The curious thing is that the Pass itself
is often quiet when the tumultuous
blasts over the shoulder of Observation
Mountain prohibit travel.

While we were camped in the Valley
of Ten Thousand Smokes, we depended
for our subsistence on supplies which
had to be brought up almost daily from
the sub-base camp in Katmai Valley.
Several times the pack-train encoun-
page 146.

Digitized by

Google

274

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by J. W. Shipley
THE TRAIL THROUGH THE TALL GRASS ALONG THE WEST BANK OF KATMAI RIVER

Representative of conditions in the zone where the trees were killed but the ashfall was so light that the

grass quickly recovered.

tered storms that would have turned
back any but the hardiest of men. One
day it was impossible to face the flying
pumice. That day the wind gauge, at
the sheltered camp, registered 60 miles
p)er hour steadily, and much higher in
the gusts.

The next day they came up prepared
"to gather up the remains" of our camp
among the Smokes, for they knew we
could not weather such a gale as they
had experienced, and reasoned that it
must have been even worse with us at
our higher altitude. But among the
Smokes the weather had remained fine
throughout and we had been quite un-
able to explain their non-appearance
the day before.

The only other published account of
crossing Katmai Pass which I have found
is that given by the members of Frank
Leslie's Expeditions,^ who also crossed

• Their accounts are scattered through the
numbers of Frank Leslie's Illustrated Newspaper,
volumes 70-73, 1 890-1 891. The material per-
taining to Katmai Pass consists of an article by

in the winter, when many of the features
of the country were concealed by snow.
They speak of a magnificent toboggan
slide down the slopes of Observation
Mountain, but give little geographic in-
formation. They state that the Pass is
beside an active volcano whose summit
was covered with cloud, '*but from whose
sides bubbled steaming sulphurous mal-
odorous streams.**

The most important item from our
point of view is that, before going over,
they camped ** 3,000 feet high in the
last thin little group of trees,** which was
probably located in the middle of the
Valley of Ten Thousand Smokes under
the protection of Mount Cerberus.

OUR ONLY INFORMATION CONCERNING

MOUNT KATMAI DERIVED FROM

COAST SURVEY CHART

Of the condition of Mount Katmai,
the accounts of these travelers afford no

E. H. Wells, published Sept. 19, 1891 (p. 106),
and one by A. B. Shantz, published Nov. 28, 189 1

(p. 268).

Digitized by

Google

THE CONDITION OF THE COUNTRY BEFORE THE ERUPTION

275

information. Spurr gives no clue as to
which one of the volcanoes was re-
ported by the natives to smoke occasion-
ally.

From the trail Mageik is much more
conspicuous than Katmai, which is too
far east to be seen to advantage from
any point except the slopes of Obser-
vation Mountain. At that point, how-
ever, the anxious eyes of the pilgrims
would naturally turn in the opposite di-
rection toward their goal, the sea, which
spread out in magnificent panorama be-
fore them.

Post's sketch map, which accompanies
Spurr *s report, although giving Mageik
and Katmai Valley accurately and in
some detail, is blank in the direction of
Katmai. But on Coast Survey Chart
Number 8555, a section of which is re-
produced on page 270, Mount Katmai is
shown as having three peaks, each higher
than Mageik, i. e., 7,500, 7,360 and 7,260
feet. Around these peaks, whose posi-
tion and altitude were determined by
precise triangulation, are drawn dotted
contours indicating the configuration
of the mountain as it could have been
seen from the triangulation station at
Cape Kubugakli.

This chart -supplies absolutely all the
information we have concerning Mount
Katmai before the eruption. But, curi-
ously enough, search of the archives of
the Coast Survey has failed to reveal the
source of the information from which the
contours of the chart were drawn.

There are no sketches in the field
books of the Kubugakli party by whom
the contours appear to have been drawn.

and the head of the party does not re-
member that any observations of Mount
Katmai were made. But whoever drew
them, they correspond well enough with
the portions of the stub that remain to
inspire some confidence in their accuracy.

RECONSTRUCTION OF THE ORIGINAL
KATMAI

By combining the outline suggested by
the old chart with our more precise sur-
vey of the present volcano, it has been
possible to reconstruct the original
mountain with some degree of satis-
faction.^

The hollows in the flanks of the vol-
cano are occupied by the remains of gla-
ciers which were beheaded when the top
blew oflF. The original peak was cer-
tainly a majestic snow cap. In a gen-
eral way, it must have resembled Mageik
(see page 78), but it was higher and the
summit was at the near end of the ridge,
rather than in the middle.

We may be fairly sure that Katmai
was the dominant peak of the district
since it was customary to apply the
name of any village to each of the most
important geographic features of its
vicinity.

That this peak, rather than Mageik,
was honored with the name of the vil-
lage is especially indicative of its former
grandeur, because it was unfavorably
located to receive such a distinction.
Mageik is in full view from the village,
making a wonderfully impressive spec-
tacle. But Katmai lay behind the Bar-
rier Range, which concealed all except
perhaps- the very summit.

^ The estimate of the amount of material blown
off the mountain in the eruption which was given
in Chapter XIV was arrived at by comparing the
reconstruction of Katmai with the stub that re-
mains. The contours are 200 feet apart. By
obtaining the areas enclosed by the different con-

tours around the old summit and within the
crater, we can calculate the volume of the mate-
rial that has disappeared. The average of four
independent computations of this volume is
1 1 ,000 millions of cubic yards.

Digitized by

Google

o
;=)

276

Digitized by

Google

XXI

A HYPOTHETICAL CHRONICLE OF THE ERUPTION

The preceding chapters have given
some description of the effects of the
Katmai cataclysm and of the volcanoes
it opened up, but little attempt has
been made to relate the various phenom-
ena in the order of their genesis or to
discuss the actual events of the eruption.
It will be desirable now to attempt a
description of the eruption as it would
have appeared to an observer on the
scene.

Since there were no witnesses com-
petent to testify to the facts, we are
limited to deductions from the study of
effects on the surrounding country. *It
should be understood that our account
must be made up of reasonable infer-
ences and conclusions that follow from
the evidence left behind. It must, there-
fore, be read as a mental reconstruction
rather than as a chronicle of a series of
observed events.

IN THE BEGINNING A BEAUTIFUL
GREEN VALLEY

At the beginning, a beautiful broad
valley, just turning green after the long
winter, stretched down from Katmai
Pass. Up to nearly 1,500 feet it was
covered with a dense forest of balsam
poplar and paper birch, with occasional
clumps of spruce, which became abun-
dant at lower levels. There were numer-
ous undrained hollows occupied by ponds
or tundras, while the steeper slopes were
covered with dense stands of tall grass.

Along the edge of the forest a herd of
caribou may have been following up the
melting snows toward their summer
pastures among the alpine meadows;
lower down in the woods occasional
moose, stragglers here at the extreme
edge of their range, might have been ob-
served.

Within a few miles were undoubtedly
several of the great brown bears of the

region, busily engaged in digging out
the roots of the spring plants to fill their
empty maws after the long winter fast.
A pack of hungry wolves, followed per-
haps by a skulking wolverine, may have
hunted a herd of caribou, or, failing such
large game, contented themselves with
snowshoe rabbits and ground squirrels.

A GREAT LANDSLIDE FROM FALLING
MOUNTAIN BEGINS THE CATACLYSM

The first sign^ that anything unusual
was in prospect were probably occasional
earthquakes, which later must have in-
creased ominously in frequency and vio-
lence, for, while details are lacking, it
appears that earthquakes were the main
cause of the alarm of the natives at
Katmai Village and Savonoski. Pre-
sumably, also, the old volcanoes, one or
more of them, soon began to manifest
unusual signs of activity by way of in-
creased smoking, accompanied perhaps
by small explosions.

Although premonitory symptoms of
this sort would have made themselves
felt over a wide stretch of country, the
opening scene in the catastrophe, so far
as we know it, was more local in char-
acter. With a thunderous roar, the face
of Falling Mountain let go, precipitat-
ing several million cubic yards of rock
into the lowland in a terrific rock streami
that must run for several miles down the
Valley.

A wind started by the rush of falling
debris roared down the Valley like a hur-
ricane. Meanwhile, the sky was dark-
ened by the great cloud of dust set up by
the falling rocks, imparting a gloomy and
portentous aspect to the scene and help-
ing to make it an altogether appropriate
prelude to the great tragedy.

Similar rock falls of almost equal
magnitude may have occurred several
times during the eruption, but the later

277

Digitized by

Google

278 THE VALLEY OF TEN THOUSAND SMOKES

^ -5

ones would have attracted
little or no attention amidst

S the general cataclysm which
by that time had broken
loose. For it would have be-
come evident that the whole
region was rocking in the
throes of tremendous subter-

o ranean forces which threat-

>, ened every moment to rend

^ the mountains asunder.

w» ;;■ VALLEY FLOOR CRACKS OPEN
^ gj AND SEETHING MAGMA

S S. RUSHES OUT

c .

Probably before the dust
•g o had settled, during an earth-
ly quake of more than usual
< .^S violence, the Valley floor
Ok "SJ broke open. Immediately
06 o c the seething white-hot
> ^>. magma appeared in the
§ *2 Sf cracks, struggling for release
5 -5 i from its prison in the rocks.
w .S'g At first it probably ap-

^ %% peared like a fire among the
g ^ ^ bushes. But it spread rapidly
S £ ^ and soon its real nature be-
came apparent. Floods of
fire burst out now here, now

C/3 H -M

7) rt

W "

2 =3 3 there, until the whole Valley
^ oj-ix: opened up and the molten

S c ^ viscera of the earth came

2*5 pouring from every fissure.

-J5 2 Tremendous columns of liq-

S-u uid fire shot out, driven sky-

% J ward a hundred feet or more,

3 Xc gilding the mountains round

^ J about with their radiance.
I S The Ten Thousand

^^ Smokes, wonderful as they

•^ are at their best, can give

3 no idea of what the Valley

u must have looked like at this

2 initial stage. In addition to

I myriads of smokes, doubt-

vc less greater than any that

^ now remain, other scores or

> hundreds of volcanoes must

I have been belching forth in-

^ candescent lava in veritable

£ fountains of fire.

Digitized by

Google

A HYPOTHETICAL CHRONICLE OF THE ERUPTION

279

A reconstruction of this scene is
quite beyond the power of human imagi-
nation. The closest picture that we
may venture to present is the description
of analogous white-hot fountains which
shot up during an eruption of Mauna
Loa, as they appeared to an eye wit-
ness: '

** Each one shows a multiplied facade of
spires composed of thousands of bunched
jets of orange color, the spire to the ex-
treme left the tallest and the others —
perhaps eight or ten — diminishing to
the right. The component jets fall in-
ward, when their upward impetus is lost,
in graceful parabolas, excepting when,
at every major ejection, there is a fierce
explosive-like projection outwardly from
the main spire.

**The whole effect is that of an illumi-
nated Gothic cathedral's front. In as-
cent the velocity of the ejection is that
of a rocket multiplied. Stupendous
projective force is what impresses one,
together with the extraordinary pyro-
technical beauty of the display.

**At the bases of the fountain there is
an intermittent boiling and surging, and
a sullen roar of awful majesty rises and
falls. . . . The jets are intermingled
with a profusion of dark angular projec-
tiles . . . which I am informed is
pumice stone."

IMPENETRABLE BLACK CLOUDS
MUST HAVE VEILED THE SCENE

In the b^inning, the many vents in
different parts of the Valley probably re-
mained distinct, each pouring forth its
own contribution to the general chaos;
but soon the outpoured lava must have
begun to coalesce in a fiery flood.

It may have remained in pools of
molten liquid for a little, but before long
gases began to be evolved in enormous
quantities, forming impenetrable clouds

somewhat like the horrible burning
clouds from Mount Pel6e, that brought
destruction to St. Pierre.

Blending together, the clouds from
the separate vents made a lurid canopy
over all. Under different conditions
they assumed various aspects: murky
black, when overladen with included
solid particles; blue, where clear steam
was intensely hot ; white where their high
rollmg peaks caught the sun, and purple
or scarlet where illumined by the fiery '
abyss below. Clouds of smoke must have
rolled over the whole country round
about, carrying with them ashes, cinders,
and charred leaves and twigs.

How all the manifestations of this sort
which must have occurred could have
passed unnoticed even in that remote
country is a mystery. But there is no
record of their having been seen, except
perhaps from Cold Bay where the night
before the great explosions observers
noted a black and stormy sky in the
direction of the Valley (see page 19).

LAVA FROTHS UP INTO A FIERY FOAM

Other aspects of the eruption would
have been more interesting to us, how-
ever, than these passing aerial disturb-
ances, which left no trace behind. The
lava gave off such great quantities of gas
that it promptly lost its character of a
molten liquid and frothed up into a fiery
foam. Cooled by the expanding gases
the non- volatile constituents quickly con-
gealed into ash and pumice, the whole
becoming a suspension of solids in the
gases they themselves were evolving.^

The lower portions of the glaciers and
the heavy snowdrifts that still lay deep
in the Valley rapidly melted away before
the incandescent lava. In the face of
the great heat, the ice must have been
converted into steam with almost explo-
sive suddenness, adding great columns

^ By Daniel Logan, as quoted from C. H.
Hitchcock, "Hawaii and its Volcanoes," p. 129.
The original speaks of the "fagade of spines."
This I take to be a misprint for spires and have
so changed it.

* Dr. Zies has suggested that the properties be-

lieved to have been exhibited by the ejecta at
this stage may be imitated by igniting the pow-
der of basic magnesium carbonate. The gas
given oflF buoys up the solid particles so that it
seethes and boils, flowing about in a manner
extraordinarily like a liquid.

Digitized by

Google

o
o

o
<

280

Digitized by

Google

A HYPOTHETICAL CHRONICLE OF THE ERUPTION

281

of its own to the clouds rising directly
from the lava.

ROARING FIRES SWEEP UP THE
MOUNTAIN SIDE

Almost before the snows had been dried
up, the bushes began to crackle and soon
roaring fires were sweeping up the moun-
tain sides, consuming every living thing
far beyond the reach of the incandescent
flood.

With the smoke of the burning forest,
the steam from the melting ice, and the
clouds from the magma, the whole Val-
ley must have been effectually masked.
Yet the strong northwest wind which
may have started even at that early stage
perhaps carried off the smoke enough to
reveal something of what was going on at
the windward side of the burning pyre.

The misery of the wretched animals
trapped in the Valley could not have
lasted long. All were roasted alive, in
short order. Soon their carcasses must
have been cremated in the general con-
flagration. Under ordinary circumstan-
ces, the stench of this holocaust would
have added another horror, but here it
was lost in the fumes.

The odors of the fumaroles in the
Valley as we found it five years later were
repulsive enough. During the eruption
the fumes were intolerable a hundred
miles away. We may not attempt to
suggest the sensations that would have
smitten the olfactory nerves during the
height of the cataclysm.

THE VALLEY BECOMES A SEETHING FIERY
TORRENT

Long before the fires that consumed
the surrounding vegetation had run their
course, the masses of incandescent frag-
ments accumulating around the separate
vents had probably run together until
they covered the whole floor of the Valley,
converting it into a single fiery torrent
of seething, swirling, red-hot sand and
rock, kept in continual turmoil by the
constant escape of gas.

The gas presumably came off mostly
in quiet ebullition, but at times it let go

with violent explosions which tore wide
holes in the mass of glowing fragments.
These explosions must have been accom-
panied by the most spectacular up-
heavals of red-hot rocks hurled high
aloft.

The fiery Valley at this stage doubt-
less presented a spectacle, if it could
have been seen through the impenetrable
black cloud, far surpassing the weird-
est image of the place of everlasting tor-
ment ever conjured up by poet or
preacher. Again, in default of language
of my own, I resort to the description of
an Hawaiian eruption : '

''At half-past three p.m. I reached
the awful crater and stood alone in the
light of its fires. It was a moment
of unutterable interest. ... I was
10,000 feet above the sea, in a vast soli-
tude untrodden by the foot of man or
beast; amidst a silence unbroken by any
living voice, and surrounded by scenes of
terrific desolation.

"Here I stood almost blinded by the
insufferable brightness; almost deafened
with the startling clangor; almost petri-
fied with the awful scene. The heat was
so intense that the crater could not be
approached within forty or fifty yards on
the windward side, and probably not with-
in two miles on the leeward. . . .

"The mountain seemed to be siphun-
culated; the fountain of fusion being
elevated some two or three thousand
feet above the lateral crater, and being
pressed down an inclined subterranean
tube, escaped through this valve with
a force which threw its burning masses
to the height of four or five hundred feet.

"The eruption first issued from a de-
pression in the mountain, but a rim of
scoriae two hundred feet in elevation had
already been formed around the orifice
in the form of a hollow truncated cone.
This cone was about half a mile in cir-
cumference at its base, and the orifice at
the top may be three hundred feet in
diameter.

• From description by Titus Coan, quoted from
C. H. Hitchcock, "Hawaii and its Volcanoes,"
pp. 87-89.

Digitized by

Google

Photograph by E. C. Kolb
FRYING BACON OVER A FUMAROLE

It is hard to appreciate the situation from a still picture. The steam is so hot and dry as it rushes
forth that it is perfectly clear. The pressure lifted the frying pan high in the air. It had to be held
down against the outrushing steam (see page 225).

Photograph by R. F. Griggs
A BONFIRE KINDLED BY WATER (SEE PAGE 22 1 )

One of the fumaroles was so hot and dry that shavings burst into flame after being plunged for a
moment into its hot vapor. Since this consisted of almost pure steam — that is to say, water — what we
really did was to kindle a fire by poking a stick into water. Temperature measurements by the
geophysicists showed that it was 645** C. (nearly 1200** P.).

282

Digitized by

Google

A HYPOTHETICAL CHRONICLE OF THE ERUPTION

283

"I approached as near as I could bear
the heat, and stood amidst the ashes,
cinders, scoriae, slag and pumice, which
were scattered wide and wildly around.
From the horrid throat of this cone vast
and continuous jets of red-hot and some-
times white-hot lava were being ejected
with a noise that was almost deafening,
and a force which threatened to rend
the rocky ribs of the mountain and to
shiver its adamantine pillars.

*'At times the sound seemed subter-
ranean, jdeep and infernal. First, a
rumbling, a muttering, a hissing or deep
premonitory surging; then followed an
awful explosion, like the roar of broad-
sides in a naval battle, or the quick dis-
charge of pack after pack of artillery on
the field of carnage.

'* Sometimes the sound resembled that
of 10,000 furnaces in full blast. Again
it was like the rattling of a regiment of
musketry; sometimes it was like the
roar of the ocean along a rock-bound
shore; and sometimes like the booming
of distant thunder. . . .

**The eruptions were not intermittent,
but continuous. Volumes of the fusion
were constantly ascending and descend-
ing like a jet d'eau. The force which
expelled these igneous columns from the
orifice shivered them into millions of frag-
ments of unequal size, some of which
would be rising, some falling, some shoot-
ing off laterally, others describing grace-
ful curves; some moving in tangents, and
some falling back in vertical lines into
the mouth of the crater.

** Every particle shone with the bril-
liancy of Sirius, and all kinds of geomet-
rical figures were being formed and
broken up. No tongue, no pen, no pen-
cil can portray the beauty, the grandeur,
the terrible sublimity of the scene. To
be appreciated it must be felt. . . .

'* During the night the scene surpassed
all power of description. Vast columns

* It should be recognized that little, if any, of
the material was transported the whole 15 miles.
At frequent intervals along the way there were
new vents which added their quota to the fiery
flood. One of the most striking features of the

of lava at a white heat shot up contin-
uously in the ever varying forms of pil-
lars, pyramids, cones, towers, turrets,
spires, minarets, etc., while the descend-
ing showers poured in one incessant cat-
aract of fire upon the rim of the crater
down its burning throat and over the
surrounding area.*'

Had the quantity of gas evolved been
less, the material might have remained
liquid lava, as in the Hawaiian flows, and
hardened into solid rock on cooling; but
as it was, the rock-forming constituents
of the magma were completely disrupted
by the bubbles of escaping gas and
blown into fine fragments which, on cool-
ing, became ash and pumice similar to
that of the typical explosive eruption.

RED-HOT ASH ROLLS DOWN THE VALLEY
LIKE A LIQUID

Soon after reaching the surface, the
tumbling rolling mass of glowing frag-
ments must have begun to move down-
grade toward the foot of the Valley.
Aided by the quantity of gas that was
still escaping from its substance, the
fiery sand behaved like a liquid, adjust-
ing itself closely to the gradient of
its bed.

A dozen miles down the Valley the red-
hot flood encountered a belt of morainic
hills that to a certain extent dammed its
advance. But it quickly found a way
around among these, seeking everywhere
the lowest notch and flowing through it
in response to gravity, almost as readily
as a flood of water would have done.

In this way it continued down the
Valley for 15 miles (counting its starting
place as Novarupta and the Pass) before
it finally came to a stop.* After its for-
ward motion had ceased, it continued to
be agitated for a time by internal ex-
plosions, as evidenced by the craters
which in places honeycomb its surface.

Valley is the close coincidence between the area
of activity and the boundaries of the sandflow.
It is not unlikely that the contribution of fresh
magma from the lower vents supplied an impor-
tant part of the material in the lower portion of
the Valley.

Digitized by

Google

284

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by J. D. Sayre

A LITTLE SPUTTERING MUD POT

Such phenomena are common in the Yellowstone Park, but by reason of the high temperatures in the
Valley of Ten Thousand Smokes, they occur only where rain water that has run into small fumaroles
is churned into mud. After a few days they become fumaroles again (see text, page 239).

THE EXPLOSION OF NOVARUPTA

The largest of the vents contributing
to the Valley flood was probably Nova-
rupta. Its action may perhaps have
been more violently explosive than the
others; at any rate, before the hot sand
flow had fully settled down, Novarupta
passed over into the paroxysmal activity
of the typical explosive volcano.

Great blocks of red-hot lava ten feet
long were hurled high in air, while glow-
ing bombs nearly a yard in diameter were
fired thousands of feet aloft clear over
Falling Mountain to the far slopes of
Mageik. Whether these large bombs
could have been seen through the black
cloud of vapor and fine ash given off with
them is uncertain, but if they were vis-
ible their sky-rocket courses must have
formed one of the most spectacular fea-
tures of the whole eruption.

OPENING OF KATMAI AND THE OTHER
VOLCANOES

What was going on at Katmai and at
the other old volcanoes during this early
part of the eruption, we have no means of
knowing. It is not likely that they could
have been altogether quiet, but they were
not active enough to leave behind dis-
cernible evidence of the fact among the
mass of ejecta from Novarupta and the
other Valley vents.

It is not improbable that the craters of
Mageik and Martin had already opened,
presumably with some explosive accom-
paniments which would have been re-
corded as considerable eruptions if not
overshadowed by other activities of
greater magnitude.

At all events Katmai must certainly
have been preparing for its great explo-
sions. Of the nature of these prepara-

Digitized by

Google

A HYPOTHETICAL CHRONICLE OF THE ERUPTION

285

tions, we have some inkling from the
character of the material thrown out, but
many of the questions we most wish to
answer are beyond our reach.

Among the most interesting and signif-
icant of these questions are the reasons,
first, for the delay of the great explosions
until after the Valley eruptions, and,
second, for their beginning promptly
after the extrusion of the incandescent
sand.

We should have expected that Katmai
would either have been set off by the
shock of the bursting of the Valley or,
having survived that cataclysm, would
have waited until some new cause pulled
the trigger to set off its magazine of
accumulated destructiveness. But the
mixture of the ejecta shows that before
Novarupta closed its period of explosions,
Katmai began its paroxysm with a vio-
lence that far eclipsed anything that had
preceded.

SUMMARY OF THE EVENTS OF THE
ERUPTION

We do not understand the mechanism
of such explosions well enough to add
much to the account given in the early
chapters. In'closing, therefore, we shall
merely summarize the events of the erup-
tion in the order of their occurrence, for
convenience of reference:

(i) A peaceful valley, covered with
trees and grass, lying among dormant
volcanoes.

(2) Premonitory symptoms, especially
earthquakes, sufficient to warn the
natives.

U, (3) Rock fall from Falling Mountain,
early June, 1912.

(4) Pouring out of the great hot sand-
flow from fissures in the Valley of Ten
Thousand Smokes and its branches, cov-
ering 53 square miles to an estimated
average depth of 100 feet.

(5) The Smokes of the Valley begin
their operation, perhaps coincident with
the sandflow.

(6) Explosive activity begins at Nova-
rupta; about one-half of a cubic mile of
coarse pumice thrown out over an area
10 to 15 miles in diameter, overlapping
with the explosion of Katmai.

(7) Opening of craters of Martin and
Mageik. (Sequence doubtful.)

(8) The Mageik and other landslides.
(Sequence doubtful.)

(9) June 6th, i p.m. The first of the
major explosions of Mount Katmai; re-
sponsible for first layer of gray ash.

(10) June 6th, 11 p.m. Second major
explosion; ^cond layer of ash (terra-
cotta).

(11) June 7th, 10.40 p.m. Third ma-
jor explosion; third layer of ash (gray).

(12) Fourth major explosion. Fourth
layer of ash (yellow).

(13) Cap layers of fine red mud thrown
out by Katmai.

(14) Katmai Mud Flow. (Sequence
doubtful.)

(15) A condition of great but gradu-
ally subsiding activity, the interior of the
crater retaining an incandescent heat,
manifested by red reflections on the
clouds as late as July 21st.

(16) The end of the explosive stage,
followed by a quiet evolution of vapor in
great quantity from Katmai, Martin,
Mageik, Trident, and the Valley until
the present time.

Digitized by

Google

•c

< to

S "^

fa •«

cn .2

i-i

w "a
« S

286

Digitized by

Google

XXII

WHAT BECAME OF THE TOP OF KATMAI?

Beyond a reconstruction of the events
of the eruption lies the problem of inter-
preting the subterranean disturbance re-
sponsible for the outbreaks observed at
the surface. This is the real task of the
volcanologist.

Until he has a clear idea of what goes
on in the depths, it is impossible for him
to understand, much less to predict, the
outbursts at the surface. It is not the
province of this book to develop an ex-
tensive theoretical discussion. Yet there
are certain very significant facts that are
apt to be overlooked unless they are con-
sidered in the light of the deductions to
which they give rise.

Along with the impression of tremen-
dous size which dominates everyone who
looks into Katmai Crater, there comes to
any thoughtful observer a question as to
the fate of the enormous quantity of ma-
terial that once occupied the stupendous
hole in the top of the volcano. This
may strike the reader as a sup)erfluous
query. Were not the great explosions
powerful enough to blow the old moun-
tain to atoms? Probably they were, but
even a cursory examination shows that
what happened was far more complex
than the mere blasting away of the old
rock.

ROCK OF THE ORIGINAL MOUNTAIN TOP
DISAPPEARED UTTERLY

If the old mountain top had simply
been blown off by explosions beneath,
the slopes would be littered with the de-
bris and one would find great chunks of
it scattered about the crater. The two
cubic miles of rock missing from the
mountain, if spread evenly around the
vent for lo miles in every direction,
would make a layer i6 feet deep.

No such quantity of material could be
concealed among the ash layers. As a
matter of fact, the total thickness of the

entire deposit of ejecta of all sorts at the
base of the mountain is less than that
amount.

We had expected, as a matter of course,
that when we reached the volcano we
should find great rocks, which had been
heaved out by the explosions, and per-
haps the most surprising thing observed
on our first visit was the fine grain of
the deposits, right up to the volcano.
Everything was '* blown to smithereens.*'
In the ash and pumice, which compose
the bulk of the deposit, the fragments
seldom exceed six inches in diameter,
the largest observed hardly reaching a
foot in length.

Of the old mountain top no large
pieces remain. Fragments of any size
are comparatively scarce and seldom
exceed a hen*s egg in dimensions. All
told, they would make only an insignif-
icant fraction of the former summit.

ROCK OF OLD MOUNTAIN VERY DIFFERENT
FROM NEW LAVA

It happens that the rock of the old
mountain can be distinguished from the
new lava at a glance. The mountain
was built up entirely of flows of dark
colored basic andesite, containing 60
per cent of silica. But the lava of this
eruption was a white, acid rhyolite
with 75 per cent silica (see page 290).
This conspicuous difference makes it
easy to follow in detail the action of the
new magma on the old mountain.

There can be little doubt that the
present hollow crater was originally oc-
cupied entirely by dark colored rock
relatively poor in silica. The whole stub
of the mountain as it stands to-day
consists exclusively of the ends of an-
cient lava flows radiating from the
former summit.

As the lava at the head of a flow must,
almost necessarily, be the same as at

287

Digitized by

Google

288

THE VALLEY OF TEN THOUSAND SMOKES

its tip, the stub furnishes certain evi-
dence of the character of the original
summit. All the old lava is the same
dark colored rock and there is no trace
of the new white lava in the crater wall.
The whole of the two cubic miles must,
therefore, once have been filled by old
dark lava.

This conclusion brings out the diffi-
culty of the situation. If there were two
cubic miles of dark rock in the beginning,
and if only a fraction of it is to be found
in the deposits, the rock of the old moun-
tain was not merely removed, it has to
a large extent utterly disappeared. It
could not have vanished into thin air.

Only two other methods of displace-
ment seem possible: the whole summit
may have caved in, slumping down into
the void left by the extrusion of new
magma; or, the new magma rising in
the throat may have melted up the old
rock.

COULD THE OLD MOUNTAIN TOP HAVE

BEEN SWALLOWED UP IN THE

BOWELS OF THE EARTH?

For a whole mountain to slump away
and be swallowed up into the bowels
of the earth seems a catastrophe of
incredible magnitude. Such an event
would undermine our instinctive con-
fidence in the solidity of the earth's
crust on which we live. Yet just that
process has been evoked in explanation
of the famous Crater Lake of Oregon.
And certainly it is quite conceivable that
the emission of a vast quantity of ash
might leave a cavity below big enough to
receive the superincumbent mountain.

Subsidence of exactly this character
occurred on a small scale in the case
of Novarupta, after its paroxysm. As
described in Chapter xix, the rock for
half a mile round about is all broken to
pieces by faults traversing it in many
directions. In every one of these, the
ground sankon the side toward the crater.

This means that Novarupta has settled
below its former level, perhaps more than
loo feet. Subsidence in this case was
delayed, as would be expected, till

after the close of the eruption. This is
proved by the fact that all the ash layers,
as well as the rock, are cut by the faults.

NO EVIDENCE OF SUBSIDENCE

But, according to eye witnesses, the
top of Katmai had already disappeared
in the early stages of its explosion. Of
this we have independent testimony from
two sources, in the stories of the natives
of Katmai Village and of Savonoski,
as detailed on page 17.

The faults around Novarupta are
straight lines, which crisscross over a
considerable area. If subsidence of this
type had occurred in Katmai, the result
would have been a ragged irregular hole,
not a clean-cut circular crater. Quite
apart from subsidence, it would be ex-
pected that the terrific impact of the
explosion would have cracked open the
old mountain, opening fissures that
would be seen radiating in every direc-
tion.

But not a break occurs in the crater
walls and there is no sign of any fracture
of the original contours of the mountain
slopes. It seems improbable that sub-
sidence of the proportions necessary to
form an abyss like Katmai could have
taken place without the formation of
some faults that would radiate beyond
the crater walls.

Another difficulty with the hypoth-
esis of crater subsidence in the case of
Katmai is that it involves definite and
rather improbable assumptions regard-
ing the shape of the magma chamber
beneath the volcano. It would mean
that the shape as well as the size of the
cavity created by the eruption corre-
sponded with the present crater; also
that the reservoir lay close to the surface.
Otherwise subsidence could not have pro-
duced a sharp, clean-cut hole like the
present crater.

If the cavity lay deep down below the
mountain, the strong roof would sag
gently rather than break sharply as a
thin roof might. A magma chamber of
the proportions necessary does not
accord with the type generally supposed

Digitized by

Google

Natural Color Photograph by Frank I. Jones
HARD BAKED CRUST AROUND A SMALL FUMAROLE
The issuing steam, being perfectly clear, is invisible in the picture.

XIII

Digitized by

Google

Natural Color Photograph by Frank I. Jones
THE INTERIOR OF A FUMAROLE LAID BARE
A vent like that on the preceding page dissected by a stream. The darkening colors measure the increasing

heat toward the throat.

XIV

Digitized by

Google

WHAT BECAME OF THE TOP OF KATMAI?

289

to underlie volcanoes. The reservoir
commonly assumed is rather of the kind
indicated by the subsidence around
Novarupta.

We must, therefore, dismiss the theory
of crater subsidence as highly improb-
able, for it seems hardly possible that
such a process could have occurred with-
out leaving more tangible evidence of
the fact.^

MUCH OF OLD MOUNTAIN EATEN
AWAY BY NEW MAGMA

The third possible explanation may
likewise appear improbable, yet it has
the merit of being supported by a great
deal of evidence of the most unequivocal
sort. This is, that the old mountain was
corrosively attacked from beneath by the
new magma and in a large measure melt-
ed away or, better, dissolved by it.

All stages in the process of digestion are
abundantly represented in the fragments
blown out by the explosions. Almost
every piece of pumice one picks up shows
evidence of a remarkable, composite
character. Everywhere, black fragments
of the old mountain top are embedded
in a matrix of the new white lava.
Most commonly these remnants of the
original rock appear as conspicuous
black bands alternating with the light-
covered matrix (see page 291).

The condition of these fragments
shows that digestion of the old mountain
was surprisingly thorough. Large pieces
of unaltered rock are nowhere to be found ;
seldom do they exceed an inch in diam-
eter, commonly they are much smaller.

ALL STAGES IN THE SOLUTION OF

OLD ROCK SHOWN IN THE

FRAGMENTS EJECTED

When examined carefully either with
the naked eye or, after sectioning, under

^ Slumping away of the unstable walls left by
the explosion has undoubtedly played a large
part in determining the present shape of the pit,
and it is still occurring at frequent intervals.
But this process is of insignificant proportions as
compared with the disappearance of the whole
mountain top in the depths of the earth, as con-
templated by the hypothesis.

the microscope, the included fragments
show all possible stages in the: penetra-
tion of the old rock by the new magma.
In some blocks only the corners are
rounded off where dissolved. Others
are penetrated by narrow tongues of
white magma. In still others the new
magma and old rock are mixed together
in the most intimate fashion, until the
whole consists of the old minerals lying
in a ground-mass which, while retaining
the dark color of the original rock,
nevertheless has an index of refraction
characteristic of the new magma.

In other cases, all that remains are
some of the old crystals. The faces of
these are corroded in such a way as to
show plainly that they were in process
of solution when thrown out and cooled
down.^

ASH AT KODIAK CARRIES MANY CRYSTALS
FROM OLD MOUNTAIN

The same duplex character is main-
tained with astonishing consistency
throughout the ejecta wherever found.
This remarkable feature was first rec-
ognized by Dr. W. J. McCaughey, a
petrologist particularly expert in the
identification of dust particles, to whom
I submitted samples of the various types
of ejecta.

It developed that McCaughey was
suspicious for a time that the samples of
ash I had given him were ** fixed.*' He
found that the most abundant compo-
nents were fragments of rhyolitic glass
(from the new white lava) and crystals
of labradorite. As these do not belong
together, McCaughey felt very much
as a naturalist would, if told that some
newly discovered island was inhabited
by giraffes and polar bears.

Except that all the samples from what-

* Even where none of the matrix of the rock
remains with these crystals, there can be no ques-
tion of their derivation from the old wall rock.
They are for the most part plagioclase feldspar
of the variety known as labradorite, exactly like
those found in the andesite of the old mountain.
The feldspar of the new magma, on the other
hand, is of very different character, not to be con-
fused with the labradorite of the old mountain.

Digitized by

Google

290

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by £. L. Wishard

NEW AND OLD LAVAS FROM KATMAI

Left: A very large piece of white pumice composed of new rhyolitic magma, silica content 75 per
cent. Right : Two views of a column of old black andesitic lava from a flow near the base of the vol-
cano, silica content 60 per cent. One-half natural size.

ever source, whether Kodiak, Katmai
Village, or the foot of the volcano, showed
the sane mixture, he would probably
have rejected them unceremoniously as
frauds. My reputation was not cleared
until thin sections of the larger pUmice
stones, which n eanwhile had been pre-
pared, were examinini with the micro-
scope. Ihen the whole story of the

dissolution of the old mountain top in
the new magma was laid bare.

MUCH OF THE OLD ROCK PERMEATED WITH
GASES FROM THE NEW MAGMA

In the larger pieces of pumice there is
a marked difference in the density of the
white ground mass and of the inter-
vening black bands. The new lava is

Digitized by

Google

WHAT BECAME OF THE TOP OF KATMAI?

291

Photograph by E. L. Wishard

MIXED PUMICE FROM KATMAI

The white material is new lava, the black, old. Portions of the old lava in process of solution when
solidification occurred are strewn through the white mass. Natural size.

generally much more porous, indicating
that it had a much greater content of
expansible gases. There is evidence also
that the new component was more fluid.
This is as would be expected. Compare
the fluted cornice structure of lava in
Novarupta (see page 296).

Yet even the black bands are very
much more vesicular than the old lava
from which they were derived. These
dark masses as well as the light material
often frothed up into a pumice. When
this happened they also must have been
heavily charged with gases that had
diffused into them from the new magma.

The thoroughness with which the
fragments are torn apart points in the
same direction. If the mass had been
less completely permeated, the heavier,
inert, unaffected portions would have
been thrown out in large unaltered
blocks, such as we had expected to find
j)n the mountain sides. . '

The fact that unaltered pieces of the

old mountain are so small and so scarce
is eloquent testimony of the complete-
ness with which the exploding mass was
penetrated by the gas from the new
magma.

In the thoroughness with which its
products were disrupted, Katmai stands
at the very extreme of explosive volca-
noes. Very rarely do all the products
of an eruption show anything like such
a fine degree of subdivision as the ash
of Katmai.

TEMPERATURE OF THE NEW MAGMA
LOW — LESS THAN I,000°C.

The thoroughness of permeation by
the new magma raises the question of
the means by which that penetration was
accomplished. As will be seen, part of
the old rock was dissolved in the pool of
lava that occupied the top of the moun-
tain during the eruption. This will be
discussed in its turn. But before the
magma reached the crater it had to eat

Digitized by

Google

292 THE VALLEY OF TEN THOUSAND SMOKES

I'botograph by Jasper U. bay re
THE INTERIOR OF ONE OF THE CRATERS

The distorted perspective makes it impossible to get a satisfactory photograph of the interior of

these yawning pits.

A LINE OF SMALL CRATERS ALONG A FISSURE

These miniature craters show the relations of the chains of volcanoes along great world fissures.
The presence of numerous lines of craters strung along fissure lines strongly suggests that the fractures
are fundamental, reaching down to the magma chamber below.

Digitized by

Google

WHAT BECAME OF THE TOP OF KATMAI?

293

its way through an unknown but cer-
tainly considerable thickness of old rock,
forming the solid summit of the moun-
tain.

Was the rising magma so hot that it
could melt away the old rock without
being chilled to the point of solidification?
Assuredly not. There is good evidence
that its temperature was low, probably
lower than the melting point of the old
rock.

It so happens that the new lava con-
tains crystals of quartz, which is a first-
class ** geological thermometer.*' In con-
sequence of the properties of quartz, '* we
can be fairly sure that a temperature of
less then 1 ,000° prevailed " ^ in the magma.

As the melting point of the old rock can
hardly have been much below i,ooo°c.,
and may have been considerably higher,
it is clear that melting (in the strict
sense) can not have occurred to any
great extent. What happened was,
rather, a process of solution. It was the
enormous surcharge of gas carried by the
magma that permitted it to attack the
old rock, dissolving it somewhat as a
lump of sugar dissolves in water or in an
unsaturated sirup.*

ALL THE ASH COULD BE PACKED INTO
THE CRATER

A comparison of the material thrown
out from Katmai with the size of the hole
in the mountain gives further significant
information concerning the nature of the
eruption.

The volume of all the material ejected,
as it lies on the ground, is estimated as
4.75 cubic miles (see page 29). Deter-
minations of the specific gravity of a

' Fenner, Jour. Geology, vol. 28, p. 602, 1920.
Under atmospheric conditions quartz cannot ex-
ist above 870** C. At that temp>erature it changes
to tridymite. The inversion point would proba-
bly be raised by the great pressure to which the
magma was presumably subjected. "A thick-
ness of 20,000 feet of rock strata might possibly
raise it 100**, but we can be fairly sure that a
temp>erature of less tharf 1,000** prevailed."

* Chemists draw a sharp distinction between
melting and solution. A mineral melts over a
fire only at a definite and constant temp>erature,
but it may be dissolved in greater or less degree

Photograph by E. L. VV'ishard

FROTH FROM THE LAVA LAKE AND AN
EQUAL WEIGHT OF GRANITE

The two have the same chemical composition
and were derived from similar magmas. If the
pumice had solidified in the depths of the earth it
would have become granite. Comparison of the
densities of the two show that the pumice liber-
ated twelve times its own volume of gas.

number of representative samples of dif-
ferent degrees of fineness average 1.03.
That is to say, the ash is slightly heavier
than water. Samples of the rock of the
old mountain uniformly have a spe-
cific gravity of about 2.5. If we may
assume that this represents the original
density of the mountain top, we can
compute the volume the ash would

at any temperature, depending on the composi-
tion of the solvent. The new magma was in
reality a mutual solution of many rock-forming
constituents, together with quantities of volatile
components, steam, acid gases, and other sub-
stances. Such a complex solution remains liquid
at a temperature far below the melting point of
the solid lava formed after it has been extruded
and its vapors have been jjermitted to escape.
Similarly, an excess of gases in a magma would
give it a power of dissolving old rock far beyond
what would be expected if its temperature alone
were taken into account.

Digitized by

Google

294 THE VALLEY OF TEN THOUSAND SMOKES

pnotograph Dy k. r. urtggt

FUMAROLES AT THE NORTHWEST CORNER OF BAKED MOUNTAIN

One of the most active areas in the Valley. The temperature as measured was not especially high,
but the quantity of heat delivered was enormous. The smoke of one of these f umaroles had a curious
pink tinge, visible in every light.

A GROUP OF THE SMALL FUMAROLES KNOWN TO THE BOYS AS PIMPLES

The moisture of the emanations cSatches blowing sand and builds up the pile. The forested slopes
beyond the limits of the sandflow are the base of Katolinat.

Digitized by

Google

WHAT BECAME OF THE TOP OF KATMAI?

295

occupy if compressed to the density of
the old rock, and compare directly the
ash with the amount of rock missing
from the mountain. The difference will
thus give us a rough estimate of the
quantity of new magma that rose from
the depths during the eruption.

A simple multiplication brings out the
astonishing information that the mass of
the material thrown out was substan-
tially equivalent to that of the old rock
missing from the mountain.* In other
words, the whole of the ash could be
packed into the crater, if it were com-
pressed to the density it occupied before
exploding!

If this comparison is valid, it must
mean that little additional magma came
up from the depths during the explosion.
The greater part of the huge amount of
new material in the ejecta must already
have been present in the mountain before
the explosion began.

A PORTION OF THE OLD MOUNTAIN SEEMS
TO HAVE SUNK INTO THE DEPTHS

If the ash deposits as they stand are
the substantial equivalent of the material
that formed the old mountain top just
before the explosion, the problem of the
disappearance of the old mountain top
takes on a new aspect.

It is clear that the greater part of the

» The computation is as follows: Cubage, ash,
4.75 cubic miles— 25,800,000,000 cubic yards of
density 1.03. Rock blown off the old mountain:
11,000,000,000 cubic yards of density 2.5.

II .O X 2.5 27.5
25.8Xi.03"56:6"'^3 per cent.

That is to say, the hole actually appears 3 per
cent bigger than the ash that came from it.

The fact that these figures so nearly correspond
is, however, of no particular significance, because
each term of the proportion is subject to large
and uncontrollable errors. Nevertheless, the
estimates are unbiased. They were made inde-
pendently by different persons, with no thought
of their ever being compared in this fashion.
Each was, moreover, checked in every way
possible.

The estimate of ashfall was made by summing
the contours of ash of different depths. The data
were not sufficient to permit more than a very
rough estimate.

ash is derived from new magma, decided-
ly less than half being directly traceable
to the old dark rock. That is to say, a
considerable portion of the old rock
must have been replaced before the ex-
plosion. What became of it?

How did the new magma squeeze into
a solid mountain? Dissolving up the old
rock does not eliminate its materials.
After solution was complete the complex
must have occupied nearly as much space
as the two components had before inter-
mingling.

Even a half-and-half mixture would
have required the addition of two cubic
miles of new magma, enough to double
the height of the mountain. That the
outer configuration of the mountain was
not materially altered before the begin-
ning of the explosions, we may be sure.

There seems only one possibility: old
rock to an amount equivalent to the rising
magma must have been displaced down-
ward. The new gas-charged liquid mag-
ma was undoubtedly less dense than
the old solid rock. Consequently, any
blocks of the latter which might become
detached from the roof or wall of a
chamber filled with the new liquid mag-
ma would certainly sink, if given time
enough.

A mass of magma pressed upward with
great force, and penetrating the joint
planes of the adjacent rock, would almost

The cubage of the rock carried off the moun-
tain top originally estimated by Maynard has
been independently computed by Sayre, Hagel-
barger, and myself from the reconstruction of the
mountain. Maynard's figure was found to be
close to the mean of the other estimates.

The specific gravity assumed for the old moun-
tain top would probably hold, if the mountain at
the time of the eruption had been composed en-
tirely of old rock. Since, as will appear, it must
have been to a large extent filled with new
magma, the figure given is undoubtedly too
high. Allowance for this error would make the
quantity of ash somewhat larger than the hole —
perhaps as much as 10 per cent. But it would
not vitiate the general conclusion.

It cannot be held, therefore, that nothing at all
was thrown out except the disrupted fragments of
the mountain, yet it is safe to conclude that
extrusion of Additional matter played at most
only a small part in the eruption.

Digitized by

Google

296 THE VALLEY OF TEN THOUSAND SMOKES

PhotORraph by R. F. Griggs
CORNICE STRUCTURE IN A BLOCK OF LAVA ON THE PLUG OF NOVARUPTA

It is evident that flowage took place after the block was broken loose. It is noticeable that the
darker bands of old lava were less fluid than the alternating light bands of new lava, for they alone
oozed out.

MOUNT CERBERUS LYING ACROSS THE HEAD OF THE VALLEY LIKE A CROUCHING

ANIMAL GUARDING THE ENTRANCE TO HADES

In the foreground a closer view of one of the lines of fumaroles shown on page 234.

Digitized by

Google

WHAT BECAME OF THE TOP OF KATMAl?

297

inevitably break off loose blocks. In
this way, the whole interior of the moun-
tain could have been gradually replaced
without any external disturbance.*

There is no means of estimating how
large a part such "overhead stoping*'
may have taken in the removal of the
old mountain. But it can hardly be
doubted that a significant fraction of the
missing dark rock is to be thus accounted
for. Perhaps it was the major factor in
its disappearance.

It is interesting to recall in this con-
nection that the banded lava in the
plug of Novarupta (see page 240) and
the sand of the Valley show the same
dual character as the ash from Katmai.

Novarupta and the other Valley vents
were undoubtedly fed from the same
magma chamber as Katmai. The syn-
chrony of the eruptions leaves no doubt
of that. What would be more probable
than that the andesitic components in
their lava were due to contamination by
fragments of old rock from Katmai in
process of sinking into the depths ? ^

EXPLOSIONS APPEAR TO HAVE STARTED AT

THE SUMMIT OF THE MOUNTAIN — NOT

IN THE DEPTHS

The remarkably fine fragmentation of
the ash from Katmai leads to another
deduction of much importance to an

•This theory of "overhead sloping" has been
elaborately worked out by R. A. Daly (see his
"Igneous Rocks and their Origin"). As this is
still a matter of controversy in some quarters, I
have hesitated to follow it in a book of this sort,
whose aim is to avoid taking sides on disputed
questions. But the facts observed in the field
almost compel a statement in terms of the theory
of overhead stoping.

^ Fenner thinks that the admixture of basic
materials in the deposits of the Valley was prob-
ably due to the digestion of glacial bowlders which
covered the Valley floor. But this explanation
cannot be applied to the banded lava in the
throat of Novarupta which, though never sub-
jected to contamination from outside sources,
shows fully as much admixture of andesitic com-
ponents from old lava as any other.

Fenner's reason for supposing that the basic
components were derived from glacial bowlders
is that, "The limited degree of mixing of solutions
shown by these specimens hardly permits us to

understanding of the volcanic mecha-
nism. My preconception of a volcanic
explosion was that the magma, heavily
charged with gases, bored its way up-
ward until it reached a place where the
weight of rock above was less than the
expansive force of the gases confined
below, when the lid was blown off and
all the magma accessible to the newly
opened orifice frothed up explosively.

The series of explosions which make
up an eruption was supposed to be due
to the rise of successive portions of the
magma body through the conduit. This
is the view generally held.

It is plain that no such theory fits the
facts here. If the magma rising in the
throat of the volcano had possessed any-
thing like the explosive energy it later
manifested, the inert crust over the top
of the mountain would have been thrown
off in short order. If this had occurred,
remnants of the shell would certainly be
found beneath the ash layers. But as a
matter of fact, *'the bottom of the deposit
seems actually to contain less of the old
rock than any succeeding stratum.'**

A MOLTEN LAKE IN THE CRATER
AT ONE STAGE

Not only were the explosions delayed
until after the crater was unroofed, but

suppose that the solvent action was long con-
tinued." (Op. cit. p. 583.)

As a matter of fact, however, we have no real
knowledge of the time element in the case. We
only know that digestion was incomplete. Many
able authorities deny that assimilation can occur
at all, except perhaps on a very small scale.

It is only the absolutely inescapable character
of the evidence that has induced both Fenner
and myself to assign so much importance to it
here. If, then, a magma is but rarely sufficiently
charged with surplus energy to attack foreign
rocks on a large scale, it might be argued a
fortiori that when assimilation does occur it pro-
ceeds slowly. Yet it is clear that if all the as-
similation observed occurred during the brief
period of the eruption, it must have been a rapid
process.

But, below ground, no time limits are assign-
able, for no one can guess how long the present
outbreak may have been brewing. The digestion
of the old rock may have been going on below
the region for centuries, for aught we know.

" Fenner, op. cit. p. 603.

Digitized by

Google

298

Digitized by

Google

WHAT BECAME OF THE TOP OF KATMAI?

299

there is evidence that even then the
magma lay for a time as a comparatively
quiet pool in the hollowed-out summit.
This is shown by the condition of some of
the fragments in the ash deposits, in
which the bubbles are spherical and very
large in comparison with the solid matter.
Such pieces of pumice, though of course
composed of the same constituents as
the dense lava, are extremely light (see
page 293).

Their spherical bubbles show that
the escaping gases were free to expand
equally in every direction. They must
have formed on the surface of the molten
magma, representing, in fact, frozen foam
from the lava lake. Their extreme in-
flation makes them exceedingly brittle —
so brittle that they are difficult to pre-
serve. These pieces are rather hard to
find, but their scarcity is perhaps due
merely to the ease with which they were
destroyed in the hurtling hail of falling
pumice.*

Besides froth, there are materials of a
denser character, representing scum and
chilled crusts on the surface of the lava
lake. In these are frozen many inclu-
sions — pumice, separate crystals, old
rock, other pieces of volcanic glass, and
even fragments of sandstone that must
have come up from the depths. The
xondition of these materials shows that
the lava was still actively assimilating
foreign bodies at the time it was thrown
out and chilled.

This permits another step in the in-
terpretation of the eruption. The crater
walls shaken by the initial explosions
seem to have collapsed and tumbled into
the lava lake, there to be broken to bits

• Highly inflated pumice of this sort furnishes
an illuminating index to the quantity of gas that
may be liberated from a surcharged magma. It
will be remembered that this magma has the
same composition as granite; if it had solidified
deep down in the earth it would have become
granite. A comparison of the volume of the
piece pictured with that of the equal weight of
granite shows that the magma liberated about a
dozen times its own volume of gases! No wonder
a violent explosion ensued when the gas let go.

" Fenner supposes that most of the assimila-
tion of the old rock occurred in the open lake,

and stirred through the mass by the
turbulence of the swirling lava.^®

MAGMA DID NOT EXPLODE IMMEDIATELY
ON REACHING THE SURFACE

The even stratification of the ash lay-
ers shows that activity was intermittent,
waxing and waning many times. This
fact is likewise established by the testi-
mony of those who heard the explosions,
which kept up for about 60 hours. What
now is the explanation of this succession
of explosions? Why did not the whole
lava lake go off at once in a single gigan-
tic blast?

Evidently the magma did not froth
up immediately with explosive violence,
on its emergence from the depths. It
paused long enough (before and after
reaching the open air) to take up the
great quantity of foreign rock which we
find in the ash layers. Clearly the
enormous pressure of the included gases
developed at the time of the explosions
was not present at first.^*

This again runs contrary to our pre-
conceptions of a volcano — at least to
my own. I had supposed that explo-
sions were largely a simple physical proc-
ess: that the magma, heavily charged
with gases under enormous pressure,
rising from the depths to a region of less
pressure, suddenly released the gases
with a violent explosion — exactly as a
bottle of Vichy pops when the cork is
pulled.

Now evidence is supplied showing that
this analogy breaks down. Instead of
exploding immediately upon rising from
the depths, the magma paused long
enough to complete the process of assimi-

i,e. during the period of eruption. It seems to
me, however, that the slow stoping away of the
old rock before the crater was unroofed must
have been at least as important an agency.

It is difficult to see how a very great quantity
of foreign matter could be taken into the molten
lake in the few hours of visible activity without
chilling it to the point of solidification. The
question, however, turns largely on the supply of
energy necessary and this is a matter that is as
yet obscure.

" This idea was developed by Fenner, op. cit.
pp. 602-604.

Digitized by

Google

3 C

•5E
^ o

o s

« H

t- .2 ^

^ :5

0U|

•u

'i-

•g

c«

•o

>
.2

rt-c

♦^

,__

"O-bfl

300

Digitized by

Google

WHAT BECAME OF THE TOP OF KATMAI?

301

lation, before the gases began to come
off explosively. The bottle did not pop
until some time after the cork was pulled,
proving that the liberation of gases must
be due to factors other than those sup-
posed.

Here we must stop. The discovery
of the mechanism of gas liberation in
the exploding magma is the next great
problem awaiting solution by the vol-
canologist. At present, we can not
discuss it intelligently. Nevertheless,
recognition of the occurrence of a delay
in reaching the explosive condition gives
a new point of attack on the problem of
the volcano.^

WHAT IS THE CHARACTER OF THE MAGMA
CHAMBER BENEATH THE REGION?

One other question remains to be con-
sidered, and to the geologist this is the
most important of all. What were the
movements of the magma in the depths,
that resulted in the observed eruptions
at the surface? What is the nature
of the magma chamber below the region?
Our knowledge upon these points is so
limited that we can do little more than
state the problem.

Outbreaks so extensive imply that
the magma approached the surface
closely over a large area. The activity
from Katmai to Martin, from the axis of
the range to the foot of the Valley, covers
a territory about 15 miles square. The
magma itself certainly came within range
of the surface over most of this area.
What was the character of the intrusion?

SILL OR BATHOLITH?

So far as the Valley of Ten Thousand
Smokes is concerned, there are two
sharply contrasted possibilities: (i) The
magma body may be a relatively thin
sheet, an offshoot from the main body
beneath the range of volcanoes, squeezed
in between the sedimentary rocks under-
lying the region — an intrusion of the
sort known among geologists as a sill;

"•Sec a suggestive paper entitled "The Devel-
opment of Pressure in Magmas as a Result of
Crystallization," by George W. Morey, Journal

(2) it may be a very large body eating
its way up directly from the depths —
a batholith.

Both these types of intrusion are
common and important the world over.
Molten magma is often pushed in be-
tween strata of sedimentary rocks for
long distances. The sills thus formed
are sometimes of huge proportions,
vertically as well as horizontally. Some
are known which reach 1,000 feet in
thickness and extend many miles.

Solidified batholiths likewise play a
very large part in the structure of the
earth's crust. Most of the extensive
granite regions of the world are batho-
liths pushed up from profound depths.
Familiar instances are the great moun-
tain mass of Southeastern Alaska and
British Columbia, which is a single bath-
olith, and the Sierra Nevada in Califor-
nia, which is another.

The country directly adjacent to the
present region contains a smaller ex-
ample of just such a batholith in a gra-
nitic area that extends northeastward for
a number of miles from Iliuk Arm of
Naknek Lake.

In favor of the sill hypothesis, which
has been suggested by Fenner, it may be
said that this minimizes the subter-
ranean disturbance which must be
assumed.

Most sills, to be sure, are different in
composition from the lava here out-
poured, being far less silicious — basalt-
ic rather than granitic — but granite
sills are not unknown.

It is manifest that only a sill of un-
usual thickness could have furnished
the energy and materials for the erup-
tion that occurred in the Valley. Not
only was the magma that gave rise to the
great sand flow poured out — of the
order of a cubic mile, after inflation —
but enough remained to keep all the
fumaroles of the Valley going ever since.
Sills big enough to sustain such tremen-
dous drains with scant diminution are
rarely met with.

of the Washington Academy of Sciences. Vol.
12, No. Q. IQ22. pp. 219-2.^0.

Digitized by

Google

e
s

1«

o
o

(4
U

PC
n

>
o
u

<
:s

•-)
o
u

fa
o

302

Digitized by

Google

WHAT BECAME OF THE TOP OF KATMAI?

303

How a sill of sufficient thickness could
make room for itself in the old rocks
constitutes a puzzling problem. Such an
intruded body must, of course, displace
older rocks by an amount equal to its
own volume. When the sill is close to
the surface, as this body certainly is, it
would be expected that the surface
would be upheaved, and probably much
broken, when the sill was squeezed in.

But there is not the slightest indica-
tion of surface disturbance around the
Valley. The only faults certainly de-
monstrable in the old rocks are those
surrounding Novarupta, which occurred
after the eruption, i.e., after the intru-
sion was complete. As has been seen,
they are connected with subsidence, not
with upheaval.

No appreciable changes of altitude oc-
curred at the time of the eruption. The
precise triangulation of the Coast and
Geodetic Survey gives us information as
to the elevation of the high peaks in
1 908. These altitudes have been checked
against many observations by our sur-
veyors and found to be unchanged with-
in the limits of a few feet.

A further question arises as to the
nature of the parent body from which
the supposed sill originated. The mag-
ma chamber giving rise to a sill must be
of vastly greater proportions than its off-
shoot. Otherwise sufficient propulsive
force to crowd the intrusion into the old
solid crust could not be developed.

Finally, the most natural supposition
concerning the dynamics of intrusion
would be that the magma, rising in the
fundamental fissure underlying the vol-
canic axis, came to a place where resist-
ance to lateral intrusion was less than
to further vertical rise, and that a part of
its mass was accordingly pushed out to
form the sill imagined. Such action
might be aided by the weight of the mag-
ma column in the core of the volcano.

The difficulty here is, that if we liken
the interior of Katmai to a standpipe
wherein a liquid magma was held at a
level higher than the surrounding valleys,
a break at the base would inevitably

draw off the liquid and lower the level
in the standpipe. The outflows in the
Valley comprise about one quarter of the
whole quantity of material thrown out.

It would seem that the withdrawal of
so large a fraction would almost nec-
essarily relieve the pressure and drain the
higher levels in the reservoir if it were
small, thereby forestalling an outbreak
from the summit. The fact that the ex-
plosions were practically simultaneous
would seem to forbid such an explana-
tion of this case.

But if we hypothecate a batholith
rather than a sill, all these difficulties
vanish. If an extremely large body of
magma, tens of thousands of feet deep
and covering many square miles, were
eating its way toward the surface, out-
breaks even of the magnitude of the
Valley eruptions would not consume an
appreciable fraction of the total energy
and might not draw off the magma
accumulated under Katmai.

It would be much easier to think of the
enormous amount of energy necessary
for the observed assimilation of the old
rock and for the explosions as concen-
trated from a large, rather than a small,
magma body.

In the top of a batholithic mass of
magma extending downward to immense
depths, there would gradually accumu-
late a heavy charge of gas due to rising
bubbles. An eruption in such a sys-
tem, once begun, would continue until
the whole of the accumulated g£is had
blown off.

If it be granted that a batholith could
gradually work its way upward by over-
head stoping, it would approach close to
the surface before giving any indication
of its presence and then it would break
through in depressions without disturb-
ing the mountains round about, exactly
as hiippened in the formation of the
Valley.

DECISION ON THEORETICAL QUESTIONS
BEST DEFERRED

It is unnecessary, however, for us to
decide between these two hypotheses.

Digitized by

Google

304

THE VALLEY OF TEN THOUSAND SMOKES

Any opinion given now might look ri-
diculous in the light of subsequent de-
velopments. For the decision as to
what may underlie the Katmai region
will in reality be given by its behavior in
the future.

If the Valley of Ten Thousand Smokes
was produced merely by the injection of
a sill, even of a very thick one, its fu-
maroles will be short-lived. If, on the
other hand, a batholith is really rising
beneath the region, the activity will con-
tinue indefinitely. We must, therefore,
rest our case and leave the verdict to the
future.

With respect to the explosion of
Katmai we are not so fortunate. That
is a closed chapter; we cannot expect
future developments in this crater to
throw light on the many important
problems whose solution is necessary to

an understanding of the mechanism of a
volcanic explosion. Yet a more in-
tensive study of the data available would
certainly give us a much clearer picture
of the inner workings even of Katmai
than we now possess.

The prime purpose of the deductions
suggested in the preceding paragraphs
is not to set forth final conclusions con-
cerning the points brought up, but rather
to indicate the possible bearing of cir-
cumstances which might otherwise pass
unnoticed.

The thesis of the chapter is, therefore,
not to support any theory of the erup-
tion, but to bring together a number of
facts which must be taken intoaccount by
any theory which is adopted, and to point
out the unique opportunity presented by
this eruption for gaining an understand-
ing of the volcanic process in general.

Digitized by

Google

Natural Color Photograph by Frank I. Jones

TAKING THE TEMPERATURE OF MOTHER EARTH
The geophysicists with their thermo-couple working along a fissure.

Digitized by

Google

Natural Color Photograph by Frank I. Jones
PEACEFUL FOREST AND PELLUCID LAKE WITHIN A DOZEN MILES OF THE SMOKES
The head of Naknek Lake from the shoulder of Katolinat.

XVI

Digitized by

Google

XXIII

GEOGRAPHICAL AND BIOLOGICAL OBSERVATIONS
IN THE SURROUNDING COUNTRY

While the principal object of the ex-
peditions was, of course, the study of
the volcanic phenomena discussed in
previous chapters, as time permitted
exploration was extended in to the sur-
rounding country. These side trips were
very profitable, for the region is alto-
gether unknown — a blank on the map —
and we were enabled to discover a num-
ber of important geographic features, as
well as to make observations on the bi-
ology and geology of the country.

It will be a matter of surprise to many
readers to learn that there are fine forests
within a few miles of the Ten Thousand
Smokes. These escaped with little in-
jury in the eruption, because the strong
northwest wind which prevailed during
the explosions carried the ash cloud off
in the opposite direction toward Kodiak.
Beyond the terminus of the sand flow the
devastation rapidly decreases, and within
five miles one enters a fine forest which is
rapidly returning to normal conditions.

THE CONIFEROUS FOREST-BORDER
MIGRATING WESTWARD

The Katmai district lies at the extreme
western limit of the great Hudsonian
coniferous forest, which stretches un-
broken clear across the continent from
Nova Scotia and Quebec.

The dominant and most characteristic
tree is the white spruce {Picea Cana-
densis), but here, as in the Maritime
Provinces, balsam poplar (Populus baU
samifera) and paper birch (Betula Ken-
aika), the Alaskan representative of the
eastern canoe birch, are almost as abun-
dant as the spruce. Around the head
of Naknek Lake the spruces reach a con-
siderable size (see pages 267 and 310),
but within 25 miles to the west they thin
out and disappear.

The forest edge both here and at
Kodiak challenged our interest, and led

us to study the growth and reproduction
of the trees, in an effort to' determine
the factors which fixed the limit of tree
growth at this particular point.

In another paper I hope to present in
detail the observations both on the main-
land and around Kodiak. Here there is
space only for the conclusions reached.
There is no evidence that the forest mar-
gin in this region is fixed by the rigors
of the climate.

The trees grow vigorously where es-
tablished. They produce copious seed.
Seedlings are abundant, and young trees,
though greatly handicapped in starting
by competition with other vegetation
in possession of the soil, give clear
evidence of being able to grow out be-
yond the forest border.

In short, the evidence indicates that
the edge of the forest is not fixed but
moving, that it has not reached its limit
but is migrating out into the treeless
country. The advance of the Hudso-
nian forest on the northern slope of the
peninsula is slow, but the coast forest
around Kodiak is traveling westward
very rapidly, so rapidly that it has
advanced conspicuously since the Amer-
ican occupation of the country.

NAKNEK LAKE

Naknek Lake is very irregular in shape,
with steep shores and many picturesque
rocky islands. It is between 40 and 50
miles long and twelve miles wide. Unit-
ed at the base into a broad expanse, it
sends out from its head two long finger-
like extensions, separated by a high
mountain.

One of these lobes, the Bay of Islands,
is comparatively shallow and contains in
its upper end scores of forest-clad granite
islets. Another chain of islands is drawn
across its mouth, effectually shutting out
the seas which at times run up the lake.

305

Digitized by

Google

3o6

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by Frank I. Jonee
MOUNT LA GORGE FROM THE COVE AT SAVONOSKI BASE CAMP

Iliuk Arm, the other lobe, is cut off
by a pair of long spits built out from the
opposite shores at a period when the
water stood about 60 feet higher than
at present. With the recession of the
lake level, these barriers, originally
low sand bars, remain as high dikes, all
but cutting it off as a separate body of
water.

Iliuk Arm is extremely deep. The
boys were unable to find bottom with a
300-foot line anywhere except close to
shore. Since its surface is only 35 or
40 feet above the sea, it is evident that
the bottom of its bed lies far below sea
level.

As a consequence of its great depth,
Iliuk Arm has clear-cut, rocky shores
without the disfigurement of a wide
fringe of marsh, which mars the beauty
of many a fine lake. This portion of the
lake, lying within the boundaries of the
park, is set in between two beautiful
mountains that rise directly out of the
water. On the north tower the pink
granite slopes of Mount La Gorce,
3,000 feet above the water.

VIEW FROM MOUNT KATOLINAT

On the south. Mount Katolinat sweeps
upward for almost a mile (4,700 feet)
in a graceful concave curve, ending in
a series of remarkable castellated pin-
nacles. In contrast with its neighbor,
Katolinat is carved out of sedimentary
rocks. At the base are fine-grained sand-
stones and shales, but the pinnacles are
cleaved out of a remarkable conglome-
rate made up of cobbles of all sorts of
igneous rocks, many of them as large as
one's head.

The origin of this great mass of con-
glomerate, which reaches the almost un-
precedented thickness of 2,000 feet, forms
a very interesting geological problem,
whose solution would afford much in-
formation as to the past history of the
country round about.

Ivan Petrof, whose account of the
country before the eruption is quoted
above (page 265), was so impressed
with the beauties of this combination of
mountain and lake scenery that he in-
serted in his report a colored picture

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY

307

MOUNT KATOLINAT, STANDING 47OO FEET OUT OF THE BLUE WATERS OF NAKNEK

LAKE

showing Iliuk Arm with Katolinat, the
only plate devoted to scenery in an
account of travels that reached almost
every comer of Alaska.

On scaling the pinnacled ridge of Ka-
tolinat, a feat which appears well nigh
impossible from below, but yet is easily
accomplished, one obtains an unsur-
passed view over wide stretches of
country. Down the Lake one may look
on a clear day out across the flat, tundra-
covered coastal plain to the waters of
Bering Sea, nearly a hundred miles away.

Behind him, to the south stands forth
the whole range of glacier-covered vol-
canoes, plainly visible for 120 miles from
Douglas to Peulik. To the east lies the
broad valley of Savonoski River, giving
easy passage behind the range to the
shores of Kamishak Bay on Cook Inlet.
To the north is a great expanse of lake and
mountain country heavily covered with
forest (see Plate xvi, facing page 305).

LAKES GROSVENOR, COVILLE,
AND BROOKS

Lying roughly parallel with Naknek
Lake, one descries three other large

lakes which were quite unknown to the
outside world until news of them was
brought back by the Geographic ex-
peditions.

Two of them, which lie together,
joined by a short but swift river, we
named for the two men through whose
vision and support the expeditions were
made possible, the president and the
chairman of the Research Committee of
the National Geographic Society, Gilbert
Grosvenor and Frederick V. Coville.
The third we named after the dean of
Alaskan explorers, Alfred H. Brooks,
who likewise has had a large share in
helping forward the work of the expe-
ditions (see map, back cover).

Lake Grosvenor, which is 28 miles
long, is perhaps even more beautiful than
Naknek Lake. On all sides it is shut in
by high, forest-clad mountains which
give it a charm that will certainly make
it a favorite place of retreat, when the
Park becomes a popular resort (see page
308). 'It may be reached by an easy
portage of a mile and a half from the
Bay of Islands, or by ascending its out-
let through Savonoski River.

Digitized by

Google

308

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by J. D. Say re

MORNING MISTS ON LAKE GROSVENOR

For the exploration of the lakes the expedition made use of native skin kayaks, now almost a thing of
the past. Our surveyors describe Lake Grosvenor as "the most beautiful place in Alaska," which is
high praise from men familiar with the beauties of the celebrated "Inside Passage'* of southeastern
Alaska.

Lake Coville also is surrounded by
high mountains which rise 3,000 feet,
making it likewise an extremely wild and
beautiful place. On the map it appears
almost continuous with Lake Grosvenor,
but it stands at a higher level. Al-
though it lies between Lake Grosvenor
and the coast its basin is shut off from
the sea by mountains. It is therefore the
head of the system and drains backward,
so to speak, into Lake Grosvenor by a
short but swift river.

Lake Brooks (see page 309) lies on the
other side of Naknek Lake, to the south,
separated from it by a high mountain.
It also is very picturesque and quite
large enough to be shown on the general
maps of Alaska. Its drainage, as in the
others, is reversed. The outlet is at the
''upper'* end, whence it connects by a
short swift river with a bay of Naknek
Lake. In its course it tumbles over an
obstructing ledge about seven feet high,
forming a fine waterfall. Here are re-

vealed the enormous numbers of salmon
which find in Lake Brooks congenial
breeding grounds, as described below.

This group of lakes recalls the cele-
brated "Finger Lakes** of central New
York. Besides the three that come with-
in our immediate territory, there are five
others farther to the northward, all
longer than wide, and very deep. The
northernmost. Lake Clark, was reported
by its discoverer to be more than 600
feet deep.

Their basins were once occupied by
extensive glaciers, which may have played
a large part in their excavation. The
shores of the southern members of the
series which we examined show remark-
ably fine examples of glacial sculpturing.
The valleys above the water level are full
of conspicuous "roches moutonn^es, **
rounded and smoothed by the ice. Con-
spicuous evidences of heavy planing and
deep grooving by overriding ice are like-
wise abundant (see page 312).

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY

309

Photograph by J. D. Sayre

THE OUTLET OF LAKE BROOKS

This beautiful body of water, fifteen miles long, is one of the discoveries of the expeditions. The
lake is named after Alfred H. Brooks, Alaskan explorer and for many years chief of the Alaskan Divi-
sion of the U. S. Geological Survey. It was to this lake that the leaping salmon were running (see il-
lustrations, pages 310, 311, and 316).

AN ANGLER S PARADISE

The fishing in these lakes and rivers
makes the region an angler's paradise.
TKeir waters are alive with giant rain-
bow trout, with such voracious appetites
that the angler never need cast more
than once or twice before he has a strike
that keeps him busy.

There is no occasion for the varied
artificial flies often necessary to lure the
sophisticated fish of civilized streams.
Our bait, a piece of bacon rind, was
snapped up so quickly as to make it
appear that anything white would have
served equally well.

The only trouble our fishermen ex-
perienced came from the great size and
weight of the fish; they were so big that
they soon broke all our tackle. The aver-
age catch measured two feet, while the
largest — caught by Wallace — was fully
32 inches from head to tail. Probably
expert anglers who read this will simply

recognize in the damage to our tackle
a confession of incompetence in the art.
Of this we must needs stand convicted,
but the success of such novices gives
ample evidence as to the quality of the
sport that awaits the skill of the expert.
Other fish undoubtedly occur in simi-
lar abundance. The natives secure large
quantities of ** white fish" in nets. But
we were too much occupied with other
duties to try catching anything except
what could be had without special effort.

ONE OF THE GREATEST OF SPAWNING
GROUNDS FOR THE SOCKEYE SALMON

The sockeye salmon, however, occurs
in such numbers that it can not be over-
looked. The group to which Lake
Naknek belongs is perhaps the greatest
breeding ground in the world for the sock-
eye (red) salmon, which is considered the
choicest species by the connoisseur of
tinned foods, though really inferior in
flavor to some of the other species.

Digitized by

Google

3IO

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by E. C. Kolb
SALMON JUMPING THE FALLS AT THE OUTLET OF LAKE BROOKS

It was hard to catch them in still pictures. The motion picture films show that sometimes six were

in the air at once.

An extensive canning industry has
developed around the outlets of these
lakes. The value of the '*pack** every
year exceeds the purchase price of Alaska.
In 191 8 the salmon taken out of Bristol
Bay was worth nearly $25,000,000,
more than three times the cost of *' Sew-
ard's Folly."

Because of almost criminally short-
sighted exploitation, overfishing has
very greatly reduced the runs in recent
years, but despite all the devices used to
take them, the salmon still run into the
lakes in incredibly large numbers.

LEAPING SALMON WITHIN SIGHT OF THE
VOLCANOES

At Brooks Falls in the outlet from the
lake of the same name, at the proper sea-
son one may watch the salmon as they
jump clear of the water and ascend the
falls. Here we stood for hours, held by the
fascination of one of the most wonderful
sights afforded by the animal kingdom,
as the endless procession of fish kept leap-
ing high in the air, up and over the falls.

Never did a second elapse between
jumps. Sometimes as many as six fish
were in the air at once. The jump ap-
peared to require their full powers; none
made the attempt except at the lowest
notch in the falls, and none jumped clear
over in a way to suggest that they could
have gone much higher if necessary.
Many of the leaps were so wide of the
mark as to give the impression that they
were not serious attempts, but rather
in the nature of reconnaissances — efforts
to learn the best place for the ascent.
Often the fish struck themselves on the
sharp rocks. Among those below the
falls were many terribly lacerated by
such accidents — so far gone that there
was little probability of their ever suc-
ceeding in the leap.

At first we were inclined to think that
very few were successful, but careful
observation showed that great numbers
were getting up. After a number of
counts at different times, we estimated
that they were ascending at the rate of
about 20 a minute, or 1,200 an hour.

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY

311

Photograph by R. F. Griggs
SPEARING OUR SUPPER AT THE FOOT OF THE SALMON FALLS
The salmon here were all the choicest reds, or "sockeyes," which command the highest price.

If the rate of asceat observed at the
time of our random visit could be taken
as a fair average of the forty-day run,
we could conclude that a million fish
were breeding in this single lake. But
it is not safe to make numerical compu-
tations from the rate observed at any
one time, for the runs are irregular and
vary greatly from day to day.

The figures are given merely to em-
phcisize the enormous numbers involved
and the great importance of properly
conserving this wonderful food resource,
that our descendants may enjoy it as
freely as do we. Nothing can be more
certain than that if the present reckless
exploitation continues, our incredibly
rich salmon fisheries will be destroyed
within a decade or two.

THE REMARKABLE HOMING INSTINCT OF
THE SALMON

Perhaps the most interesting feature
of the whole performance is the instinct
that urges the salmon over the falls. How
can they tell that there is another lake

above?' Yet by some means they are
unerringly guided to the outlets of lakes
above, for they do not run to other
streams. Why should they seek the
upper lake, when the waters they have
just passed through would serve as well,
as is attested by the fact that they are
perfectly satisfactory to other thousands
of their brethren?

Professor C. H. Gilbert of Stanford
University, who knows the habits of the
salmon better than anyone else, tells me
there is good evidence that the individual
salmon returns to the particular water
in which it w£is hatched. As indicated
by the anomalous run observed in a trib-
utary of Katmai River (see page 161),
they are apparently able to find, not
merely the same general locality, but by
some instinct are guided back to the
particular spot where they began life.
The mysterious sense by which they
choose their way among waters appar-
ently indistinguishable is quite incom-
prehensible to us.

If, like the bird migrations, it were a

Digitized by

Google

312 THE VALLEY OF TEN THOUSAND SMOKES

Phoiograph Dy R. F. Griggs
GLACL\TED SHORES OF ISLAND BAY

The smoothed surface of the granite bears the clear imprint of the glacier which scooped out the bed

of the lakes.

Photograph by R. F. Griggs
THE BASE CAMP IN THE FOREST AT THE HEAD OF NAKNEK LAKE

We were much surprised on finding such forest within a dozen miles of the Smokes. A fox used
regularly to bring her little family to our garbage pile, and Wallace shot a bear on the beach in front of
camp. The Katmai National Monument is destined to become a great game preserve.

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY

313

journey undertaken every year during
the life of the fish, it would not be so re-
markable, for the old fish could teach
the young the path. But the salmon
make the trip only once, at the end of
their allotted span. When they have
spawned they turn over and die, leav-
ing windrows of rotting carcasses on the
shore.

Their only previous experience with
the stream was when as small fry they
passed down from the spawning ground
and out to sea. Can it be that they re-
tain something analogous to memory of
the ** landmarks'* passed on their one
previous journey? And, finally, why
such frantic effort when all is to end
so soon? These questions we may not
answer. All we can say is, that these
wonderful instincts, though leading to
the death of the individual, are clearly
to the advantage of the species, whose
perpetuation is thereby assured.

ABUNDANT BIRD LIFE AROUND
THE LAKES

The lakes and ponds round about are
the abode of countless waterfowl. I
had often read of the abundance of bird
life in the north, but I was surprised at
the great numbers of birds encountered.
Many species of ducks, as well as geese
and swans, were all about us. Right in
front of camp one could often pick off
half a dozen ducks in a few minutes.

Once we saw a single flock of swans,
resting on a sand bar, which must have
numbered 500 individuals. On another
occasion the surveyors coming upon one
molting, and unable to fly, ran it down
with the power boat and picked it up out
of the water. Flapping and squawking
it made a handful, as may be judged by
the expression of the captor's face.

For game birds ashore, there are
grouse (** spruce hens'*) in the forest,
and ptarmigan on the mountains and
tundra, both of which were very scarce
after the eruption, but are now returning
in force.

Along with the game birds are numer-
ous other species. Corresponding with

Photograph by J. D. Sayre
A WILD SWAN CAPTURED ALIVE BY MILLER

Being at the height of its molt, it was unable to
fly, and so was easily overtaken and captured by
the power dory. It made such a disturbance
with its great white wings that the captor was
glad to let it go.

the difference in vegetation, there is a
marked contrast between the birds of the
lake country and those of the Pacific
slope. Around the lakes were Steller's
jays, shrikes, crossbills, three-toed wood-
peckers, and robins like those of the east
except in disposition, for instead of being
sociable they were about the shyest
creatures around the camp. These birds
and others characteristic of the interior
forest did not occur on the opposite slope.
Professor Hine, our zoologist, busied
himself with studying them and brought
back a representative collection.

Digitized by

Google

314 THE VALLEY OF TEN THOUSAND SMOKES

Of less interest to the gen-
eral reader, but more im-
portant to the specialist, are
the very large collections of
insects which Professor Hine
obtained. The insect life of
Alaska is very little known,
and he secured numerous
species new to science as
well as many others not pre-
viously reported within a
thousand miles. These will
be dealt with in the tech-
nical papers to follow.

E^ch year Hine, because
of his special interest in
the outlying undevastated
country, became the presid-
ing genius of the base camp.
In i9i7onthePacifictundra,
and in 1919 on the Lake,
the place was recognized as
"Hine's Camp/' because of
his redoubtable powers as
a woodchopper and his de-
lightful spirit as a com-
panion.

LARGE MAMMALS
RETURNING

Many kinds of big game
animals were common in
this country before the erup-
tion. Caribou and moose
were found in numbers, es-
pecially caribou, for the
moose is here at the western
edge of its raujge and not so
abundant as in some other
regions. After the eruption
these herbivorous animals
were very seriously affected
by the coating of sharp
abrasive dust on the vege-
tation, which quickly wore
out their teeth . Large num-
bers of them are said to
have perished through in-
ability to feed properly.

But since the ash has
settled down, and blows
about less than in the earlier

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY 315

years, they are returning. Our

parties saw signs of caribou and .

2 Oi

moose a number of times. For ^ "§

the latter, the forest country, ^ ^

interspersed with ponds full of •§. ^

lily pads, would seem to offer I

an ideal range. o ^

The fur-bearers — marten, ^ Xo

-otter, mink, fox, and perhaps

beaver — may likewise be ex- «

pected to reappear in force, now |

that the native villages have o

been abandoned, leaving the ^

-country an uninhabited wilder- >^
ness with no one to hunt them
<down. Foxes are common
•everywhere, though usually so

shy as to be seen rarely. One g.a

of them, however, used to bring o J

her litter regularly to our gar-
bage pile at the lake camp.
Wolverines also were frequent

o
a-

travelers along the trails we Iz^jg

used. Occasionally we would 9-2^

find the tracks of one on top of rf 'B >»

those of the bears that fol- o t> =

lowed us. Evidently he was u -^o

coming along to lick the bones « |iS

the bear might leave. Once a •-' cj.S
wolverine passed very close to ^ ^

us, for as we climbed a moun- S^

tain we found his fresh tracks I <it

on the pass at the top, and on 4, g

returning followed his trail "SS

across our own. *o g

Trappers have reported the |

capture of lynxes within the .h

Katmai National Monument •§

recently, but we had no experi- ^

ences with them ourselves. ^

Wolf tracks also were oc- ^

casionally encountered. In our >

first ascent of Katmai we found
the trail of a stray wolf that ^

had passed that way, 2,000 2

feet up the slope. It is re- *-

ported that they used to run in

dangerous packs during the §

winter. Walter Metrokin, our

packer, told us a story of how 1^

a boy, by firing through the |

door, had once driven off a ^

Digitized by

Google

O
H
en

<
o

'Ji

o
<

316

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY

317

vicious pack, which had attacked a bara-
bara in Katmai Village.

BROWN BEARS ABUNDANT

Of all the large animals, the great
brown bears are the most abundant and
most interesting. Although the Kodiak
Island species has the reputation of being
the largest bear in existence, few speci-
mens taken on Kodiak are as large as
some that inhabit the Katmai District.
They are the largest carnivorous animals
extant, far exceeding the lion or tiger in
size. They stand as high as a cow, and
their weight may sometimes exceed 1,500
pounds.

The very fact that they are the largest
of their kind has attracted hunters from
all over the world until their numbers
have been very greatly reduced. Wal-
ter, the most celebrated bear hunter of
Kodiak, has told me that when he first
went hunting in his boyhood, the bears
on the hillsides were '*just as thick as
cows.'* Those were the days of muzzle
loaders and single-shot rifles.

An old man still living in Kodiak at
the time of our visit remembered bear
hunts when the weapons were bows and
arrows. These were supplanted by the
muzzle loader of Walter *s youth, and since
then, as he says, there has been a constant
improvement in guns. First came the
breach loader, then the ** Henry repeat-
er," and then a long succession of better
and better guns which enabled the
hunter to pick off his quarry with greater
and greater ease and safety, until now a
pair of field glasses with which to find
the game is as important an item in the
hunter's equipment as his rifle.

PROTECTION NECESSARY TO PREVENT
EXTERMINATION

Thus provided with a telescope and
modern high-power rifle, the hunter may
sight his game at a great distance and
bring it down without giving it the least
semblance of a chance to defend itself,
thereby pretty nearly reducing the sport
of the old time hunting to simple slaugh-
ter induced by the passion for killing.

No prophet was needed to foretell the
result: the rapid decrease in the numbers
of the bears, which will soon face com-
plete extinction unless energetically pro-
tected. The problem of preserving such
an animal is, however, not so simple.
Congress years ago prohibited killing the
bear on Kodiak Island, but scant atten-
tion was ever paid to the law.

The trouble was that it was made
in the interest of the bear rather than of
the settler. If it were enforced, the
agricultural development of the rich
cattle country would be impossible;
bears and cattle do not mix well. Be-
sides, no adequate means of enforcing
the law was provided, even though there
are stories of men having been arrested
for defending themselves, asserting that
when it came to a choice between being
mauled by a bear or going to jail, the
latter was undoubtedly the lesser evil.

It is not only impracticable but down-
right foolish to try to maintain such an
animal in proximity with struggling
pioneer settlements. But in the Katmai
National Monument, which is bounded
on the east, at any rate, by a large area of
very rugged mountains never likely to be
densely settled, the Alaskan brown bear
should find a safe refuge in which it may be
preserved for future generations without
jeopardizing any interest of civilization.

BEAR TRAILS EVERY\VHERE

Wherever we traveled through the
devastated country, except under the
very shadow of the volcano, we found
the tracks of these great bears very nu-
merous. Everywhere we went we appre-
hensively kept a sharp lookout, for the
tracks were so thick that it seemed as
though there must be a bear crouching
behind every rock. At first we were
much concerned lest we should come up-
on one suddenly, supposing that in such a
barren country they must be ravenously
hungry. When we first landed, one
member of the party would not go for
water fifty yards from the tent without
carrying his rifle.

The reason for the abundance of bear

Digitized by

Google

3i8

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

THE FRONT TRACK OF A BEAR, NINE INCHES ACROSS

The Katmai National Monument will probably become the only preserve for these, the largest of all

carnivorous animals.

tracks is that the fine-grained surface of
the ash is a remarkable medium for the
preservation of any marks made in it.
The tracks were often so clear that we
could see the crisscross creases in their
soles. The slightest details were so faith-
fully preserved that for a time we supposed
all the tracks had been freshly made.

As we became more familiar with con-
ditions and had opportunity to examine
the same tracks repeatedly, we found
that they retained their appearance of
freshness for a long time. In fact, it is a
difficult matter to distinguish a track
freshly made in the ash from one several
days old.

THE BEARS INSPECT OUR TRAILS

Although much less abundant than
the tracks led us at first to suppose, the
bears no doubt watched us from the shel-
ter of some safe cover many times; but
they were very shy and kept out of our

way, later coming down and sniffing along
our trail out of curiosity to know what
manner of beast the intruder might be.
The country is entirely uninhabited and
the younger of them at least may have
been wholly without knowledge of man.
When we thus saw beside our own puny
footprints the trail of an enormous bear,
we could not escape a somewhat creepy
feeling.

Once, from the tracks we found, we
concluded we must have surprised a
mother and cubs that had been advanc-
ing toward us, for the trails suddenly
reversed and turned back, indicating a
hasty retreat. But the infrequency of
cubs was surprising, in view of the abun-
dance of bears and the fact that the
young stay with the mother until two
years old. Far down the Alaskan Pen-
insula, I once happened to have the
opportunity of watching such a family.
The mother had five cubs, two of them

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY

319^

Photograph by R. F. Griggs

THE HIND TRACK MEASURING loi BY 1 6 INCHES

The white marker is a six-inch rule. A man inevitably feels pretty small when he finds that such an
animal has been following his trail, especially if he is unarmed, as we were a good share of the time.

yearlings as big as heifers, the other
three young ones no bigger than dogs.
They were lying beside a stream, playing,
after having filled up on salmon. Several
sea gulls were hovering around, doubtless
attracted by the refuse. Every little
while one of the larger cubs would get
too free with the mother and she would
send it rolling with a cuff from her paw.
Even from a mile away she seemed an
enormous animal.

On one occasion, Folsom and Walter
were quietly cleaning fish beside a small
brook when they were surprised by a
loud snort, like that of a horse, followed
by much cracking of sticks as bruin made
off through the bushes. Another time,
at night, a bear came down on the camp
without suspecting it until he got with-
in twenty yards. Then, getting the
scent, he galloped off with the noise of a
frightened elephant.

Yet we never caught a glimpse of one

the first year. As the days wore on with-
out sight of the animals, we gradually be-
came as indifferent to danger from an en-
counter with them as at first we had been
apprehensive.

GOING INTO THE INTERIOR UNARMED
DESPITE THE BEARS

When we came to start up into the in-
terior for the first time, our packs were so
heavy we could not well add a rifle, and
we finally went out into the wilderness
armed only with a couple of pistols. Long
before we were through with the country,
the man who at first would not venture
away from the camp-fire without a gun
became so disgusted that he refused to
lug his rifle even on long tramps through
country in which bears were certainly
plentiful.

Bears are much more in evidence dur-
ing the early part of the season than
later, for June is their mating time.

Digitized by

Google

320

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs
PAUL HAGELBARGER ON THE TRAIL OF A BROWN BEAR

The soft mud of the tide flats took deep imprints, but it was not possible
to photograph the tracks of those which inspected the fumaroles.

During this period they wander uneasily
over the country, and are especially
curious about any strange animal mov-
ing about. If they hear any noise in the
bush they will come up to find its cause,
instead of taking to their heels as at
other times.

Once, while we were *'mushing'' up the
trail getting our outfit back into the in-
terior, we happened to be divided into
two parties, one of which traveled a few
hours behind the other. On the way up
the Valley the second party found that
within three miles as many different

bears had inspected
the trail left by the
first party in the
short interval since
it had passed. Two
of them were to-
gether, entering and
leaving the trail at
nearly the same
point. One had a
track 5 by 9 inches,
and the other 6 by
II inches. They
came up with the
trail, followed it for
a while, and then
veered off up a val-
ley. The tracks were
so fresh we could
probably have found
their makers shortly,
if we had had time
to follow.

The other bear,
which was consider-
ably larger, with a
track about 7 by 12
inches, had simply
crossed the trail in
his journey down the
mountain side, fol-
lowing it a few rods,
first one way, then
the other, and after
satisfying his curi-
osity had passed on.
We had no desire to
follow this one for
the country was rough and covered with
brush, which would have given the pur-
sued all the advantage, permitting him
to hide or to ambush the hunter at will.
Besides these, there were at that time
at least two other bears in the ten-mile
stretch of valley we were then working.
The track of one of these on hard
ground measured 9 by 14 inches. But
when he stood up on his hind legs to in-
spect his surroundings, or traveled in soft
ground so that the fur on his leg enlarged
the impression made by his sole, his track
reached the enormous dimensions of loj

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY

321

by 16 inches, more than big enough to
cover two pages of this book (see page

319).

What a sight this bear must have been !
Old *' Ivan' * in the New York Zoo is the
center of an admiring throng from morn-
ing until night because of his great size,
but the sole of his foot is only about six
inches wide. I have a skin, the gift of
Marshall Armstrong, of Kodiak, which
seems about as big as any animal could
grow. When stretched up to the ceil-
ing of the room its claws drag on the floor,
but it is by no means a large skin for a
Kodiak bear — only nine feet six inches,
with foot pads seven inches wide. Pe-
trof in the census of 1880 gives a record
of one measured by him that was four-
teen feet four inches long. This may
perhaps be about the size of our big bear.
We grew to be familiar with his tracks,
but never caught sight of him.

FEEDING HABITS OF THE BEARS

From what I had read of the habits of
the grizzly, I had supposed that bears
must necessarily have large ranges which
they patrolled in solitary state, driving
away all intruders. This is not true.
Whatever the habits of the grizzly may
be, the bears of this region have no such
habits. On the contrary, in the old days
they were almost gregarious.

Despite their enormous power and the
agility with which they can get around
in rough country, these bears do not
ordinarily hunt big game. Instead, they
eat grass like cows. Around their rest-
ing places I have found the grass close-
cropped. Roots compose a large portion
of their diet in the spring. We some-
times found places in the ash-covered
country where they had dug down into
the original soil for some root which their
nose had detected, although there was
nothing above ground to indicate its
presence. They are said to dig out mice
and ground squirrelswith great dexterity.

During the latter part of the season
they frequent the salmon streams and the

berry patches, and so are easily found
by hunters who know their habits. The
trader at Cold Bay told me that in 191 8
he saw at one time seven bears fishing in
** Moose" Creek in the southwest corner
of the Katmai reservation.

FIRST SIGHT OF A BEAR

As intimated, we were forced during
the first two years to make all our ob-
servations of the habits of the bears from
their tracks. It was not until we began
to work along the Aindevastated country
on the west side of the river, in 191 7,
that we were fortunate enough to catch
sight of one of the animals. Then one
day soon after landing, as we rounded a
rocky point, Folsom suddenly called,
' * Griggs, your glasses. There *s a bear ! * *

At first I could see nothing, but off in
the distance there was a small gray-
brown spot which, when watched closely,
could be seen to move. With the glasses
it could be made out plainly as a bear.
We could see him paddling around in the
soft marsh, digging out the grass and
eating heartily. I suppose he was get-
ting his supply of salt by eating the
saline plants of the marsh.

It was surprising how well we could
see him with the glasses. He was a
light, pale, washed-out brown, not nearly
so rich and beautiful as some of the dark
brown skins one sees. The mane on his
shoulders was long and shaggy, giving
him a curious sway-backed appearance.
He looked indescribably ungainly and
clumsy as he mushed around in the
marsh, often sinking half way up his
legs in the soft sticky mud. I could not
but compare him with some of Knight's
pictures of restorations of animals of by-
gone ages.^ He seemed no part of the
present world, but rather a relic of the
past somehow dropped down here for me
to study.

Quite naturally, I was anxious to
bring him down, but Andrean advised,
'*Wind no good.*' We could not hope
to get close, as he was already somewhat

' Familiar to readers of H. F. Osborn's "Age of Mammals."

Digitized by

Google

322

THE VALLEY OF TEN THOUSAND SMOKES

Photograph by R. F. Griggs

SKINS OF A KODIAK BEAR AND A PUMA

This 18 by no means a large skin for a Kodiak bear — 9 by 9§ feet. The track was only about seven
inches wide. Petrof records one, probably stretched, over 14 feet long (see note, page 265).

nervous, frequently raising his head to
sniff uneasily, having evidently detected
a taint in the air. Nevertheless, we de-
cided to try it, so we turned back into
cover and sneaked through the grass
and alders along the hill in the hope of
getting close. As our course took us
more and more out of the wind, we hoped
we might get to windward and approach
closer. But it was of no use. He had
had too strong a scent, and before we had
covered more than half of the distance,
he made off, now walking, now bounding
in great awkward jumps.

After all I was not sorry, for he was
only a little bear — that is, little in com-
parison with those tracks we had followed
— with foot-prints only four or five
inches wide. Still, he must have meas-
ured nearly seven feet long. I was just as
well satisfied to wait another chance in
the hope of getting a big one later, for our.
permit allowed us to take out only one.

The program of our work was so
strenuous as never to allow us the luxury
of taking a day off to organize a real
hunt and stalk a bear. Our only chance
was to stumble into an opportunity of
getting one without wasting our time in
preliminaries. Finally the opportunity
came, but not when we were altogether
prepared to accept it.

AN ENCOUNTER WITH A SLEEPING BEAR

Paul Hagelbarger and I were climbing
an unnamed mountain back of our camp
to get a view of the upper valley. The
alpine flowers were wonderful, and we
busied ourselves studying and photo-
graphing them as we went along.

We had rambled in this way up to
about 1,500 feet when, rounding a
shoulder of the mountain I saw a curious
round gray-brown object that looked
quite different from the rocks that lit-
tered the mountain side. They were

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY

323

sharp and angular, but this one was
smooth like a water-worn boulder, such
as one finds in a stream bed. *' Quick,
Hagelbarger, give me the glasses,** I
whispered.

They were buried deep in his pack-
sack, and it took some time to find them.
Meanwhile, the brown object was as still
as the rocks, and I concluded I must be
mistaken, for a bear would be moving
about. So, when the glasses were
finally fished out, I felt it almost foolish
to use them.

But one glance through the binoculars
was enough. There he was, sure enough
— a big brown bear, curled up in the
sunshine, fast asleep. His back was
toward us and his head tucked in, as
though to shield him from the blinding
sun. Fortunately the wind was right.
He was restless, turning every little
while first one way and then the other.
Part of the time he turned his head di-
rectly toward us and once he stood up,
but did not see us — bears have very
FMX)r eyesight — and lay down quietly
to sleep again.

What were we to do about it? This
was no bear hunt, but a botanizing ex-
cursion. Our rifle was at camp 1,500
feet below, and three miles away. We
had only revolvers. Should we tackle
him with these, or should we quietly
back out over the mountain with the
hope that no visions of men would dis-
turb Bruin *s slumbers? We counted
our shells. Each had enough to reload.
I was hesitant; but what were we carry-
ing those revolvers for except for pro-
tection against bears? Here was an
opportunity to see what they were good
for and under much more favorable cir-
cumstances than we could ever expect if
attacked. Besides, as compared with the
bows and spears with which the natives
had formerly hunted bears, pistols were,
indeed, formidable weapons.

GOING AFTER BRUIN WITH A PISTOL

Accordingly we worked back out of
sight and carefully picked our way
around over the shoulder of the mountain

till close upon him. The only cover as
we came up over the ridge again was a
thin square slab of stone that stuck up,
exactly like a tombstone. I remember
wondering, as I crawled up to it, if it
might be my own tombstone before I
was through. Even such bad augury
did not greatly disturb me, however, for
as I got into position and aimed, my
hand was as steady as though only shoot-
ing at a mark. I was so concerned for
fear he would escape that I could not
wait for Hagelbarger, but opened fire
before he came up.

My position could not have be^n im-
proved. The bear was still sleeping
quietly only twenty yards away, his
great shaggy head as big as a washtub
and turned directly toward me. If I
had remembered all I had been told
about the thickness of a bear's skull,
the result might have been different, but
that head was too fair a mark to be re-
sisted. I took careful aim and fired.

It hit him square; the shock knocked
him clear off the rock, and started him
tumbling down the mountain side.
I was on the point of going down to
get him. The second and third shots
glanced and went whistling off across
the valley. Their noise evidently per-
suaded him that the trouble came from
that quarter. Instead of rolling farther
down hill, he suddenly gathered himself
together and started back straight for us.
Things began to look bad, so I hurriedly
gave him another. This turned him off
along the side of the mountain below us.

By this time I was wondering what
on earth my gun was good for, anyhow,
and was more anxious to reload than to
pursue. I began to fear that the affair
might change from an offensive to a
defensive campaign. Judging from the
effect of the first five shots, we should
have been as helpless against him, if he
had charged, as though armed with
pocket knives.

I had heard many tales of the ability
of the Kodiak bears to walk off with
lead, but I had been very cocky about
my heavy service Colt, confident that

Digitized by

Google

3^4

THE VALLEY OF TEN THOUSAND SMOKES

Photograph Dy E. C Kolb
RAINBOW TROUT, NAKNEK RIV^R

It is only the little ones that get photographed. They averaged two feet. The largest was thirty-two
inches. Seldom were more than one or two casts necessary to secure a strike that kept one busy.

a shot properly placed would kill any-
thing. But now I was inclined to throw
it into the river.

No bears were killed on any of the
expeditions until 1919, when two were
secured, both the results of accidental
encounters. One morning Hine dis-
covered one fishing on the beach directly
in front of the base camp on Naknek
Lake. He fell easy prey to Wallace's
rifle. The other one was found by Cap-
tain Nyman of the Nimrod, swimming
about among the islands of Geographic
Harbor. It was a simple thing to over-
haul him with the power boat and shoot
him at close range. Either of these
bears would have been a fine subject
for motion pictures, but unfortunately
both times the camera was a hundred
miles away.

THE PROBLEM OF FINDING A HARBOR

Along with the biological and geolog-
ical work projected in 1919 was a prob-
lem of another sort — search for a better
means of access to the district than was
open to us.

The spectacular character of the vol-

canic activity naturally aroused wide-
spread public interest in the region,
coupled with a desire to visit it. But
it was clear that no great number of
people would ever brave a landing
through the surf and go up into the coun-
try by the primitive means that had to
be adopted by our expeditions. The
first requisite to opening up the new park
was the finding of a safe harbor, suitable
for the use of large vessels.

It was doubtful at the outset whether
a port such as was needed could be found.
Between Cape Douglas and Cold Bay,
harbors of any sort are singularly scarce,
considering the irregularity of the shore
line. Although the much indented coast
somewhat resembles that of southeastern
Alaska, where there is a superfluity of
harbors, protected bays with sufficient
depth are here few and far between.

There are a few bays to the eastward
which from the chart appear sheltered,
but the mountains back of them are so
very rugged as to give but poor prospect
for a road. Westward, the mountains
are separated by broad open valleys
very favorable for road building, but the

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY

325

OUR BOAT, THE NIMROD,

Photograph by L. G. Folsom
EXPLORING THE COAST IN SEARCH OF A HARBOR

bays are both shallow and exposed.
From the scenic point of view, it is im-
portant, moreover, that the projected
road should run up Katmai Valley, since
from any other direction Mount Katmai
loses the impressiveness that makes the
first sight of it from the lower Valley so
memorable.

THE POSSIBILITIES OF KUKAK AND
KINAK BAYS

The testimony of local pilots and a
study of the chart seemed to indicate
that our best prospect was Kukak Bay,
which the Russian surveys showed to be
deep as well as spacious. But it is too
big to afford altogether satisfactory pro-
tection and is exposed to bad * ' williwaws* *
in easterly gales. From its head, how-
ever, a broad open valley leads straight
through the mountains in the direction
of Mount Katmai, running practically
at sea-level across to the head of Kinak
Bay. This easy passage at once ruled
out all of the intervening bays between
Kukak and Kinak.

But, when we examined Kinak Bay,

we thought it a better entrance than
Kukak. Its head bears little resem-
blance to the chart. It is constricted
between two great cliffs into an exceed-
ingly narrow passage which opens into
an exceptionally beautiful basin, which
we named Hidden Harbor. Its water is
deep and clear of dangers close up to
the mountainous shore. It is, moreover,
nestled in among the mountains, secure
from all manner of tempests, and there is
no chance for the dreaded ''williwaws"
to reach it.

But this same mountain wall offers
the greatest objection to its use as a
port of entry, for it stands a thousand
feet high all around, and would make
road building very expensive. The same
obstacle would have to be overcome by
a road from Kukak Bay, for after trav-
ersing the level valley connecting the
two it would take the same route.

DISCOVERY OF AN IDEAL HARBOR

We did not suppose that the next bay
west, Amalik, would prove as good an
entrance -as Kinak. According to the

Digitized by

Google

a i

•M

;i<

•D

B
<0

• M

•*-•

•M

t/3

c«

■*•*

•4-»

.23

■♦-»

«*M

326

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY

327

chart, it is not nearly so deep an inden-
tation in the coast line. But upon exami-
nation we found, to our astonishment,
opening back of the outer barrier (which
is indicated as the head of the bay by the
chart), a broad deep basin forming an
ideal harbor.

When we entered the narrow opening,
we found ourselves in a long straight
canal leading back among picturesque
islands far inland behind the *' coast."
Finally, the canal opened into a broad
inner bay, itself bisected by a chain of
islands. Passing through the opening
between these, we found the upper por-
tion of the bay a perfect harbor, with
deep water throughout, except where a
strea^n had washed a delta into its head.

So far as we could detect, with the
means at our disposal, the waters are
free from hidden dangers. Captain
Nyman of the Nimrod sounded through
the channel with nine fathoms (54 feet)
of line, but found no bottom at any
point. The new harbor is, therefore,
suitable for ships of the largest size.

Its channel is straight clear sailing,
and once inside there is abundant room
for maneuvering. It is, moreover, easy
to get into, for the mouth of the outer
bay is a wide opening between con-
spicuous landmarks, readily entered in
the roughest weather. A ship coming
down the coast, picking up a light placed
on the island off the Cape (Ilktugitak)
and making a single turn would steer
right in — reaching quiet water, pro-
tected from both wind and sea, a mile
or two before entering the channel.

In honor of the Society responsible
for the exploration of the Katmai dis-
trict and for its being set aside as a
national monument, we named the new
harbor Geographic Harbor.

GEOGRAPHIC HARBOR A FITTING EN-
TRANCE TO THE VALLEY OF
TEN THOUSAND SMOKES

Not only is Geographic Harbor a secure
and commodious haven, it is a fitting
entrance to the Katmai National Monu-
ment. It is doubtful whether any of the

superlatively beautiful bays along the
Alaskan coast surpasses this one in the
grandeur of its setting. On both sides
the mountain walls tower aloft 3,000
to 3,500 feet directly out of the water, the
irregular weathering of their stratified
cliffs producing a very striking effect.
The many picturesque islands of the bay
give its shores an appearance of alluring
mystery. One has the feeling that he
could spend many days in a canoe, poking
into its recesses without seeing all that
was worth while.

The islands are the protruding hum-
mocks of a great lava flow that came
down the bay from some Tertiary vent.
Their out-shore faces are weathered into
the most striking columnar structure,
(see page 302). That the eruption re-
sponsible for these great lava masses far
antedated the recent period is well shown
by the inshore faces of the islands. They
have been all planed off into smooth,
round, gentle slopes by a glacier which
came down the Valley after the lava had
been poured out.

Back from the head of the bay, above
a sightly waterfall, is a considerable lake
whose water may be presumed to be full
of fine trout.

A FEASIBLE ROUTE FOR A ROAD
INTO THE INTERIOR

On the land side, conditions are as
favorable for the use of Geographic
Harbor as from the sea. There are two
practicable routes by which a road could
be carried over into the Valley of Soluka
Creek, the principal eastern tributary
of Katmai River. Only 50 or 60 miles
of road would be required, not only to
penetrate the Valley of Ten Thousand
Smokes, but to connect with the head of
Naknek Lake and Bering Sea as well.

Some novel engineering problems
might, to be sure, arise from the desir-
ability of keeping such a road as clear of
drifting ash and snow as possible. But
these could undoubtedly be overcome
with a little study.

It may be stated in passing that in
addition to the desirability of making the

Digitized by

Google

328 THE VALLEY OF TEN THOUSAND SMOKES

I volcanic features accessible, there is

g another important reason for the con-

^ struction of a road across the Penin-

^ sula. The canneries tributary to Bris-

a; tol Bay give the district a summer

^ population of 10,000 men. To reach

a these people, it is necessary at present

I to make a voyage of a thousand miles

I around the whole length of the Alaska
Peninsula.

This trip is so time-consuming, ex-
pensive, and dangerous that the whole

•g community is practically isolated, ex-

>» cept for the radio. In the old days,

^ when commercial operations were con-

c/) « ducted on a smaller scale than at pres-

^ § ent, it used to be considered better,

J ^ as stated by Petrof in the quotation

^ -5 given on page 265, to carry supplies

§ 2 across Katmai Pass on man-back than

^ o to attempt the voyage.

pS 2 When in 1918 and 1919 the Kat-

S g mai Expeditions desired to enter the

(/) 5 found that it was practically impos-
cTj sible to get in and out except by the

^ country from the Bering Sea side, we

H -2 kindness of the packers, who per-
^ "1 mitted us to travel to and fro on their
g .^ sailing vessels. As this required an
« ^ absence from April first to October
< J first, in order to work through June
^ g and July, it will readily be seen how
S % inadequate are the transportation fa-
% g cilities.

o % A road thrown across the base of

§ ^ the Alaska Peninsula, through the Val-
^ .S ley of Ten Thousand Smokes,^ would
"2 afford the people of Bristol Bay dis-
I trict access to the frequent steamers
^ that serve the Pacific slope of south-
western Alaska, providing adequate
mail service and that ready communi-
cation with the outside world which is
essential to the development of any
country. It would save the packers
several million dollars annually, by
permitting them to take in their crews
in June and get them out in August,
instead of holding them for six months,
largely in idleness.

* Other routes have been suggested, but no
other starts from a good harbor on the Pacific.

Digitized by

Google

OBSERVATIONS IN THE SURROUNDING COUNTRY

329

From a military standpoint, such a
road would have great strategic im-
portance. If it should ever be necessary
to dispatch a military force into Bering
Sea, such a road would be indispensable,
since without that route an enemy could
land its forces at any point of Bering
Sea as readily as we could ourselves.

SIXTY MILES OF AUTOMOBILE ROAD
WOULD OPEN UP THE WHOLE DISTRICT

For the tourist, the construction of an
automobile road, fifty or sixty miles long,
would permit the whole district to be
traversed in a single day.

Many were the times, as we plodded
up the hills with loads which, though
taxing our strength, yet contained only
a pitifully small store of provisions,
that our thoughts turned to the future
and we could see in our mind's eye
swift motor trucks rolling through the
Valley, transporting in an hour more
than we could carry in a month. How
often, as we were held miserable in camp
by a trifling storm, we thought of the
day when closed cars should run over
Katmai Pass without regard to weather,
carrying their passengers in ease and
cpmfort through weather conditions prac-
tically impossible for us.

To some these visions may seem chi-
merical. They are not. The Katmai dis-
trict is nearer to civilization and more
accessible in every way than was the Yel-
lowstone at the time of its discovery. Let
any who are skeptical of the future of the

Katmai district consider the Yellowstone.
Its vast area is tramped under foot by
innumerable campers, until there is com-
plaint that the natural wonders are in
danger of being spoiled. Almost one in
every hundred of the entire population of
the country visits the place every year!

Yet, to reach the Yellowstone in those
days, before the development of the
automobile, the expense of building a
railroad was necessary before it could
become accessible. Even after the com-
pletion of the transcontinental line, the
nearest boundary of the Park lay sixty
miles from the railroad. Contrast con-
ditions here, where ocean steamers can
land their passengers within fifteen miles
of the crater of Katmai.

It was no longer ago than 1902- that
Brooks undertook the first epoch-making
journey from Cook Inlet through the
interior to the Yukon. It was one of the
most courageous attempts to cross the
absolutely unknown that is recorded.

At the start the residents of the dis-
trict agreed that it was a foolhardy proj-
ect, impossible of success; and at the
end the natives would not believe that
the party had actually come overland
from the Pacific, as they said. Yet
within twenty years we see a railroad
and a string of settlers' cabins spanning
the trackless wilderness which Brooks
penetrated for the first time. Who can
doubt that before long the Katmai Na-
tional Monument will be made readily
accessible to all who desire to see it?

Digitized by

Google

o
US

330

Digitized by

Google

PRESIDENTIAL PROCLAMATION ESTABLISHING THE
KATMAI NATIONAL MONUMENT

Whereas, There exists upon the
southern coast of Alaska a belt of un-
usual volcanic activity which has during
the last several years exhibited at vari-
ous points energy of violence which at-
tracts the special attention of scientific
watchers,

And Whereas, Mount Katmai, one of
the volcanoes in this belt, has proved up-
on investigation to have unusual size and
character, and to be of importance in the
study of volcanism, inasmuch as its
eruption of June, 1912, was one of ex-
cessive violence, ranking in the first order
of volcanic explosive eruptions and emit-
ting several cubic miles of material dur-
ing its first three days of activity,

And Whereas, The results of this
eruption are still fresh, offering excel-
lent opportunities for studying the causes
of the catastrophe and its results and
affording a conspicuous object lesson in
volcanism to visitors interested in the
operation of the great forces which have
made and still are making America,

And Whereas, The volcanic neigh-
borhood is shown by the explorations of
the National Geographic Society to con-
tain many other striking features of an
active volcanic belt produced so recently
that they are still in the formative stage;
and in particular The Valley of the Ten
Thousand Smokes, a valley of hot
springs in a condition of development to-
ward a possible future geyser field, in
distinction from the present dying gey-
ser field of the Yellowstone,

And Whereas, This wonderland may
become of popular scenic, as well as scien-
tific, interest for generations to come, in-

asmuch as all its phenomena exist upon
a scale of great magnitude, arousing
emotions of wonder at the inspiring spec-
tacles, thus affording inspiration to pa-
triotism and to the study of nature.

Now, Therefore, I, Woodrow Wil-
son, President of the United States of
America, by virtue of the power and
authority in me vested by section two of
the Act of Congress entitled *' An Act for
the preservation of American Antiqui-
ties,** approved June 8, 1906 (34 Stat.,
225), do proclaim that there are hereby
reserved from all forms of appropriation
under the public-land laws, and set apart
as the Katmai National Monument, cer-
tain lands particularly described as fol-
lows, to wit, beginning at the United
States Coast and Geodetic Survey trian-
gulation station, latitude 57° 52' 17.040",
longitude 155° 05' 20.331", established
in 1908 about one-half west of Katmai
Bay on top of a hundred-foot bluff on the
Alaska Peninsula, named Cape Kubuga-
kli ; thence north 40° 00' west to the in-
tersection with longitude 155° 40'; thence
due north to the intersection with lati-
tude 58° 35'; thence due east to the in-
tersection with a line bearing north 60**
00' west from Cape Gull; thence south
ioUowing said line to the shore line at
Cape Gull; thence west following the
shore line of the coast to a point directly
below the triangulation station, situated
on the bluff at Cape Kubugakli; thence
up the bluff to the said station, the point
of beginning; embracing approximately
1 ,700 square miles of land, as shown upon
the diagram hereto attached and made a
part of this proclamation.

331

Digitized by

Google

332

THE VALLEY OF TEN THOUSAND SMOKES

Warning is hereby given to all unau-
thorized persons not to appropriate or
injure any natural feature of this monu-
ment or to occupy, exploit, settle, or
locate upon any of the lands reserved
by this proclamation.

The Director of the National Park
Service, under the direction of the Sec-
retary of the Interior, shall have the su-
pervision, management, and control of
this monument, as provided in the Act of
Congress entitled "An Act to establish
a National Park Service, and for other
purposes," approved August 25, 1916
(39 Stat., 535).

In Witness Whereof I have hereunto

set my hand and caused the seal of the
United States to be affixed.

Done in the District of Co-
lumbia this twenty-fourth
day of September in the
year of our Lord one thou-
sand nine hundred and
[seal.] eighteen, and of the Inde-
pendence of the United
States of America the one
hundred and forty- third.
WooDROW Wilson.
By the President:

Robert Lansing,
Secretary of State.
[No. 1487.]

Digitized by

Google

ACKNOWLEDGMENT

Ready as the National Geographic So-
ciety always is to reach out with new
fields of interest to its members, it is not
surprising that with the first news of the
eruption a representative, George C.
Martin, was dispatched to the devastated
district. The report of this first Katmai
Expedition has furnished the basis of the
earlier chapters.

There the matter might have stopped
had not Frederick V. Coville, Chairman
of the Research Committee of The Soci-
ety, recognized the importance of study-
ing the revegetation of the ash deposits.
Because of my familiarity with the region
about Kodiak as a member of the United
States Bureau of Soils Kelp Expedition
of 1913, I was selected to carry on the
investigation. To Dr. Coville the writer
owes a large debt, not only for the origi-
nal stimulus to undertake the work, but
for unceasing help and cdunsel through-
out.

The work early enlisted the enthu-
siastic and constant support of Gilbert
Grbsvenor, President of The Society.

In the field the expeditions were gen-
erously assisted by friends of the Society.
The people of Kodiak considered it a
privilege to help. To M. D. Snodgrass,
then in charge of the Experiment Station,
and to W. J. Erskine we were especially

indebted . On two occasions Mr. Erskine,
placing at the disposal of the expedi-
tions his large acquaintance with dealers
in San Francisco, purchased the whole
season's outfit for us. The expeditions
of 191 8 and 19 19 were similarly indebted
to Frank B. Petersen, President of the
Naknek Packing Company, arid his asso-
ciates. Superintendent P. Larsen, Cap-
tain George Wester, and Fred Sparks,
whose generosity made it possible to
send in parties and supplies through the
Bering Sea.

In the preparation of the topographic
map accompanying the book. The Soci-
ety has had the cordial co5peration and
assistance of the Alaska Division of the
United States Geological Survey, under
the direction of Colonel Alfred H. Brooks.
The map has had many vicissitudes, ow-
ing to interruptions incident to war and
to the oil boom which at various times
drew away the men who were working
on it.

Finally, the writer owes a great deal
to the searching and sympathetic criti-
cism of his friends, John Oliver La Gorce,
W. J. Showalter, and other members of
the Staff of the National Geographic
Magazine, to C. S. Scofield and F. L.
Ransome, of the Research Committee,
and to William R. Maxon.

333

Digitized by

Google

Digitized by

Google

INDEX

Page numbers in ordinary type (e. g. 24) refer to subjects merely mentioned or to brief comments. Figures in bold face
type (e. g. 24) refer to important comments or discussions at length. References in italic figures (e. g. 24) denote illus-
trations. Page references followed by n (e. g. 24*1) indicate footnotes. References in parentheses after Roman numerals
(e. g. PI. IX (224)] locate plates.

Abbot, C. G. 3S». 37. 39

Abrasion of trees by wind-blown ash
149

Accessibility of Katmai 327

Acid in air 11, 2 in, 25. 27. 33; in soil
209

Admiral Sampson (steamer) 25

Afognak, Alaska 7

Agriculture, bearing of volcanic ash
on 3

Airplane, po^bility of using in Kat-
mai region 75

Air wave from volcanic eruptions 35

Alaska Peninsula 15, 58. 67. 208; base
of. map. front cover; need for road
across 328

Alaskan volcanoes 57-63

Alder {Alnus sinuata) exterminated
near Katmai 83

Aleutian, volcanic chain 57-6i ; length
of 6x ; Islands 58. 6z ; volcanoes 65,
volcanoes and hot springs, map fac-
ing 60

Aleuts 265

Algae around fumaroles 209, 223, 251

Algeria, haze in from Katmai eruption
36

Alinchak Bay 17

Allen. E, T. 77, 80, 220, 229, 230, 231,
246

Alumina 46

Alum 247. PI. XI (240)

Alum root 157

Amalik Bay 69, 32s. 33o: field of
floating pumice In 30

" American Pete." story of 17

Ammonia 246. 248

Andaman Islands (Bay of Bengal) 35

Andesite loi, 181, 187. 240, 241, 287;
changes in by gases 243, 289

Andrean [Yagashofl 74. i99. 321

Animals trapped in Valley 281; her-
bivorous, effect of ash on 3x4

Animal life, in Katmai region 81, 95;
return of to Katmai Valley x6z-x6s;
in Valley of Ten Thousand Smokes
208; before eruption 265* 277* 3X4*
increase in Katmai region 315

Arctowski, Henryk 35 »

Argon 246

Arkose 271

Armstrong, (U. S. Deputy) Marshal
12, 321

Arsenic compounds, around fumaroles
247

Artemisia tiUsii 155

Asama-yama (Japanese volcano) 33

Ash 3. 7-14* 21, 23. 34, 148; views of
Kodiak immediately after fall of 4,
5, 6, 16. 20; during fall 9; depth of
8, 12, 19. 26, 55. Z69; greatest 243;
area covered by 25-27; plants grow-
ing through and in 40, 42, 47, 49, 31,
32, 34, 160; composition of 41, 158;
roots survive fall of 45* 50, 51 > I55;
on trees 45; contribution to soil 46;
analysis 46; drifts and dunes of 46,
39; effect of in stimulating plant
growth 47-49; experiments showing
growth of plants in, as compared
with sand 30; wind-blown 5i-55» 39',

effect of on tree growth 33, 83, 133;
cracks in 22, 53; moved by water

55. 81; from Bogoslof 65; slides 87*
147, 82, 86, 126; climbing slopes of
88; in streams 19. 89, 150, 151, 152,
153; caverns in 89; bridges 89, 91;
cool when it fell 95; retarded melt-
ing of snow 148; colors of 171, 285;
thrown from Katmai 285; specific
gravity of 293 ; on plants wore out
teeth of herbivorous animals 314

Ash-covered, roofs 16, 24; trees 20, 28,
38,48

Ash deposits, affected by earthworms
55; support moss 38; cross sections
of 93, 146, 149, 154, 255; erosion in
161, 298; bad lands in 298

Ashfall I, 12, 13; map showing depth
facing i; in Katmai Valley 153-155;
affected by winds iss; volume of 29;
from Tomboro34;from Iliamnaand
Redoubt 61. 62; at Kodiak in 1867,
62

Asia, volcanic festoons of, map facing

56, 57-61

Asiatic volcanoes. Katmai's relations

to 57
Astronomical observations affected by

eruptions 37
Atwood. W. W. 167
Augustine Volcano 60. 62
Austin, O. P. 341
Automobile rosid feasible in Katmai

region 329
Avalanche Mountain (Katmai region)

99
Avalanche of ash. remains of. Kodiak

14
Avalanches 147. 241. 242; of ashes 9;

wind caused small 125; from Falling

Mountain 241. 277

Bad lands in ash deposits 254. 298
Baked Mountain 233, 238. 241, 243,

253; camp 183. 224. 226, 264, wreck

of 227-231, 227, 230
Bandai San (Volcano. Japan) 145;

landslide 14S
Banks of overloaded streams 131
Baptist, Mission, Wood Island 13;

orphanage, Kodiak 45, ash-covered

garden at 48, 49
Barabaras (native houses) 17, pronun-
ciation of nn. 24, 103; interior of 29
Barrier Range (Katmai region) 90, 275
Basalt 46. 261 ; cliffs 302
Basinger. A. J. 77, 79. 80
Batholith. theory of to account for

Valley activity 301, 303
Beach. Rex 273

Beaches of old Katmai Lake 120, 125
Bear trail that sprouted 162; starts a

fumarole 224; tracks 318, 3^9, 320;

hunt with a pistol 321; skin 322
Bears 84, 85. 95. 164. 208, 223. 265.

277. 315. 317-324; feeding habiu of

321
Becharof Lake 15. 183. 265
Bell, Alexander Graham 341
Bell. Charles J. 341
Bering Sea 58, 61, 67. 69. 266, 307.

335

327. 329; expedition of 1918 from

77, 79. 328
Berries, salmon, vines appear through

ash 31, ISS
Betula kenaika 83
Bhira Gunga (gorge, India) 144
Bidarka (skin boat) 17. 268
Birch {Beltda kenaika) almost exter-
minated in Katmai region 83, 305
Birds, in Katmai region 93. 95. 163.

165, 313; injured by falling ash 11
Bitter cress 155
Blue-top grass 155. 160
Boat, canvas 97
Bogoslof (volcanic Island, Alaska) 61,

63-65, 64
Bonfire kindled by water 282
Borneo (Island. East Indies) 58
Bothwell (U. S. Coast Guard Cutter) 65
Boudry, C. L. 17. 2in
Bowlder trains (flows) 137, X42t effects

of 142, 143
Bowlders, moved by streams 96, xo8,

109, I/O, X 13; as big as houses 135;

in Bandai San landslide Z45
Bridger. Jim (comparison) 193
Bridges of ash 89, 91: over steaming

fissures 235
Bristol Bay 266. 310. 328
Brock, R. W. I39». I43»
Broken Mountain 233. 238. 240. 241.

243* 253
Brooks. Alfred H. 307, 329
Brooks' Falls 265. 310. 311* 3 10, j//
Brooks, Lake 307, 308, 309
Brown (Cap*.) 12
Buried, wood X27, 139; snowdrifu

X48-X49
Burnet (plant) isS
Buttress Mountain 237. 239, 255

Calamagrostis scabra 155, 160

California, haze in from Katmai erup-
tion 36

Camps 74; system of relay 7S» 791
views of 206, 226, 227, 230, 264, 312;
near Observation Mountain 197;
near Baked Mountain 183, 224. 226,
227-23 X, 227, 230, 264

Cameras 96. 207

Camosun (steamer) 27

Canning industry near Katmai region
3x0, 328

Canyon, of the Katmai River 69, 95,
97, 114, ij6, 118, X2X, 122. X24, X27,
'JO, r32, 273; and second valley of
the Katmai X23-X33; of Mageik
Creek X24

Carbon dioxide. 246; question of lib-
eration by. of volcanoes of extreme
importance to mxnkind 43» 44

Carbon in}noxide 246

Cardamint umbeUata 155

Carex sp. I55. i57. 160

Caribou, absence of in devastated re-
gion before 191 9. 164

Carnegie Institution. Geophjrsical
Laboratory of 78. 79. 80. 205

Carter. Chariie 267

Castle Rock (Bogosk>f Island, Alaska)
63. 65

Digitized by

Google

336

INDEX

Cattle, herd of Galloway. Kodiak 44
Caverns in ash 89, 91, 93 ; in Vall^ of

Ten Thousand Smokes aoj
Cer ostium sp. IS7
Cerberus, Mount PI. V (192) lox,

I93M. 233. 354, ^^'» 374, 296
Charcoal forest 260
Charred log. in sandflow 3S5i trees

357
Chemical work 309
Chemistry of f umaroles 346-349
Chester. C. M. 341
Chigmit Mountains (Alaska) S9
Chicken (town) 33. 36
Chickweed X57
China Sea 58
Church, D. B. 73, 7 J. 74. 76, 189. 190,

193. 316. 317
Church, Greek at Kodiak q; Russian

Z03, 267; at Katmai village 104
Clark, Lake 31, 33, 6x, 308
Clarke. F. W. 4111. 43
Cliffs of columnar basalt 302
Climate in Katmai region 90, xs8-xS9t

ao8, 309; glacial, might be caused

by volcanoes 39*4 x
Climatic effects of the eruption 33-44
Climbing ash slides 87
Clothing 307
Clouds, volcanic xs, si, 379; edge of 8;

about Mt. Katmai 167. X73. I77;

steam 168; about Mt. Martin z8i
Coan. Titus 38x11
Cold Bay (town) X7. 3X. 379. 334
Color photography 351
Colors, of Katmai cliffs 133, xa4; of

mud Z67; of ash z 71, 385; of Katmai

crater walls 174; about f umaroles

X90, 192. 304. PI. VI (193). VII

(308). VIII (309), IX (324). X (22s).

XI (240), XII (241). XIII (288).

XIV (289) ; of cliffs on Baked Moun-

tain 244; due to iron compounds 247 ;

in the Valley 350-351
Commander Islands S7
Condition of country before eruption

365-375
Conglomerate 306
Conical mounds, probably result of

landslides X37-X38, 128, 134* X39»

Z43; in Bandai San landslide X4s;

on melting snow field 156, 157
(>>nway. Sir William M. i43fi
C^ook, Captain (early explorer) 63. 63
Cook Inlet ai. 6x, 62, 369. 307. 339
Cooking 8z ; over f umaroles 196, 200,

PI. VI (193). VIII (209), 305, 334-

336, 282

Copper Center (town) 23
Cordova, Alaska 23, 25
Cornice structure in block of lava 296
Corundum in incrustations 247
Cotopaxi (volcano, Mexico) 43
Country around Katmai. geographic

observations in 305*337
Coville. Frederick V. 307, 313, 34X
Coville. Lake 307. 308, 313
Crater, of Katmai z67-z79» 387, 388;
nature of rim of z68, 171, J 73: view
into 169, 172, 174, '76' '78: lake In
169; size of 170, 173. Z76-X79; com-
pared to that of Kilauea 179; of Mt.
Martin z8x ; descent into 185
Craters, in Valley floor 236, 337* 361,

202
Crater Lake (Oregon) 388
Crust (earth's) deformation of 59
CulebraCut (comparison) X17

Daly, R. A. 297"

Dam formed by landslide in Katmai

Cansron, 1x5-1x7, 116, X3S; in Mag-

elk Valley 117: in Himalayas 145
Dangers of the Valley 30X, 303-304
David's Falls 133
Darwin, Charles 55
Dawson (city) 33. 36
Davidson, George 61 m, 63
Dead forest 87. 95
Deformation of strata 61
Deltas, process of formation 153
Dennison. Napier 33. 27
Depth of waters near Aleutians 61
Desert, as result of eruption 45
Desolation of Katmai region, 83, 94^

95. 137
Devil dub growing through ash 40
Deville. C, 24611
Digestion of old ix>ck in magma 340,

34 X, 389, 397H. 399*«
Dikes built by shifting streams 153
Disturbances, preliminary 15
Dora (steamer) 10, J3, 15. 3i, 369
Dormant, plants, under ash 5X» 53;

seeds iS7
Douglas. Cape 334
Douglas. Mount 307
Drainage in Valley 345
Drinking water 88, 89; pool 222
"Dry fog" 33
Dunes, ash 39
Dust, volcanic xx, 37, 33! fell at great

distances 27; cloud 33; from Tom-

boro 34; sun's heat absorbed by 37;

in air 8 1 ; storm in Katmai region 84 ;

clouds from landslides X45

Eagle (village) 26

Earthquakes 12, 17, 19, 21. 1x5. X40,
269, 273. 277. 278, 285: distaint
records of 33

Earthworms, in ash deposits 55

East Indies 58

Eastern Sea (Tunghai), Asia 58

Edgecombe. Mount (volcano) Alaska
S6. 36, 57H

Edson. John Joy 341

Effects of eruption, climatic 33-44; at
Kodiak 45

Ejecta, quantity of 39-3X, X44. 189,
275N. 28s. 2951; from Tomboro 34;
character of from Katmai 101, 285

Elderberry IS7

Elm Landslide (Switzerland) X43

English Harbor (Port Graham) Alaska
63

Equisetum aravense 155

Erosion, of ash 55» 171; of volcanic
cone 65; of mountains 129, 131; be-
ginning of on Mageik landslide X38;
by wind 148, 149; in ash 16 x, 298;
in sandflow 234, 238, 260

Eruption, and its significance x-5; as
experienced at Kodiak 7-X3; phe-
nomena accompanying 7-X3» I5-3Z»
33-39. 65. xis. 139, X44. 242; rec-
ords of from other localities X5-31 ;
sky phenomena following 37; condi-
tion of Valley before 19; effects of
on air pressure 35; on forests 45. 53;
on climate 33-44! on animal life
16X-X65. 3x4; at Kodiak 45; sand-
flow a new type of 361 ; condition of
country before 265-375; hjrpothcti-
cal chronicle of 377-385; summary
of events of 385

Eruptions. 3, X3, 25. 33. 34. 6X-65, I45

Expeditions to Katmai region 69; of
1912, isn, 71; of 191S. 7x, 73, 97; of
1916. 7X, 73. 97, 147. I5S. IS7, 163.
167; of 1917. 72, 73. 77. 97, 147. 174.
175. x8i, X97-3X7, 373, view of

members of 76; of X918, 17* 77; of
1919. 77, 79. 147. ISO. 181. view of
members of 77: Frank Leslie's 374;
effect of World War on 79; pxep-
arations for 74; headquarters of
Kodiak (5

Experiment Station, Kodiak, herd of
cattle belonging to 44

Explosions (volcanic) 3. 7. 15. 19. 3X,
33. 34, 35. 41. 381; heard at distant
points 33; of Augustine 63; of Bog-
oslof 65; of Novarupta 384; of Kat-
mai delaired until crater was un-
roofed 297. 399

Explosive character of Katmai 391

Ewing. Dr. U. W. 37

Eyewitnesses of eruption X5-3X

Fairbanks (city) 23. 26
Fahxhild. David 341
Falling Mountain (in Valley of Ten
Thousand Smokes) PL IV (97); 128,

X921I, 202, 3x9, 220, 333, 34X-243»

280; avalanche 377* 385

Fault scarp. Broken Mountain 244

Faults in Katmai region 132, 243, 273,
388, 303

Fenner, C. N. 77, 80, 183, X84, i8s,
329, 230, 231, 240, 293N, 39711, 39911,
30X

Ferric oxide 46n

Fertility of volcanic soils varies 46; of
ash tested 47

Fertilizer, asli from Katmai did not
act as 46; used in testing ash 47; ni-
trate 47; effect of on plants grown
in ash and sand 30

Fidde Creek 150, 151, 131

"Finger Lakes," New York (compari-
son) 308

Fire, evidence of in Katmai region 95,
357; set by lava in Valley 38 x

Fireweeds in ash-covered region 45, 46

Firewood 81

Fish in Katmai region 13Q, x6x, x63,
308, 309. 3x0, 310, 3XX, jii. 3x3, J'^

Fishing in Katmai lakes 309

Fisher. Franklin J. 34 x

Fissure Lake 212, 344* 'S8

Fissures 17. 202, 216, 335. 344. 254*
255 ; steaming 278; fundamental 303

Flies X63, i64»

Fkxxi, a freak of the 109, 1x3; the
Katmai xo3-X3x, 131; destruction
by 102, 103, 106, 107* no; the Goh-
na 119; the Mageik 131; the Johns-
town 119

Fkxxi fall of Katmai River ///, J12,
114, 155

Flowers in Katmai region 83, 322

Fluorine, surprising abundance of in
gases 347

Fogs near Alaska Peninsula 67

Folding (of strata) 27 x

Folsom. L. G. 7x. 72, 7J. 74. 76, 77, 80,
X13, 127,170. 189, 190.211,319,321

Food of expeditions 75-77. 335-336

Fording, streams 87; Katmai River
96; Soluka Creek 89; with help of
rope 97

Forests, effect of eruption on. 45, 53;
near Katmai 83, 305; dead 87, 95;
effect of Katmai flood on 107. of hot
sandflow 356, 357; in Katmai region
267, 312 PI. XVI (30s); in Valley
before eruption 277; Hudsonian 305;
border migrating westward 305

Formosa (Island, Japan) 58

Foulk, C. W. 4611, 47>»

Fowle, F. E. 3S»«. 37

Foxes 164, 3x5

Digitized by

Google

INDEX

337

Frank (Alberta), landiUde X39. X43

Ftanklin. Benjamin, on "dry fog" 33

Fujiyama (volcano) Japan 57, 63, loi

Fulmer, Elton 4611

Fulton, B. B. 7/. 73. 96

Fumaroles x, 3» 69, 71. lOo, xo6, X97«
200, 204, 309, 2x2, 2J4, 234, 243.
232, 2Q4; colors around 190, 323, Pi.
VI (193). VII (208). VIII (209), IX
(224). X (22s). XI (240). XII (241),
XIII (288). XIV (289). tempera-
tures of 77. 20s. 219, 220-322, high-
est 321 ; motion pictures of 79; the
first seen 189-190; relation to old
volcanoes X93> 209; cooking over PL
VI (193), VIII (209), 106, 200, 20S»
334-236( 282; steam from not de-
pendent on surface water 2 X9 ; slack-
ening activity of 2x9; measuring
temperatures of PI. IX (324), XIV
(389). XV (304). 221, 242: started
by a bear 324; photographing 232;
types of 23S*-339; domed 237; in a
river 34S> 243: chemistry of 246-
349; lighting fire from 282

Fumes 9. 2x11, 25,33* x8s. 192, 203. 281

Game x6s

Gaseous transfer of rock constituents
343

Gases i, 25, 221. 281, 283. 291. 293;
quantities of liberated from magma
41, 299ff; changes in rock wrought
by hot 343; collection of samples of
209; odors of 346; composition of,
unrelated to tempature 246M; prob-
lem of liberation of, from magma
30X ; heavy charges of, in magma 303

Gantier, A. 41

Geographic Harbor $, 69. X33. 302,
324, 326, 337* 328: entrance to and
islands in 330

Geographic importance of Katmai
Pass 69, 72. 74. 79. 99. 265, 367, 328

Geography of Katmai region 67-7x»
30S-337

Geological formations in Katmai re-
gion S9f xox, X34, I35t X40, 233. 238,
249. 27 X, 306; Jurassic measures of
xox, 125, 233. 249. 271

''Geological thermometer" (quartz)
293

Geophysical Laboratory of C^amegie
Institution 78. 79. 80, 205

Geranium erianthum 157

GUbert. C. H. 311

Glaciated, surfaces 124, 124. 123, X2S.
308, 327; shores of Island Bay 312

Glaciers, on Katmai Volcano 99> X67,
X7x» 173. 275; in Katmai region 124,
369. 271 largest X3X-X32; on Mount
Martin 183, 184

Glisan, Rodney L. 80

Gohna (India) Flood 1 19; landslide 144

Gore, J. Howard 341

Grand Canyon (of the Colorado) X22,
123. 249

Granite gives off water when at high
temperatures 41 ; relation of to ash,
41, 299x; 8oil formed by 47; weight
of contrasted with that of pumice
2Q3; in batholiths 301

Graph, showing relattons of tempera-
ture>un-spot numbers and radiation
intensity (pyrheliometric values) as
affected by volcanic eruptions 32;
of monthly averages of solar radia-
tion intensity 34: showing effect of
dust from Katmai on skylight polar-
ization, insolation and temperature
36

Grass comes up through ash 42:
growth following ashfall 4S; tall
growth of after eruption 274

Graves, Ralph A. 34x

Greased Hill 241

Great Khingan Mountains (Asia) 58

Greek Church. Kodiak during erup-
tion Q

Greeley, A. W. 341

"Green Kodiak": appearance of vil-
lage before eruption 2

Grewingk (Volcano) 63. 6$

Grewingk. C. 63

Griggs. Robert F. 72, 82«, 74, 76, 77*
84, 87, 132; Mrs. Robert F. 1 23. I74;
David 123; Julian 123; Rosamond
123; Ruth 123, 130

Grosvenor, Edwin P. 341

Grosvenor, Gilbert 115. 307, 333. 34i;
Foreword by XV

Grosvenor, Lake 307, 308, 308, 314;
dimensions of 3x4

Gull. Cape 33 x

Hagelbarger. P. R. 17. 74. 7^. 77 > 77.

80, 147. t8i. 205. 210, 2X1, 213. 224,

239. 295ff. 520, 322
Hagelbarger. Ralph 80, 210
Hagelbarger's Pass 133
Harbor 131. X9S, 322-^25; Geographic

S> 69. X33. 302. 324. 337> Hidden 325
Haze from eruptions 33-37
Helrlegel. Hermann 160
Helt, Richard E. 77. 79. 80
Hematite 243
Hesse. William A. 71, 86
Heuchera glabra 157
Henning, William L. 77* 80, 181, 210,

241
"High sand mark" PI. V (192), 262,

263
Hildebrand. J. R. 34 x
Hine. J. S. 74. 76, 77, 80, 208. 2/0, 214,

215. 313. 314. 324
Hissong. J. E. 29
Hitchcock. C. H. 279«, 281 m
Hobbs, W. H. 57«. S8
Homer. M. A. 71, 86, 13s
Horsetail (plant), growth in ash 47*

32, 155; as a soil binder 53
Horsehoe Pond (in Katmai region) 164
Hot and cold lake (Fissure Lake) 244
Hot ground, camp on 207
Hot springs, map facing 60, X33-X33,

197. 3X9. 339. 248, 269
Houses, effect of ashfall on 5, 14, 16, 18
Hudsonian forest 305
Humphreys, W. J. 29»«. 33. 3S". 39. 41
Hutchinson, George W. 341
Hydraulic effects of flood, peculiar

XXO-XX3
Hydrocarbons, aromatic, in fumarole

deposits 248
Hydrochloric add 246
Hydrogen sulphide 190, 246, 248
Hydrofluoric add 203. 246, 347
Hymenoptera 16411

Ice cliffs, in Katmai crater X73. I74;

climbing 184
Ice ages, relation of volcanoes to 39-4X
Iceland, volcanoes in 33
Igneous rocks, chemical make-up of 46
Illamna, Volcano X5. 61; Lake 33;

Bay 23
niuk Arm (of Naknek Lake) 69. X27,

239. 30X. 306, 307
Ilktugitak, Cape 327
Innoko 20
Innuits 266
Insects in Katmai region 95. 163. 208,

314; dead ^ upi>er Katmai Valley
262, X63-X64
Insolation graph showing effect of

dust from Katmai on (chart) 36

Iron, in incrustations 247 ; in water 27 x

Isanotsky Peaks (Unimak Island) 62

Islands, adjoining Alaska Peninsula.

map, front cover; glaciated shores of

312; Bay of 305. 307 ; in Amalik Bay,

entrance to Geographic Harbor 330

Jacob. Harry E. 79. 80
Jaggar, F. A. 6311
Japan 57; volcanoes in 33. 58, I45
Japan Sea 58

Johnstown Flood (comparison) 119
Jones, E. I^ester 341
Jones. Frank I. 80, 251
Julian's Falls 123
Juncoides sp. 160
JuncMs sp. 157
Juneau (city) 15. 23. 26
Jurassic strata in Katmai region xox,
125. 233. 249. 271

Kaflia Bay, 17; experiences of people

at 19
Kalsin Bay 52
Kamchatka 57. 58
Kamishak Bay 307
Kanatak (village) 19
Kashvik Bay X7. 74
Katalla (village) 23. 26, 29
Katmai, pronunciation of m
Katmai Bay PI. II (81), 81, 103. X33.

331
Katmai C^anyon 69. 95. 97. 1x4* xx6,

XI8. X2I, X22, X34-X27» /JO. /J2,

273; and Second Valley X23-X33

Katmai crater x67-i79> 387. 288; na-
ture of rim of x68, X71* X75'* first
view into X69, 172, 174* X76, X78;
lake within X69; size of X70, 173.
X76-X79; compared to that of Kil-
auea X79

Katmai Fk>od. X03-X2X, 13X

Katmai Lakes 115. ii7. 127

Katmai, Mount (volcano) 13. X5, 17,
57. 71. 100, 126, j68, 286, 331;
quantity of ejecta from 39-3X, 144.
X89, 27 sn, 287; comparison of with
other volcanoes 34. 39. 62; first
ascent of 72, X67; first journey to
85-xox; first sight of 99; gladers on
slopes of 99; description of xoo; pre-
vious activity of X24; deposits on
slopes of XS5> douds about 167. I73.
175; dimbing 160* 170; second as-
cent of X7x; ascent of. in 1917. X74-
X75; slackening activity of 220; re-
lations of eruption of. to Valley ac-
tivity 26X, 285; sections of. before
and after eruption 272; before erup-
tion 270, 275; beginning of eruption
of 285 ; remains of 270, 286; what be-
came of top of? 387-303 > at extreme
of explosive volcanoes 391 ; activity
of, intermittent 299

Katmai National Monument 5. 74. 80.
195. X97» 315. 317. 327; central por-
tion, map 68; topographic map,
back cover; proclamation creating
33X-333; accessibility of 339

Katmai Pass 19. 69. X89, 308, 233. 253.
265. 367-273f 375. 329; geographic
importance of 69, 72, 74. 79. 99. 265.
367, 328; views in igo, ip/. 262

Katmai region, geology of S9. xox, X34,
X3S> X40, 233. 338, 249. 37X, 304;
geography of 67^x, 305* 339* lakes
of 69. 1X5. 127. 305-3XX; desolation

Digitized by

Google

338

INDEX

in 8a, 95, 127: flowers in 83; Jurassic
measures in loi, 125, 233. 349. 371:
entrance to. discovered 133: types
of landslides in 144; conditions in,
before eruption 365-375

Katmai River 67, 69, 74. 103-104* 369.
views of ^7. too. I OS. /o<¥; fording 96;
upper reaches of 137; source of 133

Katmai Valley 23. 69, 71* 74. 268. 275;
views in PL II (81), 73, S6. 04, ill,
J 12, 114: devastation of 81-101;
features of 95, 103, 104; second 115.
117. X37, 13X ; in later years 147-153;
physiographic instability in 147-
153; ashfall in 153-185; return of
plants and animals to 155-165

Katmai Village 17. I9. 7i. 103. 155.
366. 268, 277. 288; desolation of,
after eruption 18. 104

Katolinat. Mount PI. XV^I (305). 394.
306, 307, 307

Katz. F. J. bin

Kauffmann. Rudolph 341

Kayaks (skin boaU) 308

"Kettle hole" pond 143

Kikuchi. Y. i4Sff

Kimball. H. H. 36n, 3511. 36N. 39

Kinak Bay 131. 132. 325

Knife Creek 69. 239; valley of 69, la?.
231; canyon of 254

Knife Peak 70. 233. 237. 253

Knight {Charles R.] 321

Kodiak (Island), Alaska i. 67: appear-
ance before eruption 2; scene on 2.
4* 5, 6, 0, 10, J4. 20, 22, 3S. 44, 4S,
40, 51; appearance after eruption 4,
5, 6; eruption as experienced at 7-
13; during eruption p; cattle on 44;
recovery of vegetation at 45-55* PL
I (80); waters around 61; commu-
nications with 67; birds on 165;
bears on 317

Kodiak Radio Station 8

Kolb, Emery C. 77, 80. 349, 250, 251

Korea 58

Krakatoa (volcano), Java 35. 33". 34*
35t37

Kubugakli. Cape 17. 275* 33 1

Kukak Bay 32S

Kupreanof Straits 15

Kuril Islands 57. 58

Kuskokwim River 366

Labradorite 289

Lacewing (insect) i64n

La Gorce, John Oliver 341

La Gorce. Mount 306, 306

Lakes of Katmai region 69. iiSt '20,
127. 212. 306. 307-313. 307* 308,
300, 314, 3ts

Landing, first 8x

Landslide 143 ; in Katmai Canyon ti6:
Mageik 69. 74. 135-145. 136, 138.
140, J41, 144; Noisy Mountain T38'
139, 131. X42; Frank, Alberta 139.
143; Elm, Swiuerland 143; Ross-
berg, Switzerland 144; Gohna, India
144; Bandai San, Japan 145

I landslides 285; and rivers, an inter-
mediate between 131; typical fea-
tures of 139: previous 139. X43-145;
conditions ^vorable for 140; types
in Katmai region 144

Lansing. Robert 332

Latouche (town) 35

Lava 41. 279; composition of in Kat-
mai region xox ; bowlder, moved by
stream 108; from Novarupta 238,
340. 341; laJce of 299; new and old
from Katmai 290; peculiar structure

0l2fi6

Lava flows iii. 33. 124. 123, 126, 140;

greatest during historic tiroes Ga-

pan) 33\ old, from Katmai 287
Learn. The Rev. George A. 48
Lee. Jack 17

Lu-Chu Islands (Japan) 58
Leslie. Frank 274
Lethe (stream) 239. 24S. 345
Life dependent on volcanic action

43-44; no destruction of 3, 19
Lightning 7. 8. 11, 23. 65
Lime 4611
Limestone 271
Logan. Daniel 279f(
Long Island, roof crushed by ashes j6
Loring (town) 37
Lupine in ash covered region 45. 15S,

158. X59-X60

McCaughey. W. J. 389

McConnell, R. G. I39»«. I43»«

McCulloch, (U. S. Coast Guard Cut-
ter) 65

McCulloch Peak (Bogoslof Island.
Alaska) 63. 65

McMuUen, (Capt.) C. B. is

McPherson. J. L. 15

MacDonald. T. L. 341

Mageik. pronunciation of 69M

Mageik Creek 124. 189; flood of xx?.
131

Mageik Landslide 69. 74. X35-X45,
J 36, 138, 140, 141, 144; dimensions
of 14X ; compared with other slides
X43-X45. 241, 285

Mageik, Mount (volcano) 19. 69. 78,
8s. PI. V (192), 99. 124, 160, 166,
t86, 190, 320. 233. 354. 271, 375*
284, 285; ejecta from searched for
86; relation of to Mount Martin 181

Magma, definition of m, 41 , 209. 24711,
361, 388: intrusion of 140, 193. 239;
reached surface in Valley 340, 378;
mixture of old and new 389- 39X;
ate way upward 291, 29s. 303; new,
temperature of 39X-293; lake of in
crater before explosion 399; ap-
proached surface over considerable
area 301

Magma chamber 388, 292, 301, 303

Magnesia 46M

Magnetic properties of ash 4611

Maithana. Mount (Himalayas) 144

Malbat Station (Vancouver Island) 27

Man-back transiwrtation 75

Manganese oxide 46M

Manning (U. S. Revenue Cutter) 8,
10, XI, 13; refugees on board //

Marsh gas from fumaroles 247

Martin, George C. is» .2 in. 2611, 29.
3411. 61 fi. 71. 86. 103. IS3

Martin Creek 97. 124; bridge over 92;
valley 69. 74. I3S

Martin. Mount (volcano) PI. Ill (9^),
71. 85, 180. x8x-x87. 182, 184, 190.
220, 264, 385; ejecta from searched
for 86; attempts to climb x8x, 183;
ctater of x8x* 185

Mather, Stephen T. 341

Mauna Loa (Hawaii) 279

Maynard, C F. 74. 76, in, 208. 213,
215, 295**

Meniam, C. Hart 34<

Metcalf Peak (Bogoslof Island) 63, 65

Metha Nelson (schooner) 8

Metroldn. Walter 72, 73. 74, 7^, i97.
315

Military importance of road across
Alaska Peninsula 339

Miller, August E. 77, 80, 313

Minerals, formation of 3

Moose, absence of in devastated region
164

Moose Creek 321

Morey. George W. 3011*

Mosquitoes 17. 95. 163. i64»

Moss, caught ash 38, 53; network of,
on forest floor S3. 55. 58; around
fumaroles 209. 223

Moths 164M

Motion pictures, of volcanic phenom-
ena 71. 79> X95; of pumice-choked
streams isa

Motorcycle, attempt to use. in 75

Mounds, curious conical X37, 138* 128,
'34. X39* X42; in Bandai San land-
slide 145

Mud, from Katmai Volcano 23. X55*
28s; in Katmai Valley 81. X03; flow
PL II (81), 96* 123. 253. 285; slides
X39-X3X; on slopes of Katmai 167;
stuck in the i6q; steaming 338; pots
and "pimples" 339; pot 284

Naknek, pronunciation of 69N

Naknek. Lake 17. 69. 77. 79. i55. 233.
PL XVI (30S), 26s. 266. 367. 368,
305. 307, 307. 308. 309. 324* 327

Naknek River 77. 79. 265

Naknek Village 17. 2X

National Monument. Katmai 5. 74.
80. 195. 197. 315. 317. 327; procla-
mation creating 331-333 ; accessibil-
ity of 329

National Park Service. Director of. in
charge of Katmai National Monu-
ment 332

Native huts (barabaras) 17 and m. 24,
103; interior of 26

N^v^ 131. 171, 184

Nimrod (boat) 324, 323, 327

Nitrogen compounds, absence of, in
ash 160; in fumes 209. 246. 248

Noisy Mountain (Katmai region) 74*
117, 241 ; landslides of 128, 138-139,
X3X, 142

Nome stampede, use of Katmai Pass
during 99. 267

Norman (motor boat) 13

North. S. N. D. 341

Novarupta (volcano) 39, X92, iQ4f
233. 238, 340-341, 253. 276. 280,
297 «. 300; part played by, in Valley
sandflow 261; explosion of 384;
quantity of ejecta from 285; subsid-
ence of 288. 303

Nushagak (village) 19. 266, 267. 268

Nutting, P. G. 37

Nyman, [Gus] Captain 324. 327

Observation Mountain 124. 147. 189.

253. 273. 274. 27S; camp 197
Oceans, relation to volcanoes 43
Odors 25. 167, 190. 192. 238. 346, 281
Okhotsk Sea (Asia) 58
Olga (town) 29
Oogashik (village) 265
Orloflf. Ivan, letter of 19. 21
Osbom, H. F. 32i«
Outfits for expeditions 74
"Overhead stoping" 397»», 303
Overloaded streams 150-153; channels

built up by 132
Oxygen 246

Pack train 2/0

Park. Katmai (properly Katmai Na-
tional Monument) 306, 307. 334. 339

Pass 131. 132; Katmai 19. 69, 72, 74.
79. 99. X89. IQO, IQI, 308, 333. 253.
262, 265. 367-373. 274. 329; Hagel-
barger's 133

Digitized by

Google

INDEX

339

Pavlof Volcano 66

Peat 135. I37

Pedmar. Mount 8i

Pel6e Volcano (West Indies) loi, 261,
279

Pembcrton. C. C. 27

Peneplain 167

Perry. (Capt.) K. W. 8, 9. 11. 15. 4^

Personnel of expeditions. 1915. 7i» 72;
1916. 7a» 73: 1917. 74» 76; 1918, 77;
1 9 19. 77. 77. 79. 80

Petrof. Ivan 265, 306, 321, 328

Petroleum-like emanations in Valley
247-248

Petrological studies loi. 243

Peulik (volcano) is, 183, 307

Phosphoric acid 46. 4711

Photography 207, 224. 251

Physiographic instabUity in Katmai
Valley 147-153

Physiography, curious 128, 131

Phenomena accompanying eruption
7-13. 15-31. 33-39. IIS. 139. 144.
242; sky, following eruption 37

Picea canadensis 305

" Pimples ". a type of fumaroles 240, 2P4

Plagioclase feldspar 289^

Plants, effect of ash on 3. 45-55. of
fumes 33. of reduction of average
temperature 39 J herbaceous in
Katmai region 82; grown in ash
experiments 47. so; resurrected /jj;
prevent ash blowing 5a; effect of
ash on growth 53: return of, to
Katmai Valley 155-161; practically
absent from Valley in 191 7, 209;
return of. to Valley 223

Plowing before ashfall, effect on re-
vegetation S4

Plug of Novarupta 238, 240

Poa sp. 157

Pogromni Volcano (Unimak Island)
62, 63. 63

Polarization of skylight, affected by
eruptions 36; graph of 36

PoUmoniutn coeruUum 157

Pompeii, contrast with Kodiak 13

Pond, in Katmai National Monu-
ment 266

Ponds, in upper Katmai Valley 127

Poplars, Balsam (Populus balsamifera),
in Katmai region 82, ^j, 30s

Port Graham (English Harbor)
Alaska 62

Port Moller (town) 265

Port Townscnd. Washington 27

Post. W. S. 275

Potassium compounds 46, 4711, 247

Powers, Sidney 6511

Preliminary gymptoma observed at
outlying stations 19

Preparations for expeditions 74

Pressure, air, reduced by volcanic
eruptions 35

Prince Rupert (town) 27

Prince William Sound 25

Printer (tug) 13

Property, destruction of 7

Proportions, land of stupendous 173

Pumice 17. 29. 41. 72, 75t 81, 147;
floating 30, 81. 89, 103, 151; sharp-
ness of 88; straining, out of water
88; choked streams 150-1 53( 132;
wind-blown, cuts faces 199* 229;
lightness of illustrated 250. 293;
mixed, from Katmai 2qj; highly
inflated 29911

Putnam, George R. 341

Pyrheliometric values, curve of
(graph) 32

P3rroxene crystals 243

Quartz 46, 293 and n

Quicksands 75. 88. 89, 104; fording
stream through 87; of Katmai River
J08

Radiation intensity, graph showing
32

Radio 9

Radio Station. Kodiak 8

Rains, frequent in Katmai region 52

Rampart (city) 26

Redoubt (volcano) 15, 61

Refrigerator, a natural iq8

Refugees. Kodiak 11. //

Reindeer 265. 266

Remoteness of Katmai 67

Return, of plants to Katmai Valley
155-161; of animals 161-165

Revegetation 3. 45-55. 7i; in Katmai
Valley 155-161; prevention of by
acid rains 183

Rhodiola rosea 155

Rhyolite. acid 46. loi, 240, 241. 287

Rhyolitic glass 289

Richtofen. (Ferdinand Baron von] 57K

Rishiri (Island). Japan 57

Road across Alaska Peninsula, mili-
tary importance of 329

"Roches moutonnees" 308

Rocks, "balanced" 23; hurled during
eruptions 33; slides (falls) 2 in, 128,
I28r-i29, 141-147, 148, 241; changes
in wrought by hot gases 243; old
and new mixed 289, 200

Rodents 165; absence of in devastated
regions before 191 9. 164

Rodriguez (Island, Indian Ocean)
25.35

Roots, two sets of on plants 40; sur-
vived ashfall 45. 5o. 5Z. Z55; condi-
tions encountered by 4611; "two-
storied", result from ashfall 51. 83;
of willow at ash suriace 133; re-
juvenation of 157

Rope, crossing river with 97

Rosamond's Falls 123

Rossberg Landslide 144

Rouge, C. D. 27

Rubus spectabilis 155

Ruby 26

Russian American Company 266

Russian, Church at Katmai Village
103; missionary at Nushagak 267

Ruth's Falls 123, 130

St. Michael (city) 266

St. Paul (old name for Kodiak) 268,
269

St. Pierre (Martinique) 261, 279

Sakhalin Island (Asia) 57

Salix arctica 155

Salmon in devastated district ZSO*
161, 163; in lake region 308. 309;
leaping up falls 310, j/o. 316; re-
markable instinct of 311, 3 13; spear-
ing J//; fisheries 311

Salmon-berries 51, 155; since eruption
51

Sambucus pubens 157

Sandblast 52; effect of on trees p/;
effect of on plants 158

Sandflow i. 3. I5* I7. 29. 253-261,
254f 255, 256, 257, 258, 260, 263,
284; 285. 301; dimensions of 253*
283; heat of 258; fracture in 259

Sandstone 124, 125. 140, 233. 238,
243. 249, 263, 306; fragments of in
magma 299

Sanguisorba siUhensis 155

Sargent, R. H. 65

Savonoski, River 307; Village 17. 268,
269. 271. 277. 288

Sayre, Jasper D. 74. 76, 77, 77, 79. 80,

147. 175. 205, 213. 239. 295>»
Scenery, beauties of 306-307, 32s
Sedimentary rocks in Katmai region

101, 135. 140. 233. 240. 306
Seedlings 162; slow to start 52;

timothy in ash S4; start in forest

55; start in Katmai Valley 147, 157.

158, 159; of trees abundant 305
Seeds (dormant) 157
Sekiya. S. I45n
Seldovia (town) 19. 21, 23
Seward, Alaska 15. 25
Shale 140. 233, 306
Shantz. A. B. 274ff
Shelikof Strait i, 15. 67. 75. 167, 268
Shifting streams 150-153
Shipley, J. W. 74. 76, 201, 209, 215.

246, 247, 249n
Ship Rock (Bogoslof Island, Alaska)

63
Shiras. George, 3d 341
Shishaldin (volcano) 62, 63, loi
Shoes, effect of hot ground and pumice

on 205
Showalter, William J. 341
Sierra Nevada (Mountains) 301
Silica 46». loi. 287
Sill theory to account for Valley

activity 301-303
"Silver Horde" 273
Sitka (city), Alaska 56, 57»
Skaptar JokuU (volcano. Iceland) 33
Skin boats (bidarkas) 17. 268;

(kayaks) 308
Skylight polarization, graph showing

effect of dust from Katmai on 36
Sky phenomena following eruptions

37
Slide Mountain (Katmai region) 97
Slides, ash 82, 86, 87; land 115-117.

128; rock 128, 148; the MageUc

138, 140, J 41, 144; drainage on

136
Smith, George Otis 341
Smith, W. R. 6s
Smoke (volcanic) 15. i7. 2in, 279;

from fissures 202
"Smokes" in Valley, number of

199-201
Snodgrass, M. D. 53
Snowdrifts, buried under ash 148-149,

I4Q
Snowy Mountain (in Katmai region)

131
Soda46n
Soil, formation of 3; contribution of

ash to 46; granite 47; tests of ash as

47
Solar radiation intensity, graph show-
ing monthly averages of 34
Soluka. Valley 69, 83, 33, 327: Creek

82, 87. 88-89. 00. 133
Solution of old rock in new magma

289-293, 299«
Spearing salmon j/r
Spencer, Cape 25
Springs 127; hot 132-133, 197. 2Z9»

239. 248. 269
Spruce, white 305
Spurr, J. E. 189, 190, 267. 269ft, 273.

275
Squentna Valley 61
Squhres, Grant 341
Stanovoi Mountains (Asia) 58
Steam (volcanic) 3, 5, I7i. 208; origin

of 41. 246; estimated volume of,

liberated 41; rising from Katmai

126; clouds of, continued from

Katmai 168, 169; in Valley of Ten

Thousand Smokes 190, 237; effect

Digitized by

Google

340

INDEX

of, on weather ao8; not visibly
Effected by turfaoe water a 19, 246;
predominant constituent in gasea
346; from Noyarupta 276

Steam-heated tent ao7-ao8

Steaming, mud areas a37-a38; fis-
sures 27^

Stone. Julius, Jr. 77, 80

Stone-crop (plant) iss

Stoping. overhead 2Vrn, 303.

Storm at Baked Mountain camp aa6~
a3x

Streams, ash in 19. 89. 150, 151. isa,
153; overloaded 132; views of isi,
153* 243

Streptopus amplexifolius 155

Subsidence, of Novanipta a88, 303;
no evidence of. in Katmai Crater
a88-a89

Suess. Eduard 43, S7n

Sulphur 9. 33» 371; fumes 6s, 167;
in Katmai crater 169; in Martin
crater 185. 190. xpa; dioxide 346;
around f umaroles 347

Sulphuric acid 46n; rains at great dis-
tances as; in air 37; little in Valley
20s, 247

Sumatra, volcanic chains between
Alaska and 57

Sunsets affected by volcanic dust 34.
37

Sunshine, intensity of, reduced by
eruption 33. 3S

Sun-spot numbers, graph of 52

Supplies 75; dependence on 85

Surveys of Katmai region 77

Swan. Herberts. i79»»

Swan, wild, captured alive 313

Taft. WiUiam Howard 341

Tahoma (U. S. Coast Guard Cutter) 65

Takkala Bay 133

Talus 128. 141, 176

Tatoosh Island. Washington 29

Taylor. Boyd 341

Temperature, atmospheric during ash-
faO xx; departures, graph showing
32; world, reduced by eruption 33,
34. 35. 37 (table) 39 and n; graph
showing effect of dust from Katmai
on 36; increased in high altitudes by
dust 39; of f umaroles 205, 219,
330-333 ; of new magma 391-393

Terraces, stream-carved IS3

Terrane, remarkable in Martin Valley
X3S

Tertiary lava 337

Therapeutic possibilities of the Valley
307-308

Thermocouple 33 z, Plate XV (304)

Three Forks (in Valley of Ten Thou-
sand Smokes) 339

Thunder 7, xi, 23. 6$

Thwaites, J. E. 10

TiU (glacial) 125

Timothy, seeded in ash sa* 54

Titanium oxide 46n

Tittman, O. H. 341

Tomboro, East Indies (volcano) 34

Topographic map, back cover; work
209

Tourists to Valley 80; need road 339

Trail, views along the 73, 74, igi;
through tall grass 274

Transportation in Katmai region

74-7S> 307; man-back 75 » facilities
iiuulequate 328; problems would be
solved by automobile roads 339

Tree growth, limit of 305

Trees, ash in 20, 28, 38, 53; surviving
in Katmai Valley 47-49* 8a. 83;
effects of ash on 5j. 83, 133; dead
in volcanic region qo, pi, 08^ 102,
107, 131: affected by sandflow
356-357

Trident (volcano) 08, 99> X7x, 190.
220, 233, 240, 271, 28s

Tridymite 243; 29311

Trout, rainbow, in Naknek River 324

Tuff 235. 353* 254. 256, 261

Tundra 74. 3x4

Tuimels In ash 91

Turtle Mountain (Canadian Rockies),
landslide at 143

Twilight lengthened by' volcanic dust
in air 37

Twisted-stalk (plant) 155

Ukak Valley 17, 229. a3X, 253; Ukak

River 239
Unalaska (Island) Alaska 63. 65
Unfinished work in Katmai region 80
Unimak (Volcano and Island) 62;

peaks on Island 62; Pass 63
Uplift X40
Uyak (Village) 15. X9

Valley, Katmai 33t 69, 7x, 95* X03,
I04» 163; Second (Katmai) 74. xis.
1x7. X37» X3x; Soluka*.69. 83, X33

Valley of Ten Thousand Smokes, be-
fore eruption 19; position of 69, 333;
discovery of 7a, 189-X9S; explora-
tion of 73, 74; exploration of, in X917 •
197-3x7; first adequate description
of 74; impression on fint seeing
3XX-3X7; in X9X9. 3X9-33X; descrip-
tive account of 333-351 ; dimensions
and shape of 333; entrance to 233;
unexplored parts of 335; activity in.
began before Katmai eruption 355-
356; nature of activity in 378-383;
views of 188, JQ4, 106, 200, 202, 204t
2o6t 210, 212, 214, 216, 2i8t 220,
226, 227, 230, 232, 234, 236, 238,
242, 243t 245, 248, 252, 262, 264*
276, 2Q2, 204* 206, 298, 300; a gen-
eralized section of 263

Valleys prominent topographic feature
near Katmai 67* 69, 268

Van Campen. F. R. 25

Vancouver. B. C. 27

Vegetation, effect on, of volcanic ac-
tivity 3» as, 82; recovery of at Ko-
diak 45-55* transported by slide
137. J40: recovery of in Katmai re-
gion 147; absence of in Valley 209;
return of to Valley 223

Vents, volcanic 3; reawakening of 65;
hot sand probably came from local
36x

Victoria. B. C. 27

Virginia, haze from Katmai eruption
in 35

Visitors, first to Valley of Ten Thou-
sand Smokes 80. 329

Volcanic activity, subsidence of 3X,
220; area of 301. carrying work into
67-;79; axes and structure lines, map

facing 56; axis, relations of fuma.
roles to X93; festoons 57^x

Volcanism. problems of 3. 77, 79; Kat-
mai presents unique opportunity to
study 304. 33 X

Volcano (Katmai) first journey to 8x-
xox

Volcanoes 17, 33. 34* 43. 57-65* 60, 62,
63* 64, 66, 67, 70* 78, 83* 98, j6o,
166, 168, 171, 172* 174* 175, 176,
178 1 180, J82, 184, 186, 190, 220,
333. 340, 27X, 285; dormant 7; capa-
ble of causing ice ages 39-4x; the
Alaskan waterfowl 313. 57-65;
series of, including Katmai 99; min-
iature 337

Walker Lake (properly Naknek Lake)
265

Wallace, H. M. 77, 80. 309. 3^4

Walter (Metrokinj 73. 74, 76 >

Water, drinking 8. 81. 89; liberated in
eruptions, importance of in econ-
omy of nature 4X-43

Waterfalls 81. 109. 123. 130. 265. j/o.
3x0, 31T* 316, 337

Waterfowl 313

"Water slides" X39-X3X

Waves (in sea) caused by volcanic dis-
turbances 62

Weasels 164

Weather in Katmai region 90, 97* X58-
X59» X89. 308-309* 320, 336, 369,
273; effect of eruption on 33, 368

Wells, E. H. 274»«

White, Henry 341

"WiUiwaws" (violent winds) 336-33 x,
325

Willows effect of eruption on 83

Wilson, (President) Woodrow 5, 74;
proclamation of, creating Katmai
National Monument 33X-333

Wind, caused small avalanches X35>
373; action on landscape X48-X49;
caused irregularity in ash deposits
155; chief factor in seed dispersal
X58-X59; violent in Valley X99, 336-
33 x^ violent through Katmai Pass
368, 369; protected forest near Kat-
mai 305

Wind-blown ash 5^-55* 5Q

Windy Creek 239

Wisconsin, haze from Katmai erup-
tion in 35

Wolves 164, 208. 3x5

Wolverines 164, 208, 3x5

Women's Peninsula, dunes on SQ

Wood Island 8, X3

World War. effect on expeditions 79

Wormwood X55

Wrangell. Mount (volcano) 56, 57»

Vagashoff , Andrean 76

Yellowstone Park (comparison) Z9X,

195. 197, 204, 239. 329
Vori, Charles 77, 80, 183, 184, 337

229, aap, 230
Vosemite Valley (dbmparison) 123
Yukon River 266, 329

Zay, Karl X4411

Zies. E. G. 77, 80, 221, 228, 239. 230.

231, 246. 247M
Zinc, melted in fumeroles 77, 320,

221

Digitized by

Google

NATIONAL GEOGRAPHIC SOCIETY

GEOGRAPHIC ADMINISTRATION BUILDINGS
SIXTEENTH AND M STREETS NORTHWEST, WASHINGTON, D. C.

GILBERT GROSVENOR. Preddent HENRY WHITE. Vice-President

JOHN JOY EDSON. Treasurer O. P. AUSTIN. Secretary

BOYD TAYLOR. Assistant Treasurer GEORGE W. HUTCHISON. Associate Secretary

FREDERICK V. COVILLE, Chairman Committee on Research

EDWIN P. GROSVENOR. General Counsel

EXECUTIVE STAFF OF THE NATIONAL GEOGRAPHIC MAGAZINE

GILBERT GROSVENOR, editor

wil;.iam j. showalter

Assistant Editor

JOHN OLIVER LA GORGE, Associate Editor

RALPH A. GRAVES FRANKLIN L. FISHER

Assistant Editor Chief of Illustrations Division

J. R. HILDEBRAND. Chief of School Service

CHARLES J. BELL

President American Security and
Trust Company

JOHN JOY EDSON

Chairman of the Board. Wash-
ington Loan & Trust company

DAVID FAIRCHILD

In Charge of Agricultural Ex-
plorations, U. S. Department
of Agriculture

C. HART MERRIAM
Member National Academy of
Sciences

O. P. AUSTIN
Statistician

GEORGE R. PUTNAM
Commissioner U. S. Bureau of
Lighthouses

GEORGE SHIRAS. 3l>

Formerly Member U. S, Con-
gress, Faunal Naturalist, and
Wild-game Photographer

E. LESTER JONES

Director U. S. Coast and Geo-
detic Survey

BOARD OF TRUSTEES

WILLIAM HOWARD TAFT
Chief Justice of the United States

GRANT SQUIRES

Military Intelligence Division. Gen-
eral StaflF. New York

C. M. CHESTER

Rear Admiral U. S. Navy. For-
merly Supt. U. S. Naval Ob-
servatory

FREDERICK V. COVILLE
Botanist. U. S. Department of
Agriculture

RUDOLPH KAUFFMANN
Managing Editor The Evening
Star

T. L. MACDONALD
M. D.. F. A. C. S.

S. N. D. NORTH

Formerly Director U. S. Bureau
of Census

JOHN OLIVER LA GORGE
Associate Editor National Geo-
graphic Magazine

ALEXANDER GRAHAM BELL
Inventor of the telephone
Died August 2, 1922

J. HOWARD GORE

Prof. Emeritus Mathematics. The
George Washington University

A. W. GREELY
Arctic Explorer. Major General
U. S. Army

GILBERT GROSVENOR

Editor of National Geographic
Magazine

GEORGE OTIS SMITH
Director. U. S. Geological Survey

O. H. TITTMANN

Formerly Superintendent U. S.
Coast and Geodetic Survey

HENRY WHITE
Member American Peace Com-
mission. Formerly U. S. Am-
bassador to France. Italy, etc.

STEPHEN T. MATHER
Director National Park Service

ORGANIZED for "the increase and diffusion of geo-
graphic knowledge." the National Geographic Society was
founded thirty-five years ago.

IN addition to its work of exploration and research con-
ducted in the Mount Katmai volcanic region, as set forth
in the preceding pages. The Society, at an expense of
more than I50.000. sent a notable series of expeditions
into Peru to investigate the traces of the Inca race. The
discoveries of these expeditions form a large share of our
knowledge of a civilization which was waning when
Pizarro first set foot in Peru.

THE Society also had the honor of subscribing a sub-
stantial sum to the historic expedition of Admiral Peary,
who discovered the North Pole.

NOT long ago The Society granted las.ooo, and in addi-
tion I75.000 was given by individual members through
The Society to the Federal Government when the congres-

sional appropriation for the purchase was insufficient,
and the finest of the giant sequoia trees of California were
thereby saved for the American people and incorporated
into a National Park.

THE Society is conducting extensive explorations and
excavations in northwestern New Mexico, which was one
of the most densely populated areas in North America
before Columbus came, a region where prehistoric peoples
lived in vast communal dwellings whose ruins are ranked
second to none of ancient times in point of architecture,
and whose customs, ceremonies and name have been
engulfed in an oblivion more complete than any other
people who left traces comparable to theirs.

TO achieve the purposes for which it was founded, the
National Geographic Society publishes the National
Geographic Magazine, and all the receipts of The
Society are invested in the Magazine itsdf. or expended
directly to promote geographic knowledge.

341

Digitized by

Google

Digitized by

Google

Digitized by

Google

Digitized by

Google

Digitized by

Google

__— ^^

011iJC4 07

The N'allcy of Ten Thoumid Smoko.
m4 MCZ I

3 2044 062 423 538

Digitized by

Google

\ 1

1 .

Digitized by

Google

TVw

Digitized by

Google

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "The Valley of Ten Thousand Smokes"

See other formats