ue a4
Le
et 1

ELT Tier
Reet
fate

a ead
“ Tei Ae
AR
ae

i
Pk eS HS
WG ee

*

Pa
Reais

: Nae,

i PRN
abgebme iiee

Rae

.
a

Pht 5 a
en
:

be ai ,
o

SMITHSONIAN

ONTRIBUTIONS TO KNOWLEDGE

NEON Nee eT Vi

EVERY MAN IS A VALUABLE MEMBER OF SOCIETY WHO, BY HIS OBSERVATIONS, RESEARCHES, AND EXPERIMENTS, PROCURES
KNOWLEDGE FOR MEN—SMITHSON

(No. 1739)

; CLIEXSAOF WASHENGTON
PUBLISHED BY THE SMITHSONIAN INSTITUTION
1907

lee
i he

: yee on
ae oe Ne ee ee

Tuts volume forms the thirty-fourth of a series, composed of original
memoirs on different branches of knowledge, published at the expense and
under the direction of the Smithsonian Institution. The publication of this
series forms part of a general plan adopted for carrying into effect the benevo-
lent intentions of James Smiruson, Hsq., of England. This gentleman left his
property in trust to the United States of America to found at Washington
an institution which should bear his own name and have for its objects the
“increase and diffusion of knowledge among men.’’ This trust was accepted
by the Government of the United States, and acts of Congress were passed
August 10, 1846, and March 12, 1894, constituting the President, the Vice-
President, the Chief Justice of the United States, and the heads of Executive
Departments an establishment under the name of the ‘‘SmrrHson1an Inst1-
TUTION, FOR THE INCREASE AND DIFFUSION OF KNOWLEDGE AMONG MEN.’’ The
members of this establishment are to hold stated and special meetings for
the supervision of the affairs of the Institution and for the advice and instruc-
tion of a Board of Regents to whom the financial and other affairs are
intrusted.

The Board of Regents consists of two members ex-officio of the establish-
ment, namely, the Vice-President of the United States and the Chief Justice
of the United States, together with twelve other members, three of whom are
appointed from the Senate by its President, three from the House of Repre-
sentatives by the Speaker, and six persons appointed by a joint resolution of
both Houses. To this board is given the power of electing a Secretary and
other officers for conducting the active operations of the Institution.

To carry into effect the purposes of the testator, the plan of organization
should evidently embrace two objects: one, the increase of knowledge by
the addition of new truths to the existing stock; the other, the diffusion of
knowledge, thus increased, among men. No restriction is made in favor of any
kind of knowledge, and hence each branch is entitled to and should receive a
share of attention.

The act of Congress establishing the Institution directs, as a part of the
plan of organization, the formation of a library, a museum, and a gallery of
art, together with provisions for physical research and popular lectures, while
it leaves to the Regents the power of adopting such other parts of an organiza-

tion as they may deem best suited to promote the objects of the bequest.
mI

IV ADVERTISEMENT.

After much deliberation, the Regents resolved to apportion t
income specifically among the different objects and operations of the -
in such manner as may, in the judgment of the Regents, be necessa:
proper for each, according to its intrinsic importance, and a complian
faith with the law.

The following are the details of the parts of the general plan of
tion provisionally adopted at the meeting of the Regents December

1. The memoirs thus obtained to be published in a series of vol
quarto form, and entitled ‘‘Smithsonian Contributions to Knowledg

which does not furnish a positive addition to human knowledge, re
original research; and all unverified speculations to be rejected.
3. Each memoir presented to the Institution to be submitted for e
tion to a commission of persons of reputation for learning in the bra
which the memoir pertains, and to be aecepted for publication only in
report of this commission is favorable.
4. The commission to be chosen by the officers of the Institution, «
name of the author, as far as practicable, concealed, unless a favorab’
be made.
5. The volumes of the memoirs to be exchanged for the transa
literary and scientific societies, and copies to be given to all the college
principal libraries in this country. One part of the remaining copies
offered for sale, and the other carefully preserved to form complete set
work to supply the demand from new institutions. ;
6. An abstract, or popular account, of the contents of these memoi
given to the public through the annual report of the Regents to Congress. —

II. To ivcrzass Kyowxeven.—It is also proposed to appropriate a po
the income annually to special objects of research, under the direct
suitable persons.

1. The objects and the amount appropriated to be recommended D) .
sellors of the Institution. 7 7
2. Appropriations in different years to different objects, so that in
of time each branch of knowledge may receive a share. ;

omg Y

ADVERTISEMENT. Y,

3. The results obtained from these appropriations to be published, with the
memoirs before mentioned, in the volumes of the Smithsonian Contributions to
Knowledge.

4. Examples of objects for which appropriations may be made:

(1) System of extended meteorological observations for solving the problem
of American storms.

(2) Explorations in descriptive natural history, and geological, mathe-
matical, and topographical surveys, to collect material for the formation of a
physical atlas of the United States.

(3) Solution of experimental problems, such as a new determination of
the weight of the earth, of the velocity of electricity, and of light; chemical
analyses of soils and plants; collection and publication of scientific facts, acen-
mulated in the offices of Government. |

(4) Institution of statistical inquiries with reference to physical, moral, and
political subjects.

(5) Historical researches and accurate surveys of places celebrated in
American history.

(6) Ethnological researches, particularly with reference to the different
races of men in North America; also explorations and accurate surveys of the
mounds and other remains of the ancient people of our country.

I. To prrruse Knowteper.—It is proposed to publish a series of reports, giving
an account of the new discoveries in science, and of the changes made from
year to year in all branches of knowledge not strictly professional.

1. Some of these reports may be published annually, others at longer
intervals, as the income of the Institution or the changes in the branches of
knowledge may indicate.

2. The reports are to be prepared by collaborators eminent in the different
branches of knowledge.

3. Each collaborator to be furnished with the journals and publications,
domestic and foreign, necessary to the compilation of his report; to be paid a
certain sum for his labors, and to be named on the title-page of the report.

4. The reports to be published in separate parts, so that persons interested
in a particular branch can procure the parts relating to it without purchasing
the whole.

5. These reports may be presented to Congress for partial distribution,
the remaining copies to be given to literary and scientific institutions and sold
to individuals for a moderate price.

VI ADVERTISEMENT.

The following are some of the subjects which may be embr

reports:
TI. PHYSICAL CLASS.

1. Physics, including astronomy, natural philosophy, chemis
meteorology.
Natural history, including botany, zoology, geology, ete.
Agriculture.
Application of science to arts.

Rep

Il. MORAL AND POLITICAL CLASS.

5. Ethnology, including particular history, comparative philology
uities, ete. : 7

6. Statisties and political economy.

7. Mental and moral philosophy.

8. A survey of the political events of the world; penal reform, ete.

-

Ill. LITERATURE AND THE FINE ARTS.

<6

. Modern literature.

10. The fine arts, and their application to the useful arts.
11. Bibhography.

. Obituary notices of distinguished individuals.

—
woe

Il. To pirruss Knowneper.—tIt is proposed to publish occasionally :
treatises on subjects of general interest.

1. These treatises may occasionally consist of valuable memoirs transl.
from foreign languages, or of articles prepared under the directi no}
Institution, or procured by offering premiums for the best exposition
given subject.

2. The treatises to be submitted to a commission of competent
previous to their publication.

ADVERTISEMENT. Vit

DETAILS OF THE SECOND PART OF THE PLAN OF ORGANIZATION,

This part contemplates the formation of a library, a musemn, and a gallery
of art.

1. To carry out the plan before described a library will be required con-
sisting, first, of a complete collection of the transactions and proceedings of all
the learned societies of the world; second, of the more important current period-
ical publications and other works necessary in preparing the periodical reports.

2. The Institution should make special collections particularly of objects
to illustrate and verify its own publications; also a collection of instruments
of research in all branches of experimental science.

3. With reference to the collection of books other than those mentioned
above, catalogues of all the different libraries in the United States should be
procured, in order that the valuable books first purchased may be such as are
not to be found elsewhere in the United States.

4. Also catalogues of memoirs and of books in foreign libraries and other
materials should be collected, for rendering the Institution a center of biblio-
graphical knowledge, whence the student may be directed to any work which
he may require.

5. It is believed that the collections in natural history will increase by
donation as rapidly as the income of the Institution can make provision for
their reception, and therefore it will seldom be necessary to purchase any
article of this kind.

6. Attempts should be made to procure for the gallery of art casts of the
most celebrated articles of ancient and modern sculpture.

7. The arts may be encouraged by providing a room, free of expense, for
the exhibition of the objects of the Art Union and other similar societies.

8. A small appropriation should annually be made for models of antiqui-
ties, such as those of the remains of ancient temples, ete.

9. The Secretary and his assistants, during the session of Congress, will be
required to illustrate new discoveries in science and to exhibit new objects of
art. Distinguished individuals should also be invited to give lectures on sub-

jects of general interest.

In accordance with the rules adopted in the programme of organization,
each memoir in this volume has been favorably reported on by a commission
appointed for its examination. It is, however, impossible, in most cases, to
verify the statements of an author, and therefore neither the commission nor
the Institution can be responsible for more than the general character of a

memoir.

OFFICERS

OF THE % ~

SMITHSONIAN INSTITUTION .

THEODORE ROOSEVELT,
PRESIDENT OF THE UNITED STATES,

EX OFFICIO PRESIDING OFFICER OF THE INSTITUTION.

MELVILLE W. FULLER,
CHIEF JUSTICE OF THE UNITED STATES,

CHANCELLOR OF THE INSTITUTION.

CHARLES D. WALCOTT,

SECRETARY. OF THE INSTITUTION.

RICHARD RATHBUN,

ASSISTANT SECRETARY IN CHARGE OF NATIONAL MUSEUM.

CYRUS ADLER,

VIII

ea

MEMBERS EX OFFICIO OF THE INSTITUTION.

PMODOREP TOOSEWELDy 9. oe ces ci. cen President of the United States.
Cuartes W. FarRBANKS.............-+: Vice-President of the United States.
Manayarmun We HOLGER, <2... 0. 2/5. 8s snes Chief Justice of the United States.
WhIsMU LOOT ey tae aca ce nese oe Secretary of State.

GHORGH Ds CORTERYOU. .1..0.52+2-en 6: Secretary of the Treasury.

Shiau TTT DET ADVAN cet cv. aie area eso otexare as Secretary of War.

GHarnnsi Je) BONAPARTE. - 2.20. - cles ee Attorney-General.

GMORGHEVON] 1, MinvERe.. 2.5... 560800 Postmaster-General.

Wriomton JEL, Whitey aria seed 6 eee meee oor Secretary of the Navy.

RACV Gwen GAR ETEDD se - 4 sec ane ss crams -< Secretary of the Interior.
SANTIMGMVVEIESON MG tac occc csintnclscces . « Secretary of Agriculture.

OSCAR MSM ORBAUSH iia wriels since t crs aise’ Secretary of Commerce and Labor.

Ix

Metvitte W.

IMMWIbE Re jk erke

Crarnms W. HAIRBANKS.......0-

SHetpy M. Cunnom.............

Henry Cason LopGe......... 0+.

A. O. Bacon

RicHarp OLNEY...............
Joun B. Henperson

ALEXANDER GRAHAM BELL....... Citizen of Washington City.

Grorce Gray

xX

.. Citizen of Massachusetts.

REGENTS.

Vice-President of the United Sites
Member of the Senate of the United St ;
Member of the Senate of the United Stat 5
Member of the Senate of the United St
Member of the House of Representatives.
Member of the House of Representatives.
Member of the House of Representatives. :
Citizen of Michigan. |

Citizen of New York.
Citizen of Washington City.

Citizen of Delaware.

CONTENTS.

Parts Met RevRolnaneree asta totre coh clislthas erie Gant Senee TOTEM Noe ATSC EE lil

ArticLE I (1438). A Comparison of the Features of the Earth and the Moon. By N. S.
SHALER. Published 1903. 4to, v, 130 pp., 25 plates.
ArtTicLe II (1459). On the Construction of a Silvered Glass Telescope, Fifteen and a Half
Inches in Aperture, and its Use in Celestial Photography. By Hernry
Draper, M. D. (Reprinted from Vol. XIV, “Smithsonian Contributions
to Knowledge,” 1864.) Published 1904. 4to, iii, 55 pp.
On the Modern Reflecting Telescope and the Making and Testing of Optical
Mirrors. By Grorce W. RircHey. Published 1904. 4to, vy, 51 pp., 13
plates.

Articie III (1651). Hopexins Funp. A Continuous Record of Atmospheric Nucleation.
By Cart Barus. Published 1905. 4to, xvi, 226 pp.

ARTICLE IV (1692). Glaciers of the Canadian Rockies and Selkirks. By Writutam H1'-
TELL SHeErzer, Ph. 1). Published 1907. 4to, xii, 135 pp., 42 plates.

XI

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE

PART OF VOLUME XXXIV

mc@vMPARISON OF THE FEATURES
PrtHeE EARTH AND THE MOON

BY

Wi SiS ABER

PROFESSOR, HARVARD UNIVERSITY

(No. 1438)

CITY OF WASHINGTON
PUBLISHED BY THE SMITHSONIAN INSTITUTION
1908

eye

1
'
7 .
‘
'
wr
4 i
* i
i"
™
I
‘
:
y
“
i
7 3
Ure
+

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE

PART OF VOLUME XXXIV

PaeOMPAKISON OF THE FEATURES
fof THE EARTH AND THE MOON

BY

ING Se le bsu blehIN

PROFESSOR, HARVARD UNIVERSITY

x Ss Lee

*, é ay ; 3 VY. 6 3
| Pe Ps WHS. US
Aais rae ares

See INGTON 0

(No. 1438)

CITY OF WASHINGTON
PUBLISHED BY THE SMITHSONIAN INSTITUTION
19038

Commission to whom this memoir has been referred :

GEORGE P. MERRILL,
Cc. G. ABBOT.

ADVERTISEMENT.

For more than twelve years past I have been preparing the material for the
publication of a work, on the part of the Smithsonian Institution, which it was
hoped would consist essentially of photographic views of the moon, so complete
and, it was expected (with the advance of photography), so minute, that the
features of our satellite might be studied in them by the geologist and the sele-
nographer, nearly as well as by the astronomer at the telescope. This hope has
only been partially fulfilled, for photography, which has made such eminent ad-
vances in the reproduction of nebulz and like celestial features, has indeed
progressed in lunar work, but not to the same extent as in other fields. The
expectation that such a complete work could be advantageously published for
this purpose has, then, been laid aside for the present.

It has been decided to draw from the material prepared for this larger work,

some photographs taken at the Lick Observatory and the Paris Observatory, and
P grap y y

particularly some recently obtained by Professor Ritchey at the Yerkes Observa-
tory, for which I have to express the thanks of the Institution. These illustra-
tions are attached to the present paper by Professor Shaler, and may, then, be
considered to be a separate contribution by the Institution to the study of
selenography.

Professor Shaler’s memoir gives the results of personal studies carried on for
a third of a century. He has devoted about one hundred nights to telescopic
study of the moon with the Mertz equatorial of Harvard College Observatory, his
later researches having been chiefly by means of photographs at Harvard Uni-
versity, with which he has so long been connected.

In accordance with the rule adopted by the Smithsonian Institution, the
memoir has been submitted for examination to a committee consisting of Dr.
George P. Merrill, Head Curator of Geology in the U. S. National Museum,
and Mr. C. G. Abbot of the Smithsonian Astrophysical Observatory.

S: P: LANGLEY:
SECRETARY.
Smithsonian Institution,
Washington, December, 1903.

iii

:

Plate ie
ie

Totals
IV.

V.

VI.
VII.
VIII.
IX.

Xe

XI.
XII.
XIII.
XIV.
XV.
XVI.
XVII.
XVIII.
XIX.
XX.
XXI.
XXII.
XXIII.
XXIV.
XXV.

EIST OF PRATES:

General View of Moon, Age 6 days.
Moon’s Age, 7 days.

Moon’s Age, 8 days, 4 hours.

Moon’s Age, 8 days, 22 hours.

Moon’s Age, ro days, 12 hours.

Moon’s Age, 14 days, 1 hour.

Moon's Age, 21 days, 5 hours.

Moon’s Age, 23 days, 7 hours.

Moon’s Age, 21 days, 16 hours.

Mare Crisium and Neighboring Parts.
Enlarged View of Part of Apennines.
Hyginus and the Neighboring Field.

Part of Moon Photographed with Yellow Screen and Isochromatic Plate.
Part of the Shore of the Oceanus Imbrium.
Central Portion of Moon from Mare Serenitatis to Stéfler.
Copernicus and Kepler.

Crater Region about Theophilus.

Mare Serenitatis.

Ray System about Tycho.

Copernicus and Surroundings.

Mare Nubium and Surroundings.

Mare Tranquilitatis and Surroundings.
Mare Imbrium and Surroundings.
Aristoteles, Eudoxus, and Surroundings.

Clavius, Longomontanus, Tycho, etc.

; es A it eee

PmecCOMPARISON OF THE FEATURES OF
fe EARTH AND: THE MOON,

By N. S. SHALER.

PRELIMINARY NOTE.

The object of this papér is to set forth the general results of certain
studies concerning the form and structure of the lunar surface with reference to
various terrestrial problems.

These studies were begun in 1867 with the Mertz equatorial of the Harvard
College Observatory, at the time when my lamented friend and colleague, Joseph
Winlock, was director, and have been continued in a desultory manner, from time
to time, for a third of a century. Between 1867 and 1872 about one hundred
nights were devoted to telescopic work ; since that time what has been done has
been almost altogether by means of photographs, which have of recent years
become much more convenient and for my purpose more serviceable than the
opportunities afforded by an instrument even if it were as good as the Harvard
Mertz.

It should be observed that so far as possible my task has been kept apart
from problems of selenology or selenography strictly so called. The ends sought
have been those alone which had distinct reference to geology. Certain ques-
tions, as, for instance, that concerning the antiquity of the lunar surface, necessa-
rily touch upon matters which relate to the history of the moon as an individual
sphere. In fact almost all the questions brought up by studies on the satellite
are more or less entangled with those relating to the evolution of the planet, so
that except for the detailed account of the features of either body they must needs
be considered together. These features may be compared by types, and in the
main the following essay consists of such comparisons.

If other duties permit I hope to present the matters discussed in the follow-
ing pages in a more extended form, one in which it will be possible to illustrate
the facts here set forth, as well as to discuss the conclusions attained in an ampler
manner. Almost all the points I have endeavored to make clear demand this

I

2 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

more extended treatment. As many of them are debatable, some of them,
indeed, requiring more observations and comparisons than I have been able to
give, I may hope that the criticism this paper may receive will enable me to better
the work which it describes.

GENERAL DESCRIPTION OF THE MOON.

Although the moon has been the most studied of all celestial objects, few
persons, except astronomers, have a clear idea of even the general results which
have been derived from the vast body of observations that have been made upon
it. On this account it appears desirable to preface the account of the special
inquiries which are set forth in the following pages by a statement of what is
known concerning this nearest neighbor of our earth. This account will necessa-
rily be limited to the facts which can be set forth in other than mathematical
form; fortunately, these include all that the reader needs to have in mind in order
to obtain a fairly clear understanding of the questions which are to be discussed.

The history of primitive astronomy shows that the moon, of all celestial
objects, from the beginning of man’s intellectual development has been the most
closely observed. Although the sun was doubtless recognized by the lowliest
man as the most important feature of the heavens, as the giver of life, the condi-
tions under which it is seen, especially its blinding light, long made any extended
study of it impossible. So, except for the very evident changes of its course
across the sky and the consequent succession of the seasons, little was known of
the solar center two hundred years ago, and, save its approximate distance from
the earth, its mass, and its general relations to the planets, not much knowledge
was gained until the last century. On the other hand, the moon, because of its
nearness, being only about one four-hundredth part as remote from the earth as
the sun, has in a noteworthy way entered into the records of men. Its relatively
short period of change and the very pronounced character of its alterations made
it the first index of time beyond the round of the day. It is evident, indeed, that
as soon as men began to reckon time they used the lunar month to make their
tally, rather than that of the solar year. Moreover, the surface of the moon reveals
much to the naked eye, not clearly, but sufficiently well to afford the basis for
speculation and to tempt the imagination to create there a world like our own.
It is therefore not surprising that a host of myths concerning the nature of our
satellite grew up in the days before the telescope. It is interesting to note the
fact that many of these myths have not only become fixed in the minds of unin-
structed people, but they have had a remarkable influence upon the minds of
modern astronomers, limiting their capacity to interpret what their instruments
clearly reveal to them. At every stage in the advance of selenography we note
the curious persistency of the endeavor not only to interpret the lunar features
by the terrestrial, but to warp the observed facts into accord with those seen on
the earth. There is perhaps no better instance of the extent to which prepos-
sessions and prejudices may affect the judgment of the most conscientious ob-
server, blinding him to evident truth, than the history of lunar inquiries affords.

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 3

The story of the physical conditions of the moon had best be begun by
noting that the relation of our satellite to a larger sphere is not exceptional, but
the most characteristic of all the relations of one stellar body to another. Of the
planets in the solar system, all save the two nearest to the sun, Mercury and
Venus, have one or more smaller spheres circling about them. The relation of
the sun to the several planets in a larger way repeats this plan of grouping lesser
about greater orbs.

It is generally believed by astronomers that the celestial spheres have been
formed by a process of condensation, due to gravitation, of matter which was
originally widely diffused ; that our solar system, before it was organized into the
sun and lesser bodies, was in the form of a diffused nebulous mass of spheroidal
form which extended beyond the orbit of the outermost planet. As this matter
gathered towards the center, the material now in each of the planets and its satel-
lites parted from the parent body, probably at first in the form of a nebulous
ring, or spiral, which in time broke and gathered into a spheroidal mass. In
that detached portion of the parent nebula the process of concentration was
repeated, with the result that satellites, or, as we may term them, secondary
planets, were formed substantially as the greater spheres were set off from the
sun. There are many questions and doubts concerning the details of this nebular
theory, but that the evolution of our solar system and probably of all stellar sys-
tems took place in substantially the manner indicated appears to be eminently
probable; it is, indeed, fairly well established by what we know of the distant
nebulz and by the rings of Saturn, which apparently contain the material which
normally should have formed one or more of its satellites, but which for some
unknown reason have remained unbroken.

It is not certain at just what stage in the concentration of a nebula a planet
or a satellite may be set off from the parent body; nor can the present distance
of the satellite from the main sphere be assumed as that at which the parting took
place. It is possible that the concentration of the parent body had gone so far
that the diffused or nebulous stage of its materials had been passed by and the
more advanced stage of igneous fluidity entered on. It is, however, more likely
that in all cases the separation occurred while the particles of matter were di-
vided as they are in a gas or vapor. As soon as the two spheres are separated
from one another, and so long as they remain in any measure fluid, the difference
in their gravitative attraction on the nearer and more remote part of their masses
induces tides, and the effect of these tidal movements, as has been shown by
Professor George Darwin, is necessarily to impel the two bodies farther apart.
It seems certain that before the earth and the moon became essentially rigid, as
they now are, the effect of these tides in driving them apart must have been
great enough to account for a considerable part of the interval which now
separates them.

In the present condition of the moon, it is a sphere having a computed
diameter of 2159.6 miles and its mean distance from the earth 238,818 miles. So
far as has been determined, the moon exhibits no trace of flattening at the poles

4 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON,

such as characterizes the earth, unless, as is possible, there are irregularities of
figure on the unseen part of the sphere. It is essentially globular inform. The
fact that the moon is not flattened at its poles probably indicates that if it
once rotated in the manner of the planet it ceased to do so before it became
solid.

The measure of density of the moon—z. ¢., the proportion of its weight to its
bulk—is only about six-tenths that of the earth, While the earth’s mean density
is nearly 5.7 times that of water, that of the moon is about 3.5 times as great.
Thus the total gravitative force of the lunar mass is to be reckoned as only
about ,, that of our planet.

Asthe moon revolves on its polar axis but once in about a month, and at a rate
that tends to keep the same part of its surface turned towards the earth, we should,
but for the phenomenon of librations, see no more than one-half of its superficial
area. Owing, however, to this feature, which is due to certain complications of
the moon’s exceedingly varied movements, the satellite in effect sways in relation
to the earth so that at certain times we see farther to the east and at others farther
to the west of its center, and in the succession of these movements we are able to
behold somewhat more than one-half the total area, in fact about six-tenths of it.
It is impossible to set forth in this writing the reasons for the librations of the
moon, as the matter cannot be explained without giving in mathematical form
a full account of the motion of our satellite, which is one of the most compli-
cated of astronomical problems. An excellent non-mathematical presentation of
the question, which affords a sufficient idea of it, may be found in Zhe Moon, by
Richard A. Proctor, pp. 117 e¢ seg., D. Appleton & Co., New York, 1878.

As noted below, there is some accessible information going to show that even
beyond the extreme field revealed by the librations the surface of the moon has
the same character as that which is visible. Thus we find that up to the limits
of the visible part there is no sign of change in the nature of the surface. It
is therefore reasonable to conclude that the same characteristics extend for some
distance beyond the limits of vision. We also note on the verge of the unseen
field the hither margins of certain ring-shaped structures, evidently of large size,
the so-called volcanoes, so that it is fair to conclude that these features are con-
tinued on the unseen part. Moreover, there are some light-colored bands, such
as on this side of the moon always radiate from crater-like pits, which apparently
come over from such centers on the unseen part. These several facts, taken to-
gether, make it eminently probable that the unseen four-tenths of the lunar sur-
face in no essential way differs from that we observe. It is, indeed, altogether
likely that we see every type of structure that exists on the moon, and that a
view of its whole area would add nothing essentially new to our knowledge of
the sphere.

Seen by persons of ordinarily good vision, even at a distance of about 240,000
miles, the moon reveals much of its surface shape, structure, and color ; it is evi-
dent that the color varies greatly from very bright areas to those which are rela-
tively dark, that the latter are somewhat less in total extent than the former

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 5

and that they are disposed in a general way across the northern hemisphere.!
(See plates 1. to vir. inclusive.) Persons of more than usually good vision may,
under favorable conditions, see on the edge of the illuminated area the ragged
line of the sunlight, which indicates that the surface is very irregular, the high
points coming into the day before the lower are illuminated. Such persons at
time of full moon can also note, though faintly, some of the bright bands which,
radiating from certain crater-like pits, extend for great distances over the surface.
So, too, they may see at the first stage of the new and the last of the old moon,
the light from the sunlit earth slightly illuminating the dark part of the lunar
sphere, or, as it is often termed, the old moon in the arms of the new.

With the best modern telescopes under the most suitable conditions of
observation, the moon is seen as it would be by the unaided eye if it were not more
than about forty miles from the observer. The conditions of this seeing are much
more favorable than those under which we behold a range of terrestrial moun-
tains at that distance, for the reason that the air, and especially the moisture, in
our atmosphere hinders and confuses the light, and there is several times as much
of this obstruction encountered in a distance of forty miles along the earth’s
surface as there is in looking vertically upwards.

Seen with the greater telescopes, the surface of the moon may reveal to able
observers, in the rare moments of the best seeing, circular objects, such as pits,
which are perhaps not more than five hundred feet in diameter. Elevations of
much less height may be detected by their shadows, which, because there is no
trace of an atmosphere on the moon, are extraordinarily sharp, the line between
the dark and light being as distinct as though drawn by a ruler. Elongate
objects, such as rifts or crevices in the surface, because of their length, may be
visible even when they are only a few score feet in width, for the same reason that
while a black dot on a wall may not make any impression on the eye, a line no
wider than the dot can be readily perceived. Owing to these conditions, the
surface of the moon has revealed many of its features to us, perhaps about as well
as we could discern them by the naked eye if the sphere were no more than
twenty miles away.

Separated from all theories and prepossessions, the most important points
which have been ascertained as to the condition of the moon’s surface are as
follows :

The surface differs from that of the earth in the fact that it lacks the envel-
opes of air and water. That there is no air is indicated by the feature above
noted, that there is no diffusion of the sunlight, the shadows being absolutely black
and with perfectly clean-cut edges. It is also shown by the fact that when a star
is occulted or shut out by the disc of the moon it disappears suddenly without its
light being displaced, as it would be by refraction if there were any sensible

‘It is well to note the fact that in a celestial telescope objects are seen in reverse position, or
“upside down.”’ For convenience they are usually so depicted on maps and pictures of the moon;
the north pole at the bottom, and the east where it is customary to place the west on terrestrial
maps.

6 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON,

amount of air in the line of its rays. This evidence affords proof that if there is
any air at all on the moon’s surface it is probably less in amount than remains in
the nearest approach to a vacuum we can produce by means of an air-pump.
Like proof of the airless nature of the moon is afforded by the spectroscope
applied to the study of the light of an occulting star or that of the sun as it is
becoming eclipsed by the moon. In fact a great body of evidence goes to show
that there is no air whatever on the lunar surface.

The evidence of lack of water at the present time on the surface of the
moon appears to be as complete as that which shows the lack of an atmosphere.
In the first place, there are evidently no seas or even lakes of discernible size.
There are clearly no rivers. If such features existed, the reflection of the sun
from their surfaces would make them exceedingly conspicuous on the dark back-
ground of the moon, which for all its apparent brightness is really as dark as the
more somber-hued rocks of the earth’s surface when lit by the sun. Moreover,
even were water present, without an atmosphere there could be no such circula-
tion as takes place on the earth, upward to clouds and thence downward by the
rain and streams to the ocean. Clouds cannot exist unless there be an atmos-
phere in which they can float, and even if there be an air of exceeding tenuity on
the moon, it is surely insufficient to support a trace of clouds. Some distin-
guished astronomers have thought to discern something floating of a cloud-like
nature, but these observations, though exceedingly interesting, are not sufficiently
verified to have much weight against the body of well-observed facts that shows
the moon to be essentially waterless.

The well-established absence of both air and water in any such quantities as
are necessary to maintain organic life appears to exclude the possibility of there
being any such life as that of plants and animals on the lunar surface. The reader
will find below a further discussion of this question, and it may therefore here be
passed with the statement that very few astronomers are now inclined to believe
that the moon can possibly be the abode of living forms.

Being without an effective atmosphere, for the possible but unproved rem-
nant that may exist there would be quite ineffective, the moon lacks the defense
against radiation of heat which the air affords the earth. Therefore in the long
lunar night the outflow of heat must bring the temperature of the darkened part
to near that of the celestial spaces, certainly to some hundred degrees below
Fahrenheit zero. Even in the long day this lack of air and consequent easy
radiation must prevent any considerable warming of the surface. The temper-
ature of the moon has been made the matter of numerous experiments. These,
for various reasons, have not proved very effective. The most trustworthy, the
series undertaken by S. P. Langley, indicate that at no time does the heat attain
to that of melting ice.

Turning now to the shape and structure of the moon’s crust, we observe that
it differs much from that of the earth. Considering first the more general features,
we note that there are none of those broad ridges and furrows,—the continents
and the sea basins. A portion of the surface, mainly in the northern hemisphere,

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 7

is occupied by broad plains which in their general shape are more nearly level
than any equally extensive areas of the land, or, so far as we know, of the ocean
floor of the earth, though they are beset with very many slight irregularities.
These areas of rough, dark-hued plains are the seas or mara of selenographers,
so termed because of old they were, from their relatively level nature, supposed
to be areas of water. These maria occupy about one-third of the visible sur-
face. Their height is somewhat less than that of the crust outside of their area.
The remaining portion of the moon is extremely rugged. It is evident that the
average declivity of the slopes is far greater than on the earth. This is apparent
in all the features made visible by the telescope, and it likely extends to others
too minute to be seen by the most powerful instruments. Zéllner, by a very
ingenious computation based on the amount of sunlight reflected, estimates
that the average angle of the lunar surface to its horizon is fifty-two degrees.
Though we have no such basis for reckoning the average slope of the lands and
sea bottoms of the earth, it is eminently probable that it does not amount to more
than a tenth of that declivity. This difference, as well as many others, is prob-
ably due to the lack on the moon of the work of water, which so effectively
breaks down the steeps of the earth, tending ever to bring the surface to a
uniform level.

The most notable feature on the lunar surface is the existence of exceed-
ingly numerous pits, generally with ring-like walls about them, which slope very
steeply to a central cavity and more gently towards the surrounding country.
These pits vary greatly in size; the largest are more than a hundred miles in
diameter, while the smallest discernible are less than a half-mile across. The num-
ber increases as the size diminishes; there are many thousands of them, so small
that they are revealed only when sought for with the most powerful telescopes
and with the best seeing. In all these pits, except those of the smallest size, and
possibly in these also, there is within the ring-wall and at a considerable though
variable depth below its summit a nearly flat floor, which often has a central pit
of small size or in its place a steep rude cone. When this plain is more than
twenty miles in diameter, and with increasing numbers as the floor is wider, there
are generally other irregularly scattered pits and cones. Thus in the case of Plato,
a ring about sixty miles in diameter, there are some scores of these lesser pits.
On the interior of the ring-walls of the pits over ten miles in diameter there are
usually more or less distinct terraces, which suggest, if they do not clearly indi-
cate, that the material now forming the solid floors they enclose was once fluid
and stood at greater heights in the pit than that at which it became permanently
frozen. It is, indeed, tolerably certain that the last movement of this material of
the floors was one of interrupted subsidence from an originally greater elevation
on the outside of the ring-wall, which is commonly of irregular height with many
peaks. There are sometimes tongues or protrusions of the substance which
forms the ring, as if it had flowed a short distance and then had cooled with steep
slopes.

The foregoing account of the pits on the lunar surface suggests to the

8 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

reader that these features are volcanoes. That view of their nature was taken
by the astronomers who first saw them with the telescope and has been generally
held by their successors. That they are in some way, and rather nearly, related
to the volcanic vents of the earth appears certain. The nature of this relation is
discussed below. We have now to note the following peculiar conditions of
these pits. First, that they exist in varying proportion, with no evident law of
distribution, all over the visible area of the moon. Next, that in many instances
they intersect each other, showing that they were not all formed at the same
time but in succession ; that the larger of them are not found on the maria but on
the upland and apparently the older parts of the surface; and that the evidence
from the intersections clearly shows that the greater of these structures are pre-
vailingly the elder and that in general the smallest were the latest formed. In
other words, whatever was the nature of the action involved in the production
of these curious structures, its energy diminished with time, until in the end it
could no longer break the crust.

All over the surface of the moon, outside of the maria, in the regions not
occupied by the volcano-like structures, we find an exceedingly irregular surface,
consisting usually of rude excrescences with no distinct arrangement, which may
attain the height of many thousand feet. These, when large, have been termed
mountains, though they are very unlike any on the earth in their lack of the
features due to erosion, as well as in the general absence of order in their associa-
tion. Elevations of this steep, lumpy form are common on all parts of the moon.
Outside of the maria they are seen at their best in the region near the north
pole, where a large field thus beset is termed the Alps. From the largest of
these elevations a series of like forms can be made of smaller and smaller size
until they become too minute to be revealed by the telescope; as they decrease
in height they tend to become more regular in shape, very often taking on a
dome-like aspect. The only terrestrial elevations at all resembling these lunar
reliefs are certain rarely occurring masses of trachytic lava, which appear to have
been spewed out through crevices in a semi-fluid state, and to have been so
rapidly hardened in cooling that the slopes of the solidified rock remained very
steep. As noted in more detail below, the only reliefs on the moon’s surface
that remind the geologist of true mountains are certain low ridges on the surfaces
of the maria.

The surface of the moon exhibits a very great number of fissures or rents,
which when widely open are termed valleys, and when narrow, rills. Both these
names were given because these grooves were supposed to have been the result
of erosion due to flowing water. The valleys are frequently broad, in the case of
that known as the Alpine valley, at certain places several miles in width: they
are steep-walled and sometimes a mile or more in depth; their bottoms, when
distinctly visible, are seen to be beset with crater-like pits, and show in no in-
stance a trace of water work which necessarily excavates smooth descending
floors such as we find in terrestrial valleys. The rills are narrow crevices, often
so narrow that their bottoms cannot be seen; they frequently branch and in some

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 9

instances are continued as branching cracks for a hundred miles or more. The
characteristic rills are far more abundant than the valleys, there being many
scores already described ; the slighter are evidently the more numerous: a cata-
logue of those visible in the best telescopes would probably amount to several
thousand. (See plates xu, xx1, and xxm.)

It is a noteworthy fact that in the case of the rills and in great measure also
in the valleys the two sides of the fissure correspond so that if brought together
the rent would be closed. This indicates that they are essentially cracks which
have opened by their walls drawing apart. Curiously enough, as compared with
rents in the earth’s crust there is little trace of a change of level of the two sides
of these rills—only in one instance is there such a displacement well made out,
that known as the Strait Wall, where one side of the break is several hundred
feet above the other. (See plate xxz.)

In the region outside of the maria much of the general surface of the moon
between the numerous crater-like openings appears in the best seeing with power-
ful telescopes to be beset with minute pits, often so close together that their
limits are so far confused that it appears as honeycombed, or rather as a mass of
furnace slag full of holes if greatly magnified, through which the gases developed
in melting the mass escaped. (See plates 1x, x11.)

Perhaps the most exceptional feature of the lunar surface, as compared with
that of the earth, is found in the numerous systems of radiating light bands, in
all about thirty in number, which diverge from patches of the same hue about
certain of the crater-like pits. These bands of light-colored material are gen-
erally narrow, not more than a few miles in width; they extend for great dis-
tances, certain of them being over a thousand miles in length, one of them
attaining to one thousand seven hundred miles in linear extent. In one instance
at least, in the crater named Saussure, a band which intersects the pit may be
seen crossing its floor, and less distinctly, yet clearly enough, it appears on the
steep inside walls of the cavity. In no well-observed case do these radiating
streaks of light-colored material coincide with the before-mentioned splits or
rifts. Yet the assemblage of facts, though the observations and the theories
based upon them are very discrepant, lead us to believe that they are in the
nature of stains or sheets of matter on the surface of the sphere, or perhaps in
the mass of the crust. At some points the rays of one system cross those of
another in a manner that indicates that the one is of later formation than the
other. (See plates v1, xvi, and xrx.)

Perhaps the most puzzling feature of the radiating streaks, where everything
is perplexing, is found in the way they come into view and disappear in each
lunar period. When the surface is illuminated by the very oblique rays of the
sun they are quite invisible; as the lunar day advances they become faintly
discernible, but are only seen in perfect clearness near the full moon. The
reason for this peculiar appearance of these light bands under a high sun has
been a matter of much conjecture; it is the subject of discussion in a later
chapter of this memoir, where it is shown that inasmuch as these bands appear

10 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

when the earth light falls upon the moon at a high angle, the effect must be due
to the angle of incidence of the rays on the shining surfaces. It should be noted
that the light bands in most instances diverge from more or less broad fields of
light color about the crater-like pits, fields which have the same habit of glowing
under a high illumination; in fact, a large part of the surface of the moon, per-
haps near one-tenth of its visible area, becomes thus brilliant at full moon,
though it lacks that quality at the earlier and later stages of the lunar day.

In the above considered statement concerning the visible phenomena of the
moon no account is taken of a great variety of obscure features which, though
easily seen with fairly good instruments, have received slight attention from
selenographers. As can readily be imagined, observers find it difficult to discern
obscure features which cannot be classed in any group of terrestrial objects.
Whosoever will narrowly inspect any part of the lunar surface, noting every-
thing that meets his eye, will find that he observes much that cannot be
explained by what is seen on the earth. It is evident, indeed, that while in the
earlier stages of development this satellite in good part followed the series of
changes undergone by its planet, there came a stage in which it ceased to con-
tinue the process of evolution that the parent body has undergone; the reason
for this arrest in development appears to have been the essential if not complete
absence of an atmosphere and of water.

The difference in height between the lowest and highest points on the lunar
surface is not determined. To the most accented reliefs, those of the higher
crater walls, elevations of more than twenty-five thousand feet have been assigned ;
it is, however, to be noted that all these determinations are made from the length
of the shadows cast by the eminences, with no effective means of correcting for
certain errors incidental to this method. It may be assumed as tolerably certain
that a number of these elevations have their summits at least twenty thousand
feet above their bases and that a few are yet higher. We do not know how
much lower than the ground about these elevations are the lowest parts of
the moon. My own observations incline me to the opinion that the difference
may well amount to as much as ten thousand feet, so that the total relief of the
moon may amount to somewhere between thirty and forty thousand feet. That
of the earth from the deepest part of the oceans to the highest mountain summits
is probably between fifty-five and sixty thousand feet; so that notwithstanding
the lack of erosion and sedimentation which in the earth continually tends to
diminish the difference between the sea-floor and land areas, the surface of the
satellite has a much less range of elevation than the planet. If the forces which
have built the mountains and continents of the earth had operated without the
erosive action of water there is little doubt that the difference in height between
the highest and lowest parts would now be many times as great as it is on the moon.

AGE OF THE EXISTING LUNAR SURFACE,

Several of the most important problems to be considered in this writing in-
timately depend on a determination of the age of the moon’s surface. If we

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. II

accept the commonly adopted view as to the nature of the prevailing topo-
graphical features of that sphere and regard them as essentially volcanic, z. ¢., as
due mainly to the expulsion of heated vapors or gases from the interior of the
sphere, we have a basis on which to found a determination of that age sufficiently
accurate to serve our immediate purpose.

It appears eminently probable that the lunar surface must have attained to
something like its present condition long before the earth came to the state in
which its igneously fluid mass was crusted over. And this for the following
reasons: At the time when the material of the moon and earth separated
from the previously united mass we have to believe that the amount of heat
they severally contained was in general proportionate to the mass of each
body. Now the mass of the moon is to that of the earth as one to eighty, and
its diameter about as one to four. From this, by the well-known law of cooling
bodies, it follows that the moon must have acquired a permanent rigid crust, if
indeed it did not become entirely frozen, long before the earth ceased to have a
molten surface. There are too many doubtful elements in the computation to
make any seemingly accurate reckoning trustworthy, but it appears altogether
likely that the moon cooled far beyond the point where volcanic action was pos-
sible ages before the earth’s surface could have frozen or perhaps have passed
from the gaseous to the fluid state.

At present all the volcanic action of the earth is apparently limited to the
sea-floor or regions within three hundred miles of the shore ; effectively to regions
where the central heat is brought upwards into strata containing water laid in
them when they were deposited; the rise of the heat being due to the slow
conductivity of the imposed beds. There is reason to believe that since the
earliest recorded ages the earth has mainly, if not altogether, depended on such
action for the volcanic outbreaks which have occurred upon it. While there may
in this particular matter be some reason for doubt, there is none as to the fact
that if the so-called lunar volcanoes are due to the central heat of that sphere,
they must have been shaped before the crust of the earth was formed, or long
before the earliest geological records. It has, however, been suggested by G. Kk.
Gilbert’ and others that what appear to be volcanoes on the moon are not really
such, but are, in effect, punctures caused by the falling of large meteorites or
bolides. This interesting suggestion commends itself at first sight as a possible
explanation of the pits on the moon, structures which differ in many regards from
those due to terrestrial volcanic action, in that they are often of much greater diam-
eter, have relatively much smaller encircling cones, and show little, if any, clear
evidence of lava flows, or ash showers, proceeding from them. As I propose fur-
ther on in this paper to discuss the question of their nature in more detail, I shall
now give only in brief the reasons why, as it seems to me, the hypothesis that
they were caused by bodies falling from the sky is not verified.

It is to be noted that these so-called volcanoes of the moon, vulcanoids, as I
shall term them, have generally very steep walls around their crater-like pits ;

‘See Bull. Phil. Soc. of Washington, vol. 12, p. 241, et seg.

2 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

the average outer slope, according to my estimates, exceeding forty degrees, the
inner slope being generally somewhat steeper. On this hypothesis this inner slope
must mark the path of the impinging bolide, and the cone that surrounds it be the
result of the outthrusting action of that body, such as we note when a pebble is
thrown into soft clay or a shot from a cannon enters an armor plate. We have
under Gilbert’s hypothesis to suppose that the impinging bodies came into con-
tact with the moon at something like planetary velocity. Such bodies having a
diameter of even a mile—and some of them must, on this hypothesis, have been of
fifty or more miles diameter—would, by the conversion of their momentum into
heat, have served to melt a wide field of the crust about their points of contact.’
As there is no trace of any such bolides in the bottoms of these craters, but com-
monly a floor, as of hardened lava, we have to suppose that they penetrated to a
great depth and that the lava flowed up after their entrance. But the necessary
effect of the entrance of a mass sufficiently large to have punctured these open-
ings would, if they had penetrated to a molten zone, have been to send up a
quantity of lava far more than sufficient to fill the opening they made, while in
fact with few, if any, exceptions, this lava appears at no time to have risen to the
general level of the surrounding rampart. Furthermore, if the cones about the
craters were due to outthrusts caused by such impacts on material stiff enough to
maintain the steep walls of the crater, then we should have evidence of radial
cracking in the form of open rents, such as would inevitably be developed under
the assumed conditions, but have evidently not produced in far the greater
number of the vulcanoids.

There is another and, taken alone, conclusive argument against the suppo-
sition that the lunar craters are due to the impact of bolides; this is found in the
facts presented in the series which may be traced in the sizes and distribution of
the fractures which it seeks to explain. As regards their sizes, the pits grade
from the smallest that can be discerned by the most powerful telescope, probably
not over five hundred feet in diameter, to rings that are one hundred miles across.
The steepness of the inner slopes of these cavities does not perceptibly differ, nor
is there more evidence of lava having been poured out from the larger than from
the smaller craters. Moreover, there is no better evidence of radiating fractures
in the case of the larger than in the smaller pits. Furthermore, there is no
such relation in the masses of material composing the enveloping cones or rings
as we would expect to find if they were due to the impact of bodies varying in
size as we have to suppose. In many instances the walls of a pit scores of miles
in diameter are no thicker or higher than in the case of other pits less than a
mile across.

As regards their distribution, the craters of the moon are generally placed
in such apparent lack of order as to give some warrant for the hypothesis that

* Assuming that the impinging body came upon the surface of the moon at planetary velocity,
and that all the resulting heat was applied to its mass, the resulting temperature would exceed,

according to my reckoning, 150,000 degrees. A bolide fifty miles in diameter would be likely to
melt an area many times its diameter.

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 13

they must owe their origin to other than volcanic action, for on the earth we find
volcanoes very generally disposed along lines which, in most if not all cases,
appear to be determined by faults. In many instances, however, the lunar vul-
canoids have a linear arrangement.

The vulcanoids of larger size which are arranged in linear order are not
numerous. Among these may be cited the train extending from Herschel
through Ptolemzus, and Alphonsus to Arzachel; that from Thibet to Stofler ;
that from Atlas to Franklin; and that from Vendalinus to Casatus, near the
limb in the third quadrant. (See plates 1 and xx1.) In all these instances
there are four or more pits in fairly true alignment : in alignment and in number
they appear to exclude the supposition that their order is due to chance. Pass-
ing from the examples in which the greater vulcanoids are grouped in trains
and taking the pits of smaller size, we find the instances of such arrangement be-
coming more numerous as the structures are of smaller diameter. It is, however,
in but few of the pits over ten miles in diameter that there are more than three
or four so placed in relation to one another that they can be said to be linearly
arranged.

When, in following down the series of vulcanoids as regards size, we come
to the pits less than a mile in diameter, those commonly termed craterlets, we
note that the linear order, hitherto exceptional, becomes so common that the
exceptions are rather to be found in the departures from it. The observations of
W. H. Pickering and others, as will be noted below, make it evident that there
is a causal relation between the smaller visible pits and the cracks that form on
the surface of the moon. There can be no question that there are thousands of
these smaller of the craterlets which are thus disposed in lines, some of the
series extending for hundreds of miles. (See plate xx.)

It may be taken as evident, that in the time when the larger vulcanoids were
in process of formation the conditions of strain in the moon’s crust were not such
as to determine that the points of outbreak should to any great extent be linearly
arranged and that when thus arranged they tended to follow the meridians, rather
than the parallels. In the later stages of the surface when the smaller openings
were made they obviously tended to a linear order, but the direction of the lines
was exceedingly varied, some of them being radially disposed with the greater
vulcanoids as centers, others along lines of weakness which lie in extremely
diverse positions.

Reckoning great and small, there are some hundreds of these lines of pits, a
number sufficient to make it evident that they cannot be accounted for by chance.
It is evident that to explain this linear order of vulcanoids by the hypothesis we
are considering is difficult if not impossible, for that would require us to sup-
pose the bolides to have been thus arranged during their movements through
space. It is also to be noted that in very many instances there are pits within
the larger cavities so centrally placed that they cannot be explained by the chance
in-falling of bolides. Therefore, while the relation of lunar volcanoes to those of
the earth is a perplexing question, there seem on the face of the facts to be

14 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

sufficient reasons for rejecting the suggestion that they are due to the impact of
falling bodies.

In addition to the features of the lunar volcanoes there is another though
more remote reason why such falls of celestial bodies on the moon’s surface have
not occurred. Of these we may here mention two; these are as follows: It is
evident that these vulcanoids were formed at successive times, and under some-
what diverse conditions. So far as I have been able to determine, the largest
were, at least in a general way, first produced, and the smaller, approximately, in
the order of diminishing size, the smallest in most instances being formed last.
Now, as will be more particularly noted hereafter, the light bands which radiate
from certain craters and which are clearly mere strips of material which at full
moon reflect the sun’s light more intensely than the general surface have evidently
not been covered by deposits of ordinary meteoric matter, such as falls on the
earth in considerable quantity. It thus appears that for some reason the moon,
provided its surface has anything like the antiquity it appears necessary to assign
to it, has not been the seat of such deposits; for the accumulation of a small
amount of meteoric matter would mask such stains. We would thus, according
to the Gilbert hypothesis, have to suppose a succession of showers, each sending
bolides of smaller size than the preceding, and with them no considerable amount
of ordinary finely divided meteoric material such as comes to the earth,

It is also to be noted that since the earth’s surface came to its present state
there is good reason to believe that no such falls of large bodies as are supposed
by the bolide hypothesis to have fallen upon the satellite have ever come to the
planet. There are no traces of like craters, for even the greatest calderas, such
as that which holds Lago Bolsena or Kilauea, are evidently volcanic and in no
way related to meteoric action. Moreover, the fall of a bolide of even ten miles
in diameter would, by the inevitable development of heat due to its arrest, have
been sufficient to destroy the organic life of the earth, yet this life has evidently
been continued without interruption since before the Cambrian time. The point
to be last noted is that so far as I have been able to determine from an extended
inspection of lunar craters, including several hundred of the more determinable,
they all have the axes of their pits at right angles to the surface. Now if these
pits had been formed by bolides encountering the moon in their movement, that
movement necessarily being at planetary velocity, it does not seem possible that
they could all have come upon the sphere in a path normal to its surface. Even
with the resistance of the earth’s atmosphere, which is far denser than that of the
moon ever could have been, the small meteors which enter it mostly come at high
angles to the surface of the planet, although its attractive power is more than
eighty times as great as that of the satellite. It seems, indeed, incredible that if
the lunar vulcanoids were due to bolides they should not have fallen in some-
what greater numbers on the earth because of its greater gravitative attraction.
The number received would probably be nearly in proportion to the area of the
two spheres, with a slight preponderance in the number falling on the earth
because of its greater mass and consequently the greater effect of its gravity. It

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 15

is, however, as before remarked, evident that no such falls as have formed the
hundreds of pits over ten miles in diameter which exist on the moon's surface
have occurred on the earth since the Cambrian age.

The foregoing considerations justify us in rejecting the hypothesis of falling
bolides as a means of accounting for the so-called craters on the moon. There
are, however, certain other features of lunar surface which may be explicable by
the impact of large bodies falling from space. These we will now proceed to
consider.

MARIA OR SEAS.

A large part of the surface of the moon is occupied by the so-called marza or
seas. These are extensive irregular, indistinctly circular areas of relatively level
nature and of a perceptibly darker hue than the other more rugged fields. This
dark hue is shared by the floors of a number of the craters which lie near the
seas, as for instance by that of Plato, and more rarely by craters which lie remote
from their margins. Though vulcanoids exist on the maria of the moon they are
of relatively small size, none, in my opinion, which have clearly been formed
since the material of which the maria are composed came to its present level posi-
tion, exceeding ten or fifteen miles in diameter. So far as I have been able to
reckon, the proportion of these pits on the seas does not exceed one-fifth that we
find on the other part of the lunar surface. The average discernible inclination
of the surface of the maria is relatively so small they are more nearly true plains
than any equally extensive land areas on the earth.

It is a noteworthy fact that the maria, though they occupy about one-
third of the visible part of the moon, 2. e., including what is shown by the libra-
tions, rarely, if at all, lie on the margin, in positions enabling us to infer that they
are parts of like areas on the unseen portion of the lunar surface. On the western
limb of the sphere the so-called mare Australis is generally mapped as extending
around the margin, as it in fact does at certain stages of the libration, but under
the most favorable conditions the ordinary rough surface of the satellite appears
to me to be visible beyond this small mare, so that the statement as to none of
these seas crossing the limb apparently does not admit of exception. The ill-
named mare Humboldtianum is evidently a vulcanoid. It therefore appears
probable that if such maria exist on the unseen portion they are less extensive
than on the part of the orb which we see.

The most interesting feature of the maria is found in their contact with the
higher, rougher surface areas which bound them. Whenever I have been able
to observe this contact in a sufficiently exact manner there appears to be good
evidence that the material of which their surfaces are formed flowed in against
or upon the rough ground as very liquid lava would do. Ina general way this
fact had been often noted. It fills in the lower ground forming numerous bays.
In many instances, as, for example, in the case of Doppelmeyer, it distinctly ap-
pears to have melted down the side of the crater’s wall next to it, and to have
filled the cavity to its own level. Whoever will inspect these lines of contact of

16 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

the maria with the higher parts of the moon throughout the several thousand miles
of their extent will probably come to the conclusion that they were formed by
the once fluid matter of the sea inundating firm land. Assuming, as I shall do,
that these maria are made up of vast bodies of lava, which came upon the surface
after the greater vulcanoids were made and, as we shall hereafter see, after some
of the radiating light streaks were formed, how shall we account for the produc-
tion of such bodies of igneous material? The quantity of this matter was
evidently very great and in each of the seas it seems to have appeared all at
once, there being no mark of successive flows such as compose the extensive
lava fields of the earth. So far I have not been able clearly to trace any signs
of contact or over-lapping of the lava of the several maria. The search is, how-
ever, difficult; no more has been ascertained than that the material must have
been extremely fluid, far beyond what is seen in ordinary terrestrial flows. This
is shown by the fact that although gravitative attraction is only one-sixth what
it is on the earth, there is no steep face at the front of the fields, such as oc-
curs from cooling of an ordinary stream of lava.

As for the origin of the lava of the maria there are few facts on which to
base an hypothesis. What have been gathered may be briefly set forth. First,
it is to be noted that none of the vulcanoids of the moon give forth freely flow-
ing lava streams; it is, indeed, doubtful if any true lava flows have come from
them. The features which suggest such streams are rare and rather inconclu-
sive; they justify the statement that even the greatest, in general the earliest of
the craters, and therefore those which should have had the largest amount of
molten rock beneath them, show little or no signs of a tendency to extrude free
flowing lava at the time when they were formed. Nor do any of the numerous
fissures or faults of the lunar surface, some of which evidently penetrate deeply,
distinctly give rise to lava flows. And we shall see when we come to consider
the conditions of these volcano-like openings they appear always to have retained
their lavas within or near their vents. Clearly these vulcanoid openings do not
indicate any tendency of lava to pass up to the surface in large quantities.

It is an important point that there is no evidence in any of the maria that
the lava comes from a central pipe or from an elongate fissure ; their general form
would seem to indicate that if the fluid came from within it should have emerged
as from a terrestrial volcanic pipe, for if it came from fissures these should have
been of elongate shape. But if it came either from fissured or from pipe-like
openings there should be a grade to the flow extending from the center of the
field to its margin ; owing to the slight value of gravitation this grade should be
steep. There seems to be no trace of such a slope; on the contrary, the curve of
the terminator or margin of the illumination shows that they are essentially
horizontal. It is difficult to believe that lava flowing from an opening for hun-
dreds of miles could have this absence of slope. When it flows from a terrestrial
crater the course is always short and very steep.

In view of all the facts, I am disposed to hold with Gilbert and other
inquirers that the maria are the result of large masses falling upon the surface of

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON, 17
that sphere. All the facts indicate that these vast sheets of lava did not come
from the interior, and that the interior at the time when they were formed was
not in a condition to yield any such masses of liquid rock. We are therefore
fairly driven to this working hypothesis. In its favor we may adduce the follow-
ing considerations :

The fall of a considerable body or bodies competent by the conversion of its
momentum into heat to produce an extensive melting of the lunar surface, would
be likely to develop melted lava under conditions quite different from that which
is exuded from volcanoes. Assuming that the bolide came upon the surface at
planetary velocity and that it was some miles in diameter, the heat due to the
arrest of its movement would, we may fairly suppose, convert the whole of the
body into a liquid if not into a gaseous state. A like result would occur in
the part of the sphere which received the blow. Moreover, for some distance
beyond the seat of impact the shearing strains would probably be sufficient to
convert much of the material of the surface into the fluid state, with the result
that a mass of lava at very high temperature, equal at least to the bulk of the
invading body, and probably several times as great, would be sent at the speed
determined by the gravitative value of the sphere radially from the point where
the impact took place. It seems also, perhaps, a fair supposition that a great
collision of this nature would temporarily forma heated atmosphere enveloping
the moon, which would serve to delay the cooling of the molten rock until it had
time to find its level. Yet the absence of any deposits of these temporarily volatil-
ized materials is indicated by the fact that the light streaks are not obscured.

In favor of the hypothesis above suggested, it may also be said that the evi-
dence of melting effected by the material which forms the plains of the maria ts
considerable at several points, notably in the case of the vulcanoids on the mar-
gins of the seas. It seems quite certain that the walls of these craters next the
sea have been in some manner effaced by contact with the material which came
against it. Again, as in Flamsteed in the Oceanus Procellarum, the crater wall has
been almost melted down, but still rises slightly above the surface of the appar-
ent inundation. | At many points the material forming the mare comes against
extended steep-faced cliffs, which have the same general character as the inner
slopes of the great craters, where the form of the declivity pretty certainly has
been determined by the melting action of the lava at the base. Furthermore,
where there are depressions in the area on the borders of the maria, the material
of which they are composed flows into them as a fluid would have done.

It is also to be noted that at many points where the maria come against
gently inclined slopes the material of which they are composed appears to have
at first flowed over these low but now unsubmerged areas and then retreated from
them, leaving them in a measure smoothed as if by the in-filling of their cavities
or perhaps by a partial melting of their projecting features. If such apparent
inundation really occurred, it may have been brought about by the frontal wave
of the lava which mounted, after the manner of those produced by earthquakes in
the sea, for some distance above the permanent level of the inundation.

18 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

It may further be said in favor of this hypothesis as to the origin of the maria,
that the material of which they are composed appears to have had throughout
the whole extent of the several areas a singularly uniform fluidity. As before
remarked, there are no signs of successive flows such as have always characterized
the accumulation of the relatively much less extensive lava deposits on the sur-
face of the earth. In this connection it should again be noted that none of the
vulcanoids show any tendency to send forth extended flows, and the matter which
appears to have been ejected to form the cones has evidently consolidated on
very steep slopes. Thus, if the material of the maria was fluid when it came to
rest, of which there seems no reason to doubt, it cannot have been poured forth
from the interior in the manner of volcanic effusions.

The fact that the surfaces of the maria are of a distinctly darker color than
the other and higher extended areas of the moon has some value as evidence that
they have a peculiar origin, one not connected with the interior of the sphere.
Certain of the crater floors have, it is true, about the same tint; this is con-
spicuously the case with Plato. In this, as in certain other instances, the like-
ness may be due to the penetration by subterranean passages of the material of
the neighboring mare into the cavities of the craters. There are, however, exam-
ples, as, for instance, the great vulcanoid Grimaldi, where the resemblance
cannot be thus explained. Although these exceptions weaken the value of this
evidence derived from the color of the maria, the uniformity of a tint which is
evident in all of them and the seldomness of the exceptions tend to support the
hypothesis that the rocks of which they are composed have not come from the
interior of the sphere. This point will be further discussed below.

We turn now to consider the objections which may be made to the hypothe-
sis that the maria were formed by molten rock produced by the impact of large
bodies falling upon the surface of the moon. Of these objections, the first and,
in many regards, the strongest is derived from the general consideration that like
bodies competent to generate a great deal of heat have not fallen upon the
earth’s surface in the time which has elapsed since the beginning of the geological
periods. There is indeed no geological reason for supposing that they have ever
so fallen upon the planet.

Against the above-noted objection that the geological record of our sphere
affords no trace of evidence of any such falling-in upon its surface of bodies of
sufficient mass to produce widespread melting, and the proof that no cataclysms
of this nature have occurred since the development of organic life, we may set
the following considerations: first, that the moon’s surface probably took its
shape long before the beginning of our geological record ; and, second, that even
in this late stage in the evolution of our solar system there remain bodies in that
system in order of size such as would in falling upon the surface of the larger
spheres produce the effect which we observe in the maria. Thus the group of
asteroids which lie between Mars and Jupiter, though generally of far greater
mass than would be required by impact to melt the larger of the mare fields,
probably contains many bodies which, in case of collision with our satellite,

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 19

would bring about the consequences we note. At least one such mass of
matter, Eros, apparently not to be classed either with planets or satellites, has
recently been discovered at no great distance from the earth. It is possible
that in the relatively ancient state of the solar system, when the surface of the
moon acquired its crust, these detached masses of matter were more abundant
than they are at present. The tendency would be for those near the greater
spheres to be drawn in upon them, with the result that they would become
rarer near the planets and the larger satellites.

As for the origin of detached bodies of the bolide type, we have no basis
for more than conjecture; we may, however, fairly suppose that the explosive
action, which is of not infrequent occurrence in the fixed stars, may have hap-
pened in the case of our sun or even of the planets, with the result that masses
of matter, perhaps originally gaseous or possibly in the molten state, were flung
so far away that they acquired independent orbits.

Although the direct evidence going to prove that the maria are the result of
the in-falling of large meteoric bodies is not complete, the hypothesis appears to
me to have distinct value for the reason that the cause is sufficient to produce
that evidently sudden development of large bodies of very fluid matter, which,
for reasons before given, cannot fairly be supposed to have come from the in-
terior of the lunar sphere. It is, in a word, the only working hypothesis that I
have been able to find which in any way serves to explain these remarkable
features of the lunar surface.

In considering the details of the maria it is to be noted that it is not neces-
sary to account for all of them by supposing a single falling body brought about
the melting. In several instances, especially in the case of the Mare Australis,
and sundry other indistinct patches of the mare quality, the hypothesis can best
be applied by assuming that a number of such bodies fell at about the same time
and relatively near together. In this way we can account for the fact that in
place of normal, rudely circular fields of melting, as in the case of the M. Crisium,
we find an irregular, somewhat ragged field of this nature; in some instances
with a periphery that suggests that there were several centers of dispersion of
the fluid. Gilbert has maintained that the connected seas were formed by the
in-falling of a mass upon the region occupied by the M. Imbrium. This view
seems to me to be contradicted by the fact that in the passages between the
connected maria there is no evidence of scouring action such as would have been
brought about by the swift movement of great masses of lava.

It may also be said that the evidence of melting down of the pre-existent
topography on the margin of the maria varies much. It appears most clearly in
the case of the large, distinctly circular field of the Mare Crisium, and is least in-
dicated in the irregular areas. Such are the conditions we should expect to find
brought about by the fairly supposable variations in the size and number of the
masses in any one fall. Thus, so far as my examination of the problem has
gone, the supposition that the maria have been formed by sudden melting of col-
liding bodies and of the lunar surface about the point of collision appears to be

20 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

warranted as a working hypothesis, though it has, perhaps, not been established
as a theory.

To the suggestion that the surface of the maria is in general lower than
that of the regions surrounding them, and that this fact is inconsistent with the
addition to the quantity of matter in the area they occupy, such as would be
brought about by the falling in of a bolide, the following answer may be made.
In the first place, it is to be noted that the outer part of the moon is, except in
the maria and in the crater floors, evidently characterized by a very open struct-
ure. It is prevailingly much occupied by volcanic openings, greatly rifted and
probably composed of scoriaceous materials. If any such section as that about
the Apennines were completely fused to the depth of some miles, it is likely that
we would have a subsidence of the surface quite as great as that exhibited by
the maria. In the second place, the bulk of the material brought by the bolide to
the lunar surface would be small as compared with the volume of matter which
would be melted by its impact. The proportion would probably be less than one
to ten; so that the contribution from the impinging body would be so small
that it would not be likely much to affect the general level of the melted area.
The nature of the lunar surface in the maria and on the other more extensive
regions will be further considered in the section on volcanic action.

As before noted, there is no series connecting the ordinary craters, however
large they may be, with the maria. That this is the case is well indicated by the
fact that selenographers have in only a few instances been in doubt into which
group individual examples of these two species of lunar forms should be placed.
The fields classed as seas, with the evidently related embayments thereof, termed
sinuses or paludines, have always been regarded as readily distinguishable from
the craters. This decision has not been made on the basis of well-described
categories, but on the immediately evident differences between the two groups
of forms. It is recognized that while nearly all the vulcanoids are essentially
circular, or with only moderate distortions of that outline, the seas are as gen-
erally irregular in outline. So, too, it is patent that the vulcanoids, at least
those of large size, have in all cases a fairly well-marked external slope or cone.
None of the seas are thus characterized except where their periphery in part
corresponds to some antecedent feature, such as the wall of a large pit which
they have invaded, as in the case of Fracastorius, on the margin of the Mare
Humorum, or where it encounters an elevation such as the Hemus Mountains,
on the southern border of the Mare Nectano, (See plate xxv.) This gen-
eral acceptance of an essential difference between the vulcanoid floors and the
seas, and the very slight doubt as to the classification of the level surfaces in one
or the other, is excellent evidence as to their difference in nature.

The only areas of a level surface on the moon which may not be on mere
inspection classed as maria or vulcanoid floors are a few large crater-form de-
pressions situated near the eastern limb of the moon, of which the most import-
ant and doubtful is Schickard. Even a slight examination of this feature shows
that it has a distinct continuous wall, and that the irregularities of its outline are

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 21

due to the melting down of the borders of other craters as its area was extended
in the manner which we shall hereafter see to have been common in the develop-
ment of the larger crater-form pits. In other instances, as in Ptolemzeus, the
irregularity of the crater’s shape may lead to doubt as to its classification, yet it
is regarded by Elger as one of the most characteristic walled plains, its rampart
being exceptionally good. A further analysis of the instances which at first
sight appear to lead to some doubt as to the existence of a sharp line parting
the maria from the vulcanoid floor leads to the same conclusion as the facts
previously set forth, that these groups of level areas are, as structures, completely
separated from one another, and therefore cannot have had like histories. In
the one there has been a long-continued local volcanic-like action leading to the
formation of an external rampart; in the other, a swift production of an igneous
fluid, which has swept away until it found its level and shaped its margin by
melting down the pre-existing reliefs.

Although in general the material which forms the floors of the several
maria appears to be confluent, z. ¢., to show no marks of overlapping at the lines
of junction, there is reason to believe in the opinion of many observers that
there is some diversity in the level of their floors. Thus the Mare Nectaris is
supposed to be decidedly deeper than the others. This is not inconsistent with
the view that they were all formed at nearly the same time. The greater depth
of the last-named mare may be explained by the supposition that the absorption
of the fluid matter into the ancient crust was relatively greater there than
elsewhere.

While the surfaces of the maria are, as compared with the general surface
of the moon, decidedly plain-like, they are, in fact, the seat of many irregularities.
Of these the more important are a multitude of more or less continuous low-
arched ridges, probably in no instance more than two thousand feet high, but uni-
formly of relatively great width, often several miles in transverse section. The
nature of these ridges will be hereafter discussed. There are also on the maria
numerous craters, none of them approaching in magnitude those on the old, more
elevated portions of the crust. The ratio of craters on the maria is only about
one-fifth as great as on equal areas of the original surface, and their average
size is in about the same proportion. It is also to be noted that rifts or open
cracks are apparently rarer on the maria than on the high lands and that the
light bands and patches are of relatively seldom occurrence.

CLASSIFICATION OF VULCANOIDS.

In considering the so-called volcanoes of the moon (I shall term them vul-
canoids), the first step should be a classification of their features. Selenologists
have generally agreed to distribute them in seven categories termed as follows :
walled plains, mountain rings, ring plains, craters, crater-cones, craterlets, crater
pits. Besides these groups they recognize the existence of a less characteristic
group to which they give the ill-defined name of depressions. Under the term

22 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON,

walled plains, those who use this classification include the greater pits with the
ring of high land about them. Elger selects Ptolemzeus as the type of this:
group. He states that it is the distinguishing characteristic of this group that
there is ‘no great difference in level between the outside and the inside of the
walled plain”; he proceeds to cite notable exceptions to the rule, accepting
Schmidt’s term of transitional forms for them. These many exceptions range
from Gassendi, where the interior plain lies at about two thousand feet above
the floor of the Mare Humorum, which three-fourths surrounds it, to Clavius,
where the interior is some three thousand feet below the general level of the area
in which it lies; such variations are so numerous that they include practically all
the differences in the altitude of the enclosed plain which we find in any of the
groups. Nor are the other criteria of this category more characteristic. The
irregularities in the walls, the clefts, breaches, and greater breaks, are, in propor-
tion to the length of the encircling ridges, hardly more frequent than in the
mountain rings or ringed plains. So, too, with the minor craters, cones,
and ridges on the floors and rims; they are abundant, as inspection proves
roughly, in proportion to the area and the age of the structure. A careful ex-
amination of this group of walled plains will satisfy the observer that they are
essentially like the mountain rings except for certain accidents which have be-
fallen the members of the last-named group.

Nearly all the so-called mountain rings, all, indeed, that I have been able
to group in this category, lie in the maria. They appear, as has been considered
by several selenologists, notably by Elger, to be the more or less ruined rem-
nants of what were originally to be classed as walled plains. From their posi-
tion in the maria and even more from their topographic features, they are fairly
to be regarded as akin to the first-named group in origin and general history,
save that at the time when the maria were in igneous fusion their rings were in
part melted down and it may be in part breached by the tides of lava which
surged against them. In some instances these mountain rings appear to have
been suffused by the lava when it stood at its highest level, and afterwards bared
as the surface of the fluid was lowered. The maria of the second and third
quadrant particularly abound in these structures, in every stage of assault and
demolition, from those which stood so high above the flood of lava that their
exterior slopes show only slight signs of attack, to the intermediate stage of the
broken ring immediately north of Flamsteed, and thence to sundry unnamed and
scarcely recognizable fragments of rings in other fields of the maria. There
seems, indeed, hardly any room for doubt that to establish this group we
shall have to accept the principle that the state of obliteration of lunar forma-
tions affords fit basis for their classification. It appears to me that for my
purpose this group must be rejected.

In the group of vizg plazus selenographers have grouped all the strongly
walled vulcanoid pits of the lunar surface; they find the criteria for separating
them from the walled plains in the more continuous nature of their ramparts
and the steep declivity of their inner walls. They note also that there are often

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

to
Oo

terrace-like structures on these walls such as would be produced by the succes-
sive stages of descent of the lava of the crater. Here again by the use of the
method of series we may intimately connect the vulcanoids of this group with
those of the two preceding groups. None of the students of this classification
whose writings are known to me has failed to observe that there exist examples
which may be classed as wall plains quite as well as ring plains. There is no
doubt that these ring plains have in general better defined, more volcano-like
cones than the wall plains, and that the contact phenomena of the lava of the
floor with the inner slope of the rampart are more characteristic of volcanic
action as we know it on the earth, yet these differences seem to me so to
graduate together in the two groups as to afford no basis for distinct classification.

In the group of craters selenographers have placed so far the greater
number of the vulcanoid pits. They have included in them nearly all the
distinct pits from about fifteen to about three miles in diameter. So far as I
have found, they suggest no definite criteria for the members of this group, save
that they are widely distributed, occurring even on the walls of the large
structures, and that on this and other accounts they appear to be newer than
the wall plains or the ring plains. Inspection shows that there is no structural
difference between the vulcanoids of this and the preceding groups, their rela-
tively smaller size and apparent newness of formation affording no good basis
for instituting a category in which to place them.

Following down in the order of size, the next accepted group is that of
crater-cones. he objects included in this category are all of small size. Elger
compares them to the parasitic cones of A®tna, which seems to me not a happy
comparison, for their origin is in no wise related to the A¢tna “parasites.” As
the pits are generally less than a mile in diameter it is difficult to determine the
shape of their bottoms. My own observations agree with those of the selenog-
raphers, that these pits are usually in the form of inverted cones, terminating
downward obtusely, z. ¢., with no very distinct floors, and further that they are
occasionally found with rounded, saucer-shaped bottoms, as if there had been
lava in the cups, which had withdrawn with the cessation of activity into the
deeper part of the crust. There is enough of this obscure flooring to connect
by series the crater-cones with the craters, showing clearly that the difference
between the two is one of dimensions alone and does not indicate any essential
difference in the nature of the constructive actions. As regards the distribution
of the crater-cones and craterlets, it is to be noted that they in certain instances
appear to be associated with the light streaks; of this feature we shall take
account hereafter.

The smallest of the observable pits on the surface of the moon are termed
craterlets, or crater pits. These features are extremely numerous, the actual
number on the visible part of the sphere, which might under favorable conditions
be counted, amounting to many thousands. In the most characteristic specimens
of this group there is no distinct wall or cone surrounding the pit, the opening
often being abrupt, as if it were brought about by a mere subsidence of the area

24 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON,
in which it lies. Yet here, too, there is a gradation, for in sundry instances there
is trace of a ring wall as if some material had been extruded. In many instances
these pits are not circular, but with irregular outlines, which further suggest that
in certain cases there was no explosive discharge, but an in-falling of the covering
of a pre-existing cavity. It is further to be noted that these craterlets often, per-
haps oftenest, lie upon ridges, either the walls of the larger vulcanoids or the
numerous elongate elevations which occur in great numbers on various parts of
the surface and appear not to be connected with any large vents. In general
it may be said that the craterlets are the smallest observable members of the
series which has for its largest term the ring plains, and that they are among
the newer features of the lunar topography.

Looking upon the variety of form of the vulcanoids of the moon in the light
of our knowledge concerning the shape of terrestrial volcanoes, it may be said
that the range in form is not very much greater in the case of the satellite than
in that of the planet. Between the great caldera craters, such as those of the
Sandwich Islands or the Bolsena group of Italy on the one hand, and the smaller
cones on the flanks of A¢tna on the other, we have a range in width of cup less
considerable but approaching what is found on the moon; or, comparing the
nearly coneless craters of the Eifel, the products of a single eruption, with
the peaks of the Teneriffe type or those of the Andes, we note a difference in the
ratio of the enveloping cone to the interior which is also comparable to that
exhibited by the lunar vulcanoids. It is evident that the series of lunar craters
has much ampler range in diameter than those of the earth, but the correspond-
ences are sufficiently evident to justify us in including all such features of our
satellite in one group, assuming that the conditions of their formation were prob-
ably as near alike as in the several varieties of terrestrial volcanoes. An inspec-
tion of the lunar vulcanoids shows us that the most important features which
separate them from those of the earth are to be found in the amount and nature
of their extrusions ; the order, or lack of it, in their positions on the surface ; and
the influences which have served to deform or to destroy their features. These
peculiarities will be considered below.

The presence of a level surface of frozen lava in all of the lunar vulcanoids
save perhaps the very smallest is, as compared with the volcanoes of the earth,
their most conspicuous feature. This clearly indicates the relatively languid
nature of the eruptions from those craters. There are, it is true, a number of
terrestrial volcanoes where such a floor exists, but in all cases the facts justify us
in supposing that the last eruptive action was of the milder type, as in the case
of Kilauea in the Sandwich Islands. Eruptions of even slight intensity meas-
ured by terrestrial standards result in blowing out all of the fluid rock. Thus we
are justified in regarding the level interiors of these vulcanoids as evidence that
the normal lunar crater did not discharge explosively in true volcanic fashion. If
such violent discharges took place at any stage of the history of our satellite they
appear to be unrecorded in its existing features.

Not only is the presence of lava shaped on a floor in all the hundreds if not

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 25

thousands of distinctly observable lunar pits proof of the non-explosive nature of
their eruptions, but we have other evidence to the same effect in the lack of all
signs of ejected masses and of dust-showers, such as are the most striking phe-
nomena of terrestrial outbreaks. If we select any of the vulcanoids situated in a
region of much accidented topography, which evidently existed before the vent
was formed, and examine the surface about the opening, we readily note that it is
not masked as it would be in case it had been subjected to a succession of ash
showers such as come from a normal terrestrial volcano. In many instances I
have observed that there was no trace of such ash-covering up to the very foot
of the ring wall. Like evidence of a more affirmative nature is to be had in the
very numerous instances in which one vulcanoid cuts another. So far as I have
been able to note the details of these instances, the earlier existing crater, except
where its walls have been deformed by the encroachment of its neighbor, never
suffers from any distinct obliteration. Its ring wall—craterlets, vents, terraces,
and other slighter features, which should be hidden or distinctly changed in
aspect by an accumulation of even a few score feet of ash—remains, so far as can
be discerned, unaltered. When we remember that there has evidently been no
erosive action on the moon such as has normally washed away thousands of feet
in thickness of ash about 4tna and other large terrestrial volcanoes, we see how
clear is this evidence that the lunar vulcanoids have not been the seat of ordinary
volcanic explosions.

The lack of considerable lava flows on the moon appears to be almost as
well established as the absence of ash; in but a few instances have structures
which can possibly be classed as flows of really fluid matter proceeding from
craters been reasonably suspected, and these on inspection appear to be more
than doubtful. As will be noted below, the material in the craters appears not to
have had a high order of fluidity, so that it quickly consolidated on very steep
slopes—according to my observations generally exceeding 20° of declivity—as
soon as it passed out of the cup. None of the rills or other fractures appear to
have afforded passage to the interior fluid material; they seem, indeed, to have
been formed long after the larger vulcanoids had ceased to be active.

DISTRIBUTION OF VULCANOIDS.

In considering the distribution of the lunar vulcanoids it is first to be noted
that, unlike those of the earth, they are scattered over the whole of its visible
surface. The fact that here and there all around the limb we may trace the
hither borders of great ringed plains fairly leads to the supposition that like
structures exist on the unseen portion of the sphere. Except that on the maria
there are no large vulcanoids formed since those great plains were produced,—
probably none as much as ten miles in diameter that postdate their fluid period,
_there is little to be said concerning the distribution of these features on their
surfaces. There are, it is true, considerable areas of the lunar surface outside of
the maria where the only vulcanoids are the craterlets. With slight exceptions

26 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON,

these are the regions of so-called mountains, or in fields where there exist very
many low dome-like elevations, often circular in outline but occasionally some-
what elongate. Of those regions where vulcanoids of considerable size are rare,
the most noteworthy are the field of the Hamus Mountains, the region on the
west side of the Mare Foecunditatis, and that to the northwest of the Caucasus
Mountains, though there are many others of about the same extent. (See plate
xvi.) Several of these regions are of more than fifteen thousand square miles
in area. It should be understood, however, that none of these fields entirely
lacks vulcanoids ; it is indeed doubtful if there is any part of the moon’s surface,
except it may be some portions of the maria, where craters of large or small
size may not be found in every circle of twenty miles in diameter.

In many accounts of the distribution of the lunar vulcanoids it is stated that
the greater of them exhibit a distinct train-like arrangement. As before noted, I
have been unable to find any satisfactory evidence of such order being at all
common. Here and there, as in the group of Ptolemzus, Alphonsus, and Arza-
chel, there is a trace of linear order, but a study of the facts shows that so far as
the larger structures are concerned there is no reason to believe that there is
any prevailing definite order in their placement. There is, however, good reason
to believe that the smaller vulcanoids, commonly termed craterlets, are not infre-
quently arranged in linear order. This is not true of all of them, but is clearly so
in the case of those which are in some way related to the rills or other crevices,
and to the light rays of this point I shall have more to say below.

As regards the order of distribution in time of the lunar vulcanoids, it may be
said that all the facts point to the conclusion, if they do not establish it, that the
largest of them commonly were formed first. This is shown by the fact that
in only a few instances does a large ring plain cut a decidedly smaller structure of
the same nature, while the instances in which the smaller have intersected the larger
are very numerous. So far as I have been able to apply this method of determin-
ing the relative age of the rings, it establishes the fact that the greater number, if
not all, of the vulcanoids of say over fifty miles in diameter were completely formed
before the most, if not all, of those say twenty miles in diameter were built, and
further that very many of the craterlets were opened after the greater structures
were completed. Still further it appears likely, though not certain, that before
the greater vulcanoids were formed the so-called mountain districts and the
general surface of the moon had acquired the topography we now find them to
have, at least as regards the larger features of the surface. In very many of the
great vulcanoids we find evidence that the neighboring country has had its surface
somewhat distorted by the intruding structure. In a word, there appears to have
been an ancient surface antedating the distinct ring plains, though it is possible
that this surface was itself largely made up of such rings which have been obliter-
ated by the agents of decay, which have in many instances partly demolished
structures which are still recognizable, though often but faintly. The number of
these faint rings too indistinct to be named, and rarely affording more than the
merest traces of their original form, is so great as to warrant the conjecture that

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON, 27

those now existing are but the last of a long series which has been formed and
destroyed. Close attention to these features in the moments of good seeing,
which occasionally reward the observer, will reveal a series connecting such still
distinct though extensively demolished rings with other more numerous fragments
of circles which would not be interpretable save for the connecting links.

It may here be said that the phenomena of dilapidation exhibited by the
relicts of ring walls in the fields of the maria differ essentially from what we find
on the outlying surface of the moon. In the last-named areas, the ruining of the
ancient ramparts has evidently been in large measure brought about by the
encroachment and possibly by some shearing pressure of later-formed vulcanoids,
which actions have broken down and shoved about the fragments of the once
complete circumvallations. In addition to these processes of burial and displace-
ment, there have apparently been at work some influences which have slowly
broken down the rings, so that they have lost the original steepness of their
profiles. In and on the borders of the maria we find evidence that the destruc-
tion was brought about by the immediate and swift assault of the originally fluid
material that now forms these plains of frozen lava. The rings are not deformed
but more or less broken down, in part breached, by the stroke of a tide of fluid
rock, as in the case of Doppelmeyer and Hippalus on the shores of the Mare
Humorum, or simply overflowed and melted down, as is the case with the great
unnamed ring north of Flamsteed, the more effaced ring between that structure
and Damoiseau, or the many other like instances in other maria.

As we pass from the largest rings downward in the series towards the small-
est craters which have distinct floors, we note a progressive increase in the fresh-
ness and finish of these structures. The departures from the original form
become less frequent, the walls are less breached, and the slopes of the ramparts
steeper and more even. The interference of rings of like size becomes rare, so
that with those less than five miles in diameter it does not appear to occur. All
these facts point to the conclusion which finds expression in the writings of many
selenographers, that in general the larger the rings the greater their age.

PHYSICAL HISTORY OF THE VULCANOIDS.

Comparing the lunar vulcanoids with the terrestrial volcanoes and adding to
the considerations no more than a reasonable amount of conjecture, it seems
to me that we may interpret the phenomena as set forth below. In this explana-
tion care has been taken to introduce into the interpretation nothing in the way
of action that does not appear to be warranted by the processes of our own
sphere.

It is, in the first place, evident that while the lunar vents indicate some
process of eruption it cannot be regarded as in its nature identical with that of
ordinary terrestrial volcanoes. These last-named craters are, while they remain
active, with rare and questionable exceptions, on sea-floors or near their shores.
What we observe in their action and their distribution leads us to believe that

28 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON,

they are—mainly if not altogether—the points of discharge of water-vapor or of its
dissociated gases, and that this water has been buried by aqueous sedimentation.
The result is that when heated to a high temperature the fluid commonly explodes
with a great tension, scattering large amounts of morcellated rock to great
distances from the place of escape. On the other hand, in the lunar vulcanoids,
the evidence goes to show that there were no explosions competent to drive
fragments in extended trajectories. It is evident, indeed, that the movement of
the lava in the pits was almost exclusively up and down in the cavities, often with
successive haltings ona particular level, followed by a sinking to a considerable
depth. In these stationary periods, the terraces of the frozen fluid on the inner
slopes of the ramparts apparéntly were formed. That the position of the lava
was not in all instances determined by a common interior deep level of the fluid
seems to be shown by the fact that in some of the rings its surface is several thou-
sand feet below the surrounding area, while in the case of Wargentin, just south
of Schickard, the floor apparently lies high above the surface of the surrounding
country.

That there was some kind of boiling or up-welling action in these crater lavas
is well shown by the fact that in a number of instances, more numerous than the
records show, the surface of the floor is flexed upward, so that the center is some
hundred feet above the rim of the sheet, as if the final much weakened impulse
was sufficient to arch the frozen crust but not great enough to rend it from
its adhesions to the shore. Such tumefying action is also shown by the numerous
instances in which a mountainous mass of lava has been forced up in the central
part of the crater floor. These medial heaps of lava are so common in the
vulcanoids of middle size as to be the rule rather than the exception in these
structures. In many instances they are replaced by central craters, or now and
then, as in the case of Theophilus, there is a mass spewed up, as are some terres-
trial trachytic cones, with only a faint trace of crater pipes leading downward into
the interior. (See plate xvi.)

Finding as we do evidence of some swelling and sinking process competent
to lift and lower the lava in the craters of the vulcanoids, and seeing at the same
time that this action did not take place with anything like the energy of terrestrial
eruptions, the question arises as to the nature of this eruptive force which has
operated on the crust of the moon. The only hypothesis which has suggested
itself is some kind of boiling, such as will take place in any fluid mass which is
heated below and cooled on the surface, as in molten iron, where substances in
the vaporous state, though they exist, are not present in sufficient quantities
greatly to affect the movement, or there is a circulation mainly impelled by the
escape of imprisoned vapors. Mere convection of heat in an igneous fluid
does not seem to be sufficient to account for the rise and fall of the lava in the
craters, especially as in the case of Wargentin, for there the lava floor lies at
a height of some thousands of feet above the general level of the surface. We
will therefore consider the possibility of there being materials vaporized by heat
in the lava, not enough to produce the type of terrestrial explosions, but sufficient

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 29

to lift the lava to the tops of the existing rings and to produce a circulation suf-
ficient to keep the material for a long time in a molten state. On this point we
have some direct evidence from the fact that many types of lavas that form dykes,
such as granites, are violently forced into rocks of the earth’s crust without there
being any evidence of vaporous or gaseous materials impelling them; it is more
likely, however, that what we see in the way of eruptions on the moon are the
results of extrusions brought about by the pressure of gases originally contained
in the fluid mass of the sphere.

It is commonly assumed that for a long time after any celestial sphere has
entered on its fluid state, in passing from its nebulous or fragmentary previous
condition, the process of separation of its materials volatilizable at the temper-
ature established by the concentration must necessarily go on with the result
that some such vulcanoid phenomena as appear on the lunar surface will be
likely to occur. It is a fair working hypothesis that every crater-like opening on
the moon was formed by the relatively mild outbreak of vapor such as keeps
open the terrestrial craters of the Kilauea type; in such vents there may be
vapor enough to induce some movement of the lava, but not enough to cause
very great ejections of the fluid.

It may be assumed that the lava of the moon far more than that of the
earth would tend to retain its gases and to form the viscid, slow-moving material
known as pumice, which even when near a melting temperature is of a wax-like
stiffness. The reason why the blebs of vapor could not separate from the lunar
lava as readily as from the fluid rock of our planet is to be found in the relatively
slight value of gravitation, which on the surface of the moon is only a little more
than one-sixth what it is on the earth. The tendency of bubbles to separate
from a fluid depends in large measure on the difference between the weight
of the contained vapor and that of the mass in which they lie; so that it may
well be that the lavas of the satellite were on account of their contained vesicules
of vapor less fluid and more like pumice than those we have a chance to observe
in volcanic action.

When the lavas were lifted to the edge of the encircling rampart it is evident
that they flowed out. That they were in the periods of activity so lifted and
discharged is plain from the height of the terraces in many lunar craters, and
from the elevation at which the lava floor has remained in the case of Wargentin.
The normal well-preserved vulcanoid of sufficient size to permit a study of its
features shows, in most instances, buttress-like ridges extending not more than a
few miles outwardly from its rim; these are fairly to be taken as flows which
have passed over that rim or through breaches in it. It is to be noted that all of
these buttresses have very steep slopes, both in the radial direction from the
crater and laterally from the center of the ridge. To those accustomed to the
gradual slope of lava streams, such as break forth from the base of volcanic:
cones where the angle of declivity is often not more than two or three degrees,
the twenty to thirty degrees of inclination of these supposed lunar flows may
seem to negative the hypothesis that they can be lava streams. Lyall and

30 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

others have, however, shown that lavas may, flowing over the edges of terrestrial
craters, consolidate in slopes of eighteen degrees of declivity. Now the angle at
which the stream comes to rest will, other things being equal, be determined by
the value of gravity; reckoning this as before at one-sixth that of the earth's
surface, we see that a very much increased slope may well be allowed in the case
of the lunar discharges.

The conception thus formed of the process by which a lunar vulcanoid
of the larger size was produced, a conception founded on an extended study of
their phenomena, is as follows: the first stage of the action probably consisted in
the production of a slight dome-shaped elevation such as abound on the lunar
surface, being, indeed, the commonest of the smaller features on many parts of
the areas outside of the maria. These dome-like elevations appear to be due
to some accumulation of vapors beneath the superficial layer, formed perhaps
when the whole crust was still partly softened by heat. At a certain stage of the
process this arch fell in, or was broken to pieces and thrown outwardly, leaving a
pit with lava in it. When in its oscillations of height this lava overflowed the
edge of the pit, the material so passing from the heated interior quickly consoli-
dated and began the formation of a ring-shaped rampart. With the continuance
of this action the lava would tend to melt down the interior faces of the rampart,
gradually extending the diameter of the opening, destroying and remaking the
wall as the process of enlargement went on. Finally, as the supply of melted
rock was by unknown causes reduced, the lava fell to its lowest depth and gradu-
ally froze; the last stage in the activity being usually marked by a small central
crater, a low dome, or by a spewed-out cone, such as so commonly occupies the
central part of the floors of the greater rings. It is to be noted that the present
position of the lava in the vulcanoids is not to be taken as its average height, for
practically all of the craters which preserve what seems to be a fair semblance of
their original form show the remains of terraces that indicate higher levels of
their floors.

The objection may be made that the summits of the ramparts abound in
peaks which rise far above the general level of the rings. It is evident that
these salient points present serious difficulties; in some instances they may be
accounted for on the supposition that the parts of the ridge now much lower
have been broken down by lava which has poured over its crest. In other cases
we may find the explanation in the fact that there is an obvious tendency to
form small craters on the crust of the ring wall, there being many such that are
plainly visible. Now, as we see elsewhere, particularly in the center of the
vulcanoids of middle size, sharp, irregularly shaped masses of extruded lava,
sometimes, as in Theophilus, many thousand feet high, often take the place of
small craters. (See plate xvi.) Thus these isolated peaks may be masses

of lava which have been spewed up to a great height. The origin of the small
vulcanoids on the ramparts of the greater is a difficult matter to explain; it
may perhaps be accounted for by reference to terrestrial volcanoes, where we find
some evidence of a like tendency to form secondary craters around the margins

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 31
of a plug of frozen lava which fills the cup. If we suppose a ring widening by
the process of melting and rebuilding its walls, we may conceive that the fluid
is likely to extend at points beneath the ramparts, so that when, after a period
of repose, in which the lava was frozen and had shrunk, activity was resumed,
the easiest way upward for the vapors would be by passages leading vertically
through the wall.

The curious fact may here be noted, that in no observed instance is there
distinct evidence of any lava flow which has broken under and through the ram-
part or cone surrounding a vulcanoid. When we consider that practically all the
lava streams from terrestrial volcanoes break out through the base of their cinder
cones, this condition of affairs on the moon demands an explanation. This may,
like many other of the lunar events, be explained by the fact that the weight of
the fluid, which is the impelling agent of its flowing, is only one-sixth that of
terrestrial lavas, while the cohesion of the rocks may be, and most likely is, quite
as great as on the earth ; certainly these cones, which apparently are far more firmly
built than the ash heaps of volcanoes, must have resisted the relatively slight
hydrostatic pressure of the lavas they enclose far better than the like structures
of the earth.

We may here turn aside fora moment to consider the hypothesis that the evi-
dent and often probably repeated up-and-down movement of the lava in the vulca-
noids was due to tidal action effected by the earth. While it cannot be doubted that
the effect of the earth’s attraction, at present six times as great on the moon as is
that exercised by that body on our sphere, and may of old have been yet greater,
would be competent to lift any internal united mass of fluid to a considerable
height, there are reasons why it cannot well have served to pump the lava up to the
elevations it attained in the lunar craters. To be operative, we have to suppose
that the terrestrial attraction took effect in a central mass of igneous fluid, the sur-
rounding crust being essentially rigid, not flexing to any great extent with the
pull, which seems to be an unwarranted assumption. Under these conditions the
lava would mount and descend in each lunar day, which, before the moon ceased
to have a diurnal rotation, may have been of almost any length less than what
exists at present which we have a fancy to reckon. It is, however, to be observed
that the lavas of the vulcanoids, from time to time, froze at exceedingly varied
levels, there being a range of several thousand feet in altitude in craters which
are near to one another. These stations of repose, long enough to permit the
freezing, are not to be explained on the hypothesis of incessant tidal pumping ;
nor have I been able to account for the facts by any warrantable subsidiary
hypothesis. Moreover, the smaller vulcanoids, the craterlets, which are evidently
in the same series as the greater, having little or no lava in their bases, cannot be
thus explained. Furthermore, the central cones of many of the larger vulcanoids,
the formation of which was evidently in some way connected with the actions
which built the whole structures, apparently cannot be brought under this
explanation. :

The most reasonable view as to the interior condition of the moon when its

32 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON,

vulcanoids were in activity is that it was in a state of essential fluidity with a
relatively thin crust. This fluidity may not have been that of terrestrial lavas ; it
may have been, and apparently was, more viscous or pumiceous. That such was
the case is suggested by the behavior of the extruded lavas; it is further sup-
ported by the form of those other extrusions which occur in the so-called moun-
tains, as will be further noted in the study of those structures. Thus the crust,
despite its being of greater weight than the interior lavas, may have attained a
considerable thickness ; it may have had a depth of some miles. Yet it is hard
to believe that it would have formed a sufficiently rigid enclosure of the interior
fluid to have caused the sphere to remain undeformed by the earth's attraction to
the extent necessary to bring about a great up-and-down play of the lava in the
passages leading to the surface. It is furthermore to be noted that no trace of
tidal action has been observed in terrestrial volcanoes — though this fact may be
accounted for by the difference in the nature of their origin.

I have already, in preparation for the study of the maria, considered the argu-
ments against the supposition that the vulcanoids are due to the in-falling of
meteoric bodies, the main point being that they fail to exhibit any trace of the
great melting due to the collision of bolides of sufficient size to make such pits.
The maria being, according to my view, due to such in-fallings, showing all the
evidences of a vast and sudden development of very fluid material of high tem-
peratures, it follows on this hypothesis that the vulcanoids cannot be due to like
action. The objection to this explanation in the case of all the crateriform open-
ings seems to me to be so insuperable that it may not be further discussed.

It is important to consider the group of vulcanoids which have been formed
on the surface of the maria since the lavas of the maria were produced. We note,
at the outset, that these openings are all of relatively small size. Leaving out many
doubtful cases, where it is not easy to determine whether the structure was in age
antecedent to the maria in which it lies or no, these vulcanoids, so far as I have
observed, never exceed ten miles in diameter, and even those of such width lie in
positions where the covering of lava proper to the mare may be thin. It is there-
fore possible that they are due to actions occurring beneath this marial sheet
which have manifested themselves on the new surface. The only vulcanoids
which may be with some confidence regarded as having their origin in the lavas
of the maria are the numerous small craters and craterlets, those in general of
less than a mile in diameter, which are abundantly found scattered over their fields,
though they are there less numerous than on certain other parts of the lunar
surface.

It may here be noted once again that in certain instances the likeness of
color and the relation of height of the lavas of the maria and those of large nearby
craters leans to the suggestion that the igneous fluid from the neighboring mare
passed under the ring wall, or through clefts since effaced, into the area it encloses.
This view is most distinctly suggested in the case of Plato and Grimaldi, but
there are other instances to which it would be applicable. Such a passage of lavas
by underground ways is made doubtful by the fact before adverted to, that in no

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON, 33

instance has the molten rock contained within a ring been observed to discharge
itself through the rampart, as is often the case in terrestrial volcanoes. It is per-
haps more likely that any communication with the maria was by fissures in the
walls which have since been closed, or, if remaining, are so narrow as to escape
observation.. It may be said, however, that the great heat of the marial lavas and
their evident high fluidity would have enabled them to burrow through passages
not permeable to the viscous lavas of the vulcanoids.

The evident fact that the order of succession in time of the vulcanoids ts, in
a general way at least, in the order of succession of their size, the larger being
the more ancient, enables us approximately to determine at what stage in the
lunar surface the maria were formed. All of these several areas which have
originated independently one of another appear to have about the same sizes of
minor vulcanoids on their surfaces. The small craters apparently originated after
the greater rings had been formed, but certainly before the discharge of materials
from the interior had ceased. It is possible, however, that all the vulcanoids in
the maria, except those which were situated on such elevated ground that they
were not suffused by their lavas, owe their origin to boiling action within the
liquefied zone of the seas themselves. In this case it is possible that the time
when these fields were formed was after vulcanoids ceased to be produced on
other areas of the lunar surface. The general sharpness of these structures on
the maria is in favor of their relatively recent origin, though it affords no data
for a precise determination of their age.

I have, in considering the origin of the maria, referred to what appears to
me to be evidence that the fluid of which they were originally composed had
extended upward along portions of and perhaps all of their shores, so as to pro-
duce a smudged effect on parts of the relatively low-lying ground. So far as I
have observed, this apparent effect is most evident on the southern shores of the
Mare Nubium and the Mare Humorum. (See plate xx1.) My observations
suggest that these apparently inundated fields lack craterlets, such as occur on
the areas of the distinct maria. If this observation should be confirmed, it would
make it likely that the seas were formed after the activity of the moon, asa whole,
had ceased, and that the craterlets of the maria were due, as just above suggested,
to boiling within their masses, and not to the internal fluid of thesphere. A careful
reckoning of the number of very minute craterlets on the maria, as compared with
those on other parts of the moon, will probably show that they are on the average
more numerous on them than on some other fields of higher ground, and also
that they are of prevailingly smaller size. As a group they appear to me to
grade less distinctly into the flat-bottomed craters than do those of the high-
lands. My observation on these points are, however, not sufficient to more than
suggest these possibilities. Anything like a determination of them demands
better seeing than is to be had at the Harvard College Observatory and bet-
ter sight than is now mine. Should these variations really exist, they would
tend to show that the maria had developed their vulcanoids from their own ma-
terials. In further inquiries concerning these pits on the maria, it will be well to

34 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

have them compared with like structures in the lava floors of the larger ring
plains. My inspection shows them to be very similar in aspect, as they may be
in origin, probably being both alike due to actions taking place within a moderate
distance from the surface.

MOUNTAINOUS RELIEFS OF THE MOON.

Next in topographic importance to the vulcanoids come the reliefs, which
have received the general name of mountains. In this group we find at least
three distinct categories, which probably are due to as many separate causes.
First and most important of these species of salient forms come those which have
generally been named after terrestrial ranges or orogenic systems, as, for instance,
the Alps, Apennines, Caucasus, etc. Although these groups of elevations have
a considerable local diversity in character, varying in elevation from two or three
thousand to twenty-six thousand feet or more, and in shape of their individual
peaks from seldom nearly conical forms to much extended ridges, they in general
have the character of elongate masses rudely elliptical in horizontal section, the
several units of each field showing a tendency to a rude parallelism of their axes.
These units are rarely distinct from one another, but connected at their bases, so
that the field they occupy is by their confluence considerably raised above the
general surface of the country in which they lie.

The number of these fields of mountains which have been named by sele-
nographers is about twenty-five. There are, however, probably at least twice as
many areas which exhibit this type of structure in a tolerably clear manner. One
of the most important of these is the area between Schréter on the south and
Marco Polo on the north, the area in part forming an isthmus-like barrier between
the Mare Nubium and the Sinus A‘stuum. The facts go to show that while
the tendency to form this type of topography is more evident in the northern
than in the southern hemisphere, it has existed in some measure on all parts of
the moon except those now occupied by the maria; in these fields, though there
appear to be ill-preserved remains of such structures, they are very imperfect. It
may also be said that structures of this nature seem to be more frequently
developed near the limb than elsewhere, but this may be due to errors in classi-
fication, consequent on the difficulty of determining whether elevations in that
part of the surface are the borders of vulcanoids or mountain ridges.

In considering the relation of the mountains of the moon to the vulcanoids,
it is important first of all to note the fact that where they are extensively devel-
oped there is a prevailing absence of larger crater-form structures, and that in
certain instances we may at least suspect that they have broken up such struc-
tures. At a number of points involved in these tangles of ridges there are
features which look very much like fragments of the rampart of ringed plains
which had been involved in the apparently tumultuous movements attending the
building of the mountainous reliefs. Instances of this nature occur in nearly all
the larger mountainous areas; good examples exist in the Hamus Mountains

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 35

and in the unnamed district between the Lacus Somniorum and the Mare Crisium.
As it is the habit of the ridges to be rather straight, the occurrence of curved
fragments, varying from those of a few degrees of arc to half circular, appears to
warrant the hypothesis that antecedently existing vulcanoids have been broken
up in this peculiar constructive work.

In some instances vulcanoids which were evidently once fairly perfect, as
such structures necessarily are at the time of their formation, have been appar-
ently invaded by the mountain ridges. This is the case in Marco Polo, just
above mentioned. Here an originally normal ring plain has been broken into
on its northern versant, and thereby so deformed that its original nature is not
readily perceived on casual observation. The great walled plain of Hipparchus
appears to have been in large measure destroyed by the development of mountain
ridges, which traverse its walls and in part the enclosed plain. Many other
instances could be cited to show that these mountain-building actions, whatever
their nature may be, have been very effective in deforming if not in destroying
the vulcanoids of large area. Even the generally well-preserved Plato appears to
me to exhibit in its wall evident traces of dislocation arising from the disturbance
of the moderately accidented region about it.

There is no evidence sufficient to determine the stage when the building of
lunar mountains ceased. There is, however, reason to suspect that they were not
formed after the maria came into existence. There are, it is true, a number of
groups of such structures which lie within the boundaries of the seas, but there is
some reason to believe that these are the survivals from an antecedent time, being
parts of systems which were not entirely buried by these widespread lava fields,
though they show to my eye distinct evidence of having been effected by the
inundations of liquid rock. If this judgment as to the history of the intramarian
ranges be accepted, then we may safely conclude that the mountain-building
period was passed before the seas were formed. There is some reason to suppose
that this stage of the lunar development did not extend down to the time when
the smaller vulcanoids, at least those which lie outside of the ring plains, were
produced. In no instance have I observed any of the mountainous folds break-
ing in upon craters less than ten miles in diameter, though my observations are
not sufficient to completely exclude such occurrences. In many instances, how-
ever, very well-shaped craters of several miles in diameter occur in mountain-built
areas. They often are so well preserved that we have to exclude the supposition
that they were formed before the ridges were developed.

The second group of prominences which may be termed mountains has for
its type the isolated masses which often occur in the central parts of lava floors
of the greater vulcanoids, and more rarely in excentric positions on those floors.
These reliefs were evidently produced by some action connected with the forma-
tion of small craters which they appear to replace. Such craters on the floors
of the vulcanoids are, as is well known, extremely common; in many instances
there are more than a dozen within the ring, and in the Stadius Schmidt says he
counted fifty, and forty-one have been delineated. Commonly there is either a

36 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

considerable pit or a mountain in the center of the ring, the probability of this
central feature occurring being greater with the decrease of the size of the
vulcanoid, until the diameter of the plain becomes less than about ten miles, when
it tends to disappear. The facts indicate that the central pit and mountain of the
vulcanoid floor are interchangeable features. In some cases the peak has a more
or less distinct craterlet upon its summit, or, as is shown in the central compound
structure of Theophilus, there may be traces of a crater masked in the extruded
heap.

The third group of reliefs on the lunar surface is typified by the long, low,
apparently continuous ridges which are found on all the maria, but which are
particularly well developed on the Mare Imbrium, the Mare Serenitatis, and the
Mare Nectaris. (See plates xvi1and xxiv.) The characteristic features of
these ridges are their prevailingly low-arched forms, their slight height, and their
remarkable continuity ; they very often attain a length of one hundred miles, and
in some cases of twice or thrice that extent, while the greatest elevation assigned
to them is less than two thousand feet. As their flanks grade rather indistinctly
into the general surface of the maria, their precise width cannot be stated ; it is
evidently variable, with a probable maximum of five to ten miles. So far as I
have been able to ascertain, well developed continuous ridges are limited alto-
gether to the maria and practically so to the larger fields of this nature; in the
small maria they are much less distinct, though there are instances of slight undu-
lations which may belong in the same category of structures. In fact all the
extended plains, even those of the greater vulcanoids, exhibit more or less
wrinkled surfaces, when seen with powerful telescopes under very oblique illumin-
ation, such as serves to bring out irregularities only a few score feet in height.

The distribution of the continuous ridges indicates that they belong to two
distinct groups which may be due to diverse causes, or at least to different
methods of action of some general cause. The most evident of them are often
nearly rectilinear, or with broad curves, which have no evident relations to the
outlines of the shore of the mare in which they lie. Of these, the great examples
extending from near Lambert in the Mare Imbrium, or those of the Mare Sereni-
tatis lying between Posidonius and the promontory of Acherusia, may be taken
as types. Another group, well indicated on the borders of many of the maria
and some of their embayments, has the folds following the shores and seems to
be limited to a somewhat distinct field lying near those shore lines. Elger sug-
gests that in the case of Mare Nectaris these shore-following ridges are due to
the settlement of the lava in the central part of the basin. It is undoubtedly the
fact that the lava has been lowered in the Mare Crisium since the surface has
frozen, as it probably has in all the maria; traces of like action seem to me to be
more than conjecturable in the floors of the larger vulcanoids as well; but it is
not to me clear that these shore-following wrinkles are, as Elger suggests, caving-
in steps, such as those formed on the edges of a frozen pool or stream as the
water in the basin subsides. If they are, as some of my sketches indicate,
arranged in the manner of a carpet on a stairway, as monoclinal folds of terres-

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 30),

trial rocks, we have reason to suppose that they are due to faults which skirt the
shores and which occurred in the basement rocks while the lava sheet was still in
a plastic state. This supposition has its difficulties, for there is no evident reason
why such faultings should occur ; faults with vertical displacement are very rare on
the surface of the moon, and in no case are they found in any such order as we
need to have them to account for the shore wrinkles like those curving around the
borders of the maria.

Less distinct than the typical continuous ridges, but probably to be connected
with them, as lesser phenomena of the same order, we have, as before noted, on all
the maria and on some of the greater vulcanoids’ floors, faint wrinkles of great linear
extent. The relation of these to the larger ridges appears to be confirmed by a
series in which it is impossible to determine any break. I am therefore disposed
to place all the elongate wrinkles in one group, regarding the typical examples
hundreds of miles in length as structurally related to the slight, relatively short
foldings which are barely revealed by the telescope. On close examination of the
more characteristic elongate ridges it appears likely that they are not, as they
appear at first sight to be, even arches, but in some instances at least are com-
pounded of smaller wrinkles arranged in a more or less parallel order. As these
minute features are discernible only by their shadows, it is as yet undetermined
whether they are subordinate ridges forming a kind of chain or fractured blocks.
I am inclined to think it probable that they are of the last-named nature, for the
reason that analogy with terrestrial lavas would indicate that solidified superficial
lava would fracture and not fold into arches. Some of these ridges appear to
have craterlets on their summits.

It is also to be noted that, while the systems of low elevation which we are
considering have great continuity, there is an evident tendency to break the con-
tinuity, so that the chain is composed of separate links, each parted from the other,
as in terrestrial mountain chains. Here and there these units are arranged in an
echelon order, as is the case in many terrestrial mountain chains such as the Alle-
ghanies. This arrangement makes the likeness of these lunar elevations to ter-
restrial mountains more evident than any other of its reliefs.

A third group of lunar elevations, possibly akin to the long ridges above
described, is found in the domes which abound in many parts of the surface ;
they are, according to my observations, commonest on those parts where vulcan-
oids are rare. I have suggested that certain, or perhaps all of them, may be
incipient craters. These domes are found on the maria, though here they are of
prevailingly smaller size, as well as on the older, more elevated surfaces ; in num-
ber they rival the crateriform structures. Following the plan of grouping the
lunar features, when possible, into series, I have endeavored so to connect the
domes with the elongate arches before described. There are many examples of
domes which are somewhat elongate, say with the major axis near twice the
extent of the minor, but I have not been able to unite the two groups by any
complete series of transitional steps and therefore am led to consider them as
possibly distinct.

38 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

ORIGIN OF LUNAR MOUNTAINOUS RELIEFS.

As regards the origin of the first-described groups of lunar reliefs, those
which form the massive elevated mountains, it may be said that they cannot be
placed in the category of terrestrial structures due to folding and faulting com-
bined with aqueous erosion. If there be any one certain fact concerning lunar
topography it is that it nowhere exhibits the results of water erosion. If oro-
genic action such as operates on the earth has acted on the moon, as it may
have done in the case of the elongate ridges of the maria, it could give us no
more than arches and the fractures incident on their formation. It could not
possibly have developed the steep, lofty, and extremely serrate structures such as
are found in the greater fields of the so-called mountains. So far as geology
enables us to interpret them, these elevations must be due to the ejection of
exceedingly viscous lavas, forming heaps such as we have in certain masses of
trachytic rocks on the earth. That such ejections do occur on the moon is well
shown by the very numerous and often high peaks which have evidently been
thrust up in the central part of the lava field enclosed by the greater vulcanoids.
In character of summits and slopes these tumefactions of the ring plains are to
my seeing essentially like the so-called mountains. They often attain to near
the average height of the peaks in the Alps or the Apennines or other lunar
fields of crowded elevations. The facts have led me to the following considera-
tions and to a working hypothesis based on them:

Noting that the peaks formed in the central part of the lava floors of the
greater vulcanoids clearly indicate that, after a period when tolerably fluid lavas
existed beneath the crust, there came a time when these lavas were so viscous
that while they might be extruded they would not flow, but retained the shape
in which they were spewed out; noting also that the evidence from the invasion
of the vulcanoids by mountain ridges indicates that these elevations were
among the more recently formed structures of the maria, we are led to the
suggestion that they represent a stage of the eruption when the ejected materials
were so viscous that they could no longer form vulcanoids, but poured forth
masses which not only did not flow but heaped up near the vent, just as they evi-
dently did in the central field of many craters. It is true that small craters are
here and there, though rarely, found amid these mountainous elevations; they
may represent the localized remnants of the once general fluid state, remnants
sufficient to produce slight eruptions of the earlier type.

I have already called attention to the fact that the distribution of the ex-
ceedingly numerous small bleb-like domes on the lunar surface suggests that
they are the first stage in the development of craters, the imprisoned vapors
serving to lift the surface although it was not broken through. It appears to
me likely that it is in such elevations that we have also the beginnings of the other
group of vulcanoids, the ejected peaks. In several parts of the moon, notably
in ‘the region where mountainous elevations occur, these domes abound. In
some cases small craters occur in the same field, which suggests, as before noted,

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 9

oe

that these bleb-like elevations may have been the first stage of such vents: in
other cases the cones appear to pass by a series of transitions into the mountain-
ous form. I have not been able to verify this passage from the dome to
the peak, but the indications of it appear to me to be noteworthy. In this
connection it may be remarked that the structures in the centers of the middle-
sized vulcanoids lend support to the view that a dome-shaped elevation may, by
further development, pass into a peak. When these prominences are low and
small they often have a rather evenly arched form, but when they are of consid-
erable magnitude they take on a complicated shape with serrate crests substan-
tially like the structures classed as mountains, the only evident difference being
that the masses are not so commonly elongate in horizontal section, as the indi-
vidual mountainous ridges commonly are.

The observed facts concerning the mountainous protuberances of the lunar
surface lead me to the opinion that they are classifiable in one group, of which
the simplest and most interpretable examples are found in such peaks on vul-
canoid lava plains as that of Theophilus, where we have a mass of ejected ma-
terials which shows no trace of flowing for it has very steep walls. (See plate
xvi.) This great viscid ejection covers an area of more than three hundred
square miles, and rises to a height of six or seven thousand feet above the floor
of the crater ; it is particularly interesting for the reason that while it is essen-
tially a group of peaks it retains traces of what seems to be a volcanic type, as it
has an indistinct crater on the summit of the mass. Other instances could be
cited to show this passage from the conditions of a crateriform structure to a
rugged cone. In fact the series appears to be sufficiently fairly complete to
establish the point that the last stage of activity in the craters of the vulcanoids
was that in which the interior lavas, primarily hot enough to flow in the manner
necessary to form very level surfaces, had become so viscous that they would
maintain themselves at angles of sixty degrees or more to the horizontal.

As for the ejections of viscous lava which took place outside of the craters,
forming mountain-like elevations, the evidence appears to warrant the conclusion
that they represent, as do the craterless cones within the rings, a survival of a
tendency to eruptions after the time when the lava was liquid enough to produce
the normal vulcanoid structures. In these later eruptions, because of this ex-
ceeding viscosity of the ejected material, there could be no ring wall or interior
lava plain formed. All the material which would have gone to such construc-
tions was heaped in the viscid mass which was forced out of the opening. The
natural result of these conditions is that the mountainous elevations, while less in
diameter than the larger vulcanoids and having no more material than goes to
the formation of an ordinary lunar cone and lava plain, present normally very
elevated peaks.

It may seem that if the craters and the mountains are the result of es-
sentially the same expulsive energy, with no other difference in the conditions
than the suggested variation in the fluidity of the lavas, we should find a series
of intermediate forms between the crater and the peak. Such intermediate

40 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

stages are, as I have noted, to be found in the central structures of the normal
vulcanoids. I have here and there suspected like transitional shapes among the
mountains, but can cite none that is conclusive. There are, however, in the
region of the Alps and other fields in the northern hemisphere of the moon vari-
ous instances which may be of this nature. My eyes are no longer fit for such
difficult observations, so I must leave this point, along with many others, unveri-
fied. It is well, however, to note that the passage from the state in which the
lava of the moon’s interior was sufficiently fluid to bring about the formation of
the ordinary vulcanoids, to that in which peaks only would be formed, does not
involve any great change of temperature. In terrestrial conditions, a lowering of
a few degrees in heat at the critical point in a progressive cooling would be suff-
cient to bring about the change in the nature of the eruption.

The frequently elongate shape of an individual mountain seems at first sight
to be, and perhaps really is, an objection to the above-described theory of their
origin. It is, however, to be remarked that a large part of these elevations have
rudely circular bases, and that where they depart from this figure they do not
take the shape of long, continuous ridges, the major axis rarely exceeding the
minor in the ratio of more than two to one; moreover, some of the mountains
of the crater floors show the same tendency to elongation. Later on in this
writing I shall note that the phenomena of ‘‘rills” and other rifts show that the
surface of the moon was very generally in a state of contractile tension, and this
before the formation of the smaller vulcanoids was arrested, and further that the
axis of the mountains often coincides with the direction of the rill-splitting. If
this be the case, then the extrusion of somewhat rigid materials such as formed
these cones would naturally tend to rend the crust as with a wedge, so that
an elongated opening would be formed for the extruded mass and the shape of
such opening would determine the outline of the elevation.

There is yet another class of reliefs on the lunar surface, those which are typi-
fied by the great escarpment of the Altai Mountains inthe fourth quadrant. (See
plate xv1.) In this Altai relief we find in the southeast a slight and gentle rise of
a field, which has few very noteworthy features, for a hundred miles or more to the
edge of the steep, and then a sudden fall to the northwest, the descent being on
the average at least six thousand feet. The crest of this declivity is much varied ;
one peak, at least, is said to attain the height of thirteen thousand feet above its
base. It appears likely that the northwest face of the Hemus Mountains and
the southeast face of the somewhat similar district lying between Eratosthenes
and Mt. Hadley, facing the Mare Imbrium, are structures of a related nature.
The most warrantable hypothesis, from the point of view of the geologist, is that
these reliefs are due to faulting on a large scale, accompanied by a considerable
amount of extrusion of the type that forms lunar peaks. In two of the three
evident examples of this group, those last named, the lava of the maria has
extended to the base of the declivity ; in the case of the Altai steep, the igneous
matter of the Mare Crisium, though it once extended much beyond its present
limits, did not attain the base of the escarpment. There are divers other steeps

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON, 41

which may be allied to those above described; of these perhaps the most inter-
esting is that which forms the border of the Sinus Iridum and of the Mare
Imbrium, to the northwest and southeast of that remarkable bay ; nearly the
whole eastern shore of the Sinus Roris and of the Oceanus Procellarum may be
of this nature. Though the last-named escarpment does not rise suddenly to
any great height above the mare plain, the straightness of the line suggests that
it was originally of greater vertical extent and was formed by faulting.

The principal objection to the hypothesis above stated, that the above-de-
scribed features are due to faulting, is found in the fact that clear instances of such
action are rare on the lunar surface. The most conspicuous fault, where there can
be no doubt as to the nature of the conditions, is that commonly known as the
Strait Wall on the surface of the mare between Birt and Thibet. (See plate xx1.)
Here the break has a length of at least sixty-five miles and is quite as rectilinear
as any terrestrial fault. The vertical dislocation is at least five hundred feet and
may be much greater. It is evident that this is relatively a modern feature, having
been formed after the time when the mare had cooled. It is not unlikely that in
the earlier ages, when the moon was parting more freely with its heat, the re-
sulting faults were of far greater extent than is shown in the Strait Wall. It is
to be noted that the break of the Strait Wall did not lead to the extrusion of any
considerable amount of igneous matter. Elger has observed craterlets and
mounds upon the crest of the escarpment, but it is not clear that these are
genetically connected with the break, for such features abound in the Mare
Nubium as in other seas. Thus, though there is no basis for certainty, I am
disposed to regard the Altai group of escarpments as due to faulting,

As to the age of the great escarpments above described, it may be said that
they certainly antedate the maria, which have their margins to some extent
determined by them. They seem also to antedate some of the larger vulcanoids,
for Piccolomini, which is about sixty miles in diameter, being in size among the
score of greatest structures, was formed after the Altai escarpment. Plato also,
though less clearly, appears to have been formed after the steep which bounded
the Alps on the south, now somewhat effaced by the Mare Imbrium, was devel-
oped. If this hypothesis, which seeks to account for the steep faces of highlands
by faulting, be correct, we must regard these features as among the most ancient,
perhaps the very oldest, reliefs on the lunar surface. They are now to a great
extent masked by the maria, which have found in them their natural shores, they
being, it would appear, bordered by them for near half their total coast line.
Further reference to these features will be made in the discussion of orogenic
action.

VALLEYS, RIFTS, AND RILLS.
In addition to the above-described positive reliefs of the moon, the surface

of that body presents a multitude of minor depressions which demand considera-
tion ; of these the most notable are the cavities which have received the obscurely

42 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.
defined name of valleys. The most conspicuous depression ordinarily classed in
this group is the great Alpine valley which traverses the mountainous ranges of
that name, extending in a northeast direction from near the Mare Frigoris to the
Mare Imbrium, a distance of about eighty miles. (See plate xxi.) In width it
varies from four to six miles, but at its southern extremity for about one-fourth of
its length it is somewhat narrower, being reduced at one point to about two miles
in cross-section, and at the mouth it is beset with what seem to be extruded masses,
so that it debouches by several narrow clefts into the neighboring sea. The walls
of this valley are generally nearly vertical ; from my own comparisons with other
measured objects, they appear to average more than a mile in height and to be
for the greater part of their extent almost rectilinear. The floor of the depres-
sion is approximately level, though with some obscure pits, and by its color
as well as its form is evidently covered by an extension of the Mare Imbrium.
The Ukert valley, on the east side of the crater of that name, is longer than the
Alpine and has about the same width with less depth. The fracture by which it
was formed appears to be continued in an obscure cleft, which extends from its
northern end to the vicinity of the vulcanoid called Marco Polo, the whole con-
stituting what seems to be one structure nearly two hundred miles in length. A
similar valley with a length of about eighty miles lies on the west side of
Herschel. It has a width of at least ten miles and is rather straight-walled.
Yet another notable feature of this group is that lying on the eastern side of
Rheita, which is about one hundred miles in length and about twelve miles in
diameter. Last of all we may cite the great valley on the southwest side of
Reichenbach, which extends in a rather tortuous course for about one hundred
miles and has a width of ten or twelve miles. There are many other similar,
though smaller, valleys, varying from a maximum width of ten or twelve miles
downwards, until they grade in dimensions into the group of clefts. A full list
of these structures is lacking, but they probably number several score.

As regards the distribution of the fault valleys, it is noteworthy that all the
distinct examples of the group lie outside of the maria. It is true that on those
fields there are depressions which have been classed with the vales, but, so far as I
have been able to determine, they all fail to exhibit the essential features of this
group—z. e., they lack the steep walls and the generally rather level floors char-
acteristic of the true valleys. They seem to my eye to be in their nature synclines,
or downward foldings, the counterparts of the continuous ridges which are so
characteristic of the maria, though they are not found in any definite relations to
those up-folds. As to the time of the formation of the valleys, it appears to have
been relatively late, posterior to the formation of the mountains, though before the
production of the lavas of the maria. It is not certain that any larger vulcanoids
than the craterlets were formed at a later stage in the evolution of the surface,
for only very small structures of this group appear to have been produced in their
cavities. It is also to be noted that these fault valleys are most developed in
the regions where the larger vulcanoids are not very abundant, though it must be
said that they are not lacking in the fields where these features are well developed.

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON, 43
CRATER VALLEYS.

In this group may be placed a number of curious though unnoted structures
in which one or more craters have been in some way deformed so as to make a
broad valley. The range of this action is great and the features to which it
gives rise rather obscure. The changes of shape, arising from this deforming
action, become very difficult to observe in all the vulcanoids at any distance from
the central field of the lunar surface, for the actual elongation is confused with
the apparent lengthening of the basins brought about by the obliquity of the view.

A fair sample of the crater-valley type is found in Hypatia, in the north-
central part of the fourth quadrant. (See plate xvm.) Here the crater is so far
deformed that its major axis, extending in a S. W.-N. E. direction, is about twice
as long as its minor axis ; moreover, this depression is vaguely continued as a valley
for some distance beyond the walls of the crater. There are other like depressions
in this neighborhood. Gutemberg in the same quadrant passes on the south into
a broad, extensive, ill-defined valley. Palitzch, near the western limb, is a yet
more characteristic sample, having, according to Elger, whose reckonings appear
always to be accurate, a length of sixty miles and a width of only twenty miles.
Capella also exhibits this passage into a valley, and there are, according to my notes,
six other like instances in this part of the field. It would be possible to collect
not fewer than one hundred instances of the deformation of craters into elongate
valleys, or their extension into broad vales, which are in some way evidently
connected with them. As I am not undertaking a list of lunar features I cite
only such as are needed for illustration of this point.

Besides these numerous cases, in which the craters have been so far deformed
that they have had the character of valleys imposed upon them, there are about
as numerous instances in which the greater vulcanoids have been but slightly
deformed—so little changed, indeed, that the alteration has escaped observation.
In these cases, which include a large part of the pits over twenty miles in
diameter, the northern and southern walls show a distinct, though often slight,
change of form, indicating an elongation in that axis. I find that in my rough
notes of observations I have termed this the “spooning” of the crater in that
meridional direction. This feature may be best noted in the vulcanoids of the
central part of the lunar surface. It is distinct in Hipparchus and Albatagnius
which approach being crater valleys. Alphonsus and Davy show the same
feature, and it may be noted in perhaps one-third of the greater vulcanoids which
are so placed as to make it possible to discern this feature in its slightest expres-
sion, (See plate xvii.) Without at present undertaking to discuss the condition
which has brought about the evident warping of these greater vulcanoids on the
meridional line, it may be said that its aspect suggests that they have been
involved in certain movements, tending to produce considerable synclines. I
have sought for, but failed to find, clear evidence of anticlinal folds correspond-
ing to these troughs, yet the inquiry has not been carried far enough to insure
that they do not exist.

44 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

CLEFTS AND RILLS.

The clefts and rills of the lunar surface are features which seem to me
to belong in one group, though they may reasonably be separated from one
another by certain differences. Among the clefts we may class the very numer-
ous rifts which intersect the walls of the vulcanoids, particularly those of larger
size, which often extend for considerable distances beyond the limits of the
ramparts in which they occur. In the characteristic examples of this group, the
features radiate from the crater, and are thus shown to be in some way connected
with its conditions. They closely correspond in appearance with the Val del
Bove on the eastern versant of A*tna and many like structures on other terrestrial
volcanoes. In some cases they appear to be essentially akin to the terrestrial
Graben or multiple fault depressions, as for instance the Alpine valley, in that the
ground between two fractures has been lowered. They may, indeed, be regarded
as a variety of that class of depressions determined by the strains originating in a
vulcanoid. There are very many examples of the group, ranging from those
which produce broad breaches in the crater walls to such as are shown on the
flank of Tycho, where the two parallel light streaks, which appear to follow the
path of faults, have the ground between them apparently somewhat lowered,
in the manner of a rather gentle syncline, without any evident displacement.

Related to the several fault groups of depressions in that they are alike the
results of fracturing of the crust are the remarkable features known as rills. In
this group we have a single fracture with a space separating the walls, but no
distinct indications of a floor between them. Perhaps the most characteristic
example of the group is that known as the Sirsalis Rill, so named because the
Sirsalis vulcanoid lies near to it. Elger’s description of this structure —he evi-
dently knows it well — is as follows : “‘ Commencing at a minute crater on the north
of it [= Sirsalis], it grazes the foot of the Glacis, then passing a pair of small
overlapping craters (resembling Sirsalis and its companion in miniature), it runs
through a very rugged country to a ring plain east of De Vico [ De Vico a| which
it traverses, and still following a southerly course, extends toward Byrgius, in the
neighborhood of which it is apparently lost at a ridge, though Schmidt and
Gandilot have traced it still farther in the same direction. It is at least three
hundred miles in length and varies much in width and character, consisting in
places of distinct crater rows.” It has been suggested, according to Elger, who
does not state by whom, that the rills are not in fact breaks but a series of small
craters so near to one another that the effect on the eye is that of a continuous
crevice. This view, according to my observations with the excellent fifteen-inch
Mertz refractor of Harvard University, is not maintainable ; while craterlets are
often present along the line of the rill, their nature as fractures, when clearly
seen, appears certain. The breaks are ragged, as if torn through a row of crater-
lets, not usually more than half a mile in diameter and often narrowing at one
or both ends, so that their terminations cannot be determined ; but that they are
in their essence rents seems to me beyond doubt.

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 45

As regards the number of the rill fissures on the visible part of the moon we
have no good evidence. They are probably to be numbered by thousands, and
as the fainter seem to be the more plentiful, more effective instruments may
reveal many thousands of them. As regards their distribution there are many
noteworthy features. First we observe that those which have been mapped show
an obvious tendency to be arranged in groups, and in these groups the individual
breaks show here and there a tendency to intersect one another, though they are
more often arranged in a parallel relation. The next point is that those which
are in appearance sufficiently conspicuous to be mapped lie mostly in the cen-
tral part of the visible surface between the parallels of 30° north and south of
the moon’s equator, and within 30° east and 50° west of the central meridian.
They are thus remarkably rare in high latitudes and apparently seldom near the
east and west margins of the visible part of the sphere. This apparent feature
of distribution may be due to the oblique view of those marginal fields. It is
also to be noted that all the important fractures are situated on or near the maria,
or on the floors of the greater vulcanoids. Of about seventeen examples mapped
by Elger, twelve intersect the shores of maria, and none of them lies altogether
more than one hundred miles from those lines. The great southern upland has
no mapped examples and the central parts of the larger maria are also without
them. I am aware that the floors of the greater vulcanoids abound in rills all of
small size. I am also aware of the fact that somewhere about a thousand of
these rill fractures have already been noted and that their distribution is much
wider than that indicated where only the more important are plotted, yet there
is probably some significance in the grouping of these greater specimens of the
class in or near the maria.

As to the time of the formation of the rills, it may confidently be said that
they appear to be, with the possible exception of some of the craterlets, the most
recent structural features of the moon. If narrower scrutiny than has yet been
given to the matter shows that craterlets have developed in the cracks, then the
later structures, of course, postdate the rills. If, however, as it seems to me
quite possible, the rills have merely followed lines of incipient fracture, such as
joint planes would afford, in some instances going around the pits instead of
cutting through them, the rills may be the very last of the considerable lunar
accidents. Such, indeed, they seem to me to be.

OROGENIC ACTION.—CAUSES OF DISLOCATIONS.

We turn now to consider the possible causes of the dislocations on the lunar
surface which are represented by the various kinds of valleys, clefts, rills, and
ridges which have been briefly described above. First, as to the valleys of the
Alpine type, it may be said that they appear to correspond to the Graden type of
terrestrial down-faultings, where there are two or more approximately parallel
faults, the included area having been lowered. As to the origin of geological
Graben, we have as yet no evidence of value and naturally no consensus of opinion.

46 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

It appears, however, most probable that they are due to the orogenic strains
which enter into the complex of actions involved in mountain building, combined
with some withdrawal of support ordinarily afforded by the materials of the under
earth, as would be brought about by the migration of matter seeking volcanic
vents. In the simpler and more applicable case of these down-faulted blocks of
the crust, such as occasionally occur about terrestrial volcanoes, we may fairly
assume that the sinking was due to the ejections which had made the under earth
unable to support the load. That such deficiencies of support would have locally
resulted from the lunar eruptions is highly probable. To this action then, with
fair probability of its truth, we may for the present refer the valleys of the Alpine
type. The minor cleft valleys radiating from the vulcanoids are evidently to be
most reasonably explained on the same hypothesis. They are, indeed, so far as
I can see, comparable to the Val del Bove of A¢tna.

The rills, where we have relatively narrow crevices, which seem to extend
indefinitely downward, with no distinct floors, may be regarded as due to the
secular refrigeration of the superficial parts of the lunar sphere at a time so late
that they found their way to no bodies of lava. They are evidently contraction
cracks formed on a very extensive scale. Where they are limited, as is often the
case with the smaller of them, to the lava field of a large vulcanoid, they may
represent no more than the contraction of that body of lava. When, however,
they are on the maria, an indefinitely extended sheet of the frozen material may
find relief in the fracture. The predominance of the greater rills on and about
the maria may be due to the fact that, whatever was the origin of those vast
bodies of once igneously fluid rock, the consequence of their appearance on the
moon’s surface was, when they cooled, a great necessity for contraction. Not
only were the lavas of the maria originally at a high temperature, but they must
have communicated this heat to their shores and to the high country near them,
with the result that new and extensive readjustments due to cooling would be
required in those portions of the crust which had been thus affected. Thus the
rills and the Alpine valleys appear to be distinctly diverse in origin, the former
being due to loss of temperature of the crust in general, the latter to more com-
plicated action.

As regards the rare instances of true displacement faults such as the Strait
Wall, they appear to be due to ordinary faulting such as so abundantly occurs
on the earth. They may in their first stage have been rills where there was
some lack of support which caused the rocks on one side of the fracture to
change their level with reference to those on the other. The only peculiar
feature about them, from the point of view of geology, is that they are so rare
and apparently so unconnected with compressive strains. If the surface of the
earth as it has been affected by faulting, but without the effects of erosion,
could be examined under the conditions in which we behold the moon, the fault
dislocations would appear by the hundred thousand and with vertical displace-
ments of miles in height. Nothing, indeed, so well illustrates the very great
difference in the history of these two neighboring spheres, the moon and the

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 47

earth, as the diversity in the development of this group of structures which they
exhibit.

The next question is as to the group of lunar reliefs to which the continu-
ous ridges of the maria belong. It seems clear that, whatever be the detailed
structure of these ridges, they indicate compressive actions of the terrestrial
mountain-building type. The great linear extent of these compression ruptures
shows that they are due to no local strains but are the result of stresses which
pervaded wide fields of the maria. Their narrowness and lack of considerable
height may be taken as evidence that they are the result not of deep stresses
but of such as resided in the superficial parts of the crust, probably within the
lava of which these fields are composed. As to their age they of course post-
date the formation of the maria and apparently the larger vulcanoids — none,
indeed, of great extent — which have developed on their plains. It is obviously
important to determine the time of their uplifting in relation to that when the
rills were formed. This I have been unable to do ina satisfactory manner. I
have no notes of good examples in which either of these groups of structures are
found in intersection; nor does my limited acquaintance with the literature of
the subject supply such instances. It appears, however, likely from the fresh
aspect of both groups of dislocations that they are not of very diverse age, but
that the rills are the newer.

The problem presented to us is the existence in the same field of the rills
which indicate the shrinkage of the material of which the maria are composed,
together with that of the continuous ridges which even as clearly show that this
portion of the moon’s surface has been in a state of compression that compelled
the rocks to buckle upwards and, if we have rightly interpreted the structures,
brought about the formation of corresponding synclinal forms, the shallow
troughs which exist on these plains. If it is proved, as seems likely to be
the case, that the rills on the maria were formed after the continuous ridges,
then we might conceive that the cooling of the interior of the moon brought
about a compressive strain on the already cold outer crust, and that the limited
diameter of those wrinkles was due to the fact that there was still some measure
of viscosity in the lower part of the lavas of the maria which made it possible
for the hard upper part of the sheet to act independently of the subjacent
portions of the section, so that this upper part of the sheet as a whole received
the compressive stress as a thrust from the shores against which it lay.

There is another way in which we may consider this problem of associated
compression and shrinkage in the maria. It is to be noted that the most distinct
examples of each action lie in fields remote from each other, the rills near the
shores and the continuous ridges remote from them, the one in fields where the
lava is presumably rather shallow, the others where it is deep. With this
difference in conditions it might come about that contraction of the deeper parts
of the marial sheet in the process of cooling would be sufficiently strong to
fold the surface, while in the quickly-cooling shallow parts of the maria the
only effect would be the formation of shrinkage cracks, It is to be noted that

48 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

something like these diversities of action is to be seen in terrestrial lava fields,
though it is not certain that they are due to like causes. On any frozen expanse
of lava we are apt to find at once ridges which cannot well be attributed to the
roping of the solidified crust, along with cracks which are evidently due to super-
ficial cooling. There are other possible explanations of these contracted dislo-
cations of the maria, but I shall here take leave of the subject, for it is one on
which I have not been able to form a satisfactory opinion.

ADJUSTMENTS OF THE SURFACE TO CONTRACTION,

Looking over the whole of the lunar structures, the geologist is naturally
surprised to find so little in the way of adjustment of the crust of the sphere
to a nucleus diminished by the loss of heat. On the earth he sees in the ample
folds of the sea-basins and of the continents, as well as in very many folded
mountain chains, what he takes to be evidence of a long-continued accommodation
of an anciently cooled crust to a central mass which is ever losing heat. On the
moon he finds what, in proportion to the size of that sphere, is surely not the
hundredth part of such action. The folding of the marial ridges and furrows
is trifling and is probably due to action set up in the lavas of those fields. The
features of the crater valleys and the deformed vulcanoids appear to indicate
some small measure of folding, but that may have been brought about by the
loss of the moon’s rotation through tidal action, and the consequent disappear-
ance of an equatorial bulging due to that rotation. In any event it does not
appear to represent any considerable readjustment of the crust to the interior.
It is true that the moon has only one-fourth the earth’s diameter, and the fold-
ing caused by shrinkage should only be in about that ratio to like action on the
earth. Yet on the satellite the process of cooling is probably at an end, while
in the case of the earth, reckoning from the time when the crust was formed,
it cannot well be more than half accomplished. What then is the meaning of
this startling diversity in the orogenic history of the two spheres ?

In considering the difficult problem which has been just above suggested,
the first question that comes before us is as to the value of the evidence concern-
ing the antiquity of the general surface of the moon. We may ask whether
the original sphere may not have cooled in its time to a low temperature,
making in the process the necessary adjustments of its outer crust to the dimin-
ished interior, and whether after that was all done the mass may not have been
added to by the in-falling of meteoric bodies, such as has been hypothesized to ac-
count for the maria. By such in-fallings a general outer coating of lava might
have been formed, only a few-score miles in thickness, and to this may be due all
the vulcanoid phenomena down to the time when the later coming of other such
bodies formed the maria. On the basis of this conjecture we would not have to
look for any extensive marks of readjustment of crust to central mass. It cannot
be denied that the body of any celestial sphere is liable to be added to by in-fall-
ing masses, at least until it has cleared its path of them; and the fact that it has

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 49

been found necessary to account for the maria by such action lends a certain
countenance to this view. Yet it seems to me safer to suppose that the moon
has, as a whole, had essentially the same experience in space as our earth. As
before noted, the earth, since its organic life, at least in the present series of
forms, began to exist, has evidently had no such impacts of foreign bodies as
formed the maria. It is, of course, among the possibilities that the earth has been
subjected to invasions of large meteoric bodies, as the moon appears to have
been, and that an ancient organic period was not only destroyed but the records
of its existence entirely effaced. There is, however, no other known evidence on
which to found such a conjecture, except what we find on the moon.

As regards the failure of the moon to exhibit the marks of adjustment of its
crust, which first hardened, to an interior diminished by the loss of heat, it may
at first appear that as the value of gravity is only about one-sixth what it is on the
surface of the earth the stress which would impel the superficially cooled section
to accommodate itself to the lessened bulk of the interior would be proportion-
ately smaller, so that the outer shell might remain unsupported while the inner
portion shrunk away from it. This view seems inadmissible, for the reason that
in the case of the earth, as has been well reckoned, a shell less than a mile thick
would, if unsupported, crush and fall in of its own weight, so that in the moon
the crust would in alike manner crush at less than six miles of depth. It is thus
evidently necessary to form some other hypothesis which will account for the
lack of adjustment. I have essayed several of these, which I will now briefly set
forth with the reasons why they seem adequate or otherwise.

At first it seemed possible that the aggregate wrinkling and crushing exhib-
ited in the larger ridges and furrows, as well as in the host of small ridges which
are seen with the greater telescopes, might have been sufficient to provide for
the necessary contraction through the buckling and shoving of the crust. Yet
on carefully examining selected areas of the crust where these features are best
shown it does not seem possible that the accommodation or “take up” thus
effected can amount to many miles of length. Moreover, the phenomena are
not those which would be produced by the folding of a thick crust, as it sank
upon a diminished nucleus, but only such as superficial strains would induce on
a thin outer layer. It appeared conceivable that for some reason an accommo-
dative folding might have taken place on the portion of the moon which is never
seen, but this supposition is supported by no evidence whatever ; all we see on
the extreme margin of the visible surface leads to the conclusion that the hidden
side is essentially like that we behold. Again, it appeared possible that the whole
mass of the satellite remained in the boiling condition until it had been brought
to a state where the cooling quickly induced rigidity throughout the sphere, all
parts down to the center having attained somewhere near the same temperature.
In this way we could explain the small amount of internal contraction which has
apparently occurred since the most ancient features on the lunar surface, the

larger vulcanoids, were formed.
Although in a general way we know the law of cooling bodies, we are not

50 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

yet certain as to their application to celestial spheres. It is, however, evident
that the earth did not cool down to anything like an equal temperature through-
out this sphere before a crust was formed. But in the lighter mass of the moon,
when gravity tended less to promote interior solidity than it has probably done
in the case of the earth, it is possible that boiling went on so long and effectively
that when it ceased the whole was at a temperature not much above the heat
of lava, so that the further cooling would be uniform, and the undiminished
crust would not have in any considerable’measure to conform to the diminished
interior. There are difficulties with this hypothesis, as with the others which
have been suggested. If we could suppose that the moon had been during its
cooling stage deeply wrapped with a vaporous envelope, as was probably the case
with our earth at the corresponding stage of its development, it would be easier
to conceive a process of slow cooling which would permit the exterior part to
attain about the same temperature as the central portion, so that they would
solidify at the same time. But it is likely, for reasons given below, that through
its whole history as a sphere it has lacked such a covering and has been exposed
to the temperature of space. Yet for all these objections it appears probable
that the hypothesis last above suggested is the most tenable, and that the greater
part or possibly the whole mass of our satellite became solidified at nearly the same
time and at nearly the same temperature.

To the geologist, the action of the lunar surface under the limited com-
pressive stresses to which it appears to have been subjected is of especial interest,
because it shows clearly that rocks, which certainly are not stratified, apparently
may warp into rather sharp up-and-down folds. The student of the earth has come
to recognize that, in a limited way, foldings may take place in crystalline rocks
where there is no stratification on which the separate parts of the mass may slip,
nor even schistose planes that may facilitate such action, but that such extensive
and far-reaching movements as are apparently shown in the continuous ridges and
furrows of the maria or in the crater valleys may occur, has not been appreciated.
So, too, the lunar phenomena suggest to the geologist that the variations in the
action of a sphere under conditions other than those now existing on this planet
may be exceedingly great.

DIVERSITIES IN HUE ON THE LUNAR SURFACE,

Under this head I shall consider the differences in the amount of light and its
color which the surface of the moon sends to us, taking first the permanent hue
of its several parts and then the variations which occur in the various angles of
illumination. Beginning with the observations of Sir John Herschel at the Cape
of Good Hope, there have been a number of studies on the light of the moon.
Herschel, by comparing the color of the moon with that of the face of Table
Mountain, came to the conclusion that the hue of the satellite did not perceptibly
differ from that of weathered sandstone; that it was rather a dark than a bright
object. It is easy to make an equivalent observation when the old moon is seen

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 51

in the day-lit sky. The evidence, in a word, goes to show that the surface of the
moon is, as a whole, quite as dark as the average lavas of the earth’s surface
when they are lit by a vertical sun.

Although the moon's surface, taken as a whole, must, according to Zéllner,
be regarded as nearer black than white, there is little doubt that parts of it
under certain conditions of illumination are as white as any portions of the earth’s
surface ; as white as the chalk cliffs of Dover, probably ; or as white as new-fallen
snow would appear to an observer looking upon it from the moon. Although
the range in the scale of tint between black and white is probably nearly as great
on the moon as upon the earth, it is most noteworthy that there is no distinct
trace of the other colors so abundantly exhibited in the terrestrial minerals and
rocks. There are no greens or yellows, and it may be doubted if there is any
trace of red. Schréter, whose scale of hues ranges from the black shadows to
the whitest illuminated objects in the moon, selects ten gradations in that scale,
but makes no provision for the prismatic colors; he evidently did not find them.
I have a fairsense of color and have only to confirm this judgment. The geologi-
cal importance of this point is considerable, for it clearly indicates uniformity
in the lithological composition of the moon, or at least in the aspect of its rocks,
which differs widely from that we have on the earth. It appears to me that the
value of this uniformity in the color scale of our satellite may fairly be set forth
as follows :

It is a reasonable supposition that the chemical elements of which the moon
is composed are essentially like those of the earth, for such identities are indi-
cated by the spectroscope in the sun and the remoter stars. It is, indeed, alto-
gether likely that all the elements of the terrestrial rocks would be found in
those upon the lunar surface. Is there any reason why they should not present
us with a like range of color? It seems not improbable that this difference may
be due to the lack of water or air on the satellite. In the terrestrial rocks almost
all the prismatic colors are due to processes of oxidation which water brings
about. Those which are thrown out by volcanoes commonly are without such
hues, and only exhibit them when they have been subjected to oxidation on the
surface. So subjected, they acquire, by that process acting on various substances,
particularly on the iron they contain, a considerable varietyjof tint, including yel-
lows, blues, and reds. Thus it seems to me the lack of color range on the moon
confirms the supposition that there neither is nor has been water or free oxygen
on its surface.

Within the range of tints recognizable on the moon we have room for some-
thing like as ample a scope of petrographic variation as may be supposed in the
varied volcanic rocks of the earth if they were precluded of oxidation. Accord-
ing to Proctor, the darker parts of the lunar surface are of the tint which would be
reflected by dark syenite. The whiter are probably as bright as the lightest of
our volcanic rocks or the encrustations formed by solfataric action. In a word,
there is no reason to suppose that the lunar volcanic rocks are any less varied
than are those that come from the depths of the earth. As before noted, how-

52 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

ever, there is a striking difference in the behavior of lunar and terrestrial lavas ;
the lunar, except in the maria, where the evidence of high and continued liquidity
seems to me plain, appear to have become stiff almost as soon as they escaped
from their craters, a fact which may be accounted for by their viscidity or perhaps
by the swift cooling to which they were exposed on'the airless sphere.

It is noteworthy that the most important differences of hue on the lunar
surface are found in the maria and certain of the greater vulcanoids. The maria
are without exception much darker than the higher ground. The lavas within
the craters are likewise rather dark, but less conspicuously so ; but in the case of
certain of the great rings near the eastern limb, notably Grimaldi, they are quite as
somber hued as any of the seas. In the neighboring vulcanoid, Riccoli, there is a
patch on the floor which is perhaps the darkest-colored of any part of the lunar
surface. If these dark lavas were altogether peculiar to the maria, it would be
easy to account for their color by the supposition that the material imported by
the bolides, which I have supposed to have caused their formation, was of a dif-
ferent constitution from the materials of which the general surface of the moon is
composed. The frequent incoming to the earth of considerable meteoric masses,
composed in large part of iron, would warrant the hypothesis that the bolides
which produced the lavas of the maria were largely made up of this metal. Even
if not the tenth part of the lavas were of this foreign material, it might serve to
effect the ‘darkening of the resulting sheet. The occurrence of a like hue in lavas
which lie on the central floors of distinct vulcanoids appears to negative this
supposition.

Although for the reasons given above I cannot at present strongly maintain
the hypothesis that the hue of the seas is due to the color-producing action of
the bolides which produced them, it is perhaps hasty to dismiss the view without
some consideration. It may be urged that the in-falling bodies were probably of
varied sizes. Thus the mass or masses which I have supposed to have produced
the isolated Mare Crisium were probably smaller than the mass or masses which
brought about the formation of the great system of connected maria. It is fairly
supposable that a fragment large enough to have given the lava of Grimaldi its
peculiar hue fell within that vulcanoid, and that a small fragment likewise affected
a part of the floor of Riccoli. So numerous and crowded are these great vul-
canoids near the eastern limb of the moon that there is more than an even chance
that two such falls would both lodge within some of them and not in the inter-
vening country. As before noted, Plato and other less conspicuous vulcanoids
situated near the maria have dark floors, but in these cases there is a fair chance
that the external bodies of lava may, while fluid, have penetrated into their
enclosures ; its evident exceeding fluidity would probably enable it to burrow its
way in, though the more viscid lavas of the craters in no case appear to have
been able to flow out through the cones.

Thus while the facts do not warrant us in concluding that the color of the
lavas in the maria is due to mineral peculiarities imported by bolides which formed
them, it strongly suggests that explanation. Progress towards an interpretation

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 53

of this point may possibly be made by a careful study of sundry parts of the lunar
surface where outside of the vulcanoids there are features which may have to be
accounted for on the hypothesis of meteoric falls—of masses great enough to
produce some local melting but not sufficient to create distinct maria.

AREAS OF VARIABLE HUE.

Of all the diversities of hue observable in the lunar surface, those which vary
from time to time are the most curious and the most baffling to the inquiry.
The objects of this class may conveniently be divided into two groups, of which
the first should include the irregular patches of light generally capping the flanks
and ramparts of the vulcanoids and the cones they enclose, together with the
bands of light color which in most instances radiate from vulcanoids or origi-
nate near them. It is characteristic of the objects in this group that they are
invisible or nearly so when the sun is just rising on them, that they commonly
are not noticeable, indeed, until the sun is high, and that they disappear when the
illumination becomes again very oblique. The other group contains sundry
examples where the fields are lighter colored in low than in high illumination, in
this regard reversing the conditions of the first named series. There are no
features on the moon’s surface which have been the subject of more inquiry,
though mostly of a discursive kind, than the first-named group of colored areas.
The hypotheses and speculations concerning them have been numerous, but have
led to no accepted judgment concerning them. It appears to me that the best
way to approach the problem they afford is that indicated below.

First let us note that by far the greater area of the fields, which suddenly
become very white as the lunar day advances, lies on the higher part of the vulca-
noids, on their slopes and the summits of their enclosed cones. It is evident,
therefore, that the whiteness is most likely due to some quality of the surface
imparted by the vulcanoid action to which these regions have been exposed, a
quality which is developed only under a rather high sun. Under these conditions
the measure of whiteness is roughly proportional to the approach of the illumina-
tion to verticality, perhaps not absolutely so, for it is held by most observers that
probably the brightest point on the moon’s surface is the central peak of Aristar-
chus which lies about twenty-three degrees south of the equator. I am inclined,
however, to. believe that the apparent extreme brightness of this object is due to
the contrast afforded by the dusky fields of the mare in which it lies, and that the
fields of extremest lucency are all nearer the central part of the moon.

That the brightness of the very shiny parts of the moon, the patches and
the rays alike, is not due to any change in their constitution brought about by
the action of the sun during the monthly fourteen days of illumination, is proved
by the fact that these features distinctly appear on the moon's surface when, in its
newest stage, it is receiving a like vertical earth-light. I noted this fact many
years ago, though I did not then perceive its full significance. I am now assured
that my observations were trustworthy for the reason that negatives of the dark

54 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

part of the new moon taken in the earth-light clearly show these differences of
hue; they are, indeed, plain enough to enable one to map the more brilliant rays
of the Tycho system. We may therefore dismiss the idea that these features
are evolved in the progress of one lunar day to be reconstructed in the next, and
regard them as permanencies made visible when they may reflect to us the light
which comes to them at a high angle.

As to the conditions which bring about the large amount of reflection under
a high sun from those parts of the moon which appear very white when it is full,
the experience of geologists suggests the following working hypothesis: First,
that the bright area may be covered by an incrustation of a smooth nature such as
ice or other material which forms a sheet. It cannot be frozen water, but various
volcanic emanations may be conceived as forming like surfaces of glassy smooth-
ness. Or we may suppose that some part of the material which came forth during
eruptions was distributed as vapor to become crystallized on the surface. Such
solfataric action is common enough in terrestrial volcanic districts; it would often
be sufficient to cover extensive fields were it not for the erosive agents which
scour the surface. It appears to me, however, that the suggestion of a smooth
surface, such as an incrustation, is insufficient to meet the facts, for the reason
that such coating could not be formed save of frozen water or of materials laid
down by fluid water. With the low temperature of the moon’s crust and the lack
of an atmosphere, the idea of a quick crystallization of mineral substances from a
vaporous state seems more consonant with the known facts.

It is possible that the sudden-coming brilliancy of the bright patches and
streaks is due to the fact that these shining areas are covered with crystals which
have their planes so arranged that they are prevailingly parallel with the surface
on which they lie, so that they reflect their light toward the earth only when the
sun is high. This hypothesis has some support in the appearance of certain steep
slopes, as those of the cones in the greater vulcanoids, where the face of the cliffs
may be observed to shine brightly, when the sun’s rays strike them, some time
before the adjacent nearly horizontal surfaces gain the intensity of light which
they afterwards acquire. A close study of this matter may afford data for a
determination as to the nature of the action. So far I have been able to do no
more than prove that the brilliancy is due mainly to the angle of illumination, by
noting that it appears in earth-shine as well as sunshine, though the brilliance
of the glow on the margins of the moon suggests that there is also an element of
fluorescence or other action in the phenomenon.

Although light rays distinctly appear to be connected by series with the light
patches, there are certain peculiarities about the former which demand explana-
tion. Their exceeding length and their generally slight width make them very
puzzling features. It has been frequently suggested that they are due to certain
dust-like emanations from the craters which have been blown by the wind which
bore them and lodged in crevices or in the lee of projecting points. The current
of air which bore them is conceived as produced by the gaseous emanations from
the crater. This view appears to me to be ill-founded, because the volume of

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 55

emitted gases required to produce a sufficient current even in a vacuum would
have to be impossibly great to make such a wind at the distance of hundreds (in
one instance 1700) of miles from the vent. Moreover, as W. H. Pickering has
well shown, these bands do not, in all cases, point to any large crater, but in the
case of the most remarkable group — that of Tycho— appear to originate not in
the main vent but in certain small craterlets somewhat on one side of that open-
ing. Moreover, as well observed by Pickering, these bands are not definitely
continuous but made up of relatively short strips of bright-colored surface, each
of which appears to originate in a craterlet and to fade as it extends to another
in the same line, and that this arrangement probably continues to the end of
each streak.

It is also to be noted that in some instances the bright rays of the moon show
a tendency to be parallel, or approximately so, to one another, they being in some
way causally related to rows of small vulcanoids. I have already called attention to
the existence of such near approach to parallelisms in the case of the two striking
examples in the Tycho system. There is another equally good example in the case
of Messier, where the two streaks of this system, though slightly divergent, show
an evident departure from the normal radial order. Many other instances could
be cited to show that, while these bands of lighter color obviously tend to be
placed in radial position with reference to a vulcanoid, they are here and there
affected by some conditions which warp them from that position and force them
to become parallel. This later condition is much more common in the numerous
faint streaks which cannot be referred to any group radiating from a large vul-
canoid. To my eye, this tendency to parallelism affects a considerable part of
the rays which appear to be of the older origin.

It is obviously important to determine whether the rays of bright color on
the lunar surface are due to superficial conditions alone, or whether they are the
result of some action affecting the crust beneath the surface. On this point we
have little information but that of a highly indicative kind. A glance at these
features when they are best presented shows the observer that they extend
across the irregularities of the broken country they traverse. In at least one
instance, a ray emanating from the Tycho center crosses the lava plain in the
bottom of another crater (Saussure) and apparently traverses the steep slopes
of its wall, while another ray of this group seems to have been deflected from its
normal course by the ramparts of this vulcanoid. I have personally verified the
observations on the passage of this streak over the lava plain of Saussure and
have, though imperfectly, traced its passage up the inner wall of the rampart.
Other more skilled observers appear to have no doubt that it exists.

The facts just above noted make it evident that the light rays are not
purely superficial features, but are in some way connected with the structure of
the crust; from the point of view of the geologist, they have to be accounted for
by supposing that they are the superficial expression of an action essentially
solfataric in its nature, wherein vapors of:some crystallizable substance, or sub-
stances, have passed through crevices of joint-like nature from the deeper parts

6 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

on

of the sphere, either to form a coating on the surface about a vulcanoid, or
to stain a belt of rocks on either side of the rift, so that a strip of country,
perhaps a mile or more in width, extending to the top of the crust, was impreg-
nated with the material—the deposit being perhaps more extensively accumulated
on the surface.

As already noted, there is reason to suspect that, besides the large reflecting
power which the materials of the largest rays possess when the sun is high, these
materials have a certain fluorescent property, which causes them in some measure
to store up light which is given out after the sun has passed the angle at which they
begin to shine. That such is the case is indicated by the fact that the rays are visi-
ble on the limb of the moon when the sun’s light is considerably more oblique than
it is when they become very bright. Such a property is known to exist in many
species of minerals. A close study of fluorescence may, indeed, serve to indicate
the nature of the substance which sends us the light from the very shiny parts
of our satellite,— that of the diffused patches as well as that of the rays.

If we accept the hypothesis that the bright parts of the moon are due
to the deposition of some highly reflecting and perhaps fluorescent materials,
we may proceed to derive certain important corollaries from the proposition.
It is at first sight evident, from the extent of the shining fields on and about the
ramparts of the greater vulcanoids, that the egress of the light-reflecting materials
was there by so many paths that the resulting stains were confluent, and that the
rays marked its passage in fields where the channels were rarer, though related
to the same centers of vulcanoid action. It is also evident that, while these pas-
sages for vapors from within cut a few of the crater floors of lava and occasionally
extend on to the maria, they appear never to originate in those areas. More-
over, the great extent of these rays, some of them exceeding one thousand miles
in length, and the way in which they radiate from their several centers, are prob-
lems of no small importance.

As to the common origin of the blotches of light material on and about the
vulcanoids and the rays, the series of facts leaves no good reason for doubt.
The blotches generally pass outwardly by gradations into rays, the most of which
are short, perhaps less than a score attaining a length of one thousand miles or
more. As to the deep-seated origin of these structures, it is fairly proved by
the fact that they cross irregularities of the surface, as well as by the fact that
they occur along lines of craterlets. There is some reason for believing that
these, the smallest of the vulcanoids, were formed along the lines of the rills,
presumably before those clefts were opened. Their existence along the light
rays is of itself evidence that there is some incipient breakage along their
courses. It is a reasonable supposition that vapors were forcing themselves out
on those lines, and that sometimes they did so with explosive energy.

The existence of incipient crevices, such as jointings arranged in a general
radial order with reference to the greater vulcanoid centers and extending for
very great distances, is a feature which from the point of view of the geologist is
surprising. While in the case of terrestrial volcanoes it is common to find traces

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON, 57

of a tendency of the crust to split radially so as to permit the entrance of dike-
making lavas, these fractures are not known to extend for more than at most a
score of miles from the vent to which they center; nor is there any observed
tendency of the crust about volcanoes to become penetrated with joint-planes,
having the position of those which the before-noted facts lead us to suppose
exist on the moon. Before this evident lack of likeness between the two spheres
is weighed, it is well to note that, while in the case of the earth all the extensive
jointing of the rocks is apparent, brought about by strains due to mountain-
building action, even when the beds have not been visibly dislocated it is evident
that they have been jointed by the stresses; so that the fracture systems of the
earth may be said to depend on an action which does not appear to have been
to any considerable extent effective on the satellite.

Given a sphere in which there are no extensive strains due to the contraction
of its central part and a consequent readjustment of the crust to the nucleus,
which appears to be the case with the moon, it is not unlikely that a series of
ruptures such as we find indicated by the rays would be formed. In such an
orb, the last stage of its cooling would necessarily lead to the contraction of its
outer part. Such was evidently the case in the moon, as is shown by the late
formation of its valleys and rills. After this strain had become so slight that it
was no longer competent to open distinct fissures, it might still have been sufficient
to produce the incipient tension cracks required for the escape of vapors such as
are needed to account for the light rays.

The most difficult point to explain is the radial distribution of most of the
rays and the evident relation of nearly all of them to the greater vulcanoids or to
craterlets situated on their flanks. This, it seems to me, may be accounted for in
the way set forth below. Let us suppose that in the last stage of the expulsion
of the vapors of the lunar sphere, when the formation of vulcanoids of more than
about a mile in diameter was no longer possible, the crust was by its cooling
brought into a state of contractile tension so that it had a tendency to break.
We may then fairly assume that this tendency would be greatest in the ancient
uplands, and least in the relatively new maria and in the lava floors of the vulca-
noids. These fractures, or lines of weakness, for they do not seem to have been
defined openings of measurable extent, would naturally—indeed necessarily—
be made as radii to the large pits of the crust which plentifully occur in the
higher parts of the moon. We may have visible evidence of their necessity by
watching how shrinking clay splits in relation to holes made in its surface.
Beginning in the field about a vulcanoid, a fissure would extend radially for a
certain distance, far enough to satisfy the strain which led to its formation ; if it
afterwards happened, as W. H. Pickering has noted, that a body of vapor broke
its way to the surface, forming a craterlet at the point remotest from its origin,
then the rupture might be continued on the same line, attended by the formation
of another craterlet, until the strain was again satisfied ; and this process might
be again and again repeated until the greatest observed extension of the ray was
brought about.

55 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

Supposing rays be formed by successive developments and reliefs of cooling
strains in the manner just above suggested, we find a reason for the peculiar
shape of these features which Pickering has well observed. He finds that the
longest of them are not strictly continuous, but that originating at a craterlet they
extend for a variable distance, widening and becoming dimmer the farther from
the place of origin; then at another craterlet they again begin narrow and
bright, to fade and widen once more as they become remote from the opening.
According to my view, the first craterlet started the fracture; near it the fissure
was most passable to the vapors in question, so there the streak is narrow and
bright ; farther away the fissure was less open, so that the effusion had to force
its way through the country rock, and so made a wider and fainter deposit of the
shining material. The second craterlet developed an extension of the fracture
with the same features as the first, and so on to the end of the colored belt.
According to this hypothesis, we need not suppose any such mighty accident as
required by the view that the ray system of Tycho was formed at once; it may
have been geologic ages in developing ; the end of a great ray may, indeed, have
been formed very long after its beginning.

To those who are unfamiliar with the movements of homogeneous materials
in the process of shrinking, it may seem unlikely that the outer part of the moon
in cooling equally would tend to fracture in systems of joints arranged in radial
order. A little observation on drying clay will show that slight accidents
determine in very uniform materials the direction of the fractures due to strains
which lead to cracking. When the pull is equal in every direction and when
there are depressions on the surface, the tendency is to make these pits the
center of radiating fractures. In this way, by cracks running from many centers,
the general tendency to rupture is satisfied. On the visible surface of the moon
there are near two-score recognizable ray systems, differing much in the distinct-
ness and extent of their light streaks. As these systems are widely scattered,
they are perhaps sufficient to have satisfied all the shrinkage strains of the crust
during the time when there were still vapors seeking to pass to the surface.

As to the age when the rays were formed, it appears evident that they
were not all made at or near the same time. Those of certain systems ap-
pear to cut those of other systems. Thus, according to Nicoll as quoted by
R. A. Proctor, the rays of Copernicus, Aristarchus, and Kepler cut one
another in an order indicating that they were formed in the succession in which
they are here named. It also appears possible that the greater part of the
ray systems were formed before the maria were produced, for relatively but few
extend over their fields, though it may be that their general failure to traverse
these bodies of lava and also those contained in the craters of the greater vulca-
noids is due to some condition of the material which diminishes the shrinkage
tension existing on other and older parts of the moon. That the light rays ante-
date certain of the rills, and perhaps all of them, is shown by the fact that they
are cut by these fissures. It should, however, be noted that Trouvelot, who had
a very keen eye, noted that certain of these open crevices are continued beyond

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 59

the point where they are distinctly gaping by slender streaks of shining material,
which appear, from his description, to be like the rays. It may be that vapors
ascending through open clefts would not be sufficiently concentrated to produce
a distinct band of color on their margins, while such would be the.case when
they mounted through an incipient fissure, as I have supposed to be the fact with
the radiating streaks.

In considering the succession of the ray systems, it should be noted that,
beside those which are definitely to be observed, there are evidently others in
part destroyed by later developed groups. In the best conditions of seeing, these
faintly indicated and evidently ancient sets of rays may be seen in all stages
of obsolescence, down to the state where they are conjectured rather than
observed. This, together with the phenomena of interference of one set of rays
with another, suggests that the process of their formation may have been continued
for avery considerable time, though the development of the larger bands appears
to have been brought about only in the later state of the surface, yet, as remarked
above, not to the very latest time of activity.

It is evident that the distribution of the several ray systems is not equal on
all parts of the moon. Thus the first quadrant has thirteen recognized groups,
while the fourth, just south of it, has but six. The second quadrant has eleven
and the third eight. Thus the eastern and western halves of the surface together
have the same number, but the northern hemisphere has twenty-four and the
southern fourteen systems of rays. Moreover, the greater number of the groups
are situated on that half of the visible surface wherein lie by far the greater part
of the maria, and on the surfaces of those lava fields none of the distinct centers of
radiation are found. This predominance of the rays in the regions of high country
near the maria may possibly be due to the extensive heating of the northern half
of the moon by the lavas which formed them, and to the consequent refrigeration
which would tend to develop crevices and thus lead to the production of rays.'

THE PRESERVATION OF THE RAY SYSTEMS.

The facts already set forth clearly show that the ray systems are fairly to be
regarded as features which have been somewhat gradually developed, and are, as
a whole, of ancient origin. It is, indeed, difficult to escape the conclusion that
they are, when measured in terms of geological ages, all exceedingly old. They

‘In the earlier years of my work on the moon, the results of which are here set forth, I noted
certain very faint rays which appeared to point to centers of radiation on the unseen side of the
moon. I have been unable to find the note-book in which these observations were recorded, and
my eyes, damaged by studies on that brilliant surface, no longer enable me to trace them. Accord-
ing to my memory, these streaks, as are all others near the limb, were faintly though distinctly
traceable, in the course of some years’ observation, to the number of about a score, indicating
about half a dozen such invisible centers. ‘The impression left upon my mind is that the very best
vision and opportunity might prove the existence of at least a dozen of these groups where the rays
converged to a point in the invisible field. The studies needed to determine this matter will
be difficult to make, for the reason that all these rays are faint in the regions near the limb.

60 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

must be judged to be the result of actions essentially like those termed by geolo-
gists solfataric, z. é., due to the escape of vapors from a heated sphere, which
have colored or coated the surface on which they lie. The considerations which
lead us to.believe that this internal heat of the moon vanished long before the
earth’s surface became frozen over are very strong. Accepting the view that the
light streaks on the moon are of exceeding antiquity, the question arises as to
why they have not been obscured by the fall of meteoric matter upon the surface
of that sphere. It is a well-known fact that some hundred thousand, if not some
million, meteoric bodies come upon the earth each day. It is true that nearly all
of these bodies are so small that they are burned by their friction in the atmos-
phere, and are added to our planet only as dust that descends in the rain or as
gases contributed to the air; but on the lunar surface, which, apparently, should
receive, per unit of area, quite as many of these fragments as the earth, there is
not, and probably never has been, an atmosphere sufficient to decompose these
wanderers so that they should have attained its surface unchanged.

Estimating the average diameter of the meteorites that come into our atmos-
phere at only a millimeter, which, in view of the light they afford, is probably too
small, it is evident that even in a hundred thousand years they would, if gathered
on the surface of an airless sphere, be sufficient to form a coating such as would
give a common hue to all its features, and in a geologically brief time the mass
would attain a considerable depth. Yet we have evidence in the ample grada-
tions of light reflected from the moon that very ancient features of color are as
undimmed by foreign matter as newly fallen snow. In other words, we seem to
be compelled to the opinion, either that there has been no such in-falling of mete-
oric matter on the moon as has of late taken place on the earth, or that the
whole scheme of coloring on the lunar surface has been formed within a few
thousand years. That the latter of these suggestions is not true is clearly indi-
cated by sundry considerations. It is, in the first place, to be noted that there
is much to show the absence of any accumulation of fragmental matter since the
oldest of the lunar features were formed. A meteoric rain such as comes upon
the earth for even a million years would have masked a host of objects which,
though presumably very old, are still manifestly unaffected by any such sheet of
dust as would have enwrapped the lunar sphere. Thus the exemption from
meteoric contributions appears to have been from a very remote time. More-
over, as before noted, the rays of different systems are of diverse ages, yet there
is no indication that the newer are very much brighter than the older.

As for the other possible explanation, 2. ¢., that the moon has not long
received meteoric material in the manner in which it now comes upon the earth,
there appear to be at first sight but two diverse ways that may have brought
about this condition. In the first place, the earth and moon alike may, until
very recent times, have been exempt from such contributions. In the second
place, it may be that the matter which falls on the earth is in whole or in large
part limited to materials which have been ejected from the planet by volcanic
action. The first of these suppositions must be regarded as possible, though

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 61

rather improbable. While there is no recorded instance of any meteorite having
been found in ancient geological deposits or elsewhere, save upon the surface of
the earth, the rarity of falls sufficiently large to escape burning in the air makes
it unlikely that they would be discovered in a fossil state, or, if found, that they
would be recognized as of meteoric origin, so that this consideration has not
much weight. On the other hand, if, as seems likely, the supply of carbonic
dioxide in the air depends in any considerable measure on the burning in it of
carbon meteorites, the presence of this material in something like its existing
quantity, certainly neither much greater nor much less, from the early geologic
ages, is evidence that meteoric falls, at least those containing carbon and of the
smaller size, have during that time been at about the same rate asat present. So
far as I can discern the astronomic conditions, it seems very improbable that
the earth should now be encountering a multitude of small bodies such as had
not come to it until within a few thousand years.

The suggestion that the meteoric matter which comes upon the earth may
have been expelled from it, though possible, does not seem to me to afford a way
of escape from our difficulty. It appears not improbable that volcanic action
may be sufficiently violent to impel bodies beyond the control of the earth’s
attraction. The shining clouds which were observed for some years after the
eruption of Krakatoa in 1883, and which went upward until they appeared to
escape from the atmosphere, may be instances of this nature. Moreover, large
fragments, which have been hurled forth by great eruptions, have been known to
fall at such distances from the point of ejection as to make it likely that they had
an initial velocity near to that which would be necessary to send them into space
and to make them independent of the earth; but, if I conceive the problem
rightly, such ejections would either in very rare instances fall upon the moon or
proceed to move in elliptical orbits, one focus of which would be the sun and
the other the place in space where the earth was at the time they separated from
it. It is eminently probable that in time these fragments would be apt to return
to the earth, but it seems evident that they would be about as likely to fall upon
the moon.

If we had any evidence that the moon had been surrounded with a fairly
dense atmosphere down to the present geological period, we might account for
the absence of meteoric dust upon its surface by the supposition that the smaller
bodies had been burned in its air as they are in that of the earth, but all the facts
at hand, which will be discussed below, are distinctly against this supposition and
in favor of the view that the low gravitative value of the sphere allows the gases
which do not become solid at the low temperature which prevails there by
kinetic action to move off into space; so that the development of an aérial
envelope has been impossible.

I have but recently come upon the difficulties we have to face in this problem
concerning the preservation of the surface of the moon from meteoric matter, and
am therefore not well prepared to discuss them. As they now appear to me, they
may be met by any one of the following described hypotheses: (@) That the

62 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

meteors which burn in our atmosphere are so minute that falling at their present
rate they would not have formed a dust coating had they accumulated on the
surface of the earth for all recorded geologic time, and that the larger masses,
such as now attain the ground, have been so rare that they would not of them-
selves form a coating. (6) That the earth and moon, as members of the solar
system, and sharing in its motion through space, are now in a field where meteors
are prevalent in a measure that was not the case in earlier ages, so that the
moon’s surface, though very ancient, has not been long enough exposed to such
in-fallings to have acquired a coating of them. (c) That we have entirely mis-
judged the antiquity of the moon, and that our reckonings, based on the law of
cooling bodies and on the supposition that the planet and satellite were differ-
entiated from a common nebulous mass, are altogether erroneous. Of these
suppositions, that designated as 4 seems the least objectionable, though as before
noted it presents sundry difficulties.

In considering the effects arising from the fall of bodies from the celestial
spaces upon the surface of the moon, we should take into account the fact that in
the present airless state of that sphere they would come upon its surface at
a very much greater velocity than when they break through the atmosphere of
the earth. Owing to the resistance of the aérial envelope of our planet, it is
doubtful if even the heavier meteorites have at the moment when they touch the
ground an average velocity above a thousand feet a second. Computations
which assign them a higher speed at the moment of contact are made doubtful
by the slight amount of their penetration into the soil. On the other hand, the
meteors which fall upon the moon must be moving at the average rate of at
least twenty miles a second, or about one hundred times as rapidly. Where they
impinge on the advancing side of the moon the rate would be much greater than
where they come upon it from the other or following side.

The effect due to the great speed at which meteorites would usually fall
upon the moon cannot be accurately determined ; certain of them can, however,
within limits, fairly be conjectured. It is in the first place evident that so far as
the penetration of mass into the crust was concerned it should be very much
greater than on the earth. On the assumption which has been above made as to
the comparative velocities, it should often be about a hundred-fold as great as on
the earth. It is, however, to be noted that the increase in velocity would
lead to a proportionate increase in the evolution of heat due to the friction
of the penetrating mass in its passage through the materials it encountered and
to the shearing of its particles on one another. Assuming the rocks of the lunar
surface to have the average resistance of pumice, it seems evident that any
meteoric body such as we know to fall upon the earth would not only penetrate
to a great depth, but would probably be volatilized by the very high temperature
it would attain. We see that a certain amount of this action occurs even in the
relatively slight resistance which a meteorite encounters in passing through the
air. With a resistance sufficient to produce an effective shearing movement in a
meteor, such as would be encountered on entering matter of the solidity of

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 62

pumice, we may fairly assume that the mass would, in effect, explode, the gase-
ous products being cast forth from the opening it made. The temperature pro-
duced by the arrest of the movement at a rate of twenty miles a second would
vaporize the mass.

It is also evident that on a surface in the present airless condition of the
moon all meteoric bodies, even the smallest, would come in contact with its rocks.
As is well known, by far the greater part of the meteors which enter upon the
earth are burnt in the upper air, and pass into the gaseous state or fall to the
ground gently in a purely divided condition. Such bodies, however minute,
would enter the moon’s crust at the same speed as the larger masses. Owing,
however, to their smaller bulk, they would be more quickly dissipated by the
engendered heat. If this view as to the volatilization of meteors by the conver-
sion of the energy due to their motion into heat is true, then the effect of any
such meteoric fall as takes place on the earth in, say, a hundred thousand years
would be to produce a mass of gaseous and dust-like material which should be
somewhat widely scattered from the point of impact of each meteorite, and this
for the reason that the gases evolved by the heat would enter into what is essen-
tially a vacuum and would be radially distributed at high speed, quickly to fall
upon the ground as their temperature lowered. The effect of such action
would evidently be to give the lunar surface a uniform color, determined
by the average light-reflecting quality of the resulting deposits of condensed
vapors and dust. If, on the other hand, we assume that the material bodies
penetrated into the moon without being volatilized, then the result of the first
falls would be merely to pit the surface, the color being destroyed for the area of
each pit, but when the successively formed pits became so numerous that they
occupied the whole of the original area the color would disappear. The effect
can be the better realized by firing successive charges of shot at a white plank.
As the number of penetrations increases to a point where the total amount of
lead is equal to a continuous layer, the original material becomes, in effect, covered
with the metal and takes its hue.

The considerations just above set forth make it appear eminently probable
that in either of the conditions in which we can imagine meteoric matter to have
come upon the moon, that in which it was vaporized or that in which it remained
solid, a period in a geological sense brief would suffice to obliterate the diversi-
ties of hue such as we find in the dark maria, the light streaks and patches, and
in its general surface. Thus the best interpretation which we can give to the
facts clearly leads to the supposition that our satellite has not in recent ages
shared with us in anything approaching like measure the falls of detached masses
from the celestial spaces.

On my first consideration of this matter I was inclined to believe that the
curiously pitted or honeycombed character of the lunar surface, which becomes
more and more clear as the magnifying power of the telescope is increased or
the seeing more favorable, might possibly be explained by the supposition that
the cavities were produced by the in-fall of meteorites of considerable size.

64 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON,

Many of these pits which may be seen in advantageous conditions are not more
than four or five hundred feet in diameter, and seem to have the general shape
that would probably be given them by the sudden effectively explosive develop-
ment of gases which we have seen reason to suppose would be brought about
by the penetration of large materials into the crust. Yet as there is no indi-
cation of a peculiar coloration of the fields about those pits, such as would be
produced by the precipitation of the condensed vapors, this interpretation must
be regarded as unverified, though it remains possible. Taking into account the
fact that the best instances of the honeycomb type of pits occur in tolerably clear
relation with the larger vulcanoids, it seems most likely that this group of depres-
sions owes its origin to the escape of indigenous vapors from the depths of the
lunar sphere.

The question as to the possibility of any of the distinct vulcanoids owing
their formation to the impact of large meteoric bodies is elsewhere discussed.
It is therefore only necessary here to note that, as the size of the in-falling
body increased, the heat evolved would be augmented, so that a mass a few hun-
dred feet in diameter would inevitably bring about such a general melting of
the crust where it fell that a cavity would not be formed, but in its place a
level blotch caused by the frozen lava, substantially what we find in the maria.
There are, indeed, sundry patches on the lunar surface which may have this
origin, but so far I have not been able to find any criteria sufficient to warrant
this interpretation of them.

The eminent probability that the fall of meteoric bodies on the lunar surface
should lead to the temporary production of a high temperature, suggests that it
might be possible by photographic if not by eye observations to detect these col-
lisions, if they occur with anything like the frequency per unit of area with which
they come to the earth. It is possible, though not likely, that these observations
might be practicable on the illuminated surface of the satellite, for the reason,
elsewhere noted, that as a whole it is more nearly black than white, and even a
small meteor would at its contact with the surface be likely to produce a flash
sufficiently brilliant to make an impression on a sensitive plate. On the dark
part of the sphere or even in a lunar eclipse it would probably be easier to make
the photographic observation. It is, however, to be noted that, as meteors enter
the crust at high speed and there is no atmosphere to give the train of light such
as is exhibited by those of small bulk which fall upon the earth, the flash might
be of very brief duration—so brief, indeed, that it might escape the eye and the
camera alike.

It may well be observed that, supposing the moon’s surface to have received
extensive contributions of meteoric matter, we might thereby possibly explain the
apparent degradation of some of its older features. On the supposition that the
in-falling bodies penetrated deeply and were converted into the gaseous state so
that they produced explosions, we would have an agency competent to break
down reliefs in the manner in which many of the ancient features seem to have
been mined. Yet when we note the exceeding sharpness of outline retained by

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 65

many structures, such as the cracks, displaced faults, and the smaller vulcanoids,
all of which must, on any apparently valid supposition as to the moon’s history,
be many million years old, we are led to believe this view inadmissible.

In this connection attention is due also to the fact that on the unilluminated
part of the moon various observers have, from time to time, noted patches of
light which they have believed to indicate volcanoes in activity. I have elsewhere
suggested (see p. 53) that these objects may have been highly reflecting parts
of the lunar surface illuminated by the earth-shine. It is barely possible, how-
ever, that in some instances they can be explained on the supposition that con-
siderable meteorites had recently fallen at the point where the light was noted.
So also it seems possible that the vapors which W. H. Pickering and others
have thought they observed floating in the manner of clouds on the illuminated
area may be in this way accounted for: a large meteorite penetrating deeply into
the crust might give rise to vapors which would continue to pour forth for months
or years after it fell. The difficulty with this hypothesis is to see how vapors
could float and remain in the form of a cloud in the conditions of essential vacuum
which exist on the surface of the moon. Granting the possibility of such action,
which in the present state of our knowledge seems improbable, I should much
prefer to account for these vapors by meteoric action than to seek their explana-
tion in true volcanic activity.

EROSIVE ACTION ON THE LUNAR SURFACE.

Those who are familiar with the lunar surface as it is exhibited by a good
telescope, cannot help acquiring the impression that there is some agent which
has operated on the moon in a way partly to break down the more ancient topo-
graphical features. There is an evident difference of aspect between the walls of
the older vulcanoids and those of newer formation. Apart from the distortions
of the ancient structures and the breaches of their ramparts, which may be fairly
accounted for in other ways, there are a rounding of their steeps and a general ap-
pearance of having been smoothed over by some erosive agency which are evident
in proportion to their antiquity. It is indeed a general fact which has been re-
marked by many observers, that the newer vulcanoids have an appearance of
freshness that is never found in the earliest formed. It is therefore important
to discover, if we may, what are the actions by which such changes may be
brought about.

On the surface of the earth there are four agents of erosion, all of which,
cooperating with gravitation, serve to bring about more or less considerable
changes. These are chemical alterations, which loosen the structure of rocks;
the direct action of the wind, which removes their lighter particles when they are
not protected by vegetation ; the action of moving water by waves, streams, and
glaciers ; and last, and by far the least, the expansion and contraction of materials
arising from changes of temperature. The essential effect of all these agents is
to deliver fragments of rocks to the more or less free action of gravitation, They

66 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON,

all act to send divided matter from higher to lower positions. Except the first
and the last, they incidentally provide means of carriage by which the fragments
may be conveyed to indefinite distances; chemical decay and the increase and
decrease in bulk due to variable heat acting by themselves do no more than give
the separated bits a chance to move down declivities of considerable slope.

It is evident that all the chemical change which occurs on the earth depends
on the presence of an atmosphere containing water. This condition apparently,
I think surely, does not now exist on the moon and probably, as | shall hereafter
give reasons for believing, has never existed there; for this reason we may set
aside this agent as a possible source of changes of lunar topography. From the
same facts we are led to dismiss the possibility of wind action. The only sug-
gestion of such work has been to explain the radial light bands on the supposi-
tion that the vapors emanating from the craters by their rapid diffusion caused
winds that blew the material which forms the rays to the places it occupies. We
have seen that this hypothesis does not account for the facts, and that they are
apparently explained by a much simpler view of the matter.

The idea of water having been at some time in the past an agent of erosion
on the moon is so persistently recurring that it is worth while to set forth, in
some detail, the results of my studies of the matter. I gave over fifty nights of
observing with the Harvard College Mertz refractor, which has an excellent
glass, to the question of a possible aqueous history of the several divisions of the
lunar field. The result Was to convince me that no part of that surface, new or
old, has ever been shaped by aqueous erosion, and this for the following reasons:
Aqueous erosion by river action has one characteristic effect: it, in all cases,
except where fot holes are formed by waterfalls, brings about a system of
continuous down-grades from the heights to the lower ground. My inspection
of the moon’s surface, which, from this point of view, was carefully made,
satisfied me that the streams had never done their inevitable work on that
sphere; for I was unable to find a single case of a depression of considerable
length having a continuous down-grade, or an instance where it might be sup-
posed that a valley, so shaped, had been subsequently deformed. None of the
rills which have been supposed to be stream-like in shape are in the least so to
an eye trained in terrestrial topography. They have no gathering grounds, no
trace of that digitated system of valleys which must have been formed if they
had been water channels; moreover, they have a perverse habit of branching the
wrong way, when they branch at all. Most selenographers have quite aban-
doned the idea that any of the features of the moon are due to water action,
though some of them adhere to the notion that there may be some slight trace
of water vapor in a supposed remnant of an atmosphere lying very near the
surface.

The same arguments that exclude river action on the moon will a fortzor¢
exclude glaciers. Both these forms of water require extensive evaporation areas
and the machinery of an atmosphere for their maintenance. Now that it is gen-
erally accepted that the maria are not and never could have been seas, but are

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 67

evidently lava fields, save in origin, essentially like those within the larger vulca-
noids, there is evidently no place where waters in a sufficient extent to supply
rainfall could have been stored. In this connection it is worth while to note that
on the earth, with two-thirds of the surface covered with water and with air
currents to carry moisture, large areas are practically unsupplied with water.
Without the oceans it is evident that rainfall would cease. The little which is
evaporated from the land would readily be stored in the air, perhaps to fall as
dew. So that lunar rains or snows would be impossible without a system of
great reservoirs, such as we cannot believe to have existed in any recorded stage
of the moon’s history.

There remains but one agent of erosion which can have acted on the moon,
2. é., that arising from the expansion and contraction of rocks in the changes of
temperature which there occur. On the surface of the earth, where the average
annual variation of heat on rock faces does not exceed about twenty degrees
Centigrade, and where the maximum variation is probably not more than fifty
degrees Centigrade, the effect of the variations is evident. Excluding, as far
as we may, the concomitant influence of freezing water, we find that the expansion
of rock is competent to produce cracks and to urge detached masses of rock down
the slope on which they lie. Thus the concentric structure which develops near
the surface in certain crystalline rocks, as granite, is due to the expansion of sum-
mer heat, which often causes the slabs of stone sensibly to lift from their beds.
On the surface of the moon, according to Langley’s observations, the range of
temperature is probably not less than two hundred degrees Centigrade, so that the
measure of expansion and contraction should be fourfold what it is on the earth.
Moreover, these alterations of temperature are repeated each month. During the
fourteen days’ insolation, the heat should effectively penetrate for some meters of
depth. Though it is doubtful if the melting point of water is ever attained, the
range is as effective in promoting motion as if it occurred above that point.

The effect of the great alterations of temperature in the superficial materials
of the moon is probably twofold; in the firmly imbedded rocks it must institute
successive strains and releases which should be competent to produce certain
effects not recognizable on this planet. Supposing that at a depth of three meters
the range of temperature was one hundred degrees Centigrade, the horizontal
thrust induced, if the rock had the modules of expansion of ordinary granite, would
be sufficient to produce in a sheet fifty miles in diameter an extension of some
hundred feet. From what we see of like action on the surface of the earth, we are
justified in supposing that sheets of great width would on the declivities of the
moon become separated from the subjacent materials and move over them in the
alternations of volume. So, too, we may suppose an interminable series of
varying adjustments which would, from time to time, bring about alterations in
the direction and energy of the thrusts which were thus induced. These changes
may have continued throughout a period as long as recorded geologic time, and
they may be in process of development to-day.

Another consequence of the variation in bulk of rocks in the changes of

68 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

temperature of the lunar day is that fragments lying on steep slopes would slowly
move down the declivities. Such detached blocks would, where they expanded and
in proportion to the efficiency of the gravitative impulse, press more vigorously
against the obstructions below them than on those above; they would thus gain
a chance to creep farther downward when they were again expanded. This
process would somewhat resemble what takes place where a talus slope is knit
together by a sheet of snow ice, when we may note a creeping of the united mass
due to the changes of temperature it undergoes. I have frequently observed
taluses where this process has extruded the deposit, as in the manner of a glacier,
far beyond the limits to which masses falling from the cliffs whence they came
ever attain. This process is yet more nearly alike to that which takes place in
the lead covering of roofs, where the metal has been observed slowly to work
down the slopes on which it lies in a movement evidently due to alternating
expansions and contractions.

At first sight it may seem that the relatively small value of gravity on the
surface of the moon would limit the movement of fragments due to expansion
and contraction so that the angle of repose in the taluses they formed would be
very high; but on consideration it appears to me that this angle may be even
lower than in terrestrial conditions, for the lessened weight of a given volume of
rock would greatly diminish the amount of the friction, and the value of the adhe-
sions which tended to resist its movements would, owing to the absence of water
and chemical decay, be so slight that I see no reason why, given time enough, the
talus material should not be brought to a nearly level attitude. The coefficient
of expansion is likely to be the same in lunar materials as in the igneous rocks on
the earth, while the resistances to such motion, both in the horizontal flakes of
great width and in the detritus on steep slopes, would be but one-sixth what it
is in our sphere. Therefore we may reckon on this agent of change being of
greater value on the satellite than on its planet, and find in it an explanation of
the worn character of the ancient topography which is not evident in the newer
formations. As we shall see below, this view as to the expansion of rocks may
be of value in accounting for certain possibly recurrent as well as accidental recent
changes in the shape of structural features on the lunar surface which certain
observations appear to indicate.

There is one rather obscure group of features on the lunar surface which
may be immediately due to the expansion of the superficial materials of the crust.
These are the numerous slight ridges which intersect the ground and which are
fairly visible near the terminator ; these ridges seem to me to be very low, perhaps
not more than a score or two feet in height. They are generally rather straight-
lined and so placed that they reticulate the level fields in which they lie, dividing
them into irregular blocks of very variable area, rarely more than fifty miles across.
I have seen what seems the miniature equivalent of this structure, where a sheet
of ice on a lakelet has been affected by great changes of temperature, all below
the freezing point of water, and has been broken by the expanding process into
blocks which, at their contacts, are crushed up into rude little anticlinals, formed

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 69

of ruptured bits of ice. These ridges of ice-fields retain their shape during the
contractions of the sheet in which they lie, as the blocks of stone in the moon may
do when they have found an adjustment. These lunar features deserve careful
study, though the conditions make an inquiry into their nature very difficult. I
have rarely been able to discern them clearly, and then for only a brief time.

ON THE POSSIBILITIES OF A LUNAR ATMOSPHERE.

The apparent arguments in favor of the existence of an atmosphere on the
moon, if not now, then in some former age of that sphere, are so strong that
selenologists are hardly to be undeceived by the evident facts that militate
against this view. These facts are, in brief, as before noted, as follows: There
is no trace of clouds on the moon; there is no difference in the clearness
of the seeing as between the lowest ground and that which is about six miles
higher ; there is not the faintest sign of diffusion of light on the line between day
and night; the effect is that which would take place in what we term a vacuum, but
not in the most attenuated part of the atmosphere that lies about our earth. More-
over, the course of the light of a star which goes behind the moon’s disc shows
clearly that at a mile above the lowest part of the lunar surface the air, if such there
be, has less than the thousandth part of the density of that belonging to the earth
at the same height. So, if there be any atmosphere at all on the moon, it is in
volume, at least, quite unlike that of our planet, and very like the nearest approach
to a vacuum which we can in any way produce. There is, indeed, no other valid
reason for supposing that any kind of gas or vapor exists about the moon save
that it is deemed necessary to have it in order to explain certain changes of
color which are deemed to be evidences of organic life. The value of this
evidence I shall consider below.

There is reason to believe that the moon has had upon its surface ample
material derived from the vulcanoids out of which to form an atmosphere. Re-
garding the lunar sphere as the offspring of the terrestrial, we may fairly suppose
that it received its share of the lighter elements of the original common mass
when the separation took place. If we regard the atmosphere of a celestial body
as the gaseous remnant remaining on its surface after the more readily solidified
elements have consolidated, then the moon should have had an original covering
of this kind on a scale proportionate to its total mass, z. ¢., it should have had an
atmosphere equivalent in weight to some inches of mercury. Throughout its
recorded history there has evidently been a great efflux of vaporous or gaseous
materials from below the crust, in total amount probably enough to have provided
an envelope in quantity as great as now lies upon the surface of this planet, yet
no trace of it remains. We cannot believe that the materials which should have
formed as air on the moon have been largely taken into the crust by chemical
action, as is the case on this planet, for there are good reasons to suppose that
there is no such action going on there, nor can we accept the suggestion that the
air-making gases have been frozen, for while the temperature is at times very low

7O A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

it is for a part of each month probably high enough to permit all the elements
which form our atmosphere to return to the vaporous or gaseous state. What,
then, is the condition which makes the moon an airless realm ?

It appears to me that there is a possible explanation of the lack of an atmos-
phere on the moon, one that has not been subjected to the inquiry which it
deserves ; this is, in brief, that the kinetic movement of gases causes their atoms
to fly away from the surface into space as rapidly as they are parted from the
solid sphere. I understand that this hypothesis has been adduced to account for
the separation of certain gases from our atmosphere which are held in that
of the sun; an extension of the same view may serve to explain the failure of the
moon to retain the gaseous materials which have evidently come to its surface
but which the gravitative attraction has not been sufficient to retain against the
diffusive effect of the kinetic movement.

THE EXISTING CONDITION OF THE MOON.

The idea that the moon should be the seat of some activities such as operate
on the earth is most natural. Again and again observers with much imagination
and with poor telescopes have seen what they took to be evidence of volcanic
action or of organic life on the surface. With the advance of selenography, these
views as to changes on the moon have been by better observations limited to two
groups of events. First, changes of form of certain craters, either those of a
cataclysmic and permanent nature, such as that which appears to have occurred in
the shape of the vulcanoid Linné, and the serial changes in certain other vulca-
noids, where the structures return to their original form; second, the lightening
or darkening of color of certain patches of the surface as the lunar day advances.
There are also some assertions of minor alterations which need to be separately
considered.

Of all the observations which point to the conclusion that changes are still
going on upon the moon, those which relate to the supposed sudden alteration of
Linné are the most important. This vulcanoid lies in the Mare Serenitatis, and
was mapped and described by several observers as having a crater about six miles
wide ahd with distinct steep walls. In 1866 it was believed that the structure did
not answer to the descriptions for in place of a crater there was found to be a
white spot of nearly twice {ts recorded diameter, and in the center of this field a
minute craterlet. Subsequent observations, however, have thrown doubt on this
conclusion, and led some selenologists to the opinion that Linné is a structure
that varies much under diversities of illumination, and that its variations of
aspect, combined, perhaps, with some original bad mapping and servile copying,
may account for the seeming change. Other instances, which appear to indicate
the sudden appearance of craterlets where none were observed by skilled sele-
nographers, are easily accounted for by the same difficulties arising from the
conditions under which we behold the lunar surface. Thus it has been claimed
that the lava flow of the vulcanoid Mersenius, which on close scrutiny is seen not

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON, 71
to be flat or slightly concave, as is usual in such structures, but quite convex,
indicates a change, for it must have been originally level as Schréter, as well as
Baer and Madler, so represents it. A study of this feature will convince any’
competent observer of the moon, who has had experience with his own work and
that of his fellows, that the peculiarity might easily have been overlooked. So,
too, with the craterlets on the southwest side of Copernicus, which have not
found a place on Baer and Madler’s map, and the continuation of the same
craters and a honeycombed appearance of the ground towards Eratosthenes,
which Schréter failed to notice. An inspection of the field with a better instru-
ment than those used by the above-mentioned selenographers will show that they
may well have searched it a score of times without having a chance to note these
rarely visible features. On the whole, the evidence for and against the sudden
appearance and disappearance of craters and craterlets, or of features in their
structures considered without reference to the probabilities of such changes
based on the moon’s history, leaves us in a state of doubt as to the occurrence of
such accidents. I am inclined to think that the case of Linné is the strongest and
that the walls of that vulcanoid may have, in part at least, fallen into the original
cavity so as to leave only a small pit in its crater unfilled.

If it be the case that the originally great ramparts of Linné have disappeared,
the event may be explained without having recourse to the theory of volcanic
action. Against the hypothesis of such action may be set the fact that, though
the moon is the subject of constant scrutiny, no trace of such explosive process has
been noted. Moreover, if there was volcanic action in the case of Linné, it appar-
ently must have consisted in an outpouring of very fluid lava, which formed the
extensive white patch that took the place of the previously existing rampart and
pit. In a word, the great wall must have been melted down into the flood.
When we consider the fact that none of the other vulcanoids shows a trace of any
such flows, that the evidence points to the conclusion that the lavas coming from
the interior of the sphere never freely stream forth but consolidate on slopes of
high declivity, we see how exceptional, and therefore improbable, is the occur-
rence of any such event. To the geologist it is inconceivable that in the late
stage of the moon’s history such an effusion of extremely fluid rock could have
taken place. The explanation he would give may be set forth as follows :

Assuming that the lunar crust as the seat of high and varied tensions of
contraction and expansion brought about its night and day, and that it abounds
in cavities due to the ejection of the large amount of material contained in the
ramparts of the vulcanoids, it is conceivable that from time to time ancient but
unstable adjustments may be suddenly disturbed. The state of the lunar surface
may in a way be compared with that of a Prince Rupert drop, a globular bit of
glass greatly affected by stresses which any shock is likely to set in effective
action. Now, if on such a surface a meteorite should fall, say a body of some
tons in weight, no larger than many that have come upon the earth, the resulting
shock might lead to widespread movements that would cause the walls of a vulca-
noid to fall in. It is to be noted that there are many ill-defined pits on the moon

fiz, A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON,

which may have had this meteoric history, and also that Linné is situated on a
wide mare where such stresses are indicated by the continuous ridges and
the rills, and where they would be more likely to accumulate than in the higher-
lying, irregular country. So if the supposed destruction of this vulcanoid really
occurred, a point which will ever remain doubtful, it may thus be accounted for
by other than volcanic action. It needs to be so explained if we are to retain our
conception of the moon as a sphere which has lost heat in the ratio that the earth
has lost it.

The supposed variations in the shape of the twin craters known as Messier,
changes which appear to pass through something like a cycle in the course of a
lunar period, may possibly be due to the movement of extended masses of rock
under the influence of solar heat. Assuming, as before, that a sheet of rock on
one or more sides of the pits had, because of its expansion, developed a horizon-
tal joint a few feet below the surface, this slab-like mass might slide to and fro
with the variations of temperature. The expansion of a sheet fifty miles in
diameter might amount to several hundred feet, enough to make evident altera-
tions in the shape of the cavity. That some such migrations of rock masses
under terrestrial compressive strains are possible is abundantly proved by the
studies of geologists. Movements of ten miles or more are well ascertained ; the
only question is as to the possibility of a field of rock, such as we are considering,
returning, in the process of shrinking, to its original position. On the earth, such
a plate of stone would most likely be fractured as it cooled, so it could not return
to its first state. On the moon, however, such a mass, because its weight is
less than one-sixth what it would be on the earth, would encounter less friction in
its movements; moreover, the grinding action of the adjacent surfaces would
tend to form a mass of powdery matter between them which would readily shear
so that the frictional resistance would be relatively small. The difficulties of
this hypothesis are obviously great, but if it is finally determined that there are
recurrent changes in process on the moon, such as appear to some observers
to take place in Messier, it seems preferable to that of volcanic action, for
it does not do violence to all we know concerning the processes of a cooling
sphere.

We turn now to the changes of hue of certain fields of the lunar surface
such as have been observed by W. H. Pickering and others. These changes are
of two somewhat distinct kinds, those which appear to that observer to show the
discharge of fumes from certain small craters, and those which are thought pos-
sibly to indicate the temporary development of an extended vegetation which is
born in the brief season of a lunar day and dies in its night. As regards the
blotches of color which seem to indicate eruptions, I’have had no chance to see
them, but from the account of the phenomena it appears most likely that they
are due to peculiarities of reflection much like those which make the rays glow
when the sun attains a high angle. The arguments against the existence of any
such clouds of vapor floating above the surface of the moon are very strong ;
they seem to me, indeed, to be insuperable. The phenomena of occultation

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON, 73
prove, as already noted, that at a mile above the surface there is no trace of an
atmosphere,—surely not more than the thousandth part of our own. The law of
diffusion of gases makes it impossible that there should be any great increase in
the density of such air at its contact with the sphere. How, then, could vapors
slowly float away as clouds from a crater? If they came forth they should be swiftly
and uniformly diffused in the essential vacuum. Change of hue due to the angle
of illumination or fluorescence, or both actions combined, affords a far more satis-
factory explanation of the observed facts. This explanation has difficulties, but
they are much less serious than those we encounter in a hypothesis of volcanic
action still existing and producing clouds.

The observations which indicate that extended fields of the lunar surface
darken with its advancing day are extremely interesting for the reason that they
show a departure from the general tendency of the surface to become brighter
with the higher sun. There is no doubt that these changes are of great import-
ance, but I cannot regard them as suggesting the development of any kind of
organic life. This question as to the probability of life on the lunar surface has
never been adequately discussed, and as the suggestion is recurrent I purpose to
set forth below certain considerations which, in my opinion, make it appear to be
most improbable that anything like organic structures can possibly develop there.

It is, in the first place, to be noted that all organic forms, from the lowest to
the highest, plant and animal alike, absolutely depend for their existence on the
solvent action of water on various substances. The conditions of life are that
this water shall be readily obtainable either directly from the fluid in which the
creatures dwell, from the rain, or from the moisture of the air. In all cases
this water must contain free oxygen and carbonic dioxide, as well as certain
minerals in solution. Although it is stated that certain lichens develop in rocks
within the antarctic circle, where the temperature has never been observed above
the freezing point, it may be safely assumed that these plants have now and then
received during their growing period and have retained in their bodies water in
the fluid state, otherwise their organic processes could not go on. Wherever on
high mountains, say above the level of 20,000 feet, the surface of the rock has
been examined, no resident life has been discovered. Thus in an air which is
surely many times as dense as any that can exist on the moon, terrestrial life, for
all its ample opportunities to become reconciled to such environment, has not
succeeded in establishing itself at these great altitudes. The conditions for the
formation of organisms suited to the higher peaks of the earth are vastly more
favorable than they could have been on the moon, yet the result is that they
have failed to develop in such conditions.

Whatever were the circumstances, as yet unknown, which led to the be-
ginning of life on this earth, they were evidently of rare occurrence. The succes-
sions of organic forms suggest that they have been derived from few if not from
one original form; and, further, that these initial stages have long since been
lost. It is unlikely that fresh starts in the origination of the lowliest organisms
are now making, for with all the skill of a host of well-trained inquirers we have

74 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON,

not been able to initiate an organic form. If the existing living species of this
earth were destroyed, we do not know by what process a beginning could again
be made. So much, however, is plain: First, that all of the existing organic forms
have had the initial stages of their development in aquatic conditions, for there
alone can the earlier stages of development be attained. Second, that the
aqueous stages of the forms which now inhabit the land must have required
a very long period of such life before the creatures were ready to enter on the
more difficult conditions of the land. It may safely be presumed that a period
of development such as is represented by thousands of species of successive
forms was necessary to bring the terrestrial organisms into conditions of structure
and function where even as the lowliest plants they were fit for stations in the
air. This process of reconciliation with the environment demands, among other
things, means whereby the spores may be diffused, and with all plants of rapid
growth, such as have to be assumed if they are to give color to the surface of the
moon, it requires a soil or air for food supply.

It is a favorite assumption with selenographers who adopt the hypothesis
of plant life on the moon—a pure assumption—that there may be a thin atmos-
phere of carbon dioxide next the surface and that in such an air plants would
grow with rapidity. This is a natural view, for it is based on the well-known fact
that the carbon of plants is largely obtained by the decomposition of that gas, the
carbon being taken into the structure and the oxygen set free. But the experi-
ments made by a committee of the British Association for the Advancement of
Science clearly showed that terrestrial plants, even the lowlier cryptogams, were
not sensibly helped by an increase in the amount of C O, in the air and that any
considerable augmentation of that gas was hurtful to them.

Therefore, in view of these facts: that terrestrial plants, notwithstanding all
their ample opportunities for so doing, have never been able to reconcile them-
selves to the conditions which exist at heights where the density of the air is not
more than one-third of what it is at the sea-level ; that all organic life necessarily
had its beginning in the seas or other masses of water; that the conditions of its
origin are so peculiar that we have never been able to reproduce them; and that
the development of every organic species known to us requires a considerable
supply of water,—it appears most unlikely that the moon is now or has ever been
the seat of organic life of the sort that exists on this earth.

It cannot well be denied that there may be on the other celestial spheres
than this earth forms of association of matter in which other fluids than water
may serve as the menstruum in which vital activities develop, and that the essen-
tial results accomplished in the organic forms of our planet may be thus attained.
But, so far as we know, organic individuals are limited to very narrow conditions:
to those in which water is exposed to temperatures between the freezing point
and about sixty degrees Centigrade, and which afford air such as that of the earth
in density equivalent to not less than what corresponds to a pressure of one-third
that normally existing at sea-level. These conditions clearly do not exist, and, so
far as we can determine, have never existed on the lunar surface. It is, in fact,

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 75

very doubtful if any other body in our solar system with the exception of the
planet. Mars is now in a viable state, for it is not likely that any but the earth, and
possibly the Martial sphere, has the necessary combination of water and solar
heat which has long existed on this planet.

The foregoing considerations concerning the possibilities of organic life
on the moon show clearly that we must exhaust every valid hypothesis to explain
the occurrence of changes on that sphere before we assume that they are due to
the development of living forms. I would suggest that the patent facts of color-
change shown by the blotches and rays, which gain intensity as the sun goes
higher, lead naturally to the supposition that these.other conditions of darkening
are due to a like though somewhat diverse action. We may fairly suppose that
the regions which thus darken are covered with crystals which reflect or refract
the sun’s light in such a manner that they send us less of it when the sun is about
vertical than when it is relatively low. We have command of three certainly
warranted agents for explaining changes of color in the moon: that of reflec-
tions from crystalline surfaces; that of refraction taking place in the interior of
translucent crystals; and that of fluorescence. We havea right to combine these
actions as needs be to account for such phenomena of varying color as may be
observed, for all of them are well within the limits of what we note on the earth,
but we have not a like right to bring in hypotheses of organic life when all we know
of its conditions on this planet shows that it cannot exist on the lunar surface.

It is naturally painful to conclude that the moon is and always has been
deprived of those features of existence which we deem the nobler; that it has
never known the stir of air or water or the higher life of beings who inherit the
profit of experience and thereby climb the way that has led upward toman, That
these large gifts have been denied to the nearest companion of the earth has its
lessons for the naturalist, since it clearly shows how vast are the effects arising from
the interrelation of actions. The fate of our satellite was probably in large part
determined by the ratio between its gravitative force and the energy of the kinetic
movement of the gases such as constitute the atmosphere. If that energy had
been sufficient to retain them on the satellite, there is no reason,'at least so long
as the original rotation on its axis continued, why it should not have had the
history of a miniature earth. As it is, from the beginning it appears to have
been determined that it should have no share in the solar energy which has given
the most of the dynamic and all of the organic activities of the earth, and there
is no imaginable accident that can alter its state except some catastrophe which
may return the solar system to a nebulous mass. Just as it is, our moon is likely
to see the sun’s light go out.

SUGGESTIONS CONCERNING THE STUDY OF THE MOON.

From the point of view of the geologist and geographer I venture to make
certain suggestions concerning the future work of selenographers. In the first
place, it may be said that, while the delineation of lunar features has, within a

76 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

century, been so greatly advanced that of the visible part of the moon we have
within the limits of telescopic vision much better maps than of most parts of the
earth, the classification of features and their nomenclature are in a very crude
shape. There is nosufficient categorizing of the various features, and the names
for them, generally suggested by misconceived analogies with terrestrial objects,
are often misleading. They serve, indeed, to perpetuate grave errors as to the
real nature of the lunar surface. Many of the most conspicuous topographic feat-
ures are unnamed, as, for instance, the promontories and capes along the shores
of the maria. Much of the nomenclature is so inwoven with our records that it
would be inadvisable to disturb it, but many changes and additions could be made
which would bring some order out of the confusion. I therefore venture to sug-
gest to selenographers that a committee should in some way be formed to under-
take a revision, or at least an extension, of the system of names applied to the
topography of the moon.

As to further detailed work on the moon, it appears highly desirable that
small selected areas should be jointly studied and depicted by several well-trained
selenographers, the task being done in such a manner as will enable us to form a
judgment, first as to the effects of the personal equation of individual observers
in seeing and depicting lunar features, and second as to the effect of diverse con-
ditions of seeing, including the libration, on the aspects of lunar surface. In this
way we may hope to attain something like certainty concerning the occurrence
or non-occurrence of changes.

It is also desirable that a close comparison be made between some of the
more ancient vulcanoids and those of evidently much newer age, as determined
by their relations to one another, and this with a view to ascertaining what are
the angles of slope of their respective ramparts and those buttress-like structures
which I have assumed to be flows of viscid lava. In this way we may possibly
obtain some idea as to the effect of the expansion and contraction due to solar
heat, or other forces upon their reliefs.

A closer study as to the presence or absence of ash and other ejections of
fragmental materials than I have been able to make is desirable. I have given
reasons for believing that no such violent expulsion of broken-up lava, z. ¢., vol-
canic breccias or ash, took place in the eruptions of the vulcanoids; but the
proof of this rests necessarily on negative evidence which requires much scrutiny.
This should be given to those cases where large well-developed craters lie adjacent
to older like structures. Where there is a honeycombed structure or old ram-
parts near such newer craters, the surface should be narrowly scanned to find if
the depressions have been filled with débris.

The observation of Trouvelot, that the rills are sometimes continued beyond
their open fractures by light streaks, needs to be verified, for proof of such con-
dition would go far to show that some of these bands at least are due to the pas-
sage upward of vapors which congealed at their point of escape, and afford a fair
presumption that all of them are of this nature. This inquiry should be extended
so as to determine if any of the radiating streaks are coincident with distinct rills.

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 77

The form of the so-called mountains demands a careful inquiry. It is
asserted that in some cases the steeper slopes, in certain groups of these eleva-
tions, are all, or prevailingly, in a particular versant; this point should be
determined. It appears likely that the mountains in different fields vary in shape
in a manner which will permit them to be classified according to areas. All such
variations are sure to have meaning. As a part of this work, the cones in the
center of vulcanoids such as that in Theophilus should be compared with the
peaks in the mountain systems.

I have noted that the older vulcanoids in the central field of the moon’s
surface appear to have been elongated or ‘‘spooned” in a north and south
direction, and that this change may be due to the loss of the original rotation of
the sphere. This point needs further study. If my observations be verified,
and it be found that the newer vulcanoids are not deformed as by a collapse
of the equatorial bulge due to the loss of rotation, then the time of the change
in relation to the development of the surface features may be determined, and as
the loss of rotation would have been very gradual it would be incidentally shown
that the period during which vulcanoid processes affected the surface was very
extended.

The phenomena of contact of the maria with their shores needs close study.
I have briefly stated the facts which lead me to the opinion that the lavas of
these fields originally and for a brief time rose much above their present level
and have since withdrawn from low areas they at first flooded over. If this be
affirmed, then we have evidence that the order of fluidity of the lavas in question
was far higher than that of the vulcanoids, where, as we have seen, the material
appears to have been at a low average temperature, or at least very viscid, so
that it consolidated on very steep slopes as soon as it escaped from the craters.
Much depends on the determination of the relative temperature of these groups
of lavas, for if those of the maria were decidedly hotter than those of the vul-
canoids—hotter, indeed, than any molten material which is known to have come
forth from the interior of the moon or the earth,—then the presumption that they
were due to in-falling bodies is so far affirmed.

It is most desirable to ascertain the circumstances of contact of the lavas of
the several maria which are obviously connected. If they are the result of the
impact of one falling body, or of several which fell at about the same time and
place, then the various connected areas should be perfectly confluent. If the
bodies fell here and there, affording separate centers of melting, then there may
be a trace of juncture of the lavas where they joined their floods. My own
opinion, based on rather scanty observations, is that the confluence of the appar-
ently connected maria is complete, and that their lavas were generated by one
incident ; the distinctly separated areas, the Mare Crisium and the Mare Australe
as well as the Mare Humboldtianum, if the two last named be, indeed, true
maria, having been formed apart from the main field, which includes all the other
areas classed in this group.

The naturalist, trained in interpreting terrestrial phenomena, learns the value

78 A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON.

of series in extending his conceptions. It is important that this method of
inquiry should be applied to the features of the moon, so that we may have a
foundation for a sound knowledge as to the categories into which its structures
fall, and the limits of these groups. Thus with the group of vulcanoids a close
study of their features by making extended series of their forms will be likely
to bring out relations and diversities not now understood. Besides these notable
structures, the mountains, cones, and ragged peaks, the rills and valleys, the light
streaks and blotches of color, all bespeak the same treatment. It may, indeed,
be applied to many other groups of objects.

It is obviously desirable to gather all the information we can concerning
the unseen ;4)5 of the lunar surface. This inquiry I undertook more than
thirty years ago, but the task was left incomplete. The method of my inquiry
was as follows: on the limb of the moon in the successive extremes of libration
so-called mountazus appear. Several of these ranges have a continuity which
is found only with the ramparts of the great vulcanoids. Of these, beginning at
the north pole and passing by the west around the limb, I noted the range west
of the Mare Crisium, another near Neper, the Leibnitz range near the south
pole, the great range beyond the Doerfel Mountains, and a succession of like
ridges down to ten degrees north. These and other fainter undelineated features
appear to be resolvable into arcs of circular ramparts, such as enclose the larger
vulcanoids. Plotting these as circles, the result was to establish, by fair hypothesis,
over a considerable part of the unseen realm, the existence of a topography like
that we see.

Looking closely at the limb of the full moon, observers with good eyes may
agree with me in the opinion that certain faint light rays there discernible,
though with difficulty, apparently converge to centers on the farther side of the
moon. I brought to book enough of these to establish about half a dozen of
these centers on the invisible field. A confirmation of these uncompleted
observations would reduce the region of the entirely unknown part of the moon
to less than one-fifth of its whole surface. I cannot hope to return to this inter-
esting task of looking around the edge of the moon, but it appears to be the
most interesting of the many inquiries that demand good eyes, and opportunities
for observation when the rays are most clearly visible.

Owing to the difficulty of interpreting objects seen in very oblique con-
ditions, the fields within five degrees of the limb have been much neglected.
Among the problems there found is that concerning the existence of maria on
the margin of the observable part of the surface. Except possibly in the case of
the Mare Australe, the surface of such areas is not visible. I have never been
certain that I saw the characteristic dark plain of that mapped sea. The ques-
tion is whether it be only a little varied ancient portion of the crust or a true
mare. It is also a question whether the tips of high peaks are not to be traced
on the other side of the comparative level ; if this be the case, then it is, if a mare,
one of small area. The so-called Mare Humboldtianum also needs close atten-
tion to determine whether it be a mare or, as it seems to me, an ancient vul-

A COMPARISON OF THE FEATURES OF THE EARTH AND THE MOON. 79

canoid of large size with rather low walls. If it should be proved that these
so-called maria do not belong in that clearly-defined group of features, there will
be some reason, from their distribution, for believing that they are limited to the
hither side of the sphere.

There are many other lines of work beside that of simple delineation, to
which selenographers have so generally confined themselves, which may well
engage the attention of those who desire to advance the theory of our satellite.
Some of these have been suggested in this memoir; others will present them-
selves in the course of further inquiry. In such work it should be borne in mind
that, relatively few and simple as are the forces which have acted on the moon, in
comparison with those which have shaped the earth, they are, in their effects,
very complex. The variety of objects on that surface is very much greater than
the existing accounts of them would lead the novice to suppose. It is only as
they are compared after the manner of the naturalist that we may hope clearly to
read the wonderful record of that marvelous dead sphere.

7» eee Te aya

tote Y thy

a
Pe Oi

‘hw Ny

it Sy CI WS ap eit
eee mul} a

tc, Sos
/ Weis pe)
j pe i

j
ih
By

ve

wy
J
ih }

jab OME alts be ia
> a! of me

fan: a

DESCRIPTION OF PLATES:

The following plates have been selected with reference to the illustration of the questions
discussed in this memoir. The choice of illustrations has necessarily been limited to those features
of which it has been possible to procure good photographic negatives. On this account many
interesting structures are not pictured. As a whole, however, these delineations fairly present the
more important aspects of lunar topography as seen with good telescopes.

In accordance with the usage of selenographers these plates are printed in the reversed order
in which they appear in a celestial telescope. The top of each is the south, the bottom the north,
the right hand the east, and the left the west. This will enable the student to compare them with
the maps of the moon. Except when necessary for the immediate purposes of this memoir, the
structures depicted in the several plates are left unnamed. On many accounts this omission is to
be regretted, but an extended effort to designate by name the craters, mountains, etc., showed that
to accomplish this end it would be necessary to have key maps for the greater number of these
illustrations. If the student desires to determine the name of any of the more considerable
features, he can readily do so by comparing the plate with any of the good maps of the moon.
For this purpose the map of Elger is recommended.* The photographic atlas of the moon by W.
H. Pickering, in the Annals of the Observatory of Harvard College, vol. li., 1903, and the same
work in a more popular form entitled “ The Moon,” by Doubleday, Page, & Co., N. Y., 1903, will
be found very useful for reference. Other reference would have been made to them in this work,
but they were published after the pages which precede this were put to press.

In the description of each plate, attention is called to the more important features which it
depicts and occasionally to the place in the text where the matter is discussed. This arrangement
of necessity causes many repetitions. It is hoped that the reader will find that the convenience of
the method compensates for this awkward mode of presentation, the aim being to provide in the
illustrations a basis for a criticism of the theories of lunar structure as near as possible to that
afforded by the use of a telescope.

It is suggested that those who desire to spare their time in obtaining what value this memoir
may yield, should first read the text and then compare its statements with the facts presented in
the plates ; remembering that the matters of detail, such as those concerning the rills, the light
streaks, and the other more delicate features, can not yet effectively be rendered by photographs.

1See The Moon, by Thos. Gwyn Elger. London. Geo. Philip & Son, 1895. The map is to be had separately
from the volume. «

81

ae
‘
' ft
jv aia
ye Fs
Lj ry Ph,
ial y aan
oad ys “a
ea ob.
= al Se

re
| | test
i eae
| Niet Wie @aigny
io ~ inl
a VMS
| Wi fit, ipl if ' i

; -4
fe ays Z
if ; ? S
LSet 2 ' Lika ¥ P-s9hi4
193
6 ni
a wis
# 100 ; ii P renesel E
ebsi 0 frseGiy: rt! Sc bene ty 2
ot * ¥ r » . > c
ipmiey sth. silt ee edie, eau tal Grin? ee af
a, niiGe S0t Hide ctibisgie “eae ere FRE etitirle) sath dees :
Sit te 33.5 { STIS wea O -= 4 bas ee Sy etna Tao as cl ometesar
196 Soneieni enowor F A eR 4) BEE I l ; 9
any tern Po i a ; * ‘
Teiceeas ie es) 7 Tog ahs ,
uiG : ‘ crtt Gah dy ft f
2 P ' y
4 4 a nf foe =
iZitd 2 aes : n
i
.
ae
-
7

oe in 2 See eee =e ee te ce

PEATEA:
GENERAL VIEW OF MOON, AGE 6 DAYS. BY S. W. BURNHAM, LICK OBSERVATORY.

Plates I to VIII (inclusive) show the surface of the moon in progressive stages of illumination.
Taken at the Lick Observatory.

In plate I the moon appears nearly half full. The crater of Abulfeda is coming into illumina-
tion. .The most noteworthy features are the maria, which are evidently darker than the general
surface. The lowest of these, the M. Serenitatis, is obscurely circular with rather definite margins.
In it, on the west or left-hand side, are some faint folds of its floor. Just outside of this sea, to
the west, is a rather large distinct crater (Plinius). Horizontally eastward (to the right) in the
midst of the sea is a smaller dark crater (Bessel). The same line continued about as far still east-
wardly shows in a faint white spot the position of the crater Linné, which is supposed to have been
destroyed in 1866 (see p. 70). The mare on which Plinius stands is the M. Tranquilitatis.
Next southwardly beyond Theophilus (Plate XV) is the M. Nectaris. On the southern (upper)
margin of this sea is the crater of Fracastorius of which the northern part of the rim has evidently
been melted down by the sea. This is perhaps the most conspicuous instance of this nature among
the several score that may be noted on the margins of the several maria. The northernmost of the
maria in this view near the lunar margin is the tolerably circular M. Crisium. South of it is the
irregularly shaped M. Foecunditatis, without distinct boundaries.

The observer should note the considerable range of brightness in the field, also how the craters
and other features become fainter near the brightly illuminated margin.

82

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL

ERAL VIEW OF MOON, AGE SIX DAYS. BY S. W. BURNHAM, LICK OBSER\V ATORY.

PLATE Tt

MOON 7 DAYS OLD. BY S. W. BURNHAM, LICK OBSERVATORY.

This plate shows the moon one day older than the preceding view. By comparison with plate
III the effect of twenty-four hours’ advance in the lunar day may be perceived. On the “termin-
ator ’’ or border of the advancing sunlight, a number of large vulcanoids may be seen in a tolerably
linear order. The most important of these, beginning with that nearest the equator and reckon-
ing southwardly, are Ptolomzus, Alphonsus, and Arzachel, then with an interval come Purbach,
Regiomontanus, and Walter. Traces of a like alignment are visible in other groups of lesser
vulcanoids.

At this stage of the illumination some of the light streaks or rays begin to be visible, and may
be faintly traced on the left-hand side of the plate when the sun is highest. So, too, the bright
patches whence most of the streaks emanate, are beginning to become lucent.

E, Vc

OG

IWLE

¢

S TO KNO

ONS

SMITHSONIAN CONTRIBUTI

SERVATOI

)

LICK OI}

=
a
Zz

BI

YS. W

)

DAYS OLD

MOON SEV

Wise
Haale. cee

Wwe:
Ly,

Sam! Sele
INES

PEATE AM
AGE OF MOON 8 DAYS, 4 HOURS. SEPTEMBER 22, 1890. LICK OBSERVATORY.

In this plate the most noteworthy features are the maria of the western half of the visible
portion of the sphere. The rudely circular form of these fields is well shown, also the fact that
none of them extend to the margin or “limb” of the moon. The bright, slightly curved ridge
in the lower half of the picture facing the partly illuminated mare, the Mare Imbrium, is the
Apennines ; the large vulcanoid at its southern end is Eratosthenes. ‘The larger pit in the ocean
opposite the center of the range is Archimedes ; the two craters next to the north are: the nearer,
Autolycus, and the farther and larger, Aristillus. The larger of the two dark pits near the north-
ern end of the Apennines is Eudoxus, the smaller Aristoteles. Southeast from these craters lie
the Alps, a group of bright peaks extending in a northeast and southwest direction. A faint
dark streak shows the position of the Alpine valley. The flat, irregular area north of the range
is the M. Frigoris.

Close inspection of this plate will show that many of the vulcanoids have pits or cones on
their floors, and that these are very often in the center of these level spaces.

The radiating bands or streaks are beginning to appear.

In the Mare Imbrium, near the western end of the Alps, next north of Aristillus, is Cassini,
of which the encircling cone appears to have been partly melted down by the lava of the mare so
that it shows as a faint ridge with a distinct central crater.

86

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV PLATE III

AGE OF MOON EIGHT DAYS, FOUR HOURS. SEPTEMBER , 1890. LICK OBSERVATORY

ike REE FOE

asda skiheget | sions ese

S12

&-
7 C <
_ >
‘
Fi

Gnrbtezenr ttt

ton basligi. ons Deen Gs te

analy Br

eos

PLATE IV.
MOON’S AGE 8 DAYS, 22 HOURS. LICK OBSERVATORY, 18g0.

This plate represents the moon as it appears eighteen hours later than shown in the preced-
ing plate. The pictures were taken at different times of the year, which accounts for the difference
in the position of the terminator or illuminated margin. It will be observed that several new
features have appeared beyond the southern end of the Apennines. The light bands are more
visible and the contrast of hue between the maria and the upland country is less distinct.

88

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV

MOON’S AGE EIGHT DAYS, TWENTY-TWO HOURS. LICK OBSERVATORY, 1890

PLATE VE
MOON’S AGE IO DAYS, I2 HOURS. LICK OBSERVATORY, 18g0.

The moon as delineated in this plate is thirty-eight hours older than as shown in the preceding
plate. The most noteworthy changes are the great advance in the development of the fields of
very bright hue, and in the bands radiating from them. These are most evident in the system of —
Copernicus. The system of Tycho also begins to be evident. This vulcanoid may be identified
as the deep large crater with a central cone near the border of the illuminated area. The general
irregularity of these light bands is well shown in those about Copernicus. So, too, the fact that
they are projections from an illuminated or lucent field about the vulcanoid.

On the shores of the Oceanus Procellarum, east of Plato, near the margin of the sun-lit area,
is the Sinus Iridum. This is probably a large vulcanoid which has had the part of its wall next
the mare melted down by the lava of that field. (See p. 17.)

The relative absence of large vulcanoids on the maria is noteworthy. Those which exist lie
nearly, if not altogether, on fields of high ground which appear to have risen above the floors of
the maria and so escaped melting.

The problematical crater Linné now appears as a small white patch near the middle of the
eastern side of the M. Serenitatis. (See p. 70.)

90

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV

MOON’S AGE TEN DAYS, TWELVE HOURS. LICK OBSERVATORY, 1890

. *
= : ‘
‘ SP
led $ACES ;
TeRT 131 fa) SSP IGH aS eSReG kt Beak eee f
sil spin xii od win yiis sity {18% Shean 1 Toone rh Sh te aint} rel , s
ae es scot sistas 0151249 St2 ne ellew = =
riinlon teal 6 vd Horcwye2 nin wht eieks sat os ra
baie Pee a Abieetiice binnsstiry sel sat 2 :
~ amen out ea stash be yliken {aoa rsetal f Anon woah!
, “murs vou iz 8) se ord shell bowie
Walitoe dr 7: {how sith Saiereel sada oe thiteaa) fa
isy alt Je Hv el Si) a Vor eee psa
Sth yraisotowl Me Sdr 3o-copmre aide?
sbig Disesse SHY 1406
! to Spon sud hsyon sd itvotadtt
iat ci ut it fSsties Sli rh)
(ait Feo wis Sing Stary @t fant ;

pis) sedis Seigk tobias sepeditge 8a) Ole ai -
giedo Srl fon Bit seit fire! teres otis

+ yiveuy eis Sd hoveseto ad Hy SE

ait detnl Dis: sit oeert:
“i pe ab S70) aris

és Seon, sidtery yitowiierh steyineaein dass any

1) (ets oie lSe-eteini7 statin oy? Sin ail) roa =

Jrmse Jigihia os Uliradigeé

de

te

PLATE, Viz
MOON’S AGE I4 DAYS, I HOUR. JULY 19, 1891. LICK OBSERVATORY.

In this plate the moon is nearly full, the light being oblique enough to illuminate the crater
walls on the eastern margin alone.

The maria are well shown nearly to the eastern margin. Separated by a belt of relatively high
ground from the Oceanus Procellarum is the large vulcanoid Grimaldi. It has a small crater on
its floor near its northern side. This vulcanoid has a floor nearly as dark as the seas. It will be
noted that Plato has also a dark floor. On the margin of the Oceanus Procellarum, southwest
of Grimaldi, is a crater Letronne, the wall of which that faces the maria is, as in other instances,
ruined apparently by the lava of the sea. Other like examples are shown in this neighborhood.
On the shores of the M. Humorum, there are three similar instances of crater-walls broken down
on the seaward side.

It should be noted that none of the maria distinctly attain the margin of the moon’s surface.
On the eastern lands the O. Procellarum comes near to the border of the moon, but high rugged
land is visible on the very edge. This is more clearly disclosed at certain stages of libration.
On the southwest border some observers think there is a sea crossing the border, but, as will be
seen, the level land there has not the characteristic dark hue of the maria,

It will be observed that in this nearly vertical light, except Grimaldi and Plato, the craters on
the eastern margin only are distinctly visible. ‘Those exceptions are due to the dark color of their
floors, There are two or three craters near the south pole which, because they have rather dark
bottoms, are faintly seen,

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV

MOON'S AGE FOURTEEN DAYS, ONE HOUR. JULY 19, 1891. LICK OBSERVATORY

ity ath:

HBS ! AAS cat re 1s 4h a ‘Yoox ae

Aon app AioTS As thy PRIS 2) Noon ve aislq fade bs

kat 9 acini the stb As PSA RL Shee: afc Sib beta ws anh. “bors eet Ae —
ae me 9 i

peonolg Mii y Piso mi ese i lo dersatiioe aan ighqed W2at bar euoinaegn:

WHITH sili a) Wena aSiisd are Slory Hine ff Feet ehiongnio’ sit ris : hee
Bement oi, \viskyel et sovi Sti idiots Surevae dxiw. dant: dee ‘
Se pons swide oft jaro ni Worrizedo otf extn? take ST. thew ea
halt ett aL es yods tfaw visu’ ‘OTR ick asst agi aT ey — s ge
He} off WvOMuEe Ue vit jiods eater pati il 5 ered of to sbi Fe

7 4 yy ragiat otlds aie srediies AL ML ones 10. ysis. & Mier Oryoaabrer oT ae
te suki: 1hd40te" oh) AMOR eqtiioutte 22ont ai iractte ae

PEATE. Vit.

MOON’S AGE 21 DAYS, 5 HOURS. NOVEMBER 3, 1890. LICK OBSERVATORY.

In this plate the moon is entering on the fourth quarter. The rays of the Tycho system have
nearly disappeared. The two that are nearly parallel remain illuminated. So, too, the system of
Copernicus and that of Kepler to-the southeast of it remain in nearly full glow.

The vulcanoids near the south pole are better shown in this picture than in any other of the
series. That with several craters on its floor is Playfair. Note the craters on the inner face of its
wall. The same features can be observed in other like structures in this neighborhood.

The Alps near Plato are fairly well shown, as are also the Apennines that border the western
side of the mare. The ruined craters about the M. Humorum are fairly well shown, but are faint.

The tendency to form a crater or cone in the centers of the larger vulcanoids is fairly well
shown in those structures about the south pole.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE

MOON'S 4 > TWENTY-ONE DAYS, FIVE HOURS. NOVEMBER 3, 1890. LICK OBSI

{ rere
be) 06 | rget yOS JE Fy ZA UUIPT “AT ATER AS Yoon
‘ ¥ ;
Wiis Wogdal ‘on oy ‘ge HesiSif! teas ott! Noo neGirtigy on) ties ra emieee ‘

id thio dsser- silat Hetorodue ti isiq.ar Trt Manet Diigo Shit ae sak
eg) (ise oH ats ft setns

ane wilys oiit Joni
ia , .

wotiathat tf sot Salt Son wy 2snay yj [etre oni

wourrtsiynd cit toon Gictare) rhfanid Ip bh ei ot ek ae

iiiehe Poou- at ivehTattis—at ee Hote jen: Tee nt Tee ees TRG ee
SUTIN Olt 40) 3.58 Goll ucts veo Vile! witetooes ST 2nineat! “hidsds sess “ae
wit Figticy pry Ted aeL ieotr Sioa edi ties: grad otgniac: Jaa e
'
m ‘
F = “
e a Se 7 .
> ;
; ee
S is :
, vy
ht

PLATE, VAI

MOON’S AGE 23 DAYS, 7 HOURS. JULY 28, 1891. LICK OBSERVATORY.

At this stage of the waning moon the most interesting of its fields are no longer visible. There
are few that command attention in this plate. It may be noted that the system of light bands and
the central patches whence they proceed, that have their center in Kepler, are still very bright.
The dark mare-like floor of Grimaldi is visible near the bright margin of the sphere. The observer
may obtain something of the impression, such as is afforded by good seeing with a powerful
telescope, that the Oceanus Procellarum is a relatively shallow sea, by the number of fragments of
what seems to have been the more ancient surface that protrude through it.

96

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV.

MOON'S AGE TWENTY-THREE DAYS, SEVEN HOURS, JULY 28, 1891. LICK OBSERVATORY

“
-
a. el i
A ZY ois is : . ‘
¥ atoy (Fhice pasta EF 5° : ace - Ae ik Bessie i srege 2ictt. rl
17 v Saris atts ; iMiyihinast tiger yee :
‘ rh Ob (ii Ae eerie i 2th} i
us fete fav , Nini SAt ro aH Ny APS Nin St heager
p,bazbarcrn tls epee “1 teeta: Bet iy ' LDA we etn Teme es
‘ dnt Mi A ep ty tt 2) hun 1K Siinhe- ais dhs sis Sine tonnes sah
an 2worts Aone bivtow eral Sal wit
: ar PO dine WAS Tg9s: Hit Lorin. wei Pod le oR aan SGeeee
ut ef oosne Az131c° BOER 4 ! Bom ns i ite yt PSS ge se es te i Heley Siri tie
inpike ti nici? SOT ik hid 21 i broth tonice site a ue St vind iaaston
Oe nis warsiety elie Ti SFike es Sointt Banded? aes \~ Seta
oo: , shel ot ¥ ret A ifs AlooEra ¢ fae ator % dee stifiega
OUT A yet teh ijt ti: 4 Ao ee eos i op eel ites deetoan
fii! yeeeid asf Fs
Noein wsisin 2: if resets » care SP RG a Ee ;
: ay CIOL A ebioOkeinvasinsye Shr test nest Lats. to Aas h seer wa snus
yr bethar: al eik 21). { fen r ¥ art were ary [deel okt
f art at? chain” Hiexseede; 4 , ft see at nig akin tore
; y Sat TOeahe
f ) - 5 $ iprih OS sAt aS Zi taney. id
aay se gol soe tint jlow pr dappl eines eferdiis
Bot ois ro 4 Ssh ri
“Sid apn Aid Sail tH <a ih te
Abiniast ee 6 4 a WSs 2198 ial oy
pila giles ay! Livi 9) fl av ist tine ;
Gene tity kiko anid yee ' \
Proctor sia)! je t i aw
breads MIDS R Gays) hy jeaea ' te
inisdoe ted: )isn’ <i a i os
a
od * :
eee or sf ; E e ,
‘ wo .
sy <

ae

-

PLAGE, UX:
MOON’S AGE 21 DAYS, 16 HOURS. 1895.

In this plate is depicted an area from near the moon’s equator to near the south pole. On
the eastern margin the sunlight is passing from the surface, the evening light being so oblique that
the bottoms of the vulcanoids are more or less in shadow. Here and there, in the advancing night,
there are lofty peaks on the margin of crater-rims, which still receive a touch of sun and appear
as bright points in a black field. On the western margin the surface is still well illuminated, with
the consequent effect that the surface appears to be much smoother than it is. A view taken a
few hours later would show about as rude a margin as is here depicted.

Perhaps more effectively than any other this view shows how the general surface of the moon
outside of the maria is essentially made up of vulcanoids and ridges, the apparently smooth parts
appearing so only because the small irregularities are not visible. In this connection it should be
noted that near the dark part the surface is seen to be beset by small shallow craters, the smallest
visible being more than a mile in diameter, and probably several hundred feet deep. Such pits,
in equal numbers to the unit of surface, exist on the bright part to the left when they are observed
by the higher light.

The way in which the smaller craters cut the larger is shown at many points in this field of
view. So, too, the relative lack of sharpness of outline of the greater vulcanoids as compared with
the lesser objects of this group. The low, narrow ridges which surround the pits are insufficiently
shown because the light does not bring them out. They are best observed near the uppermost
part of the picture.

The generality of the fact that the larger craters have flat floors and that these floors are pre-
vailingly nearly level is well indicated. So, too, the fact that there is a prevailing tendency of
these floors to have either a small crater or a cone in or near the center of each circular field.
Four such craters in the central part of the area extending in an obscure line from near the base
to near the middle of the picture have cones in their centers. In all, about a dozen of the hundred
or so instances in which they would be recognizable have this feature. It will be evident that all
the craters in this region have their floors far below the level of the encircling ring, and below the
general lunar surface.

In sundry instances two adjacent vulcanoids of moderate size have their neighboring walls
broken down so that they exhibit the first stage of “crater valleys” with a general north and south
axis. There are in all about ten cases of this kind on this field, but several of them are not well-
disclosed by this illumination.

98

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL, XXXIV

MOON’S AGE TWENTY-ONE DAYS, SIXTEEN HOURS. 1895.

ete ros)
widrier

+

iKEF

ITO Tee

oft ay Haste z

: ‘ j
Vern SOGZies
{}3 B eT
om
2 re ‘

: st ar
Tone She 2tra

te
eer

i Aeistypes :
+ SiA tb crd

ee 0% 4 : Fay fez ho
ni SHC gonis Togas
cal

PLATE -xX.

MARE CRISIUM AND NEIGHBORING PARTS OF THE MOON. LICK OBSERVATORY.
ENLARGED TO TWICE THE SCALE OF PLATES I TO VIII.

This plate shows the region about the M. Crisium, the most circular of the seas, It is not
completely illuminated, a portion of the western boundary being beyond the light.

In the M. Crisium the most noteworthy feature is the ruined character of its shores as if by
the melting action of the lava of the field. There is an obscure step or bench along the shore of
the mare as if the lava had subsided, as in the larger vulcanoids.

Northeast of the M. Crisium is a large crater, Cleomedes, with a small pit on its south wall
and two craters and a cone on its floor; next farther to the northeast a vulcanoid known as Burk-
hardt. Note that this has two deformed craters beside it, one to the northwest, the other to the
southeast. These features seem to have been produced by some compressive action due to the
formation of Burkhardt. East and southeast of this point there is a remarkable confusion of
deformed vulcanoids. Near the middle of the M. Fecunditatis lie two small craters known as
Messier, whence extend to the southeast two nearly parallel bands of light. The pits of this pair
of vulcanoids have been thought to change their shape in a lunar period. By some early observers
the bands were supposed to be artificial objects, and one astronomer suggested that they were built
by the selenites to signal the people of the earth. There are several ridges on the mare to the
westward of Messier.

The large vulcanoid to the southwest of Messier with a central crater, just beyond three
smaller pits of nearly the same size, is Langrenus, next south Vendelinus, yet farther south
Petavius. The first and last of these show distinct benches on their inner walls. The last has
many pits on its crest.

On the southern margin of the M. Nectaris is Fracastorius, another vulcanoid with the sea-
ward side of its wall demolished by contact with the maria, though it is still traceable; there are
several other like instances about this mare.

100

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV,

MARE CRISIUM AND NEIGHBORING PARTS OF THE MOON. LICK OBSERVATORY
ENLARGED TO TWICE THE SCALE OF PLATES I TO VIII

Fo srsurysd: ¥ AN eit
LF tttiw Sins. nap Withop san Ws

te arbi bed

sir
fray oat

Barna) wes!

ee. pk

ATMEL!

if
BO CBRE wor Astin

#arY

He 4

< ‘ Zz Oa
Ps Oe
: =.
;
d i
3
‘
i ~~
*
*
5
>
A as
: ea
os
}
Ci a AEs .

iartroq B20

:

ie Siakey eter . na

tor Siiotichioes si

nad ai ster

re seis) sondeealE © . cqyeteginl

to safoRisn yor Dery 3 Vo 952 | ee
‘fags 4a) SEVG hss ani tombe
hink-asro sit Jo Antian aaiion ageeeees ; -
~~
ce
Vv B u
> > =
Z ;
’
: ©
~ g*
‘ =¢
Cie e
a 2 e ma
7 ae
a :
s | ; .
e oa a
a em
om c
ne =i
|
& p 4 eo
bs ad
10 . =

PLATE X1-

ENLARGED VIEW OF A PART OF THE APENNINES,.

This plate shows a portion of the Apennines near the Palus Putredinis, an embayment of the
M. Serenitatis where it breaks through the mountain wall and nearly connects with the M.
Imbrium. The three large vulcanoids are Archimedes, Autolycus, and Aristillus. The very
steep or even undercut character of the front of the Apennines is well-known. So, too, the varied
condition of the old craters, breached on the side towards the mare. These features strongly
suggest a melting action of the once-fluid lava of the mare.

102

PLATE XI.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL XXXIV

THE APENNINES.

ED VIEW OF A PART OF

Tr

ENLAR(

.
‘
=
*
”
“ {
i .
th,
’ {
.
fai
Aer
ig ,
;
:

2

PLATE XII
HYGINUS AND THE NEIGHBORING FIELD.

This plate is intended to show the general character of the area in which lie the Hyginus
clefts. It should be noted that parallel and near to those of Hyginus there is another which also
intersects a vulcanoid. It is less perfect but evidently of the same nature. A yet more indistinct
object of the same nature lies near the west wall of the large crater north of Hyginus.

The group of mountains lying near Hyginus shows the elongate character which those ridges
often assume. In other parts of the field they are distinctly conical. Near the clefts is a good
example of crater valleys. Others less distinct lie near the southern border. A large vulcanoid

near the margin of the plate has evidently
had a part of its rim broken down, probably
by the lava of the neighboring mare.
The difference between the features
shown in this plate and the drawing figured
\ herewith will serve to show the reader how
diverse are appearances of the moon’s surface
under different conditions of observation.
\ This drawing may be compared with the
photographs of the same object (Pls. XII,
j XXII) to show the relative minuteness of de-
- tail grasped by a photograph and by the eye.
. It shows the vulcanoid Hyginus with the re-
\ markable clefts which proceed from it as
exhibited in a drawing. The crater is in
no wise exceptional, except for thé fissures
which break through its encircling wall and
extend for a great distance on either side.
These are among the most instructive of this
group of lunar features.

It should be noted that the general con-
tour of the walls on either side of the clefts
indicates that a number of small craters were
first formed and then divided by the formation of the vent and the separation of its walls. That
such was the case is well shown by the fact that the cleft on the right has a part of the ring of at
jeast four of these small vulcanoids on one side of its wall and a part on the other. There is
a faint trace of the same feature in the rift on the left of Hyginus. A like separation has taken
place in the walls of the principal crater. The fact that the floor of this crater is apparently not
divided probably indicates that it was molten at the time when the rupture occurred, or that it
afterwards was so.’ The level surface of the bottom of the clefts can best be explained by sup-
posing that they, too, are floored by lava which entered them at some time after they were formed.
It is probable that this lava came from the depths, for the reason that, as elsewhere noted, there
is reason to believe that the lunar lavas were not sufficiently fluid to flow readily. (See p. 12.)
The facts appear to indicate that this crevice was formed before the interior of the moon had
ceased to be fluid.

HYGINUS FROM A DRAWING WITH THE 13-INCH EQUATORIAL
AT THE ALLEGHENY OBSERVATORY. F. W,. VERY, DEL.,
1890. THREE-EIGHTHS SIZE OF ORIGINAL DRAWING.

| Elger states that he has seen, though faintly, traces of the cleft crossing the floor of the crater. If this observation
was well made, then we have to suppose that the lava did not quite fill this part of the rift, which does not appear on
this drawing, though it exhibits features that Elger had evidently not observed. Such developments are very common in
sketches of lunar structures.

104

PLATE XII.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV.

“Ds

FIEI

ORING

2
)

HE NEIGHI

I

AND

JS

U

HYGIN

ty
pad to¥ f St a wey a ; CoE AC Ly ‘
! ! +) tt ; 13} §
a P
Peps BR tee OF sy Fy 1G Aly be Ridin aA eR

Foro Tote art fyi rh iitreaeperc:s

Ge SH akg licee ‘ ENS

part qeb i 7 25 ALE 33 waerih
firs garters } e arts Yo nite G
OI o ocr Tee} G
i755 Sep hions tie “KY bre Nrertirist
I ah Trt Nyt etel elt SL OV-HeSat
s oT SS RUSS StS

sors: doy ttn Catia

UF PEF STG re # &

Ian 48) mepeh

aN 2°96 ile Hats Die amos 3 alae aes

r » ' c- Me om 2
‘ i Bis Mah : tRi 6 pater ou ot 2h
\ natutty std Stale -brto- ten ier Abate leetit: ofs tel
; sani ; ary I
: ay arf cps iivie ping fi
! , + ew
iT 2 DU ZANMORED BAY

Pra ee Gey Ray Fi

>) Wise ee res aes 3

mp Pee ri Lows taon! iP Pas ee al Pos aia)
Hi Late Ps weir bi Teo high: wats ies Baye Sougcgay G4
* ae ae ne ‘ 7 Pes. -
FUND ph, ected! sg7, : ; "

Hethss

é i 8 t
r B30 re +e TOF RTs

uy Were eine Lets ehienk re
: ke x es AO a
eRe ei ; A ie A 1h BSE So Bae. [Poe
fPAit RS Gas Sih - teu 5 Y ong Thee 34s f RATA wit rode’.
P stitee SOLS. SED cease wee tale TEI T.

fb ofl

Mia bht hi) ae #) gest COR Ue aT Tt OO $9) 1G) | ee Pe ae

SME SE? OTT tt Wye DSi kes OAL sy RiWse jens 4 ns Sep (oowiny teat Sees
- a
iH R de 1 si? inf
ited 1 Yat ahs 3 ; ist i Gite ai
a cee is
1 S| ou Petree ' a o nauk
; feis a
a >

ee dextlat tied . ; grt. feianohwesd: io

: : PS. ik, «
¥ b tris Sener
i pate ite. Say ot
uy . og at ELF she hyqoisvs
ye ' y Sshs Se abort z
yom Del atieos : iJ Shoe torrie Re i
tik vy + Big) Pets
es seME Roy a alt ape
FA ee gD Let PAU T ep) RE ST Uieeiay a EE tee ‘Tht [pin Se emt tase ;
Mn tatiT a tae: Hida es cr de i * DoE Aa) es a0 tL SG aes
Ps Avra F ‘ya

PLAGE Ir

PHOTOGRAPHED BY RITCHEY WITH 40-INCH TELESCOPE, USING YELLOW COLOR
SCREEN AND ISOCHROMATIC PLATE.

This plate shows more than half of the fourth quadrant or the southwest quarter of the moon’s
visible surface, taken at about three-fourths full. The area extends from the equator on the lunar
margin to about 55 south latitude, and from near the polar axis westwardly two-thirds the distance
to the margin of the visible field—a district rich in instructive objects.

On the lower part of the plate is a portion of the Mare Tranquilitatis ; on the middle of the
left-hand side a portion of the Mare Nectaris. The observer should note the features of contact of
these maria with the higher ground against which they lie, especially that there are some indica-
tions of a gradual passage from the rough surface of the upland to the relatively smooth floors of
the maria, and also that several of the rings (at least five) facing the M. Tranquilitatis have the side
towards that area destroyed. The wrinkles on the floor of this sea are fair but not good examples
of the mountain-like ridges that are found on those areas, That on the margin of the M. Nectaris,
extending northward from a crater half in the shadow, is noteworthy.

About a score of the vulcanoids in this field show the tendency to “spooning” or elongation
of the crater in a general north and south direction, in some instances rather northeast and south-
west. In the northeast part of the field some of them pass into crater valleys with a distinct
northeast and southwest axis. In a few instances the axes of these deformed craters are inclined
to the southeast and northwest. So that there appear to have been three different lines of strain
developed on this part of the lunar crust.

The large, deep vulcanoid with the steep, ragged peaks rising from its floor, near the dark
margin on the left, and about one-third the distance from the bottom of the plate, is Theophilus,
one of the noblest structures on the moon. The width of the crater is about sixty-four miles ; the
greatest height from the floor to the crest of the wall eighteen thousand feet. The central mass,
composed of several sharp peaks, rises about six thousand feet above the lava plain. In the
center of these masses there appears to be an obscure crater about half a mile in diameter. The
terraces in the inner wall of the cone are indistinctly shown.

It is to be noted that Theophilus in its development has partly invaded Cyrillus, the next
large vulcanoid on the southeast, and also that the older structure seems more ancient with less
steep slopes and exhibits a generally ruined appearance. Cyrillus is also more “spooned”’ or
drawn out in a north and west direction than Theophilus. South of Cyrillus, at a distance of half
its width, is Catherina. This crater is met by another of half its diameter which has developed
on one side of its floor. From near the southeastern margin of Catherina a beautiful row of small
craters extends eastwardly for a distance of over two hundred miles to the large vulcanoid Abul-
feda. This is perhaps the most noteworthy crater row on the moon.

The long curved wall extending from Piccolomini, near the upper left hand corner (the large
crater with its floor in shadow), to the east side of Catherina is the Altai Mountains. It should be
noted that this step-like structure obscurely extends northwards to the M. Tranquilitatis, where it
forms an irregular ridge-like promontory.

It should be observed that about a dozen of the larger vulcanoids have either a crater or a
cone in the central part of their flat bottoms. In some instances on the brightly illuminated parts
those structures exist, but are not revealed by the illumination.

The larger details of the general surface of the moon on the area to the left of the Altai escarp-
ment are perhaps better shown here than in any other plate. They are rarely so well revealed in
even the best telescopes. In the best seeing the trained eye has a chance to observe perhaps one-half
more than is here shown. Note near the margin southwest of Catherina the existence of obscure
ancient craters, their walls broken and shoved about, as well as the mingling of small cones and
craters, suggesting that craters began with dome-like cones (see p. 30).

106

PHOTOGRAPHED BY RITCHEY WITH FORTY-INCH TELESCOPE, USING YELLOW
COLOR SCREEN AND ISOCHROMATIC PLATE.

-
”
‘ Ee Z
if 4 LD | en i ‘
TRAVAAP AC OA EY “4 aes SEAT eeirir i ie aes Sey Otten
: P ; Pra = Cas z he
; rasa 4y Prat Fis! volbiwAt 94) n*
SAP ivhh Pacemaster Lotte? S) Ry int
: lexaisd Canepa mle arr aris atts es FIs Ro yy EG tthe VTE 7 eK che
eve ound fehl tar arnt) ce Seeno a Tos btcnr: Tes Y Sahar Pte oes rat
A : s 2 2 .
Woon. etn Mania) jttigties eal Sigh AGRO. Feeder CU Ge
Bi Stagg brid TT yh mei as retrid x
BUUAt mT Mee mich “tit usy Sexy) yt oy neg dss th Sit te (eee eae
“od W5elate A rage yiroi ith grt) Py Gary winlk ile: ovate
Geet 73) 1G rr; eth THO: SSB Reb Le saerecns gto eee
i ae : = es ~ es 5 = is
SPP aM iioD weezng of fats i gh Hey See pa Ga
44 e ngitiwe TeUy. site sinits . 2! (Bort hie SS EAGiwe fee ee ae - Veer Hee meal pare 4
HAY yr hi ewllsiene ME us : KET TORS 73 histitnte toed Sond
PBL se La Pre pentintiy Wi <i Fue {74s asta eis ag being rer y 2h ee tle Fre ofA :
for. Ss : aa ba 5% . q
i TES eid ms Tard Tao hd OM babar is FpF. car ied resis
o aa wt asone Be felt VASES eR FSS it Tae ee Sate rs iy ot PUY S Coe
PRI Lesa od) htaset add lees) lel AS OR ie T- 2 tana. He eee
Peu nea en) olds silat dregs ache, joie eC aie Den
e «ie ta 2 » . 2s c } * ; 7 ¥ Ud - es on : 4
ol vfctievsls:410 % Bei woe vitercet at Wusseb 2277 LO Ceca Pn rea ee Te “oF
FN ange ; = RBLE Pinas) of Petshredat
eaiai a
x
- in SL oy
>
bok
> we ate
f Kee )
5 es Seay ‘
* pr
Au!
+. , " ° i

PLATE, SobvV;

PART OF THE SHORE OF THE OCEANUS IMBRIUM. BY M. HENRY, PARIS OBSERVATORY.
AGE OF MOON, 240 HOURS.

In this plate there are a number of features discernible in the others, but here better exhibited
than elsewhere in this series of illustrations. The Oceanus (or mare) Imbrium occupies the cen-
tral part of the picture, its northern, western, and a part of its southern margin being shown. The
large vulcanoid with the dark floor on the northern coast is Plato. South of it, a little way out
upon the mare, is a group of noble peaks called the Teneriffe Mountains. The loftiest rises about
eight thousand feet above the mare.

Following around the shores of the Oceanus Imbrium to the left hand, we note near Plato
the great group of the Alps where there are some hundred peaks, one rising twelve thousand feet
above the mare. Cutting across them the Alpine valley is faintly shown. Farther to the left we
find the Caucasus, a ridge-shaped mountainous district, with one of its many peaks nineteen
thousand feet high. South of this (upwards on the plate) is the passage connecting the Oceanus
Imbrium with the Mare Serenitatis. On the left hand from this strait the first white spot is Linné
(see p. 70). On the right of the strait are two craters, the lower Aristillus, the upper Autolycus.
Farther up to the right is Archimedes. It is about fifty miles in diameter. Above the last-named
structure is an unnamed mountainous district. The lower parts of these fields appear to have been
swept over by the lava of the mare, but the higher are unaffected by it. The shore to the left of
this field from the strait southward is termed the Apennines. The fine crater near the end of
their distinct line is Eratosthenes. Farther on, out in the dark field of the Oceanus Procellarum, is
the great vulcanoid Copernicus. Just below it, faintly shown, is a group of elevations termed the
Carpathian Mountains.

108

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV

PART OF THE SHORE OF THE OCEANUS IMBRIUM. BY M. HENRY, PARIS OBSERVATORY.
AGE OF MOON, 240 HOURS

ones

Winco! Vartowsor
Poa aecn oer pre wit) te
a aie ea Phu o thse sare:
Be Fett ane pits ri: hl

int! oAguorods B. embio MIN Si opis tines
q pe

ua iets cceols godng Hlsler mage ead} a

3c Pair een sbadt e 3 2
0 paper 2 pat + IORI

oft ahais

ee reek 2 dow lyn ¢
siqamces fh i ei Toes TD, Pees
Finan loier yithral ons UBS i adsl ;

Hausa oar nares

Aahaa ietare off) orktg

PLATE XV;

CENTRAL PORTION OF THE MOON FROM THE M. SERENITATIS TO STOFLER.
BY M. HENRY, PARIS OBSERVATORY.

All the more important structures shown in this plate have been displayed in the preceding
plates under different conditions of illumination. The most noteworthy features here illustrated
are the seas. The lowest or northernmost is the southern part of the M. Serenitatis, which will be
seen to have its surface apparently somewhat lower than the adjacent M. Tranquilitatis. This
latter passes on the left hand or western side into the M. Foecunditatis, which is shown only in
small part, and on the south into the M. Nectaris. At the southern end of the M. Nectaris is the
great vulcanoid Fracastorius with its northern wall broken down apparently by the melting action
of the lava of the mare.

South of the distinct crater of Menelaus, a little to the right of the uppermost part of M. Se-
renitatis, at about one-fifth the distance from the bottom of the plate towards the top, is a very
irregular vulcanoid, Julius Cesar, which is partly broken down by the neighboring mare. Touching
the northern or lower margin of Julius Cesar is a good example of a crater valley. Several others
are included in this plate. About half the width of Julius Cesar farther to the south is the
Ariadzus cleft, one of the straightest fissures on the moon.

On the most illuminated part of this plate the bright streaks begin to be traceable ; they are
most visible on the M. Nectaris.

110

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL XXXIV p
LATE XV.

CENTRAL PORTION OF THE MOON FROM THE M. SERENITATIS TO STOFLER.
BY M. HENRY, PARIS OBSERVATORY.

holt VV - USHA SOY AGIIFZA 414% COSITHAIOD

: pogangie rd ety uae er’ ait 5 W2o boilqaipetotie-sigw =abdyenss Shinoliok ner

feqay the the vols troron” ac
ppt ine Of Ovise Pst Sounhinith Me TH Ia am erst? Gani ya Nandiles apache?
/ ts

aad) yd Bsbolie soTrtaiesot- She peice

bl dee 099Ioe IGIOs hw. GiAanist

sy1b eb eh Ste
udotnicite, Bal usigs ep:
Irs8o3 sain)
SHUI TEED
soe 2 tet ob oebar yest

S dns ebeedd gintt 03-Jo, mettidial

Brill sve sex
Ditaked viInHoittnG ai ebied-vaor'? To sriusnn fodons
; font. ot jettouaplos atti a

Rear :
PRM BUONO) 26h ite: 19. Ssivozie thoLpAxion SAP ba (inh ot oF porwe Hoody

eeiict alt { oll 2 spate EOC e Cha ket
4 site Unies coding iiveodt Uob it -siit-ge aatdo gesltod
$9 Sogo sity sit ; Migostic. silt sods joo} hanes: slang te Iason ae eset

(phe eset shs wooo Warde zivol lo rete say chord willed ot tk bogbr_gignove pis
E et as 2 sisigsnie abaliaoia simetareesds 3

¥
<
- A
er ek
<s
> *
5 wer > "
hl a
«

Bone Ifoy gidieosiot abled turd letiges 2 ta ddcieynvlong wumekye topes 76 2heas goed -

beg

PLATE Xvi.

COPERNICUS AND KEPLER. PHOTOGRAPHED BY RITCHEY. SCALE, ONE-HALF METER
TO MOON’S DIAMETER.

The following ten plates were photographed by G. W. Ritchey with the forty-inch Yerkes
refractor, with color screen and isochromatic plate. As will be noted, they in part repeat the
features exhibited by the other plates of this series, yet in all instances they serve to supplement
or extend the information afforded by them.

The most important features exhibited by plate XVI are the systems of bright rays of Coperni-
cus, Kepler, and Aristarchus. These three ray systems, though less extensive than those of Tycho,
taken together constitute the greatest exhibition of the bright bands that exist over the northern
part of the surface. The complex branched nature of these bands is particularly well shown, bet-
ter, indeed, than the writer has ever been able to note with the telescope. The fact that the
bright bands of each system are prolongations of a central bright field is tolerably well shown.

Although owing to the high sun and the consequent absence of shadows, Copernicus in this
view hardly appears as an elevation, it is, under favorable conditions of illumination, perhaps the
noblest object on the moon. The wall on the eastern side, according to the estimates of Schmidt,
rises to a height of twelve thousand feet above the adjacent plain. The outer slopes of the cone
are strongly ridged as by the flow from the crater of lavas which cooled on the steep slopes ; some
of these are faintly traceable in the plate.

112

NVIC SNOOW OL Wa
Ad GAHdVNDOLOHd Aad UNV

“TAX 3LV1d P "AIXXX “OA '39037MONY OL SNOILNBINLNOD NVINOSH LING

E act 3

nf: my Biviis . to Tens.

Seusd i) thar pore

Be ea iss Lbs ye ee
fips Ht BitEA sil)
o Soil tet 5 Geran G

aie

<e sth wu Fea fb Ey ie i <7
wat ont az: A oan!
is ¢

see od Cte HID

ar) i eh it 6

re omevib ae cay eran rahe Sosfi}IL

a

re}
=
Old
* 'si5
sf.
Cs a

LT SOT SS. AOE

ETI LTE Jad: ae

AbD buat biatt ay
fe AST. Beas ara

z Ste Hic

SST hat ud fyi a4 digcda ile Rad tile
sh " re +

Chey

a ip auido ¢ resi apis |

Be, « baths sapsat? in
aos ; sitet ibis Worcs ‘
(ei eibitics sTrael yy

diphiss © reo ash

KStie- h2E SECA jiatett

wll Sy

“Hel 1 4a5

t

i aoiic iy ‘

eatin Zinn
obnsi VRE BH

Heros iF

o TORENT = 455 2
a, | i mis iit Hiei
ol patel T piste =e

had EH

nftity: Seo rt ;
LER 2 Osan
as : : ot neh
ris] >is eG ihe cigs ashy SiO) TU

j
>
.
aa
- ¥ Ae
he
oe : -
ers = A ~
os
3 —
ae
=
Fa i

Mules.
inbpanih sige

PLATE 2OViie

CRATER REGION ABOUT THEOPHILUS. PHOTOGRAPHED BY RITCHEY.
SCALE, THREE-FOURTHS METER TO MOON’S DIAMETER.

A portion of this field, including the crater Theophilus, is shown in other plates. This most
important structure lies just below the middle of the plate near the margin of the illumination.

The details of structure of the lunar surface, as shown on the margin of the illuminated field,
are better exhibited in this picture than in any other ; perhaps better than in any other photograph
that has been published. The more important of them have been noted in the descriptions of
preceding plates, but attention may well be called to certain of these features, viz., to the numer-
ous shallow craterlets near Theophilus, to sundry wrecked craters in the same field, and to the
association of small cones and small craters in the region south (upward on the plate) from
Theophilus.

The frequent deformation of craters by elongation is fairly well indicated by several vulcanoids
within the field of view. The invasion of the material of the maria is well shown in the region
about Theophilus, and, as before noted, the central peak on the crater floor of that structure with
its fairly distinct central pit is admirably depicted.

It is well to note the passage from the very distinct exhibition of the structures on the termi-
nator, the margin of the illuminated field, to the obscurity of similar features when the sun is more
than forty-five degrees above the horizon,

114

INTRIBUTIONS

SMITHSONIAN C

THEOPHII

CRATER REGION ABOUT

RTHS METI

THREE-FOl

SCALE,

© ‘
; ts
Cha LETS Sado 1408 ACES
am 2
“tlt A als ; t esi
anes OY ito 29 pol Sa jue it 3 i
| Egat yy ; ae u
VOW Sous ty ety “
we Eta) Lick las SAY : :
ee a : TLEROL Rr HSS Ay tit it Gaiart
Halon Mosh anT ae, ai. 9 dpatebe i £4 thee x ford’ 4 ©
ne apie # bo Sars Surly Pets brite Tithe. tars Ferat ; size bylonty ‘
AG Sansy vrth itt tae oA aate § Ad
whan Saksa7 3G ITS Fount Sit fit . é nos yi air ve
singin Te" ate ots Ape TL Went ealiete
+ hie on Ue Yiats UbaRReh betemany SAS ame Be Aa
; Wakes “hoe $ a0 -
~ 53 i ed bee t
f tthe -atibas Se) 16 pitah wy
SrA SPN RR Neth cis
Poel.16. wolfe eg lo Sezai act 3
ist “Shir tor Fi Avtetovaly ses
Tat het “he a uv Z j
Deiat {4295
;
‘
¥§
i. e a 2 ;
:

PLATE 21k

MARE SERENITATIS. PHOTOGRAPHED BY RITCHEY. SCALE, THREE-FOURTHS METER
TO MOON’S DIAMETER.

This plate shows the whole of the Mare Serenitatis; on the upper right-hand corner a part
of the M. Vaporum ; on the lower corner of the same side portions of the Mare (or oceanus) Im-
brium, known as the Palus Nebularis. The largest vulcanoid near the dark margin is Posidonius.
The bright patch showing no distinct structure, which lies on the parallel of Posidonius, about two-
thirds across the field, is the problematical Linné. The partly illuminated portion of the mare
below Posidonius is the Lacus Somniorum.

The most noteworthy structures exhibited in this plate are as follows: The great mountainous
ridge which traverses the mare in a general north and south direction (this structure more dis-
tinctly resembles a terrestrial mountain-chain than any other elevation on the moon); the field
abounding in conical elevations in the lower part of the plate ; the crater of Le Monnier just above
Posidonius, which has a part of its wall apparently broken down by the mare, and the crater
valleys near the upper right-hand corner of the plate. There are a number of clefts, commonly
known as rills, which are fairly well shown. A group of these lies just below Plinius, the large
crater with a bright central cone emerging from the shadow of the crater wall, situated near the
upper margin of the plate. Another notable group is found in the left-hand lower section of the
plate. Faint traces of craters may be seen in these clefts.

It may be noted that a number of the larger vulcanoids here depicted exhibit that tendency
to a development in a meridional direction which has been termed in this text “spooning.” In
Posidonius and the smaller vulcanoid, Jansen, on the margin above it, the southern (upper) walls
are thus indented.

116

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV. A
PLATE XVIII

MARE SERENITATIS. PHOTOGRAPHED BY RITCHEY. SCALE, THREE-FOURTHS
METER TO MOON’S DIAMETER

*
4
is
,
x A
- ¥ ’ ’ ’
~s) &
FAS . Youle ioe ’ ; Pete
iz y
re Bas bid ‘ ‘ ‘ nH i
de onp rt 94; wht: boi Ytit it fe Uti;

ayn at hap ye that.) girenoues wh pave tt bs aes AR Mt i:
si Bile te {wit yt ete Be yeti | 1c) tes tertiary Biri Sorje ew aH an ‘% 5, Sige fai nase
abatihe : sd neti Rl gue och in eS. 8 bb eet se Aa hh b
See, Qiwanbiitoy tt: the; Lo siti ¢

= 4 * " iis Sip
SPLIT Pre "y? 4 Mis Sstriit> WoOBy U- ao > ey Ld,

tira Sanit oie do tisetiod VOR teint Vea ce ot iyos 4 pumas fg Pane e,.

NE bot 4 og Gin th). weghukils Ph Laiop eine al brs sale SHIA RGR eR ae ih
nh ra sone! Pek ssl c iAG Ag srsita at 7 cae
a'ths SAE RIA ut QTE ft Aiztt Gio = %

* ret 2

: *
3 . r
» y f
n}
? = 3 >
‘ n
». ohn
{ .
. . ne
x - i.
$ P
¥. ° 7
< : é
4 - , &
“
a2
BA
.
ie :
a ~~ 2

PIV xix

RAY SYSTEM ABOUT TYCHO. PHOTOGRAPHED BY RITCHEY. SCALE, THREE-EIGHTHS
METER TO MOON’S DIAMETER.

This, the most extensive of the ray systems of the moon, has its origin in the field about
Tycho, the large vulcanoid to which the numerous bands apparently converge. It appears under
the high sun as a large pit with a compound central cone. ‘The rays of this system should be com-
pared with those which have their centers in Copernicus and Kepler. In these last named groups
the streaks are developed on relatively level ground, while on that of Tycho they intersect a rugged
surface.

On the right hand, some of the bands may be seen crossing the-Mare Nubium. Two of them
of great length are seen to be nearly parallel for a distance of some hundred miles.

A number of large vulcanoids, partly in shadow, are shown on the southeast margin of the
moon. Of these the largest is Schiller. Its length, which is one hundred and twelve miles, will
serve as a scale in estimating that of the rays.

118

MALANVIG SNOOW OL YALAW
SH.LH “AANHL ‘ATVOS ‘AAHOLIN Ad GHHdVNDOLOHd ‘OHDAL LNOUV WALSAS AVY

c

7

5 3 5
ei

8 Pe asa

Safe
end,
*AL-
hem
co

nat arts beveaband {
ret abties 1c,
ah) Riva ogi € yd
MELO) EREIY

Oi tighivar as

ipsa

Aish! a i

She sikh)

ond - wigivest= do est
Foto e40E) Legs St? vio)
wed cates dine = i. 97

easy, fi

x
iu é rrr
j c
: {
Tet.
a? ‘
a ,
~ Bh
= oF
‘
ay ¥
“ °

4
.
*
a Ren
Ly rat eae
{ 5 a uae
f * :
I iy Rz
t r ,
Cres En | [
4 site ote ees
ait a iF y
‘ rite 3G S
eb AT ‘
2 Mr AD nse
=. and > pa
f file t4 ¢
; ey
Pelt a4 ; js
rT > ‘a
Att hd Oo phate
£ Lect
é > > jp0 GSH =
wy =) x ‘fy
% 9
}
{
‘ aT
© =
a a
ag %
+
4
5.
~

POAT E 2x.

COPERNICUS AND SURROUNDINGS. PHOTOGRAPHED BY RITCHEY, NOVEMBER 21,
IQOI, 7 HOURS 32 MINUTES P.M., CENTRAL STANDARD. TIME. EXPOSURE,
ONE SECOND. SCALE, THREE-FOURTHS METER TO MOON’S DIAMETER,

This plate of Copernicus should be compared with the plates showing the same structure
under more nearly vertical illumination when the light bands appear.

In the plate the lower level area is a part of the Mare Imbrium. This is bordered on the left
by a portion of the high country known as the Apennines, which extend as far towards the center
of the plate as the large crater Eratosthenes. To the left, separated by a little more than the
width of Copernicus, is the faintly outlined vulcanoid known as Stadius, which appears to have been
in large part melted down by the lava of the Oceanus Procellarum which has invaded this field.
On the right hand from Eratosthenes, the margin of the mare is formed by the peaks of the
Carpathian Mountains. Immediately above Copernicus is a small, double crater, one of the
simpler crater valleys.

The area about Copernicus exhibits several very interesting types of structure. The Carpa-
thian Mountains show the mare penetrating into several rude craters, the seaward faces of which
have had their walls destroyed by the fluid lava. A broken line of small craters lies midway
between Copernicus and Eratosthenes. At either end it verges into a narrow crater valley of the
“rill”’ type. The central part of the upper half of the field abounds in very perfect cones
which are associated with small crater pits.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV. PLATE XX

COPERNICUS AND SURROUNDINGS. PHOTOGRAPHED BY RITCHEY, NOVEMBER 21, 1901, SEVEN HOURS
EXPOSURE, ONE SECOND.

THIRTY-TWO MINUTES P. M., CENTRAL STANDARD TIM
SCALE, THREE-FOURTHS METER TO MOON'S DIAMETER

meQTIA, VU MART ASOT at ATI AOS Ue GMA MUIRU Ve Baia

Me AViy eer : AOA PAUP TUE Se et 2s ea F
Ma ft r 7 BW 6 8 {7 ae 40-4 '
Geel sth -sivinm vowel ait no tone tuy own oilear ews nisqol) Sie adil at

er aG-dfier Sel) ot dotwortos seo 2 sie is1095, sat le tyes Hr Sra iain or
i

eONGH Shi wetiiwstes wore ai + pa teil! o6 Teles ar oth aii its Tin fae
ee: a 20) Sit Jo higihe: aqun ett eb ; voi oh used Widegoake at lte Gortaases sor a
7. - 4 S > s Sauk 3 : ty
mew euierit to silva sti of bins eoisd sta wlth, Stoeasy eizeoiglt, cant.

i } id * : sealiiiindt <)onos deutnae Pein «eek,
Ay Stiior ct psonmieat Witte. at ni | ery 21d Re BEI yA MO WesG! Jen oie ey
rr sg off nl ebish vist) cajdsie shiowgyiuy oft Hove yhe vine it he Sahar

wt Nika saedi 10 4 arcr raters Donia vests Sis Wisk) E Se Vials seid ego Mirae a

y ¥ . . —_
to BANOS ein fantraecun) ty ST) x * io at fen? Vey ett haves sao oe

i "

bh aionosil asl) Sst etter salt dnqaltow fnsochh oialsniw. ab sietihl os: 40 ees oheie
Mole tei txo vst Ao sldEyoims in 1jisbiety Ny ashi bo-sadce eee 7 ota s = a ?
Pe aehy sift Jo bait fase ad}ii inst a osGaroen fit dyai se icl sake Saar?

Pnaods Sd) fas wole Ment ovik foto se mit? Oat > Tye SA RL nd Pat ~
Bqs} Parodi roe Bil th srudoiinte bastards iaozy es on | “eit Treads, ai bits to seat ai

& Daewt Vd WoyGies> N500 guived iseingg oils he Aybouine of] Pattie A lotwerinior wit a tae

oy aie eee iieizo iso ei leh filo gh yd ie otter adr Fore eee
oe 5 r Ate Pe we ri oe 5 * - =
" sabe sonneb & ot bas dediis Thnsen3Ss ve cinds Vina yon, Io tie ae aa oa 7h

ato Pal Sil ot yes gietg eiii no betoigeb hip Oty 4. Fag Sin syectbertHLy ehizizd ae |

me tyigtl rae es 2D eshi nyo Palo? VEO} atidiad Oe €.

als TShiaegaeows Ct lr ROTTER ta ex i Mee eet

7 = ‘ a
¥ a
ao ‘
‘one .
‘-
nt a
ar @ . Es Fi
x 4
*
-
ay P
‘
v:* eet
= ’ <
on P < .
¥ » : «
=e . - -
. P: ~ 2 i /
man (t) Ge ‘ - 34s

i _ ~~
SURI ANT &
mt \ BN HOU

Sat

PLAGE sox,

MARE NUBIUM AND SURROUNDINGS. PHOTOGRAPHED BY RITCHEY, NOVEMBER 21,
IQOI, 7 HOURS 32 MINUTES P.M. EXPOSURE, ONE SECOND.
SCALE, ONE-HALF METER TO MOON’S DIAMETER.

In this plate Copernicus is the large vulcanoid on the lower margin. The large crater near
the upper margin, a little to the right of the center, with a cone somewhat to the right of its center
and “rill” on its floor, is Pitatus. The three great vulcanoids in a row extending in a north and
south direction, are, in succession from the lowest towards the upper margin of the plate, Ptole-
meus, Alphonsus, and Arzachel. The large deep crater below and to the right of Pitatus, with a
divided central cone, is Bullialdus.

The most noteworthy features in this plate are found in the many instances in which the lavas
of the maria have partly destroyed the vulcanoids within their fields. In the upper right-hand
fourth of the plate, there are a dozen or more of these ruined craters, some of them with their walls
almost effaced. In this part of the field there are several important rills. Some of these are evi-
dently rows of craterlets in which the adjacent walls of the pits have been broken down so as to
form a ragged cleft. A number of these lines of craterlets are traceable on the external slopes of
Copernicus. The long, dark line, sixty-five miles in length, in the upper third of the plate, a little
to the left of the center, is the Straight Wall, the most extensive fault known on the moon, The
height of its cliff is about five hundred feet. The crescent shaped structure at its southern (upper)
end is the remnant of a crater, the remainder of the margin having been destroyed by the lava of
the mare. To the right of, and near by the Straight Wall, is a rill extending in a slightly curved
course for a length of about forty miles, terminating at either end in a distinct craterlet.

The brightly illuminated part of the field depicted on this plate, that to the left of the center,
exhibits many excellent examples of crater valleys, which in their series afford something like a
passage from the condition of rills to those wider depressions.

122

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV
PLATE XXI.

MARE NUBIUM AND SURROUNDINGS. PHOTOGRAPHED BY RITCHEY, NOVEMBER 21, 1901,
SEVEN HOURS THIRTY-TWO MINUTES P. M. EXPOSURE, ONE SECOND
SCALE, ONE-HALF METER TO MOON'S DIAMETER.

¢

mid”
TEC 4

BSCE ai oils sli

rail nF uossci

oN .

SLES

ile Tver

in SAP GRY hy)

eT

1 eal barons The FO OSL

}
5
4s
13
1
ig

13 i
"
]
Tf
{ bh
1gEY
i
j
:
{ay}
ve

> > Byn3i
18) fi ;
5D 4 ' ert
sed Sih
} it ; CFs
ahw it ,

4 es rg
rows
atl ;
yet Bil
oO 9
' yt 5 z
Leer iF
IIx ry,o9t8
OT Fat Ho
itent A wea
3 ichie

be

;

oJ ait
ayrye

PLATE XX.

MARE TRANQUILITATIS AND SURROUNDINGS. PHOTOGRAPHED BY RITCHEY, AUGUST
3, 1QOI, 2 HOURS 30 MINUTES A.M., CENTRAL STANDARD TIME. EXPOSURE,
3 SECOND. SCALE, THREE-FOURTHS METER TO MOON’S DIAMETER.

This plate includes nearly the whole of the Mare Tranquilitatis and, on the lower margin,
a portion of the M. Serenitatis. The large crater near the strait connecting these maria is Plinius.
The highland nearest to it is the promontory of Acherusia. On the southern, or upper, margin
the view extends to the flanks of Theophilus.

The most noteworthy features in this plate are the mountain ridges on the maria, the manner
in which the maria come in contact with the higher ground, the numerous crater valleys, and the
great “rills.”

It may be noted that ridges on the maria exhibit little trace of corresponding troughs between
them, such as are usually found in terrestrial mountain chains.

The contact of the maria with the high ground has evidently resulted in the partial melting
of the walls of several vulcanoids. Where these structures are not thus affected they are,
apparently, in origin later than the formation of the maria. The crater valleys are abundant
on the right-hand or eastern side of the field. Some of them have been invaded by the lava of the
mare.

Some of the greater rills are very well shown. That on the extreme right side is Hyginus
(see p. 44). It will be observed that the course of these rills is at high angles to the prevailing
direction of the ridges on the mare.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV.
PLATE XXII.

MARE TRANQUILITATIS AND SURROUNDINGS. PHOT‘ \GRAPHED BY RITCHEY, AUGUST 3, 1901,
TWO HOURS THIRTY MINUTES A. M., CENTRAL STANDARD TIME. 1 XPOSURE,
THREE-FOURTHS SECOND. SCALE, THREE-FOURTHS
METER TO MOON'S DIAMETER

we Tis

mM feed 4 Toe
>

ARTAVIELAG 7

y
ie id plat
i hessigniill: £-eh8 3n7> el \
, Rahs, a :
ema s tihitt Diserm DI Ihr) cay
Aa a e pa re * : 4 ‘
eit vioee Sooh tac ie
Tagacte> es to 2h
MOH pictrory Sita
Star oH) 1:
5s citi 7
‘ rate] 1 we b
Tle ret pl} v)
Mit} { }
i
sty
VRRMLOS ait gk ‘yrs
By cee ace vsews toile

vdhi

=

thes

SF
4
» t
ae eg
;
as (9ih6
: i
rerurt
jrty

PLATE Sec ii.

MARE IMBRIUM AND SURROUNDINGS. PHOTOGRAPHED BY RITCHEY, NOVEMBER 21,
I9QOI, 7 HOURS 32 MINUTES P.M., CENTRAL STANDARD TIME. EXPOSURE,
ONE SECOND. SCALE ONE-HALF METER TO MOON’S DIAMETER.

This plate depicts the western two-thirds of the Mare Imbrium : it does not show the interest-
ing Sinus Iridum on its northern shore, nor the Harbinger Mountains on its eastern side. The most
noteworthy features are the relatively level surface of the mare and the greater vulcanoids and
peaks on its margin, or in its midst, and the Alpine valley on its northwest side.

The great crater near the lower margin of the mare is Plato, This crater has a diameter of sixty
miles, and is very nearly circular. It is separated from the M. Imbrium by little more than its own
wall, and from the narrow M. Frigoris on the north by a field of upland that declines gently to that
mare. This field is thickly beset by small cones. The interior walls of the crater of Plato rise in
general to a height of about four thousand feet above its floor. At some points, however, this
wall is over seven thousand feet in height. The floor of the crater appears in the plate to be
smooth and of a rather even, very dark hue. It is, however, the seat of rather extensive topo-
graphical and color features. There are at least six crater cones, about forty patches of peculiar
coloration. The failure of these markings and structures to appear on this admirable plate may
be taken as a measure of the difference between what is shown by the best reproductions of
photographs now obtainable and the revelations of the telescope under the most favorable
conditions.

On the sea south of Plato is a group of remarkable peaks. Those on the extreme right are
known as the Straight Range; those on the center as the Teneriffe Mountains ; the solitary peak
yet farther to the west is Pico.

The wide cleft to the left of Plato, about one hundred miles away, is the Alpine valley.
Owing to the high sun it is not well shown.

‘he three great vulcanoids near the left-hand margin of the mare are: the largest Archimedes,
the intermediate Aristillus, and the smallest Autolycus.

126

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV 5 :
PLATE XXIII.

MARE IMBRIUM AND SURROUNDINGS. PHOTOGRAPHED BY RITCHEY, NOVEMBER 21, 1901,
SEVEN HOURS THIRTY-TWO MINUTES P. M., CENTRAL STANDARD TIME.
EXPOSURE, ONE SECOND. SCALE, ONE-HALF METER
TO MOON’S DIAMETER.

bib is act

Reve 2YH ite pes ;
Sadly ibd Se Pet

Brut) a

AB yekt cs

Pe

grin

on

¥255
vai;

PLATE SOc.

ARISTOTELES, EUDOXUS, AND SURROUNDINGS. PHOTOGRAPHED BY RITCHEY, OCTOBER
13, I900, 2 HOURS 40 MINUTES A.M. EXPOSURE $ SECOND. SCALE,
THREE-FOURTHS METER TO MOON’S DIAMETER.

In this plate the large vulcanoid near the top of the lower third of the field, that which cuts
the ring of the smaller crater on the left of its wall, is Aristoteles ; the somewhat smaller structure
just above is Eudoxus ; that near the upper left-hand corner is Posidonius. On the right hand, at
the same levelas Aristoteles, the great Alpine valley is partly seen, the illumination being too nearly
vertical to show it well.

Among the noteworthy features exhibited by this plate the following are the most important :

The wall of Aristoteles evidently has broken that of the small unnamed crater adjacent to it on
the west (left-hand) side. This shows that Aristoteles was in activity since the smaller vulcanoid
was formed. The inner slopes of the first-named crater abound in rude terraces. Its limited
floor bears numerous cones.

South of Eudoxus is an extensive field of elevations known as the Caucasus Mountains. The
western portion of this field peculiarly abounds in cones and craterlets of about the same diameter
as these cones, suggesting that the two groups of structure are in origin in some way related.
Certain other good examples of these cones are exhibited in the lower part of the plate.

To the west of Eudoxus is a great, irregular vulcanoid with a large crater (Burg) somewhat
excentrically placed on its floor. On this floor are some remarkable rills.

The greater part of the upper third of the plate is occupied by the Mare Serenitatis. A por-
tion of its mountain-like ridges is well shown.

128

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VO

ARISTOTELES, EUDOXUS, AND SURROUNDINGS. PH
TWO HOURS FORTY MINUTES A. M EXPOSI E, ON
THREE-FOURTHS METI M

J

-
a
«
yi
zi
i
j ‘ 14 Fae tyre i 1 wat t rf
: ew Ee a piri Bs rH boost
2 ri? oe H WTS LEH
ey cry aod dette ser ' cramer tt eras ree at aisty tak
; i re : i Sos Shines 5 sae u
wR i ; Meauin} : (nar ily ao got Wises dete

ij fe rH eal HE OY, Tt bane}

i : *Ty | Ty ‘.c. sti os ie x NOAM Sty 4
imal baste i ‘a Pats , : ths 2 hthe}
, i . ee : ; oh bare “*
att val ' f T SHG Th MN
tic teonsidH

% mata I i TARO Wala tee oe
tron | i il cA \ wy ” :
ie i ‘ ' pvk et F bait
iin ; Si { ait? Rinit ahi) 2 zit hi “dep

PANE sri ate eas 7 fist Sd he io alt
an t, i; - ‘ , J : th mini ft z Seah

PPiaks om vmasaate wise 2 fired
.
BD
S
e

PEATE XXV:

CLAVIUS, LONGOMONTANUS, TYCHO, ETC. PHOTOGRAPHED BY RITCHEY, NOVEMBER 21,
IQOI, 7 HOURS 32 MINUTES P.M., CENTRAL STANDARD TIME. SCALE,
THREE-FOURTHS METER TO MOON’S DIAMETER.

In this plate the large crater, only partly illuminated, on the line of the terminator and cut by
the upper edge of the plate, is Klaproth. Just below Klaproth is Blanchianus, which on its lower
margin nearly touches the wall of Clavius, the largest structure in the field. Clavius is one hun-
dred and forty-two miles in diameter. North of Clavius, on the edge of the illumination, is
Longomontanus. Nearly in the center of the plate is Tycho, about which the great ray system,
visible under a very high sun, originates. This structure may be recognized by its central, sharp,
irregular cone. The large vulcanoid near the center of the lower part of the plate is Pitatus,
situated on the margin of the Mare Imbrium. It may be better identified by the “rill” on the
northeast part of its crater floor. :

The most noteworthy features of this plate are as follows: The abundance of relatively large
vulcanoids ; the difference in the nature of their floors, some being relatively smooth, others much
varied by pits and craters, and the association of small cones and craterlets of like horizontal sec-
tion, in all parts of the field where the light is favorable for their exhibition.

The effect of the lava of the mare, when it comes in contact with the high ground, also
deserves attention. It appears to have more or less completely destroyed the walls of several
vulcanoids with which it came in contact.

130

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE, VOL. XXXIV. BCE
r ATE AAV.

y 7
he if
c

Y

~,

CLAVIUS, LONGOMONTANUS, TYCHO, ETC. PHOTOGRAPHED BY RITCHEY, NOV] MBER 21, 1901,
SEVEN HOURS, THIRTY-TWO MINUTES P. M., CENTRAL STANDARD TIME.
SCALE, THREE-FOURTHS METER TO MOON'S DIAMETER

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE

i ; PART OF VOLUME XXXIV , & Wt YS ame

- TIAL PHOTOGRAPHY .

By HENRY DRAPER, M.D. .

PROFESSOR OF NATURAL SCIENCE IN THE UNIVERSITY OF NEW YORK
(Reprinted from Vol. XIV, ‘‘ Smithsonian Contributions to Knowledge,”’ 1864)

‘ AND

DN THE MODERN REFLECTING TELESCOPE
AND THE MAKING AND TESTING

G35 + on
oF OpTi@aL MIRRORS - ial
; Ee, eae
» By GEORGE W. RITCHEY
4 ASSISTANT PROFESSOR OF PRACTICAL ASTRONOMY, AND SUPERINTENDENT OF *
* INSTRUMENT CONSTRUCTION, IN YERKES OBSERVATORY
ease ebay. - - >
Dy
% <
A °
(No. 1459) ‘
CITY OF WASHINGTON ° gees e
PUBLISHED BY THE SMITHSONIAN INSTITUTION ;
1904 |
e
— ————— — ——
A
- v2 ~e>

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE

PART OF VOLUME XXXIV

» ON THE CONSTRUCTION OF A SILVERED GLass
| TELESCOPE, FIFTEEN AND A HALF INCHES
IN; APERTURE, AND ITS USE IN CrLES-
TIAL PHOTOGRAPHY
By HENRY DRAPER, M.D.

PROFESSOR OF NATURAL SCIENCE IN THE UNIVERSITY OF NEW YORK
(Reprinted from Vol. XIV, ‘‘ Smithsonian Contributions to Knowledge,”’ 1864)

0
On THE MODERN REFLECTING TELESCOPE
AND THE MAKING AND TESTING
OF OpticaAL MIRRORS
By GEORGE W. RITCHEY

ASSISTANT PROFESSOR OF PRACTICAL ASTRONOMY, AND SUPERINTENDENT OF
INSTRUMENT CONSTRUCTION, IN YERKES OBSERVATORY

(No. 1459)

CITY OF WASHINGTON
PUBLISHED BY THE SMITHSONIAN INSTITUTION
1904

a =e

7 ‘Se
: es . aioe
) <= ue < a
- i oe
a oe oe = Onlin
- a Ay
r — = -
> ic SD a a

INTRODUCTION.

For few papers published by the Institution has there been a more constant
demand than for the memoir by Professor Henry Draper, entitled “On the Con-
struction of a Silvered Glass Telescope,” originally issued forty years ago, in 1864.

The paper is of remarkable merit as a summary of, and an addition to, the
knowledge existing at the time, but during the long interval which has elapsed,
progress has been made in various directions and by various hands.

On the occasion of a new edition of this classic memoir, it was sought to give
an account of the latest knowledge on the subject, and I was gratified to be able
to obtain from Mr. Ritchey, whose labors in this direction are so well known, an
account of the processes which he has employed for making the great mirrors that
have been so effective at the Yerkes Observatory, and it has been decided to re-
publish, with the original Draper memoir, but as an entirely independent con-
tribution to the subject, the present article by Mr. Ritchey.

The great refracting instruments which have been produced in recent years
have not superseded the use of the reflector, which, on the contrary, is occupying a
more and more important place.

The reader is here presented with the most recent methods and results needed
in the construction of great mirrors for modern reflecting telescopes.

S. P. Laneuey,

Secretary of the Smithsonian Institution.

WASHINGTON, June, 1904.

QN THE CONSTRUCTION

SILVERED GLASS TELESCOPE,

FIFTEEN AND A HALF INCHES IN APERTURE,

ITS USE IN CELESTIAL PHOTOGRAPHY.

BY.

HENRY DRAPER, M.D,
PROFESSOR OF NATURAL SCIENCE IN THE UNIVERSITY OF NEW YORK.

COMMISSION

TO WHICH THIS PAPER HAS BEEN REFERRED.

Prof. Wotcorr GIBBs.
Com. J. M: ‘Ginniss, Uls: N:
JosEPpH Henry,
Secretary S. 1.

HISTORICAL SKETCH OF THE TELESCOPE.

g§1

§ 3.

§ 4.

§ 5.

§ 6.

CONTENTS.

GRINDING AND POLISHING THE Mirrors
1. Experiments on a metal speculum. Corrosion by aqua regia; voltaic grinding

2. Silvering glass. Foucault’s and Cimeg’s processes ; details of silvering a mirror ; thick-
ness and durability of silver films; their use in daguerreotyping
3. Grinding and polishing glass. Division of subject

a. Peculiarities of glass; effects of pressure; effects of heat ; oblique mirrors

b. Emery and rouge ; elutriation of emery ‘ : 3 :

ec. Tools of iron, lead, pitch; the gauges; the leaden tool; the iron tool; the pitch
polisher . , : : ,

d. Methods of examination; two tests, eyepiece and opaque screen; appearance of
spherical surface; oblate spheroidal surface; hyperbolic surface; irregular
surface ; details of tests; atmospheric movements ; correction for parallel rays
by measure; appearances in relief on mirrors . . ‘ :

e. Machines ; Lord Rosse; Mr. Lassell ; spiral stroke machine ; its construction and
use; the foot-power; method of local corrections; its advantages and disad-
vantages ; machine for local corrections ; description and use

4. Eyepieces, plane mirrors, and test objects

THe TELEscoPpE MOUNTING . : ‘ “ : : :
Stationary eyepiece ; method of counterpoising .
a. The tube; the mirror support; air sac; currents in the tube
b. The supporting frame

THe CLock MOVEMENT
a. The sliding plateholder ; the frictionless slide
b. The clepsydra; the sand-clock
c. The sun camera

THE OBSERVATORY
a. The building
b. The dome; its peculiarities
ec. The observer’s chair

Tue Puorocraputc LABORATORY

a Description of the apartment , . : 2
b. Photographie processes ; washed plates ; difficulties of celestial photography

THe PHoTroGRAPHic ENLARGER
a. Low powers; use of a concave mirror, its novelty and advantages ; of the making

of reverses : ; :
b. High powers; microscopic photography

( ii.)

MEMOIR DIVIDED INTO SIX SECTIONS :—

bo

13

26
27
27
28
3l
33
33
36
40
41
41
44
45
46
46
47

ol

_
7
: .
ied
ne minded owen
6 a . 7
ee se hat o ee
erste 7
Pa § pie a“ )
‘ .
ite al
=
el) rae
be pi snigs
0 dae a
-
;
.
‘

‘hi Meet ae } Set 2 |
ed oe oe
7 | i hy Sacee

a an ine

AIN BAS ©: © OnU Nem

OF

THE CONSTRUCTION AND USE OF A SILVERED GLASS TELESCOPE.

Tue construction of a reflecting telescope capable of showing every celestial
object now known, is not a very difficult task. It demands principally perse-
verance and careful observation of minutie. The cost of materials is but trifling
compared with the result obtained, and I can see no reason why silvered glass
instruments should not come into general use among amateurs. The future hopes
of Astronomy lie in the multitude of observers, and in the concentration of the
action of many minds. If what is written here should aid in the advance of that
noble study, I shall feel amply repaid for my labor.

A short historical sketch of this telescope may not be uninteresting. In the sum-
mer of 1857, I visited Lord Rosse’s great reflector, at Parsonstown, and, in addition
to an inspection of the machinery for grinding and polishing, had an opportunity
of seeing several celestial objects through it. On returning home, in 1858, I
determined to construct a similar, though smaller instrument; which, however,
should be larger than any in America, and be especially adapted for photography.
Accordingly, in September of that year, a 15 inch speculum was cast, and a
machine to work it made. In 1860, the observatory was built, by the village
carpenter, from my own designs, at my father’s country seat, and the telescope
with its metal speculum mounted. This latter was, however, soon after abandoned,
and silvered glass adopted. During 1861, the difficulties of grinding and polishing
that are detailed in this account were met with, and the remedies for many of them
ascertained. The experiments were conducted by the aid of three 15$ inch disks
of glass, together with a variety of smaller pieces. ‘Three mirrors of the same
focal length and aperture are almost essential, for it not infrequently happens that
two in succession will be so similar, that a third is required for attempting an
advance beyond them. One of these was made to acquire a parabolic figure, and
bore a power of 1,000. The winter was devoted to perfecting the art of silvering,
and to the study of special photographic processes. A large portion of 1862
was spent with a regiment in a campaign’ in Virginia, and but few photographs
were produced till autumn, when sand clocks and clepsydras of several kinds having
been made, the driving mechanism attained great excellence. During the winter,
the art of local corrections was acquired, and two 154 inch mirrors, as well as two
of 9 inches for the photographic enlarging apparatus, were completed. The greater
part of 1863 has been occupied by lunar and planetary photography, and the
enlargement of the small negatives obtained at the focus of the great. reflector.
Lunar negatives have been produced which have been magnified to 8 feet in

1 May, 1864. (Gale)

9 ON THE CONSTRUCTION AND USE OF

diameter. I have also finished two mirrors 155 inches in aperture, suitable for a
Herschelian telescope, that is, which can only converge oblique pencils to a focus
free from aberration. ‘This work has all been accomplished in the intervals of pro-
fessional labor.

The details of the preceding operations are arranged as follows: $1. GrinpING
AND PovisHinc THE Mirrors; § 2. THe TeLescopeE Mountine; § 3. THe Cock
Movement; § 4. THE Opservatory; § 5. THe PHorograpHic LABoratory; § 6.
THe PuHoroGRAPHic HNLARGER.

$1. GRINDING AND POLISHING THE MIRRORS.

(1.) EXPERIMENTS ON A METAL SPECULUM.

My first 15 inch speculum was an alloy of copper and tin, in the proportions
given by Lord Rosse. His general directions were closely followed, and_ the
casting was very fine, free from pores, and of silvery whiteness. It was 2 inches
thick, weighed 110 pounds, and was intended to be of 12 feet focal length. The
grinding and polishing were conducted with the Rosse machine. _Although a great
amount of time was spent in various trials, extending over more than a year, a fine
figure was never obtained—the principal obstacle to success being a tendency to
polish in rings of different focal length. It must, however, be borne in mind that
Lord Rosse had so thoroughly mastered the peculiarities of his machine as to pro-
duce with it the largest specula ever made and of very fine figure.

During these experiments there was occasion to grind out some imperfections,
3, of an inch deep, from the face of the metal. ‘This operation was greatly assisted
by stopping up the defects with a thick alcoholic solution of Canada balsam, and
having made a rim of wax around the edge of the mirror, pouring on nitro-hydro-
chloric acid, which quickly corroded away the uncovered spaces. Subsequently an
increase in focal length of 15 inches was accomplished, by attacking the edge
zones of the surface with the acid in graduated depths.

An attempt also was made to assist the tedious grinding operation by meluding
the grinder and mirror in a Voltaic cirenit, making the speculum the positive pole.
By decomposing acidulated water between it and the grinder, and thereby oxidizing
the tin and copper of the speculum, the operation was much facilitated, but the
battery surface required was too great for common use. If a sufficient intensity
was given to the current, speculum metal was transferred without oxidation to the
erider, and deposited in thin layers upon it. It was proposed at one time to make
use of this fact, and coat a mirror of brass with a layer of speculum metal by
electrotyping. The gain in lightness would be considerable.

During the winter of 1860 the speculum was split into two pieces, by the
expansion in freezing of a few drops of water that had found their way into the
supporting case.

(2.) SILVERING GLASS.

At Sir John Herschel’s suggestion (given on the occasion of a visit that my
father paid him in 1860), experiments were next commenced with silvered glass

A SILVERED GLASS TELESCOPE 3

specula. These were described as possessing great capabilities for astronomical
purposes. They reflect more than 90 per cent. of the light that falls wpon them,
and only weigh one-eighth as much as specula of metal of equal aperture.

As no details of Steinheil’s or Foucault’s processes for silvering in the cold way
were accessible at the time, trials extending at intervals over four months were
made. A variety of reducing agents were used, and eventually good results
obtained with milk sugar.

Soon after a description of the process resorted to by M. Foucault in his excel-
lent experiments was procured. It consists in decomposing an alcoholic solution
of ammonia and nitrate of silver by oil of cloves. ‘The preparation of the solutions
and putting them in a proper state of imstability are very difficult, and the results
by no means certain. ‘The silver is apt to be soft and easily rubbed off, or of a
leaden appearance. It is liable to become spotted from adherent particles of the
solutions used in its preparation, and often when dissolved off a piece of glass with
nitric acid leaves a reddish powder. Occasionally, however, the process gives
excellent results.

In the winter of 1861, M. Cimeg published his method of silvering looking-
glasses by tartrate of potash and soda (Rochelle salt). Since I have made modifica-
tions in it fitting the silver for being polished on the reverse side, I have never on
any occasion failed to secure bright, hard, and in every respect, perfect films.

The operation, which in many details resembles that of M. Foucault, is divided
into: Ist, cleaning the glass; 2d, preparing the solutions ; 3d, warming the glass ;
4th. immersion in the silver solution and stay there; 5th, polishing. It should be
carried on in a room warmed to 70° F. at least. The description is for a 155 inch
mirror.

Ist. Clean the glass like a plate for collodion photography. Rub it thoroughly
with nitric acid, and then wash it well in plenty of water, and set it on edge on
filtering paper to dry. Then cover it with a mixture of alcohol and prepared chalk,
and allow evaporation to take place. Rub it in succession with many pieces of
cotton flannel. This leaves the surface almost chemically clean. Lately, instead
of chalk I have used plain uniodized collodion, and polished with a freshly-washed
piece of cotton flannel, as soon as the film had become semi-solid,

2d. Dissolve 560 grains of Rochelle salt in two or three ounces of water and
filter. Dissolve 800 grains of nitrate of silver in four ounces of water. Take an
ounce of strong ammonia of commerce, and add nitrate solution to it until a brown
precipitate remains undissolved. Then add more ammonia and again nitrate
of silver solution. This alternate addition is to be carefully continued until the
silver solution is exhausted, when some of the brown precipitate should remain in
suspension. The mixture then contains an undissolved

Fig. 1.

excess of oxide of silver. Filter. Just before using, mix
with the Rochelle salt solution, and add water enough
to make 22 ounces.

The vessel in which the silvering is to be performed may .
: : re 2 5 : ry: The Silvering Vessel.
be a circular dish (Fig. 1) of ordinary tinplate, 165 inches

in diameter, with a flat bottom and perpendicular sides one inch high, and coated

4 ON THE CONSTRUCTION AND USE OF

inside with a mixture of beeswax and rosin (equal parts), At opposite ends of one

diameter two narrow pieces of wood, a a’, } of an inch thick, are cemented. They

are to keep the face of the mirror from the bottom of the vessel, and permit of a
rocking motion being given to the glass. Before using such a vessel, it is necessary
to touch any cracks that may have formed in the wax with a hot poker. A spirit
lamp causes bubbles and holes through to the tin. ‘The vessel too must always,
especially if partly silvered, be cleaned with nitric acid and water, and left filled
with cold water till needed. Instead of the above, India-rubber baths have been
occasionally used.

3d. In order to secure fine and hard deposits in the shortest time and with weak
solutions, it is desirable, though not necessary, to warm the glass slightly. ‘This is
best done by putting it in a tub or other suitably sized vessel, and pourimg in water
enough to cover the glass. Then hot water is gradually stirred in, till the mixture
reaches 100° F. It is also advantageous to place the vessels containing the in-
gredients for the silvering solution in the same bath for a short time.

4th. On taking the glass out of the warm water, carry it to the silvering vessel—
into which an assistant has just previously poured the mixed silvering solution—
and immediately immerse it face downwards, dipping in first one edge and then
quickly letting down the other till the face is horizontal. ‘The back of course is
not covered with the fluid. ‘The same precautions are necessary to avoid streaks
in silvering as in the case of putting a collodion plate im the bath. Place the
whole apparatus before a window. Keep up a slow rocking motion of the glass,
and watch for the appearance of the bright silver film. ‘The solution quickly turns
brown, and the silver soon after appears, usually in from three to five minutes.
Leave the mirror in the liquid about six times as long. At the expiration of the
twenty minutes or half hour lift it out, and look through it at some very bright
object. If the object is scarcely visible, the silver surface must then be washed
with plenty of water, and set on edge on bibulous paper to dry. If, on the con-
trary, it is too thin, put it quickly back, and leave it until thick enough. When
polished the silver ought, if held between the eye and the sun, to show his disk
of a light blue tint. On coming out of the bath the metallic surface should have
a rosy golden color by reflected light.

5th. When the mirror is thoroughly dry, and no drops of water remain about
the edges, lay it upon its back on a thoroughly dusted table. Take a piece of the
softest thin buckskin, and stuff it loosely with cotton to make a rubber. Avoid
using the edge pieces of a skin, as they are always hard and contain nodules of
lime.

Go gently over the whole silver surface with this rubber in circular strokes,
in order to commence the removal of the rosy golden film, and to condense the
silver. Then having put some very fine rouge on a piece of buckskin laid flat on
the table, impregnate the rubber with it. The best stroke for polishing is a motion
in small circles, at times gomg gradually round on the mirror, at times across on
the various chords (Fig. 2). At the end of an hour of continuous gentle rubbing,
with occasional touches on the flat rouged skin, the surface will be polished so as
to be perfectly black im oblique positions, and, with even moderate care, scratchless.

A SILVERED GLASS TELESCOPE a

The process is like a burnishing. Put the rubber carefully away for another
occasion.

Fig. 2.

Polishing Strokes.

The thickness of the silver thus deposited is about 55,45, of an ich. Gold
leaf, when equally transparent, is estimated at the same fraction. The actual value
of the amount on a 155 inch mirror is not quite a cent — the weight bemg less than
4 grains (239 milligrammes on one occasion when the silver was unusually thick),
if the directions above given are followed.

Variations in thickness of this film of silver on various parts of the face of the
mirror are consequently only small fractions of 5 55!;95 of an inch, and are therefore
of no optical moment whatever. If a glass has been properly silvered, and shows
the sun of the same color and intensity through all parts of its surface, the most
delicate optical tests will certainly fail to indicate any difference in figure between
the silver and the glass underneath. The faintest peculiarities of local surface
seen on the glass by the method of M. Foucault, will be reproduced on the silver.

The durability of these silver films varies, depending on the cireumstances under
which they are placed, and the method of preparation. Sulphuretted hydrogen
tarnishes them quickly. Drops of water may split the silver off. Under certain
circumstances, too, minute fissures will spread all over the surface of the silver, and
it will apparently lose its adhesion to the glass. This phenomenon seems to be
connected with a continued exposure to dampness, and is avoided by grinding the
edge of the concave mirror flat, and keeping it covered when not in use with a sheet
of flat plate glass. Heat seems to have no prejudicial effect, though it might have
been supposed that the difference in expansibility would have overcome the mutual
adhesion.

Generally silvered mirrors are very enduring, and will bear polishing repeatedly,
if previously dried by heat. I have some which have been used as diagonal re-
flectors in the Newtonian, and have been exposed during a large part of the day
to the heat of the sun concentrated by the 155 inch mirror. ‘These small mirrors
are never covered, and yet the one now in the telescope has been there a year, and
has had the dusty film—like that which accumulates on glass—polished off it a
dozen times.

In order to guard against tarnishing, experiments were at first made in gilding
silver films, but were abandoned when found to be unnecessary. A partial con-
version of the silver film into a golden one, when it will resist sulphuretted hydrogen,

6 ON THE CONSTRUCTION AND USE OF

can be accomplished as follows: Take three grains of hyposulphite of soda, and
dissolve it in an ounce of water. Add to it slowly a solution in water of one grain
of chloride of gold. A lemon yellow liquid results, which eventually becomes clear.
Immerse the silvered glass in it for twenty-four hours. An exchange will take place,
and the film become yellowish. I have a piece of glass prepared in this way which
remains unhurt in a box, where other pieces of plain silvered glass have changed
some to yellow, some to blue, from exposure to coal gas.

I have also used silvered glass plates for daguerreotyping. ‘They iodize beauti-
fully if freshly polished, and owing probably to the absence of the usual copper
alloy of silver plating, take impressions with very short exposures. The resulting
picture has a rosy warmth, rarely seen in ordinary daguerreotypes. ‘The only pre-
caution necessary is in fixing to use an alcoholic solution of cyanide of potassium,
instead of hyposulphite of soda dissolved in water. ‘The latter has a tendency to
split up the silver. The subsequent washing must be with diluted common alcohol.

Pictures obtained by this method will bear high magnifying powers without
showing granulation. Unfortunately the exposure required for them in the telescope
Is SIX times as great as for a sensitive wet collodion, though the iodizing be carried
to a lemon yellow, the bromizing to a rose red, and the plate be returned to the
iodine.

(3.) GRINDING AND POLISHING GLASS.

Some of the facts stated in the followimg paragraphs, the result of numerous
experiments, may not be new to practical opticians. I have had, however, to polish
with my own hands more than a hundred mirrors of various sizes, from 19 inches
to } of an inch in diameter, and to experience very frequent failures for three years,
before succeeding in producing large surfaces with certainty and quickly. It is
well nigh impossible to obtain from opticians the practical minutiae which are
essential, and which they conceal even from each other. ‘The long continued re-
searches of Lord Rosse, Mr. Lassell, and M. Foucault are full of the most valuable
facts, and have been of continual use.

The subject is divided into: a. The Peculiarities of Glass; b. Emery and Rouge ;
c. Tools of Iron, Lead and Pitch; d. Methods of Examining Surfaces; e. Machines.

a. Peculiarities of Glass.

Effects of Pressure.—lt is generally supposed that glass is possessed of the power
of resistance to compression and rigidity in a very marked manner. In the course
of these experiments it has appeared that a sheet of it, even when very thick, can
with difficulty be set on edge without bending so much as to be optically worthless.
Fortunately in every disk of glass that I have tried, there is one diameter on either
end of which it may stand without harm.

In examining lately various works on astronomy and optics, it appears that the
same difficulty has been found not only in glass but also in speculum metal. Short
used always to mark on the edge of the large mirrors of his Gregorian telescopes
the point which should be placed uppermost, in case they were removed from their
cells. In achromatics the image is very sensibly changed in sharpness if the flint

A SILVERED GLASS TELESCOPE u
and crown are not in the best positions ; and Mr. Airy, in mounting the Northum-
berland telescope, had to arrange the means for turning the lenses on their common
axis, until the finest image was attained. In no account, however, have I found a
critical statement of the exact nature of the deformation, the observers merely
remarking that in some positions of the object glass there was a sharper image than
in others.

Before I appreciated the facts now to be mentioned, many fine mirrors were
condemned to be re-polished, which, had they been properly set in their mountings,
would have operated excellently.

In attempting to ascertain the nature of deformations by pressure, many changes
were made in the position of the disk of glass, and in the kind of support. Some
square mirrors, too, were ground and polished. As an example of the final results,
the following case is presented: A 154 inch unsilvered mirror 14 inch thick was
set with its best diameter perpendicular, the axis of the mirror being horizontal
(Fig. 8). ‘The image of a pin-hole illuminated by a lamp was then observed to be
single, sharply defined, and with interference rings surrounding it as at a, Fig. 3.
On turning the glass 90 degrees, that is one quarter way round,

Fig. 3.

its axis still poiting in the same direction, it could hardly be
realized that the same concave surface was converging the rays.
The image was separated into two of about equal intensity, as
at b, with a wing of ight going out above and below from the

junction. Inside and outside of the focal plane the cone of
rays had an elliptical section, the major axis being horizontal Etect of Pressure on a Re-
imside, and perpendicular outside. Tuming the mirror still pete pane
more round the image gradually improved, until the original diameter was perpen-
dicular again—the end that had been the uppermost now being the lowest. A
similar series of changes occurred in supporting the glass on various parts of the
other semicircle. It might be supposed that irregularities on the edge of the glass
disk, or in the supporting arc would account for the phenomena. But two facts
dispose of the former of these hypotheses: in the first place if the glass be turned
exactly half way round, the character of the image is unchanged, and it is not to
be believed that in many different mirrors this could occur by chance coincidence.
In the second place, one of these mirrors has been carefully examined after being
ground and polished three times in succession, and on each occasion required the
same diameter to be perpendicular. As to the second hypothesis no material differ-
ence is observed whether the supporting arc below be large or small, nor when it
is replaced by a thin semicircle of tinplate Imed with cotton wool.

I am led to believe that this peculiarity results from the structural arrangement
of the glass. ‘The specimens that have served for these experiments have probably
been subjected to a rolling operation when in a plastic state, in order to be reduced
to a uniform thickness. Optical glass, which may be made by softening down
irregular fragments into moulds at a temperature below that of fusion, may have
the same difficulty, but whether it has a diameter of minimum compression can
only be derenmimed by experiment. Why speculum metal should have the same
property might be ascertained by a critical examination of the process of casting,

8 ON THE CONSTRUCTION AND US OF

and the effect of the position of the openings in the mould for the entrance of the
molten metal.

Effects of Heat—Yhe preceding changes in glass when isolated appear very
simple, and their remedy, to keep the proper diameter perpendicular, is so obvious
that it may seem surprising that they should have given origin to any embarrassment.
In fact it is now desirable to have a disk in which they are well marked. But in
practice they are complicated in the most trying manner with variations produced
by heat pervading the various parts of the glass unequally. The following case
illustrates the effects of heat :—

A 154 inch mirror, which was giving at its centre of curvature a very fine image
(a, Fig. 4) of an illuminated pin-hole, was heated at the edge by placing the right

Effects of Heat on a Reflecting Surface.

hand on the back of the mirror, at one end of the horizontal diameter. In a few
seconds an are of light came out from the image as at 6’, and on putting the left

,

hand on the other extremity of the same diameter the appearance c’ was that of
two arcs of light crossing each other, and having an image at each intersection.
The mirror did not recover its original condition in ten minutes. Another person
on a subsequent occasion touching the ends of the perpendicular diameter at the
same time that the horizontal were warmed, caused the image d’ to become some-
what like two of c’, put at right angles to each other. <A little distance outside
the focus the complementary appearances, 6, c, 7, were found.

By unsymmetrical warming still more remarkable forms emerged in succession,
some of which were more like certain nebule with their milky light, than any
regular geometrical figure.

If the glass had, after one of these experiments, been immediately put on the
polishing machine and re-polished, the changes in sur-

Fig. We

face would to a certain extent have become permanent,
as in Chinese specula, and the mirror would have re;
quired either re-grinding or prolonged polishing to get
rid of them. ‘This occurred unfortunately very fre-
quently in the earlier stages of this series of experi-
ments, and gave origin on one occasion to a surface

which could only show the image of a pin-hole as a
Effects of Heat rendered permanent. ae = : - : ees
lozenge (6, Fig. 5), with an image at each angle inside
the focus, and as an image a with four wings outside

But it must not be supposed that such apparent causes as these are required to

A SILVERED GLASS TELESCOPE. 9

disturb a surface injuriously. Frequently mirrors in the process for correction of
spherical aberration will change the quality of their images without any perceptible
reason for the alteration. A current of cold or warm air, a gleam of sunlight, the
close approach of some person, an unguarded touch, the application of cold water
injudiciously will ruin the labor of days. The avoidance of these and similar causes
requires personal experience, and the amateur can only be advised to use too much
caution rather than too little.

Such accidents, too, teach a useful lesson in the management of a large telescope,
never, for instance, to leave one-half the mirror or lens exposed to radiate into cold
space, while the other half is covered by a comparatively warm dome. Under the
head of the Sun-Camera, some further facts of this kind may be found.

Oblique Mirrors.—Still another propensity of glass and speculum metal must be
noted. A truly spherical concave can only give an image free from distortion when
it is so set that its optical axis poits to the object and returns the image directly
back towards it. But I have polished a large number of mirrors in which an image
free from distortion was produced on/y when oblique pencils fell on the mirror, and
the image was returned along a line forming an angle of from 2 to 3 degrees with
the direction of the object. Such mirrors, though exactly suited for the Herschelian
construction, will not officiate im a Newtonian unless the diagonal mirror be put
enough out of centre in the tube, to compensate for the figure of the mirror. Some
of the best photographs of the moon that have been produced in the observatory,
were made when the diagonal mirror was 6 inches out of centre in the 16 inch
tube. Of course the large mirror below was not perpendicular to the axis of the
tube, but was inclined 2° 32’. The figure of such a concave might be explained by
the supposition that it was as if cut out of a parabolic surface of twice the diameter,
so that the vertex should be on the edge. But if the mirror was turned 180° it
apparently did just as well as in the first position, the image of a round object being
neither oval nor elliptical, and without wings. The image, however, is never quite
as fine as in the usual kind of mirrors. ‘The true explanation seems rather to be
that the radius of curvature is greater along one of the diameters than along that
at right angles. How it is possible for such a figure to arise during grinding and
polishing is not easy to understand, unless it be granted that glass yields more to
heat and compression in one direction than another,

After these facts had been laboriously ascertained, and the method of using such
otherwise valueless mirrors put in practice as above stated, chance brought a letter
of Maskelyne to my notice. He says, “I hit upon an extraordinary experiment
which greatly improved the performance of the six-feet reflector”... .. It was
one made by Short. ‘As a like management may improve many other telescopes,
I shall here relate it; I removed the great speculum from the position it ought to
hold perpendicular to the axis of the tube when the telescope is said to be rightly
adjusted, to one a little inclined to the same and found a certain inclination of about
24° (as I found by the alteration of objects in the finer one of Dollond’s best night
glasses with a field of 6°), which caused the telescope to show the object (a printed
paper) incomparably better than before ; insomuch that I could read many of the
words which before I could make nothing at all of. It is plain, therefore, that this

Dy May, 1864.

10 ON THE CONSTRUCTION OF

telescope shows best with a certain oblique pencil of rays. Probably it will be found
that this circumstance is by no means peculiar to this telescope.” This very valuable
observation has lain buried for eighty-two years, and ignorance of it has led to the
destruction of many a valuable surface.

As regards the method of combating this tendeney, it is as a general rule best
to ye-grind or rather re-fine the surface, for though pitch polishing has occasionally
corrected it in a few minutes, it will not always do so. I have polished a surface
for thirteen and a half hours, examining it frequently, without changing the obliquity
in the slightest degree.

Glass, then, is a substance prone to change by heat and compression, and requiring
to be handled with the utmost caution.

b. Emery and Rouge.

In order to excavate the concave depression in a piece of glass, emery as coarse
as the head of a pin has been commonly used. This cuts rapidly, and is succeeded
by finer grained varieties, till flour emery is reached. After that only washed
emeries should be permitted. They are made by an elutriating process invented
by Dr. Green.

Five pounds of the finest sifted flour emery are mixed with an ounce of pulverized
gum arabic. Enough water to make the mass like treacle is then added, and the
ingredients are thoroughly incorporated by the hand. ‘They are put into a deep jar
containing a gallon of water. After being stirred the fluid is allowed to come to
rest, and the surface is skimmed. At the end of an hour the liquid containing
extremely fine emery in suspension is decanted or drawn off with a siphon, nearly
down to the level of the precipitated emery at the bottom, and set aside to subside
in a tall vessel. When this has occurred, which will be in the lapse of a few hours,
the fluid is to be carefully poured back into the first vessel, and the fine deposit in
the second put imto a stoppered bottle. In the same way by stirrmg up the pre-
cipitate again, emery that has been suspended 30, 10, 3, 1 minutes, and 20, 3,
seconds is to be secured and preserved in wide-mouthed vessels.

f
.

The quantity of the finer emeries consumed im smoothing a 155 inch surface is
very triflimg—a mass of each as large as two peas sufficing.

Rouge, or peroxide of iron, is better bought than prepared by the amateur. It
is made by calcining sulphate of iron and washing the product in water. ‘Three
kinds are usually found in commerce: a very coarse variety containing the largest
percentage of the cutting black oxide of iron, which will scratch glass like quartz ;
a very fine variety which can hardly polish glass, but is suitable for silver films ;
and one intermediate. ‘Trial of several boxes is the best method of procuring that
which is desired.

c. Tools of Tron, Lead, and Pitch.

In making a mirror, one of the first steps is to describe upon two stout sheets of

brass or iron, arcs of a circle with a radius equal to twice the desired focal length,

and to secure, by filing and grinding them together, a concave and convex gauge.
When the radius bar is very long, it may be hung against the side of a house. By

A SILVERED GLASS TELESCOPE. 11

the assistance of these templets, the convex tools of lead and iron and the concave
surface of the mirror are made parts of a sphere of proper diameter.

The excavation of a large flat disc of glass to a concave is best accomplished by
means of a thick plate of lead, cast considerably more convex than the gauge.
The central parts wear away very quickly, and when they become too flat must be
made convex again by striking the lead on the back with a hammer. ‘The glass is
thus caused gradually to approach the right concavity. ‘Ten or twelve hours usually
suffice to complete this stage. The progress of the excavating is tested sufficiently
well by setting the convex gauge on a diameter of the mirror, and observing how
many slips of paper of a definite thickness will pass under the centre or edge, as
the case may be. This avoids the necessity of a spherometer. The thickness of
paper is found correctly enough by measuring a half ream, and dividing by the
number of sheets. In this manner differences in the versed sine of a thousandth
of an inch may be appreciated, and a close enough approximation to the desired
focal length reached—-the precision required in achromatics not being needed.
The preparation of the iron tools on which the grinding is to be finished is very
laborious where personal exertion is used. They require to be cast thin in order
that they may be easily handled, and hence cannot be twmed with very great
exactness.

The pair for my large mirrors are 154 inches in diameter, and were cast 2 of an

3

inch thick, bemg strengthened however on the back by eight ribs ¢ of an inch high,

radiating from a solid centre two inches in diameter (a, Fig. 6). They weighed 25

Fig. 6.

The Iron Grinder.

pounds apiece. Four ears, with a tapped hole in each, project at equal distances
round the edge, and serve either as a means of attachment for a counterpoise lever,
or as handles.

After these were turned and taken off the lathe chuck, they were found to be
somewhat sprung, and had to be scraped and ground in the machine for a week
before fitting properly. The slowness in grinding results from the emery becoming
imbedded in the iron, and forming a surface as hard as adamant.

Once acquired, such grinders are very valuable, as they keep their focal length
and figure apparently without change if carefully used, and only worked on glass
of nearly similar curvature. At first no grooves were cut upon the face, for in the

12 ON THE CONST RUC DEON OF

lead previously employed for fining they were found to be a fruitful source of
scratches, on account of grains of emery imbedding in them, and gradually break-
ing loose as the lead wore away. Subsequently it appeared, that unless there was
some means of spreading water and the grinding powders evenly, rings weye likely
to be produced on the mirror, and the iron was consequently treated as follows :—

A number of pieces of wax, such as is used in making artificial flowers, were
procured. ‘The convex iron was laid out in squares of | of an inch on the side,
and each alternate one being touched with a thick alcoholic solution of Canada
balsam, a piece of wax of that size was put over it. This was found after many
trials to be the best method of protecting some squares, and yet leaving others in
the most suitable condition to be attacked. A rim of wax, melted with Canada
balsam, was raised around the edge of the iron, and a pint of aqua regia poured
in. In a short time this corroded out the uncovered parts to a sufficient depth,
leaving an appearance like a chess-board, except that the projecting squares did not
touch at the adjoining angles (4, Fig. 6). I should have chipped the cavities out,
instead of dissolving them away, but for fear of changing the radius of curvature
and breaking the thin plate. However as soon as the iron was cleaned, it proved
to have become flatter, the radius of curvature having increased 7} inches. ‘This
shows what a state of tension and compression there must be im such a mass, when
the removal of a film of metal ., of an inch thick, here and there, from one surface,
causes so great a change.

When the glass has been brought to the finest possible grain on such a grinder,
a polishing tool has to be prepared by covering the convex iron with either pitch
or rosin. ‘These substances have very similar properties, but the rosin by being
clear affords an opportunity of seemg whether there are impurities, and therefore
has been frequently used, straiming being unnecessary. It is, however, too hard as
it occurs in commerce, and requires to be softened with turpentine.

A mass sufficiently large to cover the iron { of an inch thick is melted in a
porcelain or metal capsule by a spirit lamp. When thoroughly liquid the lamp is
blown out, and spirits of turpentine added, a drachm or two at a time. After each
addition a chisel or some similar piece of metal is dipped into the fluid rosin, and
then immersed in water at the temperature of the room. After a minute or two it
is taken out, and tried with the thumb-nail. When the proper degree of softness is
obtained, an indentation can be made by a moderate pressure.

The iron having been heated in hot water is then
painted in stripes £ of an inch deep with this resinous com-
position. ‘The glass concave to be polished being smeared
with rouge, is pressed upon it to secure a fit, and the iron

is then put in cold water. With a narrow chisel straight
grooves are made, dividing the surface into squares of one
inch, separated by intervals of one-quarter of an inch (Fig.
7). Under certain circumstances it is also desirable to take

off every other square, or perhaps reduce the polishing sur-
face irregularly here and there, to get an excess of action on

The Polishing Tool.

some particular portion of the mirror.

A SILVERED GLASS TELESCOPE. 15
It is well, on commencing to polish with a tool made in this way, to warm the
glass as well as the tool in water (page 4) before bringing the two in contact. If
this is not done the polishing will not go on kindly, a good adaptation not being
secured for a length of time, and the glass surface being injured at the outset. The
rosin on a polisher put away for a day or two suffers an internal change, a species
of irregular swelling, and does not retain its original form. Heating, too, has a
good effect in preventing disturbance by local variations of temperature in the glass.
The description of ‘* Local Polishers” will be given under Machines.

d. Methods of Examining Surfaces.

I have been in the habit of testing mirrors exclusively at the centre of curvature,
not putting them in the telescope tube until nearly parabolic or finished. The
means of trial are so excellent, the indications obtained so precise, and the freedom
from atmospheric disturbances so complete, that the greatest faeilities are offered
for ascertaining the nature of a surface. In addition the observer is entirely inde-
pendent of day or night, and of the weather. I do not think that anything more
is learned of the telescope, even under favorable circumstances, than im the work-
shop. For the improvement of these methods of observation, Science is largely
indebted to M. Foucault, whose third test--the- second in the next paragraph— is
sufficient to afford by itself a large part of the information required in correcting
a concave surface.

There are two distinct modes of examination: Ist, observing with an eye-piece
the image of an illuminated pin-hole at the focus, and the cone of rays inside and
outside that plane; 2d, receiving the entire pencil of light coming from the mirror
through the pupil on the retina, and noticing the distribution of light and shade,
and the appearances in relief on the face of the mirror.

The arrangements for these tests are as follows: Around the flame of a lamp (a,

ii hi ie ii | il il | 1 i
‘ al ant ; ne ! | ! | i ! ; ) i ae

ii bi)

bi

| .
| »

i

MIN

Testing a Concave at the Centre of Curvature.

Fig. 8) a sheet of tin is bent so as to form a cylindrical screen. Through it at the
height of the pee part of the flame, as at 4, two holes are bored, a quarter of an
inch apart, one ,!, of an inch in diameter, the other as small as the point of the finest

needle will eae perhaps 545 of aninch, ‘This apparatus is to be set at the centre

14 ON THE CONSLIRUCLLON TOR

of curvature of the mirror ce—the optical axis of the latter bemg horizontal—and
so adjusted that the light which diverges from the illuminated hole in use, may,
after impinging on the concave surface of the glass, return to form an image close
by the side of the tin screen. In the case of the first test, the returning rays are
received into an eye-piece or microscope, /, magnifying 20 times, and moving upon
a divided scale to and from the mirror. In the second test the eye-piece is removed
away from before the eye, and a straight-edged opaque screen, ¢, is put m its place.
The mirror is supported in these trials by an are of wood /, lined with thick woollen
stuff, and above two wooden latches, g, g, prevent it from from falling forward, but
do not compress it. It is, of course, unsilvered. In the figure the table is repre-
sented very much closer to the mirror than it should be. In trials on the 155 meh
it has to be 25 feet distant.

The appearance that a truly spherical concave surface presents with the first test
is: the image of the hole is sharply defined without any areola of aberration around
it, and is surrounded by interference rings. Inside and outside the focus the cone
of rays is exactly similar, and circular im section. It presents no trace of irregular
illumination, nor any bright or dark circles. With the second test, when the eye
is brought into such a position that it receives the whole pencil of reflected rays,
and the opaque screen is gradually drawn across in front of the pupil, the bright-
ness of the surface slowly diminishes, until just as the screen is cutting off the last
rele of the cone of rays (Fig. 9), the mirror pre-
sents an uniform grayish tint, followed by total
darkness, and gives to the eye the sensation of a
plane.

If, however, the mirror is not spherical, but

instead gradually decreases in focal length toward
the edge, the following changes result: The
image at the best focus is surrounded by a nebu-
losity, stronger as the deviation from the sphere

Action of the Opaque Screen. is greater, and neither can a sharp focus be
obtained nor interference frimges seen. In order

to include this nebulosity in the image, it will be necessary to push the eye-piece
toward the mirror. Before the cone of rays has completed its convergence, the
mass of light will be seen to have accumulated at the periphery, and after the focus
Fig. 10. is past and divergence has commenced, the
soe accumulation will be around the axis. That

is, a caustic (Fig. 10) is formed with its
summit from the mirror. By the second
test, in gradually eclipsing the light coming
from the mirror, just before all the rays are
obstructed, a part of those which have con-
stituted the nebulosity will escape past the
screen (Fig. 11) into the eye, and cause
there an extremely exaggerated appearance

Ca ee

al Mirror.

in relief of the solid superposed upon the

A SILVERED GLASS TELESCOPE. 15

true surface beneath. The glass will no longer seem to be a plane, but to have a
section as in Fig. 12. Let us examine by the aid of M. Foucault’s diagrams why it
is that the surface seems thus curved. If the

dotted lne, Fig. 13, represents the section Fig. 11.

of the mirror, and the solid line a section of [yf

a spherical mirror of the same mean focal I ae

length, it will be seen that the curves touch a ro
at two points, but are separated by an inter- :
val elsewhere. If this interval be projected
by means of the differences of the ordinates,

Fig. 12.

Action of the Opaque Screen.

) | |
etn Mh

the section curve will seem to change, the

Apparent Section of Oblate Spheroidal Mirror. Fig. 13.
\ |
the resulting curve will be found to be the Ny Yo
same as that which the mirror apparently K 4

has. |

If the opaque screen be drawn a_ short |

distance from the mirror, the appearance of

bottom of the groove (Fig. 12) between the

centre and edge advancing inwards, and the \ (
re ee

mound in the middle growmsg smaller, If Section of Spherical and Spheroidal Mirrors.

the screen be pushed toward the mirror the

reverse takes place, the central mound becoming larger, but the edge decreasing.
The reason for these variations becomes apparent by considering the three diagrams,
Fig. 14. The dotted curve in each instance represents the real curve of the mirror
described in the last paragraph, while the

Fig. 14.

solid lines are circles drawn with radii pro-
egressionally shorter in a, 6 and c, and re-
present sections of three spherical mir-
rors whose focal lengths also progressively
shorten.

When the opaque screen is at a given
distance from the mirror under examination,
the only parts of the mirror which can offi-
ciate well are those which have a curvature

Relation of Spheres to Oblate Spheroid.
corresponding to a radius equal to the same

distance. All the other parts seem as if they were covered by projecting circular
masses. In looking at Fig. 14, it is plain, then, if the opaque screen is at a maxi-
mum distance from the mirror, that the central parts alone will seem to operate,
because the two curves (a) only touch there. If the screen is moved toward the
mirror the curves (/) will coincide at some point between the centre and edge, while
if carried still farther in only the edges touch and the appearance will be as if a

16 ON THE CONSTRUCTION. OF

large mound were fixed upon the centre. I have been careful in explaining how
a surface may thus seem to present entirely different characteristics if examined
from points of view which vary slightly in distance, because a knowledge of these
facts is of the utmost importance in correcting such an erroneous figure. It is now
obvious that the correction will be equally effectual if the mirror be polished with a
small rubber on the edge, or on the centre, or partly on each. The only difference
in the result will be, that the mean focal length will be increased in the first instance,
and decreased in the second, while it will remain unchanged in the third.

If the mirror, instead of having a section like that of an oblate spheroid, should
have either an ellipse, parabola, or hyperbola, as its section curve, the appearances
seen above are reversed. Whilst by the first test there is still an aberration round
the image at the best focus, the eye-piece must now be drawn from the mirror to
include it. The cone of rays is most dense round the axis inside, and at the
periphery outside the focus, and the
suminit of the caustic (Fig. 15) is
turned towards the mirror. The
second test shows a section as in
Fig. 16, a depression at the centre,
and the edges tumed backwards.
The nature of the movement neces-
sary to reduce the surface to a sphere

is very plainly indicated, action on a

zone « between the centre and edge.
16. If, however, a parabolic section is
ot required, the zone @ must not be

EE! Ll entirely removed, and the surface
Hee eee one eer a rendered apparently flat, but as much
of it must be left as experience shows to be desirable.

If, in still a fourth instance, the mirror is not formed by the revolution of any
regular curve upon its axis, but has upon its surface zones of longer and shorter
ene radius intermixed irregularly, a very com-

mon case, the two tests still indicate with

<= precision the parts in fault, and the correc-

tion demanded. ‘Thus the mirror seen in

= 2 SSS ae section in Fig. 17, when the principal mass

z — i of light was obstructed by the opaque screen,

i ae a would still permit that coming from certain
a parts to find its way into the eye.

Figure 18 represents an irregular mirror,

Aetonios th esOp ene eeree a that was produced in the process of correc-
Fig. 18. tion of a hyperbolic surface, which had an

Diese eee a apparent section like Fig. 16 previously.

Li Tha r > ‘ > arta n 74 ‘
‘ipareit Seotioncge MMIEEGE Gen RAS he zone a had been acted upon with a

small local polisher, and the mirror was
finished by subsequently softening down 4 and ¢ with a larger tool.

A SILVERED GLASS TELESCOPE. a

After having gained from the preceding paragraphs a general idea of the value
and nature of these tests at the centre of curvature. a more particular description
of their use is desirable. M. Foucault in his methods first brings the mirror to a
spherical surface, and then by moving the luminous pin-hole toward the mirror,
and correspondingly retracting the eye-piece or opaque screen, carries it, avoid-
ing aberration continually by polishing, through a series of ellipsoidal curvatures,
advancing step by step toward the paraboloid of revolution. The length of the
apartment, however, soon puts a termination to this gradual system of correction,
and he is forced to perform the last steps of the conversion by an empirical process,
and eventually to resort to trial in the telescope.

With my mirrors of 150 inches focal length, demanding from the outset a room
more than 25 feet long, this successive system had to be abandoned, It was wot
found feasible to place the lamp in the distant focus of the ellipse—the workshop
being less than 30 feet long

and putting the luminous source on stands outside,
introduced several injurious complications, not the least of which was currents in
the layers of variously refracting air in the apartment. In a still room the density
and hygrometric variations in its various parts only give rise to sheht embarrass-
ment. The moment, however, that currents are produced, satisfactory examination
of a mirror becomes difficult. The air is seen only too easily to move in great
spiral convolutions between the mirror and_ the eye, areolke of aberration appear
around a previously excellent image, and were it not for the second test, any de-
termination of surface would be impossible. By that test the real deviations from
truth of figure can be distinguished from the atmo-

Fig. 19.

spheric, and to a practised eye sufficient indications
of necessary changes given. Such a movement as
that caused by placing the hand in or under the line
of the converging rays, will completely destroy the
beauty of an image, and by the second test give
origin in the first case to the appearance Fig. 19.
In order to be completely exempt at all times from
aerial difficulties, it is desirable to have control of a
long underground apartment, the openings of which
can be tightly closed. As no artificial warmth is
needed, there is the minimum of movement in the
inclosed air, and conclusions respecting a surface may be arrived at in a very short
time. The mirror may also be supported from the ground, so that tremulous vibra-

Atmospheric Motions.

tions which weary the eye, and interfere with the accuracy of criticism, may be
avoided.

Driven then from observing an image kept continually free from aberration,
through advancing ellipsoidal changes, it became necessary to study the gradual
increase of deformation, produced by the greater and greater departures from a
spherical surface, as the parabola was approached. It was found that a sufficient
guide is still provided in these tests, by modifying them properly.

The longitudinal aberration of a mirror of small angular opening is easily caleu-
lated—being equal to the square of half the aperture, divided by eight times the

3 June, 1864,

18 ON THE! CONS TRUGDION AND ison

principal focal length. That is, if a 155 inch mirror of 150 inches focal length
were spherical, and were used to converge parallel rays, those from its edge would

reach a focus of an inch nearer the mirror than those from its central parts.

100
If now the converse experiment be tried, and a mirror of the same size and focal
length which can converge parallel rays, fallmg on all its parts, to one focus, be
examined at the centre of curvature, it gives there an amount of longitudinal
aberration 51°, of an inch, equal to twice the precedmg. ‘This latter, then, is the
condition at the centre of curvature, to which such mirror must be brought m order
to converge parallel rays with exactness. In addition, strict watch must be kept
upon the zones intermediate between the centre and edge, both by measurement
with diaphragms of their aberration, and better yet, by observation of the regu-
larity of the curve of that apparent solid, Fig. 16, seen by the second test.

This modification of the first test is literally a method of parabolizing by measure,
and is capable of great precision when the eye learns to estimate where the exact
focus of a zone is. The little irregularities found round the edges of the holes
through the tin screen, Fig. 8, are in this respect of material assistance. They
show, too, the increased optical or penetrating power that is gained by increase of
aperture. Minute peculiarities, not visible under very high powers with a 10 inch
diaphragm, become immediately perceptible even with less magnifying when the
whole aperture is used, provided the mirror is spherical.

In the use of the second test precautions have to be taken, as may be inferred
from page 15, to set the opaque screen exactly in the proper position. ‘The best
method for ascertaining its location is, having received the image into the eye,
placed purposely too near the mirror, to cause the screen to move across the cone
of rays from the right towards the left side. A jet black shadow begins to advance

at the same time, and in the same direction

Fig. 20.

across the mirror. If the eye is then moved
from the mirror sufficiently, this black shadow
can be made to originate by the same motion
of the screen as before, from the left or oppo-
site side of the mirror. Midway between these
extremes there is a point where the advance is
from neither side. ‘This is the true position
for the screen when it is desired to see the im-
perfections of the surface in highly exaggerat-
ed relief, as in Fig. 20, which represents the
appearance of Fig. 12.

The interpretation of the lights and shadows

upon the face of a mirror in this test is always

Adjusting the Opaque Screen.

easy, and the observer is not likely to mistake
an elevation for a depression, if he bears in mind the fact that the surface under

‘ Tn order to examine Fig. 20, the book should be held with the left side of the page toward a
window or lamp. The eye should also be at least two feet distant. The centre will then be seen to

protrude, and the surface present the apparent section engraved below it.

A SILVERED GLASS TELESCOPE. 19

examination must always be regarded as illuminated by an oblique light coming
from a source on the side opposite to that from which the screen advances, coming
for instance from the left hand side, in the above description.

In practice, the diaphragms commonly used for a 154 inch mirror have been as
small as the hight from the unsilvered surface would allow. A six inch aperture at
the centre, a rmg an inch wide round the edge, and a two inch zone midway between
the two.

e. Machines.

In the beginning of this section the difficulties into which I fell with Lord Rosse’s
machine were stated. These caused it at the time to be abandoned. A machine
based on the same idea as Mr. Lassell’s beautiful apparatus was next constructed.
[t varied, however, in this, that the hypocycloidal curve was described partly by the
rotation of the mirror, and partly by the motions of the polisher—the axes of the
spindles carrying the two being capable either of coimcidence or lateral separation
to a moderate extent. A great deal of time and labor was expended in grinding
and polishing numerous mirrors with it, but still the difficulty that had been so
annoying in the former machine persisted. Frequently, m fact’ generally, from six
to eight zones of unequal focal length were visible, although on some occasions
when the mirror was hyperbolic, the number was reduced to two. At first it was
supposed that the fault lay with the polishing, the pitch accumulating irregularly
from being of improper softness, for it was found to be particularly prone to heap
up at the centre. But after 1 had introduced a method of fine grinding with elu-
triated hone powder, which enabled the glass to reflect light before the pitch
polishing, it became evident that the zones were connected with the mode of
motion of the mechanism. Many changes were made in the speed of its various
elements, and a contrivance to control the irregular motion of the polisher intro-
duced, but a really fine and uniform parabolic surface was never obtained, the very
best showing when finished zones of different focal lengths. Although it cannot
be said that I have tried this machine thoroughly, for Mr. Lassell has produced
specula of exquisite defining power with it, and must have avoided these imperfec-
tions to a great extent, yet the evident necessity of complicating the movement!
considerably, to avoid the polishing in rings, led me to adopt an entirely different
construction, which was used until quite recently. Although it has now been
replaced by another machine, which is still better in principle, and gives fine results
much more quickly, yet as it produced one parabolic surface that bore a power of
more than 1000, and as it serves to introduce the process of grinding, it is worthy
of description. The action of machines for grinding and polishing has been
thoroughly examined in my workshop, no less than seven different ones having
been made at various times.

1 Messrs. De La Rue and Nasmyth, who used one of Mr. Lassell’s machines, as I have since
learned, met with the same trouble, and were led to make two additions to the mechanism: 1, to
control the rotation of the polisher rigorously ; and 2, to give the whole speculum a lateral motion,
by which the intersecting points of the curves described by the polisher were regularly changed in
distance from the centre of the mirror. Mr. Lassell had previously, however, introduced a contrivance

for this latter purpose himself.

2) ON THE CONSTRUCTION AND USE OF

The machine, which is a simplification of Lord Rosse’s, was intended to give
spiral strokes. It differed from the original, however, in demanding a changeable
stroke, and in the absence of the lateral motion. In another most essential feature
it varied from both that and Mr. Lassell’s, the mirror was always uppermost while
polishing, and being uncounterpoised escaped to as great an extent as possible from
the effects of irregular pressure. ‘To any one who has studied the deformations of
a reflecting surface, and knows how troublesome it is to support a mirror properly,
the advantage is apparent.

Polishing Machine.

The construction is as follows: A stout vertical shaft, a, Fig. 21, carries at its
top a circular table 6, upon which the polisher ¢ is screwed. Below a band-wheel d
is fixed. Above the table, at a distance of four inches, a horizontal bar e is arranged,
so as to move back and forward in the direction of its length, and to carry with it
by means of a screw /, the mirror m, and its iron back or chuck 7. The bar is

moved by a connecting rod /, attached to it at one end, and at the other to a pin g

The Foot Power.

moving a slot. This slot is in a crank hf, carried by a vertical shaft 7, near the
former one a. ‘The band-wheel & is connected with the foot power, Fig. 22. The

A SILVERED GLASS TELESCOPE. WN

machine, except those parts liable to wear by friction, is made of wood. The ends
oo’ of the horizontal bar e, are defended by brass tubes working in mahogany, and
have even now but little shake, though many hundred thousands of reciprocations
have been made.

The foot power consists of an endless band with wooden treads a a’. passing at
one end of the apparatus over iron wheels b 4', which carry the band-wheel ¢ upon
their axle. At the other end it goes over the rollers dd’. ‘Two pairs of inter-
mediate wheels ¢¢’, serve to sustain the weight of the man or animal working in it.
The treads are so arranged that they interlock, and form a platform, which will
not yield downwards. Owing to its inclination when a weight is put on the plat-

form a’, it immediately moves from 4 toward d and the band-wheel turns. By a
moderate exertion, equivalent to walking up a slight incline at a slow rate, a power
more than sufficient to polish a 154 inch mirror is obtained. ‘This machine,
in which very little force is lost in overcoming friction, is frequently employed
for dairy use, and is moved commonly in the State of New York by a sheep. I
have generally myself walked in the one used by me, and have travelled some days,
during five hours, more than ten miles.

In order to give an idea of the method of using a grinding and polishing machine,
the following extract from the workshop note-book is introduced :—

“A disk of plate glass 15$ inches in diameter, and 14 inch thick was pro-
cured. It had been polished flat on both sides, so that its internal constitution
might be seen.' It was fastened upon the table 4 of the machine, by four blocks
of wood as at ¢, Fig. 21. Underneath the glass were three thick folds of blanket,
15 inches in diameter, to prevent scratching of the lower face, and avoid risk of
fracture. A convex disk of lead weighing 40 pounds having been cast, was laid
upon the upper surface of the glass, and then the screw / was depressed so as to
catch in a perforated iron plate 7, at the back of the lead m, and press downward
strongly.

* Emery as coarse as the head of a pin having been introduced, through a hole
in the lead, motion was commenced and continued for half an hour, an occasional
supply of emery being given. ‘The machine made 150 cight-inch cross strokes, and
the mirror 50 revolutions per minute, ‘The grinder a was occasionally restrained
from turning by the hand. At the end of the time the detritus was washed away,
and an examination with the gauge made. A spot 11 inches in diameter, and
of an inch deep, was found to have been ground out. ‘The same process was con-
tinued at intervals for ten hours, measurements with the gauge being frequently
made. ‘The concave was then sufficiently deep. The leaden grinder was kept of
the right convexity by beating it on the back when necessary. A finer variety of
coarse emery, and after that flour emery were next put on, each for an hour. These
left the surface moderately smooth, and nearly of the right focal length. The
leaden grinder was then dismissed, and the iron one, Fig. 6, put im its stead. ‘The

1 The glass that I have used has generally been such as was intended for dead-lights and sky-

lights in ships.

29 ON THE CONSTRUCTION AND USE -OF

mirror was removed from its place, and ground upon a large piece of flat glass for
ten minutes, to produce a circular outline to the concavity. It was cemented with
soft pitch to the concave iron disk, the counterpart of Fig. 6, and again recentred
on the blanketed table 6. Emeries of 3 and 20 seconds, and 1, 3, 10, 30, 60 minutes’
elutriation were worked on it, an hour each. ‘The rate of cross motion was reduced
to 25 per minute to avoid heating, the mirror still revolving once for every three
cross strokes. ‘The screw pressure of / was stopped. ‘This produced a surface
exquisitely fine, semi-transparent, and appearing as if covered with a thin film of
dried milk. It could reflect the light from objects outside the window until an
incidence of 45 degrees was reached, and at night was found to be bright enough
for a preliminary examination at the centre of curvature.

“The polisher was constructed in the usual way (page 12), and being smeared
with rouge was fastened to the table 6, where the mirror had been. ‘The latter
warmed in water to 120° F., was then put face downwards upon the former, and
the screw / so lowered as to cause no pressure. ‘The machine was allowed to make
20 four-inch cross strokes per minute, and the polisher to revolve once for every
three strokes. The mirror being unconstrainedly supported on the polisher, was
irregularly rotated by hand, or rather prevented from rotating with the polisher.
The tendency of this method is to produce an almost spherical surface. ‘To change
it to a paraboloid, it was only necessary when the glass was polished all over to
increase the length of the stroke to 8 inches, and continue working fifteen minutes
at a time, examining in the intervals by the tests at the centre of curvature. The
production of a polish all over occupied about two hours, but the correction of
figure took more time, on account of the frequent examinations, and the absolute
necessity of allowing the mirror to come back to a state of equilibrium from which
it had been disturbed when worked on the machine.” I have seen a mirror which
was parabolic when just off the machine, by cooling over night become spherical.
And these heat changes are often succeeded by other slower molecular movements,
which continue to modify a surface for many days after.

This correction, where time and not length of stroke is the governing agent, has
once or twice been accomplished in fifteen minutes, but sometimes has cost several
hours. If the figure should have become a hyperboloid of revolution, that is, have
its edge zones too long in comparison with the centre, it is only necessary to shorten
the stroke to bring it back to the sphere, or even to overpass that and produce a
surface in which at the centre of curvature the edge zones have too short a focal
length (Fig. 12).

Very much less trouble from zones of unequal focal length was experienced after
this machine and system of working were adopted. This was owing probably partly
to the element of irregularity in the rotation of the mirror, and partly to the fact
that the surface is kept spherical until polished, and is then rapidly changed to
the paraboloid, Where the adjustments of an apparatus are made so as to attempt
to keep a surface parabolic for some hours, there is a strong tendency for zones to
appear, and of a width bearmg a fixed relation to the stroke.

The method of producing reflecting surfaces next to be spoken of, is however
that which has finally been adopted as the best of all, being capable of forming

A SILVERED GLASS TELESCOPE.

bo
wo

mirrors which are as perfect as can be, and yet only requiring a short time. It
is the correction of a surface by local retouches. In ‘the account published by M.
Foucault, it appears that he is in France the inventor of. this improvement.

The mode of practising the retouches is as follows: Several disks of wood, as a,
Fig. 23, varying from 8 inches to 5 an inch in diameter, are to be provided, and
covered with pitch or rosin of the usual hardness, in squares
as at ¢, on one side.!. On the other a low cylindrical handle NS Se
4. is to be fixed. The mirror a, Fig. 24, having been fined |
with the succession of emeries before described, is laid face
upward on several folds of blanket, arranged upon a circular
table, screwed to an isolated post in the centre of the apart-
ment, which permits the operator to move completely round it. An ordinary barrel
has generally supplied the place of the post, the head ¢, Fig. 24, serving for the
circular table, and the rim / preventing the mirror sliding
off. The other end is fastened to the floor by four cleets dd’. eee

The large polisher is first moved over the surface in straight

Local Polisher.

strokes upon every chord, and a moderate pressure is ex-
erted. As soon as the mirror is at all brightened, perhaps in
five minutes, the operation is to be suspended, and an ex-
amination at the centre of curvature made. By carefully
turning round, the best diameter for support is to be found,

and marked with a rat-tail file on the edge, and then the =
curve of the mirror ascertained. If it is nearly spherical, gection of Optician’s Post.
as will be the case if the grinding has been conducted with

care and irregular heating avoided, it is to be replaced on the blanketed support, and
the previous action kept up until a fine polish, free from dots like stippling, is
attained. ‘This stage should occupy three or four hours. Another examination
should reveal the same appearances as the preceding. It is next necessary to
lengthen the radius of curvature of the edge zones, or what is much better shorten
that of the centre, so as to convert the section curve into a parabola. This is
accomplished by straight strokes across every diameter of the face, at first with a
4 inch, then with a 6 inch, and finally with the 8 inch polisher. Examinations
must, however, be made every five or ten minutes, to determine how much lateral
departure from a direct diametrical stroke is necessary, to render the curve uniform
out to the edge. Care must be taken always to warm the polisher, either in front
of a fire or over a spirit lamp, before using it.

Perhaps the most striking feature in this operation is that the mirror presents
continually a curve of revolution, and is not diversified with undulations like a
ruffle. By walking steadily round the support, on the top of which the mirror is
placed, there seems to be no tendency for such irregularities to arise.

If the correction for spherical aberration should have proceeded too far, and
the mirror become hyperbolic, the sphere can be recovered by working a succession

2 M. Foucault used plano-convex lenses of glass, of a radius of curvature slightly less than that

of the mirror, and covered with paper on the convex face.

Q4 ON THE CONSTRUCTION AND USE OF

of polishers of increasing size on the zone a, Fig. 16, imtermediate between the
centre and edge, causing their centres to pass along every chord that can be de-
scribed tangent to the zone.

A most perfect and rapid control can thus be exercised over a surface, and an
uniform result very quickly attained. It becomes a pleasant and interesting occupa-
tion to produce a mirror. But two effects have presented themselves in this
operation, which unfortunately bar the way to the very best results. In the first
place the edge parts of such mirrors, for more than half an inch all around, bend
backwards and become of too great focal length, and the rays from these parts
cannot be united with the rest forming the image. In the second place, the sur-
face, when critically examined by the second test, is found to have a delicate wavy
or fleecy appearance, not seen in machine polishing.' Although the variations from
the true curve implied by these latter greatly exaggerated imperfections are ex-
ceedingly small, and do not prevent a thermometer bulb in the sunshine appearmg
like a disk surrounded by rings of interference, yet they must divert some undula-
tions from their proper direction, or else they would not be visible. All kinds of
strokes have been tried, straight, sweeping circular, hypocycloidal, &c. without
effecting their removal. M. Foucault, who used a paper polisher, also encountered
them. Eventually they were imputed to the unequal pressure of the hand, and in
consequence a machine to overcome the two above mentioned faults of manual
correction was constructed.

The mirror @, is carried by an iron chuck or table 6, covered with a triple

Machine for Local Corrections.

fold of blanket, and is prevented from slipping off by four cleets cc’. The vertical
shaft d passes through a worm-wheel e¢, the endless screw of which f, is driven
by a band g, from the primary shaft 2. At 7 is the band-wheel for connection to

1 By this it is not meant that there is a rippled polish, like that produced by buckskin,

A SIL VERED GLASS TELESCOPE. 25

the foot-power. At one end of the primary shaft is firmly fixed the cogwheel /,
which drives the crank-shaft 7. Attached to the horizontal part of /, is the crank-
pin m. ‘The two bolts x’ move in a slot, so that the crank-pin may be set at any
distance from 0 to 2 inches, out of line with /. Above, the crank-pin carries one
end of the bar o, the other end passing through an elliptical hole in the oak-block
p- Down the middle of the bar runs a long: slot, through which the screw-pin q¢
passes, and which permits q to be brought over any zone from the centre to the
edge of the mirror a. It is retained by the bolts 77’, which are tapped into s.
The local polisher is seen at ¢. The curve which the centre of the local polisher
describes upon the face of the mirror, varies with the adjustments. Fig. 26 is a
reduction from one traced by the machine, the overlapping
being seen on the left side. ‘The mirror is not tightly con-
fined by the cleets ¢c’, for that would certainly injure the
figure, but performs a slow motion of rotation, so that in
no two successive strokes are the same parts of the edge
pressed against them.

The local polishers are made of lead, alloyed with a small
proportion of antimony, and are 8, 6, and 4 inches in di-
ameter, respectively. The largest and smallest are most

used, the former on account of its size polishing most

Hypocycloidal Curve.

quickly, but the latter giving the truest surface. The rosin

that covers them is just indentable by the thumb nail, and is arranged in a novel
manner. The leaden basis, as seen at ¢, Fig. 25, is perforated in many places with
holes, which permit evaporation, serve for the introduction of water where needed,
and allow the rosin to spread freely. Grooves are made from one aperture to
another, and the rosin thus divided into irregular portions. The effects of the pro-
duction of heat are in this way avoided.

The mirror may be ground and fined on this machine, in the same manner as on
that described at page 21, or it may be ground with a small tool 8 inches in
diameter, as recently suggested by M. Foucault, the results in the latter case being
just as good a surface of revolution as in the former. It is best polished with the
8 inch, and a moderate pressure may be given by the screw q, if the pitch is not
too soft. This, however, tends to leave an excavated place at the centre of the
mirror, the size depending on the stroke of the crank im, which should be about 2
inches. The pin g ought to be half way from the centre to the edge of the mirror,
but must be occasionally moved right or left an inch along the slot. When the
surface is approaching a perfect polish, the warmed 4 inch polisher must be put in
the place of the 8 inch. The pin g must be set exactly half-way between the centre
and edge of the mirror, and the crank must have a stroke of two inches radius.
The polisher then just goes up to the centre of the glass surface with one edge,
and to the periphery with the other, while the outer excursion of the inner edge
and inner excursion of the outer edge meet, and neutralize one another at a mid-
way point. Wherever the edge of a polisher changes direction many times in
succession, on a surface, a zone is sure to form, unless avoided in this manner. AI]
the foregoing description is for a 155 inch mirror.

4 June, 1864,

26 ON THE CONSTRUCTION AND USE OF

By this system of local polishing the difficulties of heat, distribution of polishing
powders, irregular contact of the rosin, &c. that render the attainment of a fine
figure so uncertain usually, entirely disappear. A spherical surface is produced as
above described, and afterwards by moving g towards the edge, and at the same
time increasing the stroke, it is converted into a paraboloid. The fleecy appearance
spoken of on a former page is not perceived, and the surface is good almost up to

the extreme edge.

(4.) Eye-Pieces, PLANE Mirrors Anp ‘Test OBJECTS.

The telescope is furnished with several eye-pieces of various construction, giving
magnifying powers from 75 to 1200, or if it were desired even higher. For the
medium powers 500 and 600 Ramsden, or rather positive eye-pieces have been
adopted. ‘They differ, however, from the usual form in being achromatic, that is,
each plano-convex is composed of a flint and crown, arranged according to formulas
calculated by Littrow. In this way a large flat field and absence of color are
secured, and the fine images yielded by the mirror are not injured. For the higher
powers, single achromatic lenses are used, and for the highest of all a Ross
microscope.

With these means it has been found that the parabolic surfaces yielded by the
processes before described, will define test objects excellently. Of close double
stars they will separate such as 7° Andromedie, and show the colors of the compo-
nents. In the case of unequal stars which seem to be more severe tests, they can
show the close companion of Sirius—discovered by Mr. Alvan Clark’s magnificent
refractor—the sixth component of @' Orionis, and a multitude of other difficult
objects.

As an example of light collecting power, Debillisima between ¢ and 5 Lyre is
found to be quintuple, as first noticed by Mr. Lassell. In the 1s} inch specula of
Herschel, it was only recorded as double, and, according to Admiral Smyth, Lord
Rosse did not notice the fourth and fifth components. Jupiter’s moons show with
beautiful disks, and their difference in diameter is very marked. As for the body
of that planet, it is literally covered with belts up to the poles. The bright and
dark spots on Venus, and the fading illumination of her imner edge, and its irregu-
larities are perceived even when the air is far from tranquil. Stars are often seen
as disks, and without any wings or tails, unless indeed the mirror should be wrongly
placed, so that the best diameter for support is not in the perpendicular plane, pass-
ing through the axis of the tube.

It has been found that no advantage other than the decrease of atmospheric
influence on the image, results from cutting down the aperture of these mirrors by
diaphragms, while the disadvantage of reducing the separating power, is perceived
at the same time. Faint objects can be better seen with the whole surface than
with a reduced aperture, and this though apparently a property common to all
reflectors and object glasses is not so in reality. A defective edge will often cause
the whole field to be filled with a pale milky light, which will extinguish the fainter
stars. Good definition is just as important for faint as for close objects.

The properties of these mirrors have been best shown by the excellence of the

A SILVERED GLASS TELESCOPE. Q7

photographs taken with them. Although these are not as sharp as the image seen
in the telescope, yet it must not be supposed that an imperfect mirror will give just
as good pictures. A photograph which is magnified to 3 feet, represents a power
of 380. As the original negative taken at the focus of the mirror is not quite 14
inch in diameter when the moon is at its mean distance, it has to be enlarged
about 25 times, and has therefore to be very sharp to bear it.

The light collecting power of an unsilvered mirror is quite surprising. With a
15§ inch, the companion of a Lyre can be perceived, though it is only of the
eleventh magnitude. ‘The moon and other bright objects are seen with a purity
highly pleasing to the eye, some parts being even more visible than after silvering.

In order to finish this description, one part more of the optical apparatus requires
to be noticed—the plane mirrors. In the Newtonian reflector the image is rejected
out at the side of the tube by a flat surface placed at 45° with the optical axis of
the large concave.’ If this secondary mirror is either convex or concave, it modifies
the image injuriously, causing a star to look like a cross, and this though the curva-
ture be so slight as hardly to be perceptible by ordinary means. For a long time
T used a piece 3 <5 inches, which was cut from the centre of a large looking-glass
accidentally broken, but eventually found that by grinding three pieces of 6 inches
in diameter agaist one another, and polishing them on very hard pitch, a nearer
approach to a true plane could be made. They were tested by being put in the
telescope, and observing whether the focus was lengthened or shortened, and also
by trial on a star. When sufficiently good to bear these tests, a piece of the right

5D

size was cut out with a diamond, from the central parts.

§2. THE TELESCOPE MOUNTING.

The telescope is mounted as an altitude and azimuth imstrument, but ma manner
that causes it to differ from the usual instrument of that kind. The essential
feature is, that the eye-prece or place of the sensitive plate is
stationary at all altitudes, the observer always looking
straight forward, and never having to stoop or assume in-
convenient and constrained positions.

The stationary eye-piece mounting was first used by
Miss Caroline Herschel, who had a 27 inch Newtonian
arranged on that plan. Fig. 27. (Smyth’s Celestial Cycle.)

Subsequently it was applied to a large telescope by Mr.
Nasmyth, the eminent engineer, but no details of his con-
struction have reached me. He used it for making draw-

ings of the moon, which are said to be excellently executed.
When it became necessary to determine how my tele-

\\

scope should be mounted, I was strongly urged to make it Miss Herschel’s Telescope.

1 A right-angled prism cannot be used with advantage to replace the plane silwered mirrors, because
it transmits less light than they reflect, is more liable to injure the image, and the glass is apt to be
more or less colored. Its great size and cost, one three inches square on two faces being required
for my purposes, has also to be considered.

IS ON THE CONSTRUCTION AND USE OF

an equatorial. But after reflecting on the fact that it was intended for photography,
and that absolute freedom from tremor was essential, a condition not attamed in
the equatorial when driven by a clock, and im addition that m the case of the moon
rotation upon a polar axis does not suffice to counteract the motion in declination,
I was led to adopt the other form.

A great many modifications of the original idea have been made. For instance,
instead of counterpoising the end of the tube containing the mirror by extending
the tube to a distance beyond the altitude or horizontal axis, I introduced a system
of counterpoise levers which allows the telescope to work im a space little more than
its own focal length across. ‘This construction permits both ends of the tube to be
supported, the lower one on a wire rope, and gives the greatest freedom from tremor,
the parts coming quickiy to rest after a movement. In the use of the telescope for
photography, as we shall sce, the system of bringing the mass of the instrument to
complete rest before exposing the sensitive plate, and only driving that plate itself
by a clock, is always adopted.

The obvious disadvantage connected with the alt-azimuth mounting—the difh-
culty of finding some objects—has not been a source of embarrasment. In fact
the instability of the optical axis in reflecting instruments, if the mirror is uncon-
strainedly supported, as it should be, renders them unsuitable for determinations of
position. <A little patience will enable an observer to find all necessary tests, or
curious objects.

The mounting is divided into: a. The Tube; and b. The supporting frame.

a. The Tube.

The telescope tube is a sixteen sided prism of walnut wood, 18 inches in diameter,
and 12 feet long. ‘The staves are 2 of an inch thick, and are hooped together with
four bands of brass, capable of beg tightened by screws. Inside the tube are
placed two rings of iron, half an inch thick, reducing the internal diameter to about
16 inches. At opposite sides of the upper end of the tube are screwed the per-
forated trunnions a, Fig. 28 (of which only one is shown), upon which it swings.
Surrounding the other end is a wire rope 6 6’ 6’, the ends of which go over the
pulleys ¢ (¢ not shown) on friction rollers, and terminate in disks of lead dd’.
These counterpoises are fastened on the ends of levers ee’, which tum below on a
fixed axle /.

By this arrangement as the tube assumes a horizontal position and becomes, so
to speak, heavier, the counterpoises do the same, while when the tube becomes
perpendicular, and most of its weight falls upon the trunnions, the counterpoises are
carried mostly by their axle. A continual condition of equilibrium is thus reached,
the tube being easily raised or depressed to any altitude desired. It is necessary,
however, to constrain the wire rope 6b’ b’’, to move in the arc of the circle described
by the end of the tube and ends of the levers and hence the twelve rollers or guide
pulleys gg’ g’. Over some of the same pulleys a thin wire rope hh’ runs, but while
its ends are fastened to the lower part of the tube at 4, the central parts go twice
around a roller connected with the winch 7, near the eye-piece, thus enabling the
observer to move the telescope in altitude, without taking the eye from the eye-piece.

29

A SILVERED GLASS TELESCOPE.

*£IOPVALASYO JO MALA [BMOLPIIg

SS... WW, ie
KK KK

WELEELELEEEEEEE™ VILLE LE, WEE” FILED \
— \ ~
Yy \N \ SG lo
y \ aa \ De Cae I o/b ae ot See a ag en ey fo BES Soe ‘ SN S ty)
VV Se = Yoo \ N \ \ \ 1)
S ; __ —— Sa \
V, \ ay | SG SG S
y \ f ] , — SQ S \
Y i Le \ S SE_
y \ — SS
Yj <---~- SSS y
y = Si = - Y S S
= JY \
| = J

30 ON THE CONSTRUCTION PsNeD sso

The iron wire rope required to be carefully made, so as to avoid rigidity. It
contains 24 miles of wire, ;4, of an inch in diameter, and has 300 strands. Each
single wire will support 7 pounds It is, however, more flexible than a hempen
rope of the same size, owing to its loose twisting.

At the lower end of the tube, at the distance of a foot, and crossing it at right
angles, held by three bars of iron 77’ 7’, Fig. 29, is a circular table of oak e, which

The Mirror Support.

carries an India-rubber air sac d, and upon this the mirror / is placed. The edge
support of the mirror is furnished by a semicircular band of tin-plate a, lined inside
with cotton, and fastened at the ends by links of chain 4, (/’ not seen) to two screws
cc; g and h are the wire ropes, marked 6 and / in Fig. 28.

Instead of the blanket support which Herschel found so advantageous, M. Fou-
cault has suggested this use of an air sac. In his instrument there is a tube going
up to the observer, by which he may adjust its degree of inflation. It requires
that there should be three bearings ¢c¢' ¢’, in front of the mirror, against which it
may press when the sac behind is inflated, otherwise the optical axis is altogether
too instable, and objects cannot be found. The arrangement certainly gives beau-
tiful definition, bringing stars to a disk when the glass just floats, without touching
its front bearings. ‘The first sac that I made was composed of two circular sheets
of India-rubber cloth, jomed around the edges. But this could not be used while
photographing, because the image was kept in a state of continuous oscillation if
there was a breeze, and even under more favorable circumstances took a long time
to come to rest. It was not advisable to blow the mirror hard up against its three
front bearings, in order to avoid the instability, for then every point in of an object
became triple. ‘To the eye the oscillations were not offensive, because the swaying
image was sharp.

Subsequently, however, an air chair cushion was procured, and as the surface was
flat instead of convex the difficulty became so much less, that the blanket support
was definitely abandoned. It is necessary that the mirror should have free play in

A SILVERED GLASS TELESCOPE. Bl

the direction of the length of the tube when this kind of support is used, and that
is the reason why the tin edge hoop must terminate in links of chain.

The interval, eight or ten inches, which separates the face of the mirror from the
tube, is occupied by a curtain of black velvet, confined below by a drawing cord
and tacked above to the tube. This permits access to the mirror to put a glass
cover on it, and when shut down stops the current of air rushing up. When the
instrument is not being used this curtain is left open, because the mirror and tube
are in that case kept more uniform in temperature with the surrounding air.

In spite of such contrivances there is still sometimes a strong residual current
in the tube. I have tried to overcome it by covering the mouth of the tube with
a sheet of flat glass, but have been obliged to abandon that because the images
were injured. At one time, too, when it was supposed that the current was partly
from the observer’s body, heated streams of air going out around the tube, the
aperture in the dome was closed by a conical bag of muslin, which fitted the mouth
of the telescope tightly. The only advantages resulting were mere bodily comfort
and a capability of perceiving fainter objects than before, because the sky-lght
was shut off.

b. The Supporting Frame.
The frame which carries the preceding parts is of wood, and rests on a vertical
axis a, Fig. 30, turning below in a gun-metal cup 6, supported by a marble block

Fig. 30.

Section of Azimuth Axis.

resting on the solid rock. The upper end of the axis is sustained by two collars,
one cc’ above, and the other below an intermediate triangular box ee’ from the
sides of which three long beams f/f 12x38 inches diverge, gradually declining
till they meet the solid rock at the limits of the excavation in which the observatory

32 ON THE CONSTRUCTION AND USE OF

is placed. These beams are fastened together by cross-pieces g gg, Fig. 31, and
go through the floor in spaces hhh, so contrived that the floor does not touch them.
At the ends they are cased with a thick leaden sheathing, to deaden vibration
and prevent the access of moisture.

Plan of Observatory (lower floor).

This tripod support in connection with the sustaining of the telescope by the wire
rope, gives that steadiness which is so essential in photography. Only a slight
amount of force, about two pounds, is required to move the instrument in azimuth,
though it weighs almost a thousand pounds.

The plan of the frame centrally carried by the axis a is as follows: From the
corners of a parallelogram 77 (213 feet) of wooden beams, eight inches thick
and three inches broad, perpendiculars nv’, Fig. 28, rise. At the top they are
connected by lighter pieces to form a parallelogram, similar to that below, and just

A SILVERED GLASS TELESCOPE 33
large enough to contain the tube of the telescope. At right angles to the parallelo-
gram below, and close upon it, a braced bar 00’, Fig. 28, crosses. From its ex-
tremities four slanting braces as at p p’, Fig. 28, go to the comers of the upper
parallelogram, and combine to give it lateral support. At the top of one close
pair of the perpendiculars n’, Fig. 28, are bronze frames carrying friction rollers
upon which the trunnions move, while similarly upon the other pair ” are two
pulleys, also on friction rollers, for the wire rope coming from the counterpoises.

Movement in altitude is very easily accomplished, and with the left hand upon
the winch 7, under high powers, both altitude and azimuth motions are controlled,
and the right hand left free. The whole apparatus works so well, that in ordinary
observation the want of a clock movement has not been felt. Of course for pho-
tography that is essential.

§3. THE CLOCK MOVEMENT.

The apparatus for following celestial bodies is divided into two parts; a. The
Sliding Plate-holder ; and b. The Clepsydra. In addition a short description of the
Sun-Camera, c, is necessary.

a. The Sliding Plate-holder.

Mr. De La Rue, who has done so much for celestial photography, was the first
to suggest photographing the moon on a sensitive plate, carried by a frame moving
in the apparent direction of her path. He never, however, applied an automatic
driving mechanism, but was eventually led to use a clock which caused the
whole telescope to revolve upon a polar axis, and thus compensate for the rotation
of the earth, and on certain occasions for the motion of the moon herself. In this
way he has produced the best results that have been obtained in Europe. Lord
Rosse, too, employed a similar sliding plate-holder, but provided with clock-work
to move it at an appropriate rate. I have not been able as yet to procure any
precise account of either of these instruments.

The first photographic representations of the moon ever made, were taken by
my father, Professor John W. Draper, and a notice of them published in his quarto
work “On the Forces that Organize Plants,” and also in the September number,
1840, of the London, Edinburgh, and Dublin Philosophical Magazine. He _ pre-
sented the specimens to the New York Lyceum of Natural History. The Secretary
of that Association has sent me the following extract from their minutes :—

“ March 23d, 1840. Dr. Draper announced that he had succeeded in getting a
representation of the moon’s surface by the Daguerreotype ..... The time
occupied was 20 minutes, and the size of the figure about 1 inch in diameter.
Daguerre had attempted the same thing, but did not succeed. ‘This is the first time
that anything like a distinct representation of the moon’s surface has been obtained.

“Rost. H. Brownne, Secretary.”

As my father was at that time however much occupied with experiments on the

Chemical Action of Light, the Influence of Light on the Decomposition of Car-
D June, 1864,

34 ON THE CONSTRUCTION AND USE OF

bonie Acid by Plants, the Fixed Lines of the Spectrum, Spectrum Analysis, &c., the
results of which are to be found scattered through the Philosophical Magazine,
Silliman’s Journal, and the Journal of the Franklin Institute, he never pursued this
very promising subject. Some of the pictures were taken with a three inch, and
some with a five inch lens, driven by a heliostat.

In 1850, Mr. Bond, taking advantage of the refractor of 15 inches aperture at
Cambridge, obtained some fine pictures of the moon, and subsequently of double
stars, more particularly Mizar in Ursa Major. The driving power, in this mstance,
was also applied to move the telescope upon a polar axis.

Besides these, several English and continental observers, Messrs. Hartnup, Phillips,
Crookes, Father Secchi, and others, have worked at this branch of astronomy, and,
since 1857, Mr. Lewis M. Rutherfurd, of New York, has taken many exquisite
lunar photographs, which compare favorably with foreign ones.

But in none of these instances has the use of the sliding plate-holder been per-
sisted in, and its advantages brought into view. In the first place it gets rid com-
pletely of the difficulties arising from the moon’s motion in declination, and in the
second, instead of injuring the photograph by the tremors produced in moving the
whole heavy mass of a telescope weighing a ton or more, it only necessitates the
driving of an arrangement weighing scarcely an ounce.

My first trials were with a frame to contain the sensitive plate, held only at three
points. ‘Two of these were at the ends of screws to be turned by the hands, and
the third was on a spring so as to maintain firm contact. ‘This apparatus worked
well in many respects, but it was found that however much care might be taken,
the hands always caused some tremor in the instrument. It was evident then
that the difficulty from friction which besets the movements of all such delicate
machinery, and causes jerking and starts, would have to be avoided in some other way.

I next constructed a metal slide to run between two parallel strips, and ground
it into position with the greatest care. ‘This, when set in the direction of the moon’s
apparent path, and moved by one screw, worked better than the preceding. But
it was soon perceived that although the strips fitted the frame as tightly as practi-
cable, an adhesion of the slide took place first to one strip and then to the other, and
a sort of undulatory or vermicular progression resulted. ‘The amount of deviation
from a rectilinear motion, though small, was enough to imjure the photographs.
At this stage of the investigation the regiment of volunteers to which I belonged
was called into active service, and I spent several months in Virginia.

My brother, Mr. Damel Draper, to whose mechanical ingenuity I have on several
occasions been indebted for assistance in the manifold difficulties that have arisen
while constructing this telescope, continued these experiments at imtervals. He
presented me on my return with a slide and sand-clock, with which some excellent
photographs have been taken. He had found that unless the slide above mentioned
was made ungovernably long, the same trouble continued. He then ceased catch-
ing the sliding frame h, Fig. 32, by two opposite sides, and made it run along a
single steel rod a, being attached by means of two perforated plates of brass b, 0’.
The cord ¢ going to the sand-clock, was applied so as to pull as nearly as possible in

the direction of the rod, A piece of cork c, gave the whole steadiness, and yet

A SILVERED GLASS TELESCOPE. 20
softness of motion. ‘The lower end of the frame was prevented from swinging back
snd forward by a steel pin d, which played along the glass rod e. All these parts
were attached to a frame /, fitting on the eyepiece
holder, and permitting the rod @ to change from
the horizontal position in which it is here drawn, to
any angular one desired. The thumb-screw / re-
tamed it in place; g and gy’ are pulleys which per-
mit the cord to change direction.

Subsequently, a better method of examining the
untformity of the rate, than by noticing the sharp-
ness of the photograph produced, was invented. — It
consists im arranging a fixed microscope, magnify-
ing about 40 times, at the back of the ground glass
plate, which fits in the same slide as the sensitive

plate. By watching the granulated appearance Bee ences
pass before the eye, as the slide is moved by the ‘i
clock, the slightest variation from uniformity, any pulsatile or jerking movement is
rendered visible. By the aid of this microscopic exaggeration, it was seen that
occasionally, when there had been considerable changes in temperature, the steadi-
ness of the motion varied. This was traced to the irregular slipping of 6, 6’.

A different arrangement was then adopted, by which a lunar crater can be
kept bisected as long as is necessary, and which gives origin to no irregularities, but
pursues a steady course. The principle is, not to allow a slipping friction anywhere,
but to substitute rolling friction, upon wheels tuming on points at the ends of their
axles. The following wood-cut is half the real size of this arrangement.

Frictionless Slide (front view). Sectional view.

36 ON THE CONSTRUCTION AND USE OF

A glass rod a, a’, Fig. 33, is sustained by two wheels 6. 6’, and kept in contact
with them by a third friction roller ¢, pressed downward by a spring. This rod
carries a circular frame d, d’, wpon which at e, e, e’, are three glass holders and
platinum catches. A spring / holds the sensitive plate in position, by pressing
against its back. The circular frame d is kept in one plane by a fourth friction
roller g, Which runs on a glass rod /, and is kept against it by the ward pressure
of the overhanging frame ¢. The cord 7 is attached to the arm /, and pulls im the
direction of the glass rod a. From am to a fixed point near }, a strip of elastic India-
rubber is stretched, to keep the cord tight. The ring of brass 1, 7’ carries the
whole, serving as a basis for the stationary parts, and in its tun being fastened to
the eyepiece holder, so as to allow the glass rod a to change direction, and be
brought into coincidence with the apparent path of the moon. At o is a thumb-
screw or Clamp Through the ring », 2’, a groove p is cut, into which a piece of
yellow glass may be placed, when the actinic rays are to be shut off from the plate.

Since this contrivance has been completed, all the previous difficulties have
vanished. ‘The moving of a plate can be accomplished with such precision, that
when the atmosphere was steady, negatives were taken which have been enlarged
to three feet in diameter.

The length of time that such a slide can be made to run is indefinite, depending
in my case on the size of the diagonal flat mirror, and aperture of the eyepiece
holder. I can follow the moon for nearly four minutes, but have never required
to do so for more than fifty seconds. At the mouth of the instrument, where no
secondary mirror is necessary, the time of running could be increased.

‘The setting of the frictionless slide in angular position is accomplished as follows :
A ground glass plate is put into it, with the ground face toward the mirror. Upon
this face a black line must have been traced, precisely parallel to the rod, This
may be accomplished by firmly fixing a pencil point against the ground side, and
then drawing the frame ¢ and glass past it, while the rest of the slide is held fast.
As the moon passes across the field, the position of the apparatus must be changed,
until one of the craters runs along the line from end to end. A cross line drawn
perpendicular to the other, serves to adjust the rate of the clepsydra as we shall
see, and when a crater is kept steadily on the intersection for twice or three times
the time demanded to secure an impression, the adjustment may be regarded as
complete.

It is necessary of course to expose the sensitive plate soon after, or the apparent
path of the moon will have changed direction, unless indeed the slide is set to suit
a future moment.

b. The Clepsydra.

My prime mover was a weight supported by a column of sand, which, when the
sand was allowed to run out through a variable orifice below, could be made to
descend with any desired velocity and yet with uniformity. In addition, by these
means an unlimited power could be brought to bear, depending on the size of the
weight. Previously it was proposed to use water, and compensate for the decrease
in flow, as the column shortened, by a conical vessel; but it was soon perceived that

A SILVERDED GLASS THLUSCOPRT 37

as each drop of water escaped from the funnel-shaped vessel, only a corresponding
weight would be brought into play. This is not the case with sand, for in this
instance every grain that passes out causes the whole weight that is supported by
the column to come into action. In the former instance a movement consisting of
a series of periods of rest and periods of motion occurs, because power has to accu-
mulate by floating weight lagging behind the descending water, and then suddenly
overtaking it. In the latter case, on the contrary, there is a regular descent, all
minor resistances in the slide being overcome by the steady application of the whole
mass of the weight.

When these advantages in the flow of sand were ascertained, all the other prime
movers were abandoned. Mercury-clocks, on the principle of the hydrostatic para-
dox, air-clocks, &c., in great variety, had been constructed.

The sand-clock consisted of a tube a (Fig. 34), eighteen inches long and one and
a half in diameter, nearly filled with sand that had been raised to a bright red heat
and sifted. Upon the top of the sand a leaden
weight 6 was placed. At the bottom of the Fig. 34.
tube a peculiar stopcock, seen at (2) enlarged,
regulated the flow, the amount passing depend-
ing on the size of the aperture d. This stop-
cock consisted of two thin plates, fixed at one
end and free at the other. The one marked e
is the adjusting lever, and its aperture moves
past that in the plate gy. The lever f serves to
turn the sand off altogether, without disturbing
the size of the other aperture, which, once set
to the moon’s rate, varies but slightly in short
times. A movable cover h, perforated to allow
the cord 7 to pass through, closed the top, while
the vessel /: retained the escaped sand, which at
suitable times was returned into the tube «, the
weight 4 bemg temporarily lifted out. From Te cece
the clock the cord i communicated motion to
the frictionless slide, as shown in Fig. 33. This cord should be as inelastic as

possible, consistent with pliability, and well waxed.

One who has not investigated the matter would naturally suppose that the flow
of sand in such a long tube would be much quicker when the tube was full than
when nearly empty, and that certainly that result would occur when a heavy weight
was put on the shifting mass. But in neither case have 1 been able to detect the
slightest variation, for, although by shaking the tube a diminution of the space
occupied by the sand may be caused, yet no increase of weight tried could accom-
plish the same reduction. These peculiarities seem to result from the sand arching
as it were across the vessel, like shot in a narrow tube, and only yielding when the
under supports are removed. In blasting, a heavy charge of gunpowder can be
retained at the bottom of a hole, and made to split large masses of rock, by filling
the rest of the hole with dry sand.

38 ON THE CONSTRUCTION AND USE OF

I believe that no prime mover is more suitable than a sand-clock for purposes
where steady motion and a large amount of power are demanded. The simplicity,
for instance, of a heliostat on this plan, the large size it might assume, and its
small cost, would be great recommendations. In these respects its advantages over
wheelwork are very apparent. The precision with which such a sand-clock goes
may be appreciated when it is stated, that under a power of 300 a lunar crater can
be kept bisected for many times the period required to photograph it, ‘To secure
the greatest accuracy in the rate of a sand-clock, some precautions must be taken.
The tube showld be free from dents, of uniform diameter, and very smooth or
polished inside. Water must not be permitted to find access to the sand, and
hygrometric varieties of that substance should be avoided, or their salts washed out.
The sand should be burned to destroy organic matter, and so sifted as to retain
grains nearly equal in size. ‘The weight, which may be of lead, must be turned so
as to go easily down the tube, and must be covered with writing paper or some
other hard and smooth material, to avoid the proneness to adhesion of sand. A
long bottle filled with mercury answers well as a substitute.

T have used in such clocks certain metallic preparations: Fine shot, on account
of its equality of size, might do for a very large clock with a considerable opening
below, but is unsuitable for a tube of the size stated above. There is, however,
a method by which lead can be reduced to a divided condition, like fine gunpowder,
when it may replace the sand. If that metal is melted with a little antimony, and
while cooling is shaken in a box containing some plumbago, it breaks up at the
instant of solidifying into a fine powder, which is about five times as heavy as sand.
If after being sifted to select the grains of proper size, it is allowed to run through
a small hoie, the flow is seen to be entirely different from that of sand, looking as
if a wire or solid rod were descending, and not an aggregation of particles. It is
probable, therefore, that it would do better than sand for this purpose. T have not,
however, given it a fair trial, because just at the time when the experiments with the
sand-clock had reached this point, I determined to try a clepsydra as a prime mover,

‘The reason which led to this change was that it was observed on a certain occa-
sion when the atmosphere was steady, that the photographs did not correspond in
sharpness, being in fact no better than on other nights when there was a consider-
able flickermg motion in the air. A further investigation showed that in these
columns of sand there is apt to be a minute vibrating movement. At the plate-
holder above this is converted into a series of arrests and advances. On some
occasions, however, these slight deviations from continuous motion are entirely
absent, and generally, indeed, they cannot be seen, if the parts of the image seem
to vibrate on account of currents in the air. By the aid of the microscopic exag-
geration described on a former page—which was subsequently put m practice—
they may be observed easily, if present.

When the negative produced at the focus of the great mirror is intended to be
enlarged to two feet or more in size, these movements injure it sensibly. A variety
of expedients was resorted to in order to avoid them, but none proved on all
occasions successful.

It is obvious that im a water-clock, where the mobility of the fluid is so much

A SILVERED GLASS TELESCOPE. 39

greater than that of solid grains, this difficulty would not arise. The following
contrivance in which the fault of the ordinary clepsydra, in varying rate of flow as
the column shortens, is avoided, was next made. With it the best results are
attaimable, and it seems to be practically perfect.

Fig. 35.

The Clepsydra.

It consists of a cylinder a, in which a piston 6 moves watertight. At the top of
the piston rod is a leaden five-pound weight ¢, from which the cord 7 goes to the
sliding plateholder g. ‘The lower end of the pointes terminates in a stopcock «,
the handle of which carries a strong index rod e, moving on a divided are. At f
a tube with a stopcock is attached. Below, a Si receives the waste fluid.

In using the clepsydra the stopcock of fis opened, and the piston being pulled
upwards, the cylinder fills with water from h. ‘The stopcock is then closed, and if
d also is shut, the weight will remain motionless. ‘The string 7 is next connecte od
with the slide, and the telescope turned on the moon, As soon as the slide is
adjusted in angular position (page 36) the stopcoc k d is opened, until the weight ¢
moves downwards, at a rate that matches the moon’s apparent motion.

In order to facilitate the rating of the clepsydra, the index rod ¢ is pressed by a
spring / (2), against an excentric 1, As the excentric is turned round, the stopcock
d is of course opened, with great precision and delicacy. The plug of this stop-
cock (3) is not perforated by a round hole, but has a slit. ‘This causes equal move-

40 ON THE CONSTRUCTION AND USHIOE

ments in the rod e, to produce equal changes in the flow. The ratmg requires
consequently only a few moments.

The object of the side tube / is to avoid disturbing ¢ when it becomes necessary
to refill the cylinder, for when it is once opened to the right degree, it hardly
requires to be touched again during a night’s work. In order to arrest the down-
ward motion of the piston at any point, a clamp screws on the piston rod, and can
be brought into contact with the cylinder head, as in the figure.

That this instrument should operate in the best manner, it is essential to have
the interior of the brass cylinder polished from end to end, and of uniform diameter.
If any irregularity should be perceived in the rate of going, it can be cured com-
pletely by taking out the piston, impregnating its leather stuffing with fine rotten
stone and oil, and then rubbing it up and down for five minutes in the cylinder, so
as to restore the polish. The piston and cylinder must of course be wiped, and
regreased with a mixture of beeswax and olive oil (equal parts) after such an
operation. In replacing the piston, the cylinder must be first filled with water, to
avoid the presence of air, which would act as a spring.

Although it may be objected that this contrivance seems to be very troublesome
to use, yet that is not the case in practice. Even if it were, it so far surpasses any
prime mover that I have seen, where the utmost accuracy is needed, that it would
be well worth employing.

ce. The Sun Camera.

In taking photographs of the sun with the full aperture of this telescope, no
driving mechanism is necessary. On the contrary, the difficulty is rather to arrange
the apparatus so that an exposure short enough may be given to the sensitive plate,
and solarization of the picture avoided. It is not desirable to reduce the aperture,
for then the separating power is lessened. ‘The time required to obtain a negative
is a very small fraction of a second, for the wavy appearance produced by atmos-
pheric disturbance is not unfrequently observed sharply defined in the photograph,
though these aerial motions are so rapid that they can scarcely be counted. Some
kind of shutter that can admit and cut off the solar image with great quickness is
therefore necessary.

in front of an ordinary camera a, Fig. 36, attached to the eyepiece holder of the
telescope, and from which the lenses have been removed, a spring shutter is fixed.
It consists of a quadrant of thin wood 6, fastened by its right

Fig. 36.

angle to one corner of the camera. Over the hole in this

quadrant a plate of tin d can be adjusted, and held in position

by a serew moving in a slot so as to reduce the hole if desired

to a mere slit. It may vary from 1} inch to less than ,, of

an inch, The quadrant is drawn downwards by an India-
1

rubber spring g, 1 inch wide, § of an inch thick, and 8 inches

== long. ‘This spring is stretched when in action to about 12
Pheepring Shane! inches, and when released draws the slit past the aperture ¢ m
the camera. ‘Two nicks in the edge of the quadrant serve with the assistance of a

pin e, which can easily be drawn out by a lever (not shown in the cut), to confine

A SILVERED GLASS TELESOOPE. 4]

the slit either opposite to or above c. A catch at / prevents the shutter recoiling.
The sensitive plate is put inside the box as usual in a plate-holder. When a photo-
graph is taken, the spring shutter is drawn up so that the lower nick in the edge
of the quadrant is entered by the pin e, and the inside of the camera obscured.
The front slide of the plateholder is then removed in the usual manner, and the
solar image being brought into proper position by the aid of the telescope finder,
the trigger retaining ¢ is touched, the shutter flies past c, and the sensitive plate may
then be removed to be developed.

To avoid the very short exposure needed when a silvered mirror of 188 square
inches of surface is used, I have taken many solar photographs with an unsilvered
mirror, which only reflects according to Bouguer 25 per cent. of the light falling
upon it, and should permit an exposure 37 times as long as the silvered mirror.
This is the first time that a plain glass mirror has been used for such a purpose,
although Sir John Herschel suggested it for observation many years ago. But
eventually this application of the unsilvered mirror had to be abandoned. It has,
it is true, the advantage of reducing the light and heat, but I found that the moment
the glass was exposed to the Sun, it commenced to change in figure, and alter in
focal length. This latter difficulty, which sometimes amounts to half an inch,
renders it well nigh impossible to find the focal plane, and retain it while taking
out the ground glass, and putting in the sensitive plate. If the glass were supported
by a ring around the edge, and the back left more freely exposed to the air, the
difficulty would be lessened but not avoided, for a glass mirror can be raised to 120°
F. on a hot day by putting it in the sunshine, though only resting on a few points.
Other means of reducing the light and heat, depending on the same principle, can
however be used. By replacing the silvered diagonal mirror with a black glass or
plain unsilvered surface, as suggested by Nasmyth, the trouble sensibly disappears.

oD

I have in this way secured not only macule and their penumbrie, but also have
obtained faculie almost invisible to observation. On some occasions, too, the precipi-
tate-like or minute flocculent appearance on the Sun’s disk was perceptible.

It seems, however, that the best means of acquiring fine results with solar photo-
graphy, would be to use the telescope as a Cassegranian, and produce an image so
much enlarged, that the exposure would not have to be conducted with such rapidity.
Magnifying the image by an eyepiece would in a general way have the same result,
but in that case the photographic advantages of the reflector would be lost, and it
would be no better than an achromatic,

§ 4. THE OBSERVATORY.
This section is divided into a, The Building; 6, The Dome; and c, The Observer's
Chair.
a. The Building.
The Observatory is on the top of a hill, 225 feet above low water mark, and is
in Latitude 40° 59’ 25’ north, and Longitude 73° 52’ 25’ west from Greenwich,
according to the determinations of the Coast Survey. It is near the village of

Hastings-upon-IHudson, and is about 20 miles north of the city of New York. The
6 July, 1864.

42 ON THE CONSTRUCTION AND USEROR

“

surrounding country on the banks of the North River is occupied by country seats,
on the slopes and summits of ridges of low hulls, and no offensive manufactories

Dr. Draper’s Observatory.

vitiate the atmosphere with smoke, Our grounds are sufficiently extensive to exclude
the near passages of vehicles, and to avoid tremor and other annoyances.

An uninterrupted horizon is commanded in every direction, except where trees
near the dwelling house cut off a few degrees toward the southwest. The advantages
of the location are very great, and often when the valleys round are filled with
foggy exhalations, there is a clear sky over the Observatory, the mist flowing down
like a great stream, and losing itself in the chasm through which the Hudson here
passes.

The foundation and lower story of the building are excavated out of the solid
granite, which appears at the edge of the hill. This arrangement was intended to
keep the lower story cool, and avoid, in the case of the metal reflector, sudden
changes of temperature. ‘The eastern side of the lower story, however, projects
over the brow of the hill, and is therefore freely exposed to the air, furnishing, when
desired, both access and thorough ventilation through the door. The second story
or superstructure is of wood, lined inside with boards like the story below. They
serve to inclose in both cases a non-conducting sheet of air.

The inside dimensions of both stories taken together are 174 feet square, and 22
fect high, to the apex of the dome. ‘This space is unnecessarily large for the tele

A SILVERED GLASS TELESCOPE. 43

e

scope, which only requires a cylinder 13 feet in diameter and 13 feet high. A gene-
ral idea of the internal arrangement is gained from Fig. 28. In Fig. 38, a a’ is the

Fig. 38.

Plan of Observatory (upper floor).

floor of the gallery, 6 6’ 6” the circular aperture in which the telescope ¢ c’ turns.
The staircase is indicated by d. The Enlarger, § 6, rests on the shelf e, the helios-
tat being outside at The door going into the photographie room is at g, h I’ are
tables, 7 the water tank, / the tap and sink, / the stove, m a heliostat shelf, n the
door, 0 the window.

The building is kept ventilated by opening the door in the lower part, and the
dome shutter, seen in Fig. 37, for some time before using the instrument. On a sum-
mer day the upper parts, and especially those close under the dome, become without
this precaution very hot, and this occurred even before the tin roof was painted.
Bright tinplate seems not to be able to reflect by any means all the heat that falls
upon it, but will become so warm in July that rosin will melt on it, and insects which
have lighted in a few moments dry up, and soon become pulverizable. A knowledge
of these facts led to the abandonment of wooden sheathing under the tin, for without
it when night comes on the accumulated heat radiates away rapidly, and ceases to
cause aerial currents near the telescope.

The interior of the building is painted and wainscoted, and the roof is ornas
mented partly in blue and oak, and partly with panels of tulip-tree wood.

There are only two windows, and they are near the southern angles of the roof.
While they admit sunshine on some occasions, they can on others be closed, and
the interior be reduced to darkness. In the southeast corner a small opening e¢

44 ON THE CONSTRUCTION AND USE OF

may allow a solar beam three inches in diameter to come in from a heliostat outside.
The greatest facilities are thus presented for optical and photographical experiments,
for in the latter case the whole room can be used as a camera obscura.

b. The Dome.

The roof of the observatory is 20 feet square. The angles are filled in solid, and
a circular space 15 feet in diameter is left to be covered by the revolving dome.
Although such a construction is architecturally weak and lable to lose its level, yet
the great advantages of having the building below square, and the usefulness of
the comers, determined its adoption, the disadvantages bemg overcome by a very
light dome.

‘The dome is 16 feet in outside diameter, and rises to a height of 5 feet above its
base. It is, therefore, much flatter than usual, in fact, might have been absolutely
flat, with this method of mounting. It would then have been liable, however, to
be crushed in by the deep winter snows.

It consists of 32 ribs, arcs of a circle, uniting at a common centre above. Each
one is formed of two pieces of thin whitewood, 6, Fig. 39, fastened side by side,
with the best arrangements of the grain for strength. They are three inches wide
and one inch thick at the lower end, and taper gradually to 25 by 1.

Over these ribs tinplate is laid in triangular strips or gores, about 18 imches wide
at the base, and 10 feet long. Where the adjacent triangles of tin a@ a’ meet, they
are not soldered, but are bent together. This allows a certain
amount of contraction and expansion, and is water-proof. — It
strengthens the roof so much, that if the ribs below were taken
away, this corrugated though thin dome would probably
sustain itself. The tin is fastened to the dome ribs 6 by
extra pieces ¢ imserted in the joint and doubled with the
other parts, while below they are nailed to the ribs. In the
figure the tin is represented very much thicker than it is in
reality.

Tenis ini TincoeDanie: This dome, although it has 250 square feet of surface, only
weighs 250 pounds. ‘That at the Cambridge (Massachusetts)
Observatory, 291 feet in diameter, weighs 28,000 pounds.

The shit or opening is much shorter than usual, only extending half way from
the base towards the summit. It is in reality an inclined window, 2+ feet wide at
the bottom, 15 wide at the top, and 4 feet long. It is closed by a single shutter,
as seen in Fig. 37, and this when opened is sustained in position by an iron rod
furnished with a hinge at one end and a hook at the other.

The principal peculiarity of the dome, the means by which it is rotated, remains
to be described. Usually in such structures rollers or cannon balls are placed at
intervals under the edge, and by means of rack work, a motion of revolution is
slowly accomplished. Here, on the contrary, the whole dome 6 6’ b” (Fig. 40) is
supported on an arch hh’ h’’, carrying an axis @ at its centre, around which a slight
direct force, a pull with a single finger, will cause movement, and by a sudden push
even a quarter of an entire revolution may be accomplished. It is desirable, how-

A SILVERED GLASS TELESCOPE. 45

ever, to let it rest on the edge 6 6’, when not in use. At c there is an iron catch
on the arch, by which the lever e, that raises the dome, is held down. The fulerum

Fig. 40.

The Dome Arch.

is at d. The lever is hinged near c, so that when by being depressed it should have
come in the way of the telescope below, the lower half g can be pushed up, the
part from ¢ toward d still holding the dome supported.

The arch can be set across the observatory im any direction, north and south, east
and west, or at any intermediate position, because the abutments where the ends
rest, are formed by a ring //' 1’, fastened round the circular aperture, through the
stationary part of the roof.

When the telescope is not in use, and the dome is let down, so that there is no
longer an interval of a quarter of an inch between it and the rest of the roof, it
is confined inside by four clamps and wedges. Otherwise, owing to its lightness, it
would be lable to be blown away. ‘These clamps a,
Fig. 41, are three sides of a square, made of iron one Pee
inch square. They catch above by a point in the
wooden basis-circle of the dome 6, and below are
tightened by the wedge e.

When the dome is raised it is prevented from moving
laterally and sliding off by three rollers, one of which
isseen at f, Fig. 40. These catch against its imer edge, —
and only allow slight play. At first it was thought ne-
cessary to have a subsidiary half arch at right angles to
the other to hold it up, but that is now removed. A Dome Cian

All the parts work very satisfactorily, and owing to
the care taken to get the roof-circle and basis-circle flat and level, no leakage takes

place at the joint, and even snow driven by high winds is unable to enter.

c. The Observer's Chair.

This is not a chair in the common acceptation of the word, but is rather a movable
platform three feet square, capable of carrying two or more persons round the
observatory, and maintaining them in an invariable position with regard to the tele-
scope eyepiece.

46 ON THE CONSTRUCTION AND USE OF

Its general arrangement is better comprehended from the sketch, Fig. 42, than
from a labored description. Below, it runs on a pair of wheels a (one only is

The Observer’s Chair.

visible) 9 inches in diameter, whose axles point to the centre of the circle upon
which they run. ‘They are prevented from shifting outwards by a wooden railroad
b, b’, and inwards by the paling /,/’. Above, the chair moves on a pair of small
rollers ¢, which press against a circular strip or track d, d’, nailed around the lower
edge of the dome opening. Access to the platform is gained by the steps e¢, é’.
Attached to the railing of this platform, and near it on the telescope, are two
tables (not shown in the figure) for eyepieces, the sliding plateholder, &c.

§ 5. THE PHOTOGRAPHIC LABORATORY.

This section is divided into a, Description of the Apartment; and 6, Photographic
Processes.

a. Description of the Apartment.

The room in which the photographical operations are carried on, adjoins and
connects with the observatory on the southeast, as is shown in Figs. 28 and 38.
It is 9 by 10 feet inside, and is supplied with shelves and tables running nearly all the
way round, which have upon them the principal chemical reagents. It is furnished,
too, with an opening to admit, from a heliostat outside, a solar beam of any size, up
to three inches in diameter.

The supply of water is derived from rain falling on the roof of the building, and

A SILVERED GLASS TELESCOPE. 47

running into a tank 7, Fig. 38, which will contain a ton weight. The roof exposes
a surface of 532 square feet, and consequently a fall of rai equal to one inch in
depth, completely fills the tank. During the course of the year the fall at this place
is about 32 inches, so that there is always an abundance. In order to keep the
water free from contamination, the roof is painted with a ground mineral compound,
which hardens to a stony consistence, and resists atmospheric influences well. The
tank is lined with lead, but having been in use for many years for other purposes, is
thoroughly coated inside with various salts of lead, sulphates, &c. In addition the
precaution is taken of emptying the tank by a large stopcock when a rainstorm is
approaching, so that any accumulation of organic matter, which can reduce nitrate
of silver, may be avoided. It has not been found feasible to use the well or spring
water of the vicinity.

The tank is placed close under the eaves of the building, so as to gain as much
head of water as is desirable. From near its bottom a pipe terminating in a stop-
cock i, Fig. 38, passes into the Laboratory. In the northeast corner of the room,
and under the tap is a sink for refuse water and solutions, and over which the
negatives are developed. It is on an average about twelve feet distant from the
telescope. In another corner of the room is a stove, resembling in construction an
open fireplace, but sufficient nevertheless to raise the temperature to 80° F. or higher,
if necessary. Asa provision against heat in summer, the walls and roof are double,
and a free space with numerous openings above is left for circulation of air, drawn
from the foundations. The roof is of tinplate, fastened directly to the rafters, with-
out sheathing, in order that heat may not accumulate to such an extent during the
day as to constitute a source of disturbance when looking across it at night.

For containing negatives, which from being unvarnished require particular care,
there is at one side of the room a case with twenty shallow drawers each to hold
eighteen. ‘They accumulate very rapidly, and were it not for frequent reselections
the case would soon be filled. On some nights as many as seventeen negatives have
been taken, most of which were worthy of preservation. Not less than 1500 were
made in 1862 and ’63.

b. Photographic Processes.

In photographic manipulations I have had the advantage of my father’s long
continued experience. He worked for many years with bromide and chloride of
silver in his photo-chemical researches (Journal of the Franklin Institute, 1837),
and when Daguerre’s beautiful process was published, was the first to apply it to
the taking of portraits (Phil. Mag., June, 1840) in 1839; the most important of all
the applications of the art. Subsequently he made photographs of the interference
spectrum, and ascertained the existence of great groups of lines JM, N, O, P, above
H, and totally invisible to the naked eye (Phil. Mag., May, 1843). The importance
of these results, and of the study of the structure of ilames containing various
elementary bodies, that he made at the same time, are only now exciting the
interest they deserve.

In 1850, when his work on Physiology was in preparation, and the numerous
illustrations had to be produced, I learnt microscopic photography, and soon after

48 ON THE CONSTRUCTION AND USE OF

prepared the materials for the collodion process, then recently invented by Scott
Archer. We produced in 1856 many photographs wider a power of 700 diameters,
by the means described in the next section.

At first the usual processes for portrait photography were applied to taking the
Moon. But it was soon found necessary to abandon these and adopt others. When
a collodion negative has to be enlarged—and this is always the case in lunar photo-
graphy, where the original picture is taken at the focus of an object glass or mirror
—imperfections invisible to the naked eye assume an importance which causes the
rejection of many otherwise excellent pictures. Some of these imperfections are
pinholes, coarseness of granulation in the reduced silver, liability to stains and mark-
ings, spots produced by dust.

These were all avoided by washing off the free nitrate of silver from the sensitive
plate, before exposing it to the light, and again submitting it to the action of water,
and dipping it back into the nitrate of silver bath before developing. ‘The quantity
of nitrate of silver necessary to development when pyrogallic acid is used, is how-
ever better procured by mixing a small quantity of a standard solution of that salt
with the acid.

The operation of taking a lunar negative is as follows. The glass plates 27 x 34
inches are kept in nitric acid and water until wanted, ‘They are then washed under
a tap, being well rubbed with the fingers, which have of course been properly cleaned.
They are wiped with a towel kept for the purpose. Next a few drops of iodized
collodion are poured on each side, and spread with a piece of cotton flannel. ‘They
are then polished with a large piece of this flannel, and deposited in a close dry
plate box. ‘This system of cleaning with collodion was suggested by Major Russel,
to whose skilful experiments photography is indebted for the tannin process. It
certainly is most effective, the drymg pyroxyline removing every injurious impurity.
‘There is never any trouble from dirty plates.

The stock of plates for the night’s work, a dozen or so, being thus prepared, one
of them is taken, and by movement through the air is freed from fibres of cotton.
It is then coated with filtered collodion being held near the damp sink. ‘The coated
plate, when sufficiently dry, is immersed in a 40 grain nitrate of silver bath, acidified
with nitric acid until it reddens litmus paper. The exact amount of acid in the bath
makes in this “* Washed Plate Process” but little difference. When the iodide and
bromide of silver are thoroughly formed the plate is removed, drained for a moment,
and then held under the tap till all greasiness, as it is called, disappears. Both front
and back receive the current in turn.

It is then exposed, being carried on a little wooden stand, Fig. 43, covered with
filtering paper to the telescope, and deposited on the sliding plateholder which has
been set to the direction and rate of the moon, while the plate was in the bath.
‘The time of exposure is ascertained by counting the beats of a half-second pen-
lulam.

The method by which exposure without causing tremor is accomplished, is as
follows: A yellow glass slides through the eyepiece-holder, Fig. 33, just in front
of the sensitive plate, and is put in before the plate. ‘The yellow-colored moon is
centred on the collodion film, and the clepsydra and slide are set in motion, the

A SILVERED GLASS TELESCOPE. 49

mass of the telescope being at rest. A pasteboard screen is put in front of the
telescope, and the yellow glass taken out. After 20 seconds the imstrument. re-
maining still untouched and motionless, the screen is withdrawn, and
as many seconds allowed to elapse as desirable. ‘The screen is then Rig. 43.
replaced and the plate taken back to the photographic room.

After being again put under the tap to remove any dust or impurity, it
is dipped into the nitrate bath for afew seconds. ‘Two drachms of a solu-
tion of protosulphate of iron 20 grains, acetic acid 1 drachm, and water

1 ounce, is poured on it. As soon as the image is fairly visible this is
washed off, and the development continued if necessary with a weak plate Carrier.
solution of pyrogallic acid and citro-nitrate of silver—pyrogallic and

citric acids each ! grain, nitrate of silver ;!, grain, water 1 drachm. In order to
measure these small quantities standard solutions of the substances are made, so
that two drops of each contain the desired amount. ‘They are kept in

bottles, through the corks of which pipettes descend to just below the Fis: 4
level of the liquid. ‘This avoids all necessity of filtermg, and yet no
blemishes are produced by particles of floating matter.

During the earlier part of the development, when the protosulphate
of iron is on the film, an accurate judgment can be formed as to the pro-
per length of time for the exposure in the telescope. If the image
appears in 10 seconds, it will acquire an appropriate density for enlarge-

ment in 45 seconds, and will have the minimum of what is called pipette Bottle.
fogging and the smallest granulations. If it takes longer to make its
first appearance the exposure must be lengthened, and vice versa.

The latter part of the development, when re-development is practised, is purposely
made slow, so that the gradation of tones may be varied by changing the propor-
tion of the ingredients. As it would be tiresome and un-
cleanly to hold the plates in the hand, a simple stand is used
to keep them level. It consists of a piece of thin wood «,
Fig. 45, with an ordinary wood screw, as at 4, going through

each comer. Four wooden pegs, as at ¢, furnish a support

Developing Stand.

for the plate d. By the aid of this contrivance and the
washing system, I seldom get my fingers marked, and what is much more important,
rarely stain a picture.

When the degree of intensity most suitable for subsequent enlargement is reached,
that is, when the picture is like an overdone positive, the plate is again flooded with
water, treated with cyanide of potassium or hyposulphite of soda, once more
washed and set upon an angle on filtering paper to dry. It isnext morning labelled,
and put away unvarnished in the case.

To the remark that this process implies a great deal of extra trouble, it can only
be replied that more negatives can be taken on each night than can be kept, and that,
even were it not so, one good picture is worth more than any number of bad ones.

Although the above is the method at present adopted, and by which excellent
results have been obtained, it may at any moment give place to some other, and is
indeed being continually modified. The defects it presents are two—first, the time

7 July, 1864.

50 ON THE CONSTRUCTION AND USE OF

of exposure is too long, and second, there is a certain amount of lateral diffusion
in the thickness of the film, and in consequence a degree of sharpness inferior to
that of the image produced by the parabolic mirror, ‘The shortest time im which
the moon has been taken in this observatory has been one-third of a second, on the
twenty-first day, but on that occasion the sky was singularly clear, and the intrinsic
splendor of the light great. The full moon under the same circumstances would
have required a much shorter exposure, A person, however, who has put his eye
at the focus of such a silvered mirror will not be surprised at the shortness of the
time needed for impressing the bromo-iodide film ; the brillancy is so great that it
impairs vision, and for a long time the exposed eye fails to distinguish any moder-
ately illuminated object. The light from 188 square inches of an almost total
reflecting surface is condensed upon 2 square inches of sensitive plate.

Occasionally a condition of the sky, the reverse of that mentioned above, occurs.
The moon assumes a pale yellow color, and will continue to be of that non-actinic
tint for a month or six weeks. ‘This phenomenon is not confined to special localities,
but may extend over great tracts of country. In August, 1862, when our regiment
was encamped in Virginia, at Harper’s Ferry, the atmosphere was in this condition
there, and was also similarly affected at the observatory, more than 200 miles dis-
tant. As to the cause, it was not forest or prairie fires, for none of them of suffi-
cient magnitude and duration occurred, but was probably dust in a state of mimute
division. No continued rain fell for several weeks, and the clay of the Virginia roads
was turned into a fine powder for a depth of many inches. ‘The Upper Potomac
river was so low that it could be crossed dry-shod. On a subsequent occasion when
the same state of things occurred again, I exposed a series of plates (whose sensi-
tiveness was not less than usual, as was proved by a standard artificial flame) to the
image of the full moon in the 155 inch reflector for 20 seconds, and yet obtained
only a moderately intense picture. This was 40 times as long as common.

Upon all photographic pictures of celestial objects the influence of the atmosphere
is seen, being sometimes greater and sometimes less. To obtain the best impres-
sions, just as steady a night is necessary as for critical observations. If the image
of Jupiter is allowed to pass across a sensitive plate, a streak almost as wide as the
planet is left. It is easily seen not to be continuous, as it would have been were
there no atmospheric disturbances, but composed of a set of partially isolated images.
Besides this planet, I have also taken impressions of Venus, Mars, double stars, &c.

An attempt has been made to overcome lateral diffusion in the thickness of the
film by the use of dry collodion plates, more particularly those of Major Russel and
Dr. Hill Norris. ‘These present, it is true, a fine and very thin film during exposure,
but while developing are so changed by wetting in their mechanical condition that
no advantage has resulted. It was while trying them, that I ascertained the great
control that hot water exercises over the rapidity of development, and time of expo-
sure, owing partly no doubt to increase of permeability im the collodion film, but
also partly to the fact that chemical decompositions go on more rapidly at higher
temperatures. I have attempted in vain to develop a tannin plate when it and the
solutions used were at 32° F., and this though it had had a hundred times the exposure
to light that was demanded when the plate was kept at 140° F. by warm water.

A SILVERED GLASS TELESCOPE. 5]

Protochloride of palladium, which I introduced in 1859, is frequently employed
when it is desired to increase the intensity of a negative without altering its thick-
ness. This substance will augment the opacity 16 times, without any tendency to
injure the image or produce markings. It is only at present kept out of general
use by the scarcity of the metal.

§ 6. THE PHOTOGRAPHIC ENLARGER.

Two distinct arrangements are used for enlarging, a, for Low Powers varying
from 1 to 25; and 6, for High Powers from 50 to 700 diameters.

a. Low Powers.

The essential feature in this contrivance is an entire novelty in photographic
enlargement, and it is so superior to solar cameras, as they are called, that they are
never used in the observatory now. It consists in employing instead of an achro-
matic combination of lenses, a mirror of appropriate curvature to magnify the
original negatives or objects. ‘The advantages are easily enumerated, perfect coinci-
dence of visual and chemical foci, flat field, absolute sharpness of definition. If the
negative is a fine one, the enlarged proofs will be as good as possible.

Fig. 46.

The Photographic Enlarger.

The mirror is_of 9 inches aperture, and 115 inches focal length. It was polished
on my machine to an elliptical figure of 8 feet distance between the conjugate foci,
and was intended to magnify 7 times. At first the whole mirror was allowed to
officiate, the object being illuminated by diffused daylight. But it was soon ap-

52 ON THE CONSTRUCTION AND USE OF

parent, that although a minute object placed in one focus was perfectly reproduced
at the other, seven times as large, yet a large one was not equally well defined in
all its parts.

I determined then to produce the enlarged image by passing a solar-beam 15 inch
in diameter through the original lunar negative—placed in the focus nearest to the
mirror—and allowing it to fall on a portion of the concave mirror, 15 inch im
diameter, at one side of the vertex. Being reflected, it returns past the negative,
and goes to form the magnified image at the other focus of the ellipse.

In Fig. 46, a is the heliostat on a stone shelf outside; 6 a silvered glass mirror,
to direct the parallel rays through c, the negative; d is the elliptical mirror; e an
aperture to be partly closed by diaphragms; fa rackwork movement carried by the
tripod g; the curtain / /’ shuts out stray light from the interior of the observatory.
The aperture 7 is also diaphragmed, but is shown open to indicate the position of
the heliostat, the shelf of which joins the outside of the building at 7. The dotted
line points out the course of the light, which coming from the sun falls on the
heliostat mirror «, then on 4, through ¢ to d, and thence returning through e to the
sensitive plate in the plate holder &.

The distance of this last can be made to vary, being either two feet or twenty-
eight feet from d. In the latter case a magnifying power of about 25 results, the
moon being made three feet in diameter. ‘The sensitive plate is carried by a frame,
which screws to the side wall of the building, and can be easily changed in position.
The focussing is accomplished by the rack /, Where so small a part (15 inch) of
the surface of the mirror is used, a rigid adherence then to the true foci of this
ellipse is not demanded, the mirror seeming to perform equally well whether magni-
fying 7 or 25 times. ‘Theoretically it would seem to be limited to the former
power.

If instead of placing a lunar photograph, which in the nature of the case is never
absolutely sharp, at c, some natural object, as for instance a section of bone, is attached
to the frame moved by /, then under a power of 25 times it is as well defined as in
any microscope, while at the same time the amount of its surface seen at once is
much larger than in such instruments, and the field is flat. If the intention were,
however, to make microscopic photographs, a mirror of much shorter focal length
would be desirable, one approaching more to those of Amici’s microscopes.

By the aid of a concave mirror used thus obliquely, or excentrically, all the diffi-
culties in the way of enlarging disappear, and pictures of the greatest size can be
produced in perfection. I should long ago have made lunar photographs of more
than 3 feet im diameter, except for the difficulties of manipulating such large sur-
faces.

In order to secure a constant beam of sunlight a heliostat is placed outside the
observatory, at its southeast comer f, Fig. 38. This beam, which can be sent for
an entire day in the direction of the earth’s axis, is intercepted as shown at 6, Fig.
46, and thus if needed an exposure of many hours could be given. The interior of
the observatory and photographic room being only illuminated by faint yellow rays,
no camera box is required to cut off stray light. The eye is by these means kept
in a most sensitive condition, and the focussing can be effected with the critical

A SILVERED GLASS TELESCOPE. D3

.

accuracy that the optical arrangement allows, no correction for chromatic aberra-
tion being demanded.

I have made all the parts of this apparatus so that they can be easily separated
or changed. he flat mirrors are of silvered glass, and are used with the silvered
side toward the light, to avoid the double image produced when reflection from both
sides of a parallel plate of glass is permitted. ‘The large concave mirror happens
to be of speculum metal, but it can be repolished if necessary by means of a four
inch polisher, passed in succession over every chord of the face. A yellow film of
tarnish easily accumulates on metal specula if they are not carefully kept, and de-
creases their photographic power seriously.

Of the making of Reverses.— In addition to the use of the Enlarger for magnify-
ing, it is found to have important advantages in copying by contact. ‘The picture
of the image of the moon produced in the telescope is negative, that is, the lights
and shades are reversed. In enlarging such a negative reversal again takes place,
and a positive results. ‘This positive cannot, however, be used to make prints on
paper, because in that operation yeversing of light and shade once more occurs. — It
is necessary then at some stage to introduce still another reversal. This may be
accomplished either by printing from the original negative a positive, which may be
enlarged, or else printing from the enlarged positive a negative to make the paper
proofs from. In either case a collodion film, properly sensitized, is placed behind
the positive or negative, and the two exposed to light.

If diffused light or lamplight is used, the two plates must be as closely in contact
as possible, or the sharpness of the resulting proof is greatly less than the original.
This is because the light finds its way through in many various directions. If the
two plates, however, are placed in the cone of sunlight coming from the Enlarger.
and at a distance of fifteen or twenty feet from it, the light passes in straight lines
and only in one direction through the front picture to the sensitive plate behind.
I have not been able to see under these circumstances any perceptible diminution

; of an inch apart. It is perfectly feasible

in sharpness, though the plates had been 5!
to use wet collodion instead of dry plates, no risk of scratching by contact is incurred,
and the whole operation is easily and quickly performed. ‘The time of exposure, 5
seconds, is of convenient length, but may be increased by putting a less reflecting
surface or an unsilvered glass mirror in the heliostat. A diaphragm with an aper-
ture of half an inch if placed at e, Fig. 46, to shut out needless light, and avoid
injuring the sharpness of the reverse by diffusion through the room. In enlarging
other diaphragms are also for the same reason put in the place of this one. For a
half moon for instance, a yellow paper with a half circular aperture, whose size may
be found by trial in a few minutes, is pinned against e.

The enlarged pictures obtained by this apparatus are much better than can be
obtained by any other method known at present. The effect, for instance, of a
portrait, made life-size, is very striking. Some astronomers have supposed that
advantages would arise from taking original lunar negatives of larger size in the
telescope, that is, from enlarging the image two or three times by a suitable eye
piece or concave achromatic, before it reached the sensitive plate. But apart from
the fact that a reflector would then have all the disadvantages of an achromatic,

54 ON THE CONSTRUCTION AND USE OF

the atmospheric difficulties, which in reality constitute the great obstacle to success,
would not be diminished by such means. ‘The apparent advantage, that of not
magnifying defects in the collodion, is not of much moment, for when development
of the photographs is properly conducted, and thorough cleanliness practised,
imperfections are not produced, and the size of the silver granules is not objection-
able,

b. High Powers.

Although negatives of astronomical objects have not as yet been made which
could stand the high powers of the arrangement about to be described, yet they
bear the lower powers well, and give promise of improvement in the future.

Photography of microscopic objects as usually described, consists in passing a
beam of light through the transparent object into the compound body of the micro-
scope, and receiving it on its exit from the eyepiece upon a ground glass or sensi-
tive plate. ‘The difficulty which besets the mstrument generally, and interferes
with the production of fine results, arises from the uncertainty of ascertaining the
focus or place for the sensitive plate. For if the collodion film be put where the
image on ground glass seems best defined, the resulting photograph will not be
sharp, because the actinic rays do not form their image there, but either farther
from or nearer to the lenses, depending on the amount of the chromatic correction
given by the optician, Practically by repeated trials and variation of the place of
the sensitive compound, an approximation to the focus of the rays of maximum
photographic intensity is reached.

Fig. 47.

SSS EEE

Microscope for Photography.

During my father’s experiments on light, and more particularly when engaged
in the invention of portrait photography, he found that the ammonio-sulphate of
copper, a deep biue liquid, will separate the more refrangible rays of light, the rays

A SILVERED GLASS TELESCOPE. Dd
concerned in photography, from the rest. If a beam of sunlight be passed through
such a solution, inclosed between parallel plates of glass, and then condensed upon
an object on the stage of a microscope, a blue colored image will be formed on the
ground glass, above the eyepiece. If the place of best definition be carefully ascer-
tained, and a sensitive plate put in the stead of the ground glass, a sharp photograph
will aiways result.

Besides, there is no danger of burning up the object, as there would be if the
unabsorbed sunlight were condensed on it, and hence a much larger beam of heht
and much higher powers can be used. The best results are attained when an image
of the sun produced by a short focussed lens is made to fall upon and coincide with
the transparent object. In 1856 we obtained photographs of frog?’s blood disks,
navicula angulata, and several other similar objects under a power of 700 diameters,
excellently defined. Since then several hundreds of microscopic pictures have been
taken.

In the figure, @ is the heliostat, / a lens of three inches aperture, ¢ the glass cell
for the ammonio-sulphate of copper, d the object on the stage of the microscope e,
f the camera for the ground glass or sensitive plate. Above the figure the course
of the rays is shown by dotted lines.

In concluding this account of a Silvered Glass Telescope T may answer an inquiry
which doubtless will be made by many of my readers, whether this kind of reflector
can ever rival in size and efficiency such great metallic specula as those of Sir
William Herschel, the Earl of Rosse, and Mr. Lassell? My experience in the
matter, strengthened by the recent successful attempt of M. Foucault to figure such
a surface more than thirty inches in diameter, assures me that not only can the four
and six feet telescopes of those astronomers be equalled, but even excelled. It is
merely an affair of expense and patience. I hope that the minute details I have
given in this paper may lead some one to make the effort.

Hasrinas, WESTCHESTER COUNTY,
New York, 1863.

Postscript.—Since writing the above T have completed a photograph of the moon
50 inches in diameter. The original negative from which it has been made, bears
this magnifying well, and the picture has a very imposing effect.

PUBLISHED BY THE SMITHSONIAN INSTITUTION,
WASHINGTON CITY,

suty, 1864.

PP (Dy ae
ih st ith
Du Hi
+n ah nn i
“10 ea
<a ein ca

ye Ae

‘Se

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE

PART OF VOLUME XXXIV

ONS DAE MODERN REFEECTING
PearEScCOPE, AND THE MAKING AND
PeShNC Or OPTICAL MIRRORS

BY

GEORGE W. RITCHEY

ASSISTANT PROFESSOR OF PRACTICAL ASTRONOMY, AND SUPERINTENDENT OF INSTRUMENT
CONSTRUCTION, IN YERKES OBSERVATORY

(No. 1459)

CITY OF WASHINGTON
PUBLISHED BY THE SMITHSONIAN INSTITUTION
1904

CONTENTS.

Besta IMG LO LS Peet Yee ee cca eS sia aoa yee, Nie. io bo eto One ee am
CHAPTER I.
PeeOl Gilaiss 1Or Optical, MITrOrs oc ier.0 ss Aes ac ws Wis vee aclee com -€e0odae ea sake eon 1
CHAPTER II.
The Optical Laboratory of the Yerkes Observatory..............0cccecececeecsececocs 2
CHAPTER III.
Sarmidin candor olishin ee MAGHINeS |. oie nyajneecc «asses 6a \ara ayers <i ars secrete oe a ee 3
CHAPTER IV
BET POLS re eget ne Yeo Ye Solo ooo Fait x FinjeBlq8 eel ate apoE ON Ee 6
CHAPTER V.
Re Oli shiny eal Ol s pysrse Metre ars fey sital cree 3) Sol sia oia seine alesis) eves a cree ape a ae ee 7
CHAPTER VI.
Rough-grinding the Face and Back of a Rough Disk of Glass, and Making the Same
ara © likey cay Peres stays ste yet sok hcsaie euceel Si \e).s atoiS7el »os)o)c'ss 5 love “cr fe ays erstrenotele toe aioe teste TC ere 10
CHAPTER VII.
Grinding dee ot Glass:— Rounding of Comers: 22... 2. scsedese ese scee se ecen ue le
CHAPTER VIII.
Hine-grinding and Polishing the Back of the Mirror ................scccecees cece vece 13
CHAPTER IX.
rc om N eMC ONCAVE MSU LAC Catpatete/ctercfelaialeterelviersrole(atareietanoeroleickeletekslelaieietel elie ieters pete eteraieis 14
CHAPTER X.
idea ites Lair opmre potato ey stale et-tavstevelc) ace! stel oiel(c{«/ela/a(o'e’s:elsliaielnfole,aishatn elelefstelofeieieleleveletstelchaietelelerelerenceres 17
CHAPTER XI.
Westinoganduhouringsy spherical Marnors¢s:, 0. -1s)ecle a cian «nivel see idferelole sheta sierelei pa ciate Saver Nore 21

1V CONTENTS.

PAGE
CHAPTER XII.
Grinding, Figuring, and Testing Plane Mirrors...............2+6+ .aness.chsecouererese wieeierees 25
CHAPTER XIII.
Testing and Figuring: Paraboloidal Mirrors 7222 2. .n.-otcircis selon stele aetotrstorelenersTeretstererener= 30
CHAPTER XIV.
Testing and Figuring Convex Hyperboloidal Mirrors ......... a iets sie reittoneletsicla Buclspero moka 38
CHAPTER XV.
SUlvie rin Ouetemereie clerks oss. sie eie wiaiilalels le winilete veyahevoreve enstetellelorel eet aim tevelevereRetlietstersiekeruer tenors 40
CHAPTER XVI.
A Support-system for Large Mirrors..............2. Societe SystePeleteyeteeial <i tterayaeterans 43
CHAPTER XVII.
A Mounting for a Large Reflecting: Telescope:s.. 22 .- «.sscceeneeaaeieee rr eee 46

PLATE I.

PLATE II.

PLATE ITI.

PLATE IV.

PLATE V.
PLATE VI.
PLATE VII.

FIGURE 1.
FIGURE 2.
FIGURE 3.
FIGURE 4.
FIGURE 5.
FIGURE 6.
FIGURE 7.
FIGURE 8.
FIGURE 9.
FIGURE 10.
F1GURE 11.
FIGURE 12.
FIGURE 13.
FIGURE 14.
FIGURE 15.
Puate VIII.
PLATE IX.
PLATE X.
PLATE XI.
PLATE XII.
PLATE XIII.

LIST OF ILLUSTRATIONS.

CentralbPart/ofithe! Great, Nebulavin Andromedas...sec. 0. sscieseecsccleceeiiesieerreeneinicecics

Five-foot Mirror and Grinding Machine, Showing Method of Tipping Glass on Edge for
MOS Gim Orryepayass crete <i. etever = svaVvote\aFutefa,eialaiels,s 5, Sioretalayels;s sieletet i cvavetenctotarswevarera: evel cls miiatolalets) stcrsmetanete!

Five-foot Mirror and Grinding Machine, Showing Lever for Handling Heavy Grinding and
ROMSHIN Gs LOOMS hao oaters ciclencscisieinkelats stolen cise sieve shore sisiercrasistecslebsieteoteteaiecete tare erie sre

Five-foot Mirror and Grinding Machine, Showing Half-size Grinding Tool Suspended on
Menara PAs nae ayetotetsiorerelessrster=ie¥=ietelelcieisje/sloletetoheselosevalaroYererelsieze staverel tale/clesaketevetelefersralatatstatey sstetetere (eter

Large Spherometer on Twelve-inch Glass Grinding Tool............-.....-ceeeeee es eeeee
Five-foot Mirror and Grinding Machine, Showing Method of Grinding Edge of Glass......

Five-foot Mirror, with Front and Back Polished Approximately Flat. Looking through the

(GAGS Byeteraye reds stereo ole oie einai shal clovelo,csaLelorevevors o/e)s,exeicekere\ ave! eYafareielete cfetwiche sYa¥elatatetst iat -an-Tetst taper T sia
Arrangement by which Artificial Star is Used very close to Optical Axis...............+45
Diagram Illustrating Testing of a Plane Mirror............0ceeccer ee see nce sseecerecetcns
Diagram Illustrating Testing of a Plane Mirror...........cseeesceececeesccceccecescerens
NormalWeolishin oe Oolaieyeyaterstctoveiere eietekerelerctelsiotelere/slalcleialstere ota sereltiofetetstetetalelelnveletel=fereleietat= fora nereie
Concaving Polishing Tool for Figuring Plane Mirror...........:..eee sees eee tee eee cece
Convexing Polishing Tool for Figuring Plane Mirror. ........+.....:seeeee cece eee eeeees
Diaphragm Used in Testing a Paraboloidal Mirror at its Center of Curvature............-.
Testing a Paraboloidal Mirror at its Focus. ........ 6.0. eee cee eee eee ee te eee eee teen eee
Full-size Polishing Tool for Parabolizing...........0.se scence ee ete ete eee e renee eens
Diagram Illustrating Testing of Hyperboloidal Mirror.....-...++...seseeee sere renee ees
FIGURES 16 and 17. Support System for Large Mirror........... .++--+++- Face Plate IX.
FIGURES 18 to 21. Support System for Large Mirror, ...-....--..+++++55 Face Plate VIII.
Large Coelostat with 30-inch Plane Mirror ....-...-++++++eeeesee secre ees Oe SS) eee
Design for Mounting of Five-foot Reflecting Telescope... ..-.+..+++0++0e05 Face Plate XIT,

FIGURES 22 to 27. Design for Mounting of Five-foot Reflecting Telescope. Face Plate XT.

Large Double-slide Plate-carrier Attached to 40-inch Refractor : Yerkes Observatory.....

4

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE — RITCHEY PLATE |

CENTRAL PART OF GREAT NEBULA IN ANDROMEDA. PHOTOGRAPHED BY G. W. RITCHEY,
YERKES OBSERVATORY. TWO-FOOT REFLECTOR.

Enlarged 5.26 times
from original negative

ON ThE MODERN REFLECTING? TRE:
SCOPE, AND THE MAKING. AND TESTF-
PNG Or OPPICAT MIRRORS:

By G. W. Rrrcney.

INTRODUCTION.

Tue present paper describes the methods employed by the writer in the optical
laboratory of the Yerkes Observatory in making and testing spherical, plane, para-
boloidal, and (convex) hyperboloidal mirrors. On account of the very great import-
ance of supporting mirrors properly in their cells when in use in the telescope, a
chapter is devoted to the description of an efficient support system for large mirrors.
Intimately related to this, and equally important, is the subject of the mounting—the
mechanical parts—of a modern reflecting telescope; accordingly, the final chapter
is devoted to a consideration of this subject.

CHAPTER I.
DISKS OF GLASS FOR OPTICAL MIRRORS.

No greater mistake could be made than to assume that cheap and poorly
annealed disks of glass, or those with large striz or pouring marks, are good enough
for mirrors of reflecting telescopes. While I am not prepared to say that optical
glass of the finest quality must be used for mirrors in order to secure the best
attainable results, it is evident that a very high degree of homogeneity and freedom
from strain is necessary in order that the figure of mirrors shall not be injuriously
affected by changes of temperature. If it were not necessary to consider the ques-
tion of cost, I should advise the use of the finest optical (crown) glass always, in
order to be as free as possible from risk; usually considerations of cost would, in
the case of large mirrors, make it necessary to choose between such an optical disk
of a given size and a somewhat larger one of the kind furnished by the St. Gobain

I

2 THE MODERN REFLECTING TELESCOPE.

Company, for example. The diagonal plane mirror of a Newtonian, and the convex
mirror of a Cassegrain reflector, should always be made of the best optical glass,
since the expense for these is comparatively slight.

The writer has used many disks made at the celebrated glass-works of St.
Gobain, near Paris, of sizes from 8 inches in diameter and 1} inches thick, to
the great one shown in the plates accompanying this article, which is 5 feet in
diameter and 8 inches thick, and which weighs a ton. All of these disks are
beautifully free from bubbles and large striz, and are fairly well annealed, con-
sidering their great thickness. It is a most encouraging fact that the quality of
the 5-foot disk is not inferior in any respect to that of disks of 8, 12, 20, 24, and
30 inches diameter which I have used. The makers of the 5-foot disk have re-
cently expressed their readiness to undertake for us a 10-foot disk, one foot thick,
which they think could now be made as perfect in all respects as the 5-foot disk.
In ordering these disks it is always specified that great care be given to thorough
stirring and thorough annealing. I have no doubt that in the case of very large
and thick disks the makers could be prevailed upon to give even greater care to
these points than is now given.

A very important point is in regard to the best thickness of optical mirrors.
As a result of experience in making and using many mirrors of 24 and 30 inches
diameter, in which the thickness of the several disks varies from one-twelfth to
one-sixth of the diameter, I have no doubt that the thicker disks are always prefer-
able, provided that they are as homogeneous and well-annealed as the thinner ones.
The thinner mirrors suffer much greater temporary change of curvature from the
very slight heat generated during the process of polishing; and they are undoubt-
edly more liable to suffer temporary disturbance of figure from changes of tem-
perature when in use in the telescope. In the cases of the large paraboloidal mirror
of a reflecting telescope, and the large plane mirror of a coelostat or heliostat, which
should always be supported at the back to prevent flexure, the thickness should not
be less than one-eighth or one-seventh of the diameter; in the writer’s opinion the
latter ratio leaves nothing to be desired. In the cases of the small diagonal plane
mirror and the small convex mirror, which cannot easily be supported at the back,
the thickness should be not less than one-sixth of the diameter.

All mirrors should be polished (not figured) and silvered on the back as well
as on the face, in order that both sides shall be similarly affected by temperature
changes when in use in the telescope; for the same reason the method of supporting
the large mirror at the back, in its cell, should be such that the back is as fully
exposed to the air as possible.

CHAPTER II.
THE OPTICAL LABORATORY OF THE YERKES’ OBSERVATORY.

A LarGE, well-lighted room, 70 feet long by 20 feet wide, in the north
basement of the Observatory, was designed for the optical laboratory. The floor,

XL YO4 JOGA NO ssv1y9 ONIddIL JO GOHLAW DONIMOHS
“ANIHOVW ONIGNIND ANV YOUMIN LOOA-AATA

I] 3LW1d

ASHOLIY — 3D037MONY OL 8NOILNAINLNOD NVINOSHLINS

THE MODERN REFLECTING TELESCOPE. 3

which is nearly on a level with the ground outside, is of cement and is heavily
painted. The walls are of brick, are about two feet thick, and are covered with
two layers of heavy ceiling paper arranged so as to give two tight one-inch air-
spaces for constant-temperature purposes. All joints of the paper are lapped and
are nailed down with strips of wood. The ceiling of the room is heavily varnished.

The large room is divided into three rooms connected by large doors; these
doors are so arranged that the entire length of the large room and of a wide hall
opening from it, making an apartment 165 feet long, can be utilized for testing.
The east and middle rooms of the three are used for grinding and polishing. The
large windows of these rooms are fitted with storm sash on the inside ; these are built
in permanently and are made air-tight by means of ceiling paper. The west room
contains the motor which supplies power to the grinding and polishing machines in
the inner rooms ; power is transmitted by a long shaft which runs the entire length
of the rooms; this shaft is built in air-tight (to prevent dust) beneath the long
work-bench which runs along one side of the rooms.

With these arrangements temperature, moisture, and freedom from dust can be
controlled in the grinding and polishing rooms with all necessary refinement. In other
respects, however, three great improvements could be made in planning an ideal op-
tical shop; two of these relate to the comfort and health of the optician. First, the
rooms should be arranged so that direct sunlight could be admitted to them during
all parts of the optical work in which this would not be injurious to the work itself.
Second, provision should be made for supplying to the rooms an abundance of
fresh air, of a definite temperature, and washed free from dust. Third, for constant-
temperature purposes, walls and partitions covered with a heavy layer of asbestos
plaster (commercially termed Asbestic) would be preferable, on account of the
superiority of the insulating and fire-proofing qualities of this material, to those of
ceiling paper with air-spaces.

CHAPTER III.
GRINDING AND POLISHING MACHINES.

Tue grinding and polishing machines used by the writer are somewhat similar
in principle to Dr. Draper’s machine, shown in Fig. 25 of his book, but are more
elaborate. I shall describe here the machine used in making the 5-foot mirror,
both because it embodies most of the essential features of a grinding and polishing
machine, and also because it is the only one of my machines of which I have a
series of photographs for illustration. A good idea of this machine may be gained
from the views of it shown in Plates u, rm, rv, and vt.

The massive turntable upon which the glass rests consists of a vertical shaft
or axis five inches in diameter, carrying at its upper end a very heavy triangular
casting, upon which, in turn, is supported the circular plate upon which the glass
lies. This plate is of cast-iron, weighs 1,800 pounds, is 61 inches in diameter,

4 THE MODERN REFLECTING TELESCOPE.

is heavily ribbed on its lower surface, and is connected to its supporting triangle
by means of three large leveling screws. The surface of the large plate was
turned and then ground approximately flat; two thicknesses of Brussels carpet are
laid upon this, and the glass, with its lower surface previously ground flat, rests
upon the innumerable springs formed by the looped threads of the carpet. No
better support for a glass during grinding and polishing could be desired.

Three adjustable iron arcs at the edge of the glass serve for centering the
latter upon the turntable, and prevent it from slipping laterally.

The entire turntable, with the heavy frame of wood and metal which supports
it, can be turned through 90° about a horizontal axis, thus enabling the optician.
to turn the glass quickly from the horizontal position which it occupies during
grinding and polishing, to a vertical position for testing. This is shown in Plate m.

The turntable is slowly rotated on its vertical axis by means of the large pul-
ley below (Plate m1). This rotation is effected by means of belting from the main
vertical crank-shaft on the east end of the machine; this shaft is well shown at the
left in Plate rv. At the upper end of this shaft is the large crank, with adjustable
throw or stroke, which moves the large and strong main arm to which the grinding
and polishing tools are connected, and by means of which they are moved about upon
the glass. This I shall always refer to as the main arm. It is a square tube of
oak wood, and is strong enough to carry the counterpoising lever shown in Plate rv,
and the weight of any of the grinding tools, when fully or partially counterpoised.
This main arm also carries the system of pulleys and belts by which the slow rota-
tion of the grinding and polishing tools is rigorously controlled; these, and the
manner in which this rotation is effected, are well shown in Plate rv.

The west end of the main arm consists of a strong steel shaft which slides in
a massive bronze swivel-bearing which corresponds to the “elliptical hole in the
oak block p” of Dr. Draper’s machine (see his Fig. 25). But this bearing is not
stationary as in Draper’s machine; it is not only mounted on a long slide (which I
shall refer to throughout this article as the transverse slide), so that it can be
slowly moved for several feet across the west end of the machine by means of a
long screw, but this bearing and slide are carried upon a secondary strong arm,
which is moved by a secondary crank at the southwest corner of the machine.
Unfortunately there is no photograph which shows this part of the machine as it
appears when in use; Plates m and mt show the secondary crank well, but the secon-
dary arm is shown swung around with one end resting on a bracket on the wall, in
order to have it out of the way.

The arrangement of the west end of the machine is the result of experience
with several machines, and is found extremely serviceable and convenient. The
long transverse slide on the secondary arm allows the grinding and polishing tools
to be placed so as to act on any desired zone of the glass, from the center to the
edge; and this setting can be changed as desired while the machine is running.
The secondary crank, which turns at the same speed as the large one which drives
the main arm, enables the optician to change as desired the width of the (approxi-
mately) elliptical stroke or path of the tool with reference to the length of this

‘STOOL ONIHSITOd GNV ONIGNIYS AAVAH ONITANVH 2
“AaNIHOVW ONIGNIYD AGNV YOUUNIN LOO:

JT UAAAT ONIMOHS
AAT

I] 34v1d AB3HOLIY — 3D031TMON} OL SNOILNEIYLNOD NVINOSHLINS

THE MODERN REFLECTING TELESCOPE. 5

stroke; this change is especially desirable when figuring the glass; it is, of course,
impossible when only one driving-crank is used.

I regard the transverse slide, or something equivalent to it, as absolutely neces-
sary to the success of a grinding and polishing machine; it will be noticed that its
purpose corresponds, in some measure, to that of the long slot in the main arm of
Draper’s machine; I have used both arrangements and have found the transverse
slide to be far more effective and convenient in use; its use will be described in
the chapters on grinding and polishing.

The secondary crank, while very desirable and convenient, for the reason given
above, is not indispensable ; I have used several smaller machines which have given
good results without it.

The manner in which the grinding and polishing tools are connected to the
main arm is shown in Plate rv. A vertical shaft, 12 inches in diameter and 24
inches long, both rotates and slides (vertically) freely in bronze bearings attached to
the main arm. The grinding and polishing tools are connected to the lower end of
this shaft through the medium of a large universal coupling,—a gimball or Hooke’s
joint,—with two pairs of horizontal pivots at right angles to each other; this allows
the tools to rock freely in all directions in order to follow the curvature of the glass.
The tools are lifted, for counterpoising them, by the lever above (see Plate rv),
through the medium of the vertical shaft and the universal coupling. In the case
of very massive grinding tools of moderate size, like that shown in this illustration,
the universal coupling is connected directly to the back of the tool; but in the
case of all large tools which are to be used for fine work this connection is made
through the medium of a system of bars and triangles, so that the tools are counter-

poised without the slightest danger of changing their shape. A small coupling with
ball bearings at the upper end of the vertical shaft allows the latter to rotate freely
with reference to the link which connects it to the counterpoise lever.

To recapitulate briefly: this method of connecting the grinding and polishing
tools allows them to be controlled in all of the following ways simultaneously :
(1) the stroke of the tool is given by the motion of the main arm; ‘(2) the slow
rotation of the tool is rigorously controlled by the belting above; (8) the tool is
allowed to rock or tip freely by means of the universal coupling, in order that it
may follow the curvature of the glass; (4) the tool rises and falls freely by means
of the sliding of the 13-inch vertical shaft in its bearings, in order that it may fol-
low the curvature of the glass; (5) the tool is counterpoised by means of the lever
on the main arm, through the medium of the same vertical shaft and universal
coupling.

In Plate mt is shown the large lever by which the 5-foot glass, which weighs
a ton, is lifted on and off the machine, and by means of which, also, the large grind-
ing tools are handled. One of the full-size grinding tools, weighing 1,000 pounds,
is shown suspended by the lever. The arrangements are so convenient that the
optician alone can do all parts of the work.

o>

THE MODERN REFLECTING TELESCOPE.
CHAPTER IV.

GRINDING TOOLS.

Wuite grinding tools of glass were used in much of my earlier work, and are
still used for small work, I now use cast-iron grinding tools for all large work.
These are cast very heavy, with ribs on the back; the ribs are made heavy, but not
deep (or high). For large work iron tools are cheaper than glass ones; they are
more easily prepared; they are more easily and safely counterpoised, which is
always necessary in the fine-grinding of large work; and they produce on the glass
a fine-ground surface fully as smooth and perfect as can be obtained with glass tools.

An important question is in regard to the size of grinding tools,—whether they
should be of the same diameter as the mirror. For mirrors up to 24 or 30 inches
in diameter full-size tools are generally used. For concave mirrors larger than 30
inches in diameter I use grinding tools whose diameter is slightly more than half
that of the glass, 7. ¢, a 16-inch tool for a 30-inch glass; a 32-inch tool for a 60-inch
glass. These I shall refer to as half-size tools. Full-size tools are, of course, much
more expensive and difficult to make; they are many times heavier than half-size
tools of equal stiffness; and they require a much stronger grinding machine to
counterpoise them properly; grinding can be done with them, however, more
quickly than with the smaller tools. Half-size tools are economical and are quickly
prepared ; they are easily counterpoised ; and a much greater variety of stroke can
be used with them, so that with a well-designed grinding machine I have found it
easier to produce fine-ground surfaces, entirely free from zones, with half-size than
with full-size tools. If temperature conditions and uniform rotation of the glass
are carefully attended to, the surface of revolution produced by the smaller tools is
fully as perfect as that given by the larger ones; I always take the precaution,
however, to work a full-size approximately flat tool on the glass before beginning
to excavate the concave, so as to start out with a surface of revolution.

Grinding tools for concave and convex mirrors are always made in pairs, one
concave, the other convex. Grinding tools for plane mirrors are made in triplicate.
These iron tools, when being cast, are “poured” face down, so that the faces will
be clean. I shall describe the preparation of a pair of iron tools for a concave
mirror, leaving the description of tools for plane mirrors until the making of plane
mirrors is discussed. The convex and concave tools are turned in a lathe to the
proper curvature as shown by templets. The convex tool, which is, of course, to
be used on the concave glass, is now placed on a planing machine, and has a series
of grooves cut across the convex surface. These grooves are usually + inch wide,
and run in two directions at right angles to each other; these divide the surface into
squares, which are usually made about one inch on a side. These grooves serve to
distribute the grinding material uniformly, and entirely obviate the tendency of the
tools to cling to the glass in fine-grinding. No grooves are cut in the concave
tool. A number of holes are now bored through both tools, in such positions that
wooden cups or funnels can be inserted into the holes from the back or ribbed side

‘NUV NIVW NO GACNAdSAS IOOL INIGNINOD AZISATVH DNIMOHS

T-HATA

Al 3Lv1d AFHOLIY — 3DG31MONY OL SNOILNSINLNOD NVINOSHLINS

- -
‘ <3 ie
4

opie Mee ae
os Peper

THE MODERN REFLECTING TELESCOPE. 7

of the tool, without interfering with the ribs; these cups serve for the introduction
of the grinding material during the process of grinding; they should be thoroughly
varnished.

The convex and concave tools are now ground together on the machine, with
fine grades of carborundum (which is much more effective for this purpose than
emery) and water. This eliminates the circular marks left by the lathe, and
enables the optician to secure the exact curvature desired. A very important point
is that by grinding with the concave tool on top, the radii of curvature of both
tools can be gradually shortened ; when the convex tool is used on top the curva-
ture of both is gradually flattened. By this means, and the use of very fine grades
of carborundum, a most perfect control of the curvature of the tools may be had.

The curvature of the tools and of the glass is measured by means of a large
spherometer ; this is shown in Plate v, resting upon a 12-inch glass grinding tool.
The spherometer is of the usual three-leg form; the legs terminate in knife-edges,
the lines of which are parts of the circumference of a 10-inch circle. The central
screw is very carefully made; it was ground in its long nut (which was made adjust-
able for tightness) with very fine grades of emery such as are used in optical work ;
screw and nut were then smoothed and polished by working them together with
rouge and oil. The screw is of 4 millimeter pitch, and the head, which is 4 inches
in diameter, is graduated to 400 divisions. On fine-ground surfaces settings can
be made to one-half or one-third of a division, corresponding to a depth of =>
sven of an inch, approximately.

1 5
d00 Ol

CHAPTER V.
POLISHING TOOLS.

AFTER experience with polishing tools of various kinds, the tools which I now
use exclusively for large work consist of a wooden disk or basis constructed in a
peculiar manner, and covered on one side with squares of rosin faced with a thin
layer of beeswax. The wooden disk may be replaced, in the case of small polish-
ing tools up to 12 or 15 inches diameter, by a ribbed cast-iron plate so designed as
to be extremely light and rigid; the bases of larger tools may be made of east
aluminum, but this, in order to be strong and rigid, must contain 15 % or more of
other metals; such a basis for a 80-inch polishing tool weighs about sixty pounds,
and the rough casting alone costs about fifty dollars. It is possible that a metal
basis possesses an advantage over a wooden one in that its surface is less yielding.
Tools properly constructed of wood, however, are light and extremely rigid, are
easily made, and are economical in cost. As their proper construction is a matter
of the utmost iniportance, I shall describe, somewhat in detail, the method of
making wooden bases of from 15 to 40 inches diameter.

A large number of strips of dry and straight-grained pine wood 1} inches wide
and 55, inch thick are prepared; the wooden basis is built up of successive layers

8 THE MODERN REFLECTING TELESCOPE.

of these strips. The strips in all layers except the two outer ones are laid just 4
of an inch apart. Those of each layer are placed at right angles to those of the
next layer below, and are glued and nailed down with long wire brads. The best
cabinet-maker’s glue is used, and the strips are warmed before the glue is applied.
Each crossing of the strips in the successive layers (7, é., each of the 14-inch squares),
is nailed with at least two nails. The upper surface of each layer is carefully
planed flat before the next layer of strips is apphed. For a 20-inch tool six layers
of pine strips (each 55, inch thick) are used; for a 24-inch tool, seven layers; for :
36-inch tool, ten layers. Next, one layer of thoroughly seasoned strips of hard
straight-grained cherry wood about 2 inch thick and slightly less than 15 inches
wide is added, to form the outer layer at the back of the tool; these strips are laid
almost touching each other. In the case of tools for flat mirrors, a precisely similar
layer of cherry strips is added to form the outer layer at the front or face of the
tool. But in the case of tools for concave or convex mirrors the strips composing
the front layer must be made thicker, to allow for the curvature of the face of the
tool. If this curvature is great, the cherry strips forming the front layer are made
of double width (¢. ¢., slightly less than 3 inches wide), in order that the width of
their bases shall be greater as compared with their thickness; this is usually done
when the depth of the curve is greater than } inch. The gluing and nailing of
the outer layers of strips are done with the greatest thoroughness, four of the long
fine nails being driven through into each of the squares of pine wood beneath.
For tools less than 20 inches in diameter thinner strips and a larger number of
layers are used. The entire thickness of the wooden disk or basis built up in this
way should be between one-tenth and one-eighth of its diameter.

This wooden basis is next placed in a large lathe, the edge is turned smooth
and to the proper diameter, and the face is turned to fit the curvature of the glass
to be polished.

A round pan of galvanized iron large enough to contain the wooden disk hav-
ing been prepared, enough hard paraffin is melted in it so that the disk can be
soaked in the liquid paraffin; the latter must not be hotter than 150° Fahrenheit,
otherwise the strength of the glue-joints will be injured. It is best to melt the
paraffin on a gas or gasoline stove, so that the degree of heat can be easily controlled.
The tool should soak for several hours, being moved continually and turned over
often. Since the construction of the wooden basis is such that a great number of
openings extend entirely through it, the melted paraffin permeates the entire tool
thoroughly. The wooden tool prepared in this way is lighter than any metal tool
of the same degree of stiffness, and is entirely impervious to the moisture which is
necessary in the polishing room. ‘The question of lightness is a most important
one, as will be seen when the work of polishing is described later.

The front or face of the wooden basis is now lightly scraped with a wide,
sharp chisel, to remove any excess of paraffin, and is then marked off for 14-inch
squares of rosin, with grooves } inch wide between them; the grooves should come
exactly above the 41-inch spaces left between the pine strips beneath.

The preparation of the rosin squares is usually a very troublesome matter, but

“TOOL DONIGNIYD SSVIDO HONIUATAML NO YALANOUAHe

ADUVT

A 341d ABHOLIY — JOGSIMONH OL SNOILNEIYNLNOD NVINOSHLINS

THE MODERN REFLECTING TELESCOPE. 9

becomes easy when the following directions are observed. A clean, flat board,
having an area about twice that of the polishing tool, is prepared. One face of
this is covered with clean paper. Long strips of wood } inch square are fastened
upon the paper by means of fine brads; these strips are placed just 14 inches apart,
and the ends of the grooves thus formed (grooves 1} inches wide, } inch deep,
and of any convenient length) are closed with strips of wood. The board is now
carefully leveled. The rosin, when melted and softened to the proper degree, is to
be poured into these grooves, which serve as moulds.

A quantity of rosin sufficient to fill all of the grooves is melted in a clean pan.
Even when only a small quantity is needed it is best to melt at least ten pounds
of rosin, since the entire process of “tempering” and pouring is more easily and sat-
isfactorily carried on with large quantities than with small. Only lumps of clear,
clean rosin should be used. A gas or gasoline stove is very convenient for melting
the rosin, since the degree of heat can be easily controlled. When the rosin is
melted the pan is removed from the stove and a quantity of turpentine, equal in
weight to about 5. of the rosin used, is added, and the mass thoroughly stirred.
A tablespoon tul of the liquid is now dipped out and immersed for several minutes
in a bucket of water at the temperature of the polishing room, which should be
about 68° Fahrenheit. The spoonful of rosin is now taken out, and its hardness tried
with the thumb-nail. If the rosin is brittle at the thin edges it is still too hard, and
a little more turpentine must be added; if, however, it is soft like wax or gum, it is
too soft, in which case the pan of rosin must be hardened by boiling for a few min-
utes ; this drives off the excess of turpentine. When the rosin is of the proper hard-
ness an indentation about } inch long can be made in it by moderate pressure of the
thumb-nail for five seconds. When the proper degree of hardness has been ob-
tained it is often necessary to heat the pan of rosin again so that it will not be too
thick to pass readily through the strainer; this is a long, narrow bag of cheese-
eloth through which the rosin is strained as it is being poured into the grooves or
moulds previously described. If such heating is necessary it must be done gently
and without boiling; otherwise the rosin will be hardened. Enough is poured
into each groove to just fill it.

After being poured, the rosin should cool for six or eight hours. Then the
nails which held the quarter-inch strips of wood to the board below are removed,
and the layer of rosin, wooden strips, and paper is carefully lifted from the board,
when the paper is easily stripped from the rosin, to which it does not adhere
closely. With care the thin strips of wood can now be removed, one after the
other, and the long strips of rosin, 1} inches wide and 4 inch thick, are secured
without chipping or breaking. These are now readily eut into squares with a
hot knife.

The squares are attached to the previously marked wooden basis by quickly
warming one face of each square over a flame and then pressing it gently against
the tool with the fingers. The tool is now ready for rough-pressing.

Three strong eyes are screwed into the back of the tool, and it is suspended,
face down (by means of wires connected to the ceiling of the room), so as to hang

10 THE MODERN REFLECTING TELESCOPE.

about two feet above the flame of a gas or gasoline stove. The tool can now be
swung about so that the rosin squares are warmed uniformly. When the squares
are slightly soft and very slightly warm, but not hot, to the touch, the tool is
placed upon the previously ground glass which is to be polished, the glass having
just previously been thoroughly wet with distilled water so that the rosin will not
stick to it. Slight pressure may be exerted to assist in pressing the rosin surface
to fit the glass. The tool will have to be slightly warmed and pressed several times
before good contact is secured all over. I always prefer to “rough-press” the rosin
tools on an iron grinding tool having the same form as the glass, if a sufficiently large
one is available; but the precaution is always taken to cover the iron tool with
clean wet paper.

The rosin squares will have spread somewhat irregularly during the rough-
pressing ; so the surface is marked with a straight-edge and knife, and the edges of
the squares are trimmed so that the grooves between them are straight and of uni-
form width. This trimming is best done with a sharp knife, held so as to make an
angle of about 60° with the surface of the tool, and drawn quickly toward the
workman.

The rosin squares are now ready for coating with wax. A pound of best bees-
wax is melted in a large clean cup and is very carefully strained through several
thicknesses of cheese-cloth into a similar clean cup. A brush is made by tying sev-
eral thicknesses of cheese-cloth around the end of a thin blade of wood 14 inches wide.
Each rosin square is now coated with a thin layer of wax, by a single stroke of the
brush; the wax should be very hot, otherwise the layer will be too thick.

The tool is now ready for “cold-pressing” or “ fine-pressing,” a matter of the
most vital importance, which will be more properly described later, in connection
with the work of polishing the glass.

The work of making a large concave mirror will now be described in detail.

CHAPTER VI.

ROUGH-GRINDING THE FACE AND BACK OF A ROUGH DISK OF GLASS, AND
MAKING THE SAME PARALLEL.

Tue rough disk of glass is placed upon the carpeted turntable, and a long
strip of thin oilcloth is drawn around its edge; the upper edge of the oilcloth is
securely fastened to the glass by means of a strong cord, and the junction between
oilcloth and glass is made water-tight by means of water-proof adhesive tape. The
oilcloth strip is wide enough to hang several inches below the edge of the iron
plate on which the glass rests, so that the circular trough of galvanized iron, which
can be seen in Plates tv and vi, catches all of the emery and water which are
washed over the edges of the glass during grinding; this circular trough is station-
ary, has two holes in its bottom above the buckets, which can be seen in the plates,

, ie

THE MODERN REFLECTING TELESCOPE. ile

and is kept scraped clean by two scrapers which reach down into it from the
revolving turntable. Several important results are thus secured : the carpet cushion
under the glass is kept dry; the entire machine is kept perfectly free from the
dripping of the grinding material; and all of the latter material is caught in
buckets and is used again and again in the later and finer grinding.

The large irregularities of surface of large rough disks are usually ground
away with coarse emery and a heavy, flat, half-size iron tool without grooves, the
surface of which is rounded up considerably at the edge, so that the tool may rise
easily over obstructing irregularities without breaking them. The grinding machine
is set so that the half-size tool moves over the glass well out to one side of the lat-
ter; the rotation of the turntable of course brings all parts of the glass in succession
under the tool; if the setting of the machine is such that the half-size tool passes
in much beyond the center of the glass at every stroke, the surface of the latter will
become concave.

When the marked irregularities of surface are ground away, the full-size, flat,
grooved iron tool is put on. A tool of this kind is almost indispensable in making
a mirror. Emery and water are supplied through the cups at the back of the tool,
and the glass is quickly ground approximately flat. The glass is now turned over,
and the other side is ground in a precisely similar manner.

The thickness of the glass is now tried, all around, by means of calipers. The
approximately flat surfaces will probably be found to be far from parallel. If this
is the case, the thick side may be ground down as follows: The belt which drives
the turntable is loosened, until it will just rotate the latter, and a brake is arranged
so that the workman can stop the rotation of the turntable at any desired point by
pressing on the brake with his foot. A flat, half-size grooved tool is put on, and
set so as to work far out to one side of the glass. A medium grade of emery (No.
70) is used, and the machine started. As the thick side of the glass, which has
been marked, comes beneath the moving tool, the turntable is slowed down or
stopped, so that a great excess of grinding is done on the thick side at each revolu-
tion. By distributing the grinding carefully, and trying the thickness often with
the calipers, the upper surface is easily made parallel to the lower one. When this
‘is done the full-size tool is again used for a short time. The glass is then ready for
edge-grinding.

CHAPTER VII.
GRINDING EDGE OF GLASS.—ROUNDING OF CORNERS.

In order that an efficient edge-support, which will be described later, may be
given to the glass, it is desirable that the edge of the latter be ground truly cireu-
lar and square with the face. The manner in which this is accomplished is shown
in Plate v1. The glass lies upon three large blocks of wood, which hold it several

12 THE MODERN REFLECTING TELESCOPE.

inches above the surface of the circular iron plate. Thin oileloth is arranged
about the blocks and over the iron plate, to keep them dry. A smooth, flat, iron
face-plate is mounted (so as to rotate in a vertical plane) on a heavy lathe head-
stock; the latter is carried upon a strong slide which can be moved toward, the
glass by means of a fine pushing-screw. The lathe and face-plate are driven at a
high rate of speed by means of a belt. In the case of the 5-foot glass the face-
plate used was 24 inches in diameter and made 1,000 revolutions per minute. For
a 24-inch glass, 34 inches thick, a face-plate 11 inches in diameter, making 1,800
revolutions per minute, is used. A frame of wood, covered with oilcloth, is built
around the face-plate, so that the grinding materials will not be thrown about the
room. The glass rotates slowly with the turntable, as usual. Emery and water,
or sand and water, are heaped upon the horizontal surface of the glass, and are
slowly scraped toward and over the edge, so as to come between the revolving face-
plate and the glass; a small jet of cold water, brought from the hydrant by means
of a rubber tube, greatly assists in the uniform feeding of the emery, and also in
preventing the generation of heat. But there is in reality no danger of heating, for
the revolving face-plate never actually touches the glass. As the irregularities of the
edge are ground away the face-plate is gradually moved forward by means of
the slide and pushing-screw.

If the edge of the rough disk be very irregular, as is usually the case, the sur-
face of the iron face-plate will have a circular groove worn in it, by the time the
rough-erinding of the edge is done; in this case the face-plate should be turned flat
and true again, and smoothed on a flat iron grinding-tool, before the edge of the glass
is fine-ground. Several fine grades of emery are now prepared by the process of
washing to be described later, and the edge-grinding is finished by the use, in sue-
cession, of three such grades of emery as flour, three-minute washed, and ten-minute
washed. Care should be taken throughout the process that the edge of the glass
is ground square with the face; any error in this respect can be corrected by slightly
raising or lowering the outer end of the slide which supports the lathe head-stock.

Edge-grinding is accomplished very quickly in the manner described. The
edge of a 24-inch disk four inches thick, even when very rough and irregular, has
been ground and smoothed in ten hours of actual grinding. Despite the great
speed of the rotating face-plate, I have never had any chipping of the glass or
accident of any kind occur.

Before beginning the fine-grinding of the face and back it is well to round the
corners at the edge of the glass. This is done by means of a smooth flat strip of
sheet-brass of the size and shape of a large flat file; this is worked over the corners
of the glass by hand, while the disk rotates slowly, emery and water being used for
cutting. A “quarter-round” corner is usually made. Finer and finer grades of
emery are used for smoothing the quarter-round. This rounding and smoothing
are very necessary, as particles of glass from a sharp or rough edge are lable to be
drawn in upon the surface by the action of the grinding tool during fine-grinding,

The wooden blocks are now removed and the glass replaced upon the carpeted
turntable.

THE MODERN REFLECTING TELESCOPE. 138

CHAPTER VIII.

FINE-GRINDING AND POLISHING THE BACK OF THE MIRROR.

Berore discussing the work of fine-grinding I shall describe briefly the making
of the fine grades of emery. I never buy finer grades than “flour.” The latter
grade is used with the full-size flat grooved tool to give a moderately fine surface
to the glass after the rough-grinding previously described has made the front and
back approximately flat and parallel. The residue of emery, fine ground glass, and
water, resulting from the grinding with flour emery, is caught in buckets, as_pre-
viously described. This residue is mixed with an abundance of water, in (for a
large mirror) three or four clean granite-ware buckets, which are marked A. The
contents of these buckets are thoroughly stirred, and are allowed to settle for two
minutes; during this time all coarse particles will have settled to the bottom, and
“two-minute” emery and finer grades remain in suspension in the water. The
liquid is now quickly siphoned off, by means of a rubber tube, into other clean
granite-ware buckets marked 4, from which the handles have been removed. The
contents of the latter are allowed to settle for four minutes, when the greater part
of the liquid in each is carefully poured back into the buckets A. The contents of
the latter buckets are reserved. The sediment remaining in the buckets B is the
“two-minute” washed emery, with which the fine-grinding of the back is begun.
After the grinding with this grade is finished, the residue from this grinding is
mixed with what was reserved in the buckets A, the whole is stirred again and
allowed to settle for five minutes, the liquid is siphoned off, and thus “five-minute
washed ” emery is secured. In a similar manner emeries which have remained in
suspension in water for 12, 30, 60, 120, and 240 minutes are secured. In this way
the large quantities of the finer grades which are necessary for large work can be
secured as the work progresses. If accumulations of residues from previous work
are available, some time will be saved by washing out all of the fine grades desired
before the fine-grinding is begun.

Plane and concave mirrors are finished approximately flat on the back, as this
form is most convenient for the application of the support-system. Fine-grinding
of the back is usually done with the full-size, flat grooved tool, as this works
rapidly. In this part of the work, in which the greatest refinement is not neces-
sary, it is my custom to use the fine grades of emery (when these have all been
prepared in advance) in succession, without stopping the machine or taking off the
tool between grades for the purpose of cleaning the tool and the glass. The emery
and water are supplied through the wooden cups at the back of the tool.

For a 24-inch mirror and its full-size tool, strokes varying from 6 to 8 inches
in length are used with the 2-, 5-, and 12-minute washed emeries; shorter strokes,
from 4 to 6 inches in length, are used with the finer grades. Considerable lateral
displacement of the tool, amounting at the greatest to 2 or 23 inches on the glass, is
given at short intervals, by means of the transverse slide, On an average 20 double
strokes per minute are given in fine-grinding a 24-inch mirror with full-size tool.
Between 7 and 8 double strokes occur for each revolution of the glass and turntable.

14 THE MODERN REFLECTING TELESCOPE.

With regard to counterpoising the tools during fine-grinding, the following
statements may be made: My full-size iron tools for a 24-inch mirror weigh about
150 pounds, or } pound for each square inch of area. This weight, or even $
pound to the square inch, is not objectionable with emeries down to 5-minute or
10-minute washed; but when this weight is allowed with finer emeries, scratches
are liable to occur; indeed, with 30-minute washed and all finer grades they are
almost certain to occur. The pressure on the glass is therefore decreased, by coun-
terpoising the tool, to approximately + pound to the square inch for 12- to 20-minute
emeries, } pound per square inch for 30- to 60-minute emeries, and about 54, pound
per square inch for 120- and 240-minute emeries. This rule is followed, approxi-
mately, in all fine-grinding, whether of back or face. ‘This obviates, to a great
extent, the danger of scratches in grinding, provided that thorough cleanliness is
practiced in the preparation and use of the fine emeries. ‘The apparatus by which
the counterpoising is effected has already been described (page 5).

In fine-grinding a 24-inch class, the 2-minute and 5-minute emeries are used for
three-quarters of an hour each; the 12- and 30-minute emeries for one hour each,
and the 60-, 120-, and 240-minute emeries for one and one-half hours each. The
fine-ground surface resulting is so exquisitely smooth that it takes a full polish very
readily.

The back of the glass is now ready to be polished. This is done with a half-
size or two-thirds size polishing tool, which is moved about on the glass by the
action of the machine precisely as a half-size grinding-tool would be. Optical rouge
and distilled water are used, instead of emery and water. The work of polishing
will be described in detail later, in connection with the work of finishing the face
of the glass.

It is an excellent plan to fine-grind and polish the front surface of a disk
also, approximately flat, as has been described for the back; the optician is then
able to examine carefully the internal structure of the disk. Usually there is no
choice as to which side shall be used for the face of the mirror, but this can readily
be determined when both sides are polished. Plate vi shows the 5-foot disk with
both sides ground and polished in this manner.

CHAPTER. IX.

GRINDING THE CONCAVE SURFACE.

As before stated, it is my practice to use full-size grinding tools for concave
mirrors up to 24 or 30 inches in diameter. For larger concave mirrors half-size tools
are generally used. I shall first describe the grinding of a 24-inch concave.

The glass must be carefully centered by means of the three adjustable ares
attached to the supporting plate; these arcs must not be screwed tightly against
the glass, lest: the latter be strained; several thicknesses of heavy drawing paper
are used between ares and glass.

kl.

pl S ale art BAe?

od

-;
rr

PLATE VI

ee See O18 oe

GLASS.

OF

Z
<
=
Z
Zz
Zz

QA
as
es
we
rt 6a
ze
eZ fy
SV
=

THE MODERN REFLECTING TELESCOPE. 15

The glass must also be carefully /evc//ed (by means of the three large adjusting
screws of the turntable) so that its upper surface is accurately at right angles to
the axis of rotation; this is determined by rotating the turntable, and trying the
surface with a surface-gauge. The band of thin oilcloth is securely bound around
the edge of the glass, to keep the polished back and the cushion clean and dry.

The excavation of the concave is begun with moderately coarse emery (if the
concave is to be quite deep) and a lead tool; this is a lead disk about 10 inches in
diameter and 14 or 2 inches thick; it is easily turned in a lathe to the proper
curvature; it is used on and near the center of the glass until a depression of
approximately the desired curvature (as determined by the spherometer) and of 12
or 13 inches diameter is produced. A heavy iron tool about 13 inches in diameter,
which has been turned and ground to the proper curvature, is now put on with
about No. 90 emery. By giving careful attention to the length of stroke, and to
the position of the tool on the glass as determined by the setting of the transverse
slide, and by frequent trials of the curvature of the excavation with the spherom-
eter, the diameter of the excavation is gradually increased, while its curvature is
continually kept very near that which is desired for the finished mirror; this keeps
the iron tool of proper curvature also.

The stroke used in this work should vary from 6 to 10 inches in length. As
the size of the excavation increases, the setting of the transverse slide is continually
changed so that the tool acts farther and farther to one side of the center of the
glass; otherwise the radius of curvature will be shortened. When the diameter of
the excavation has increased to about 22 inches, flour emery is substituted for the
No. 90, and the grinding is continued as before. Care is now taken to make the
curvature read exactly right with the spherometer. When the excavation becomes
about 23 inches in diameter, the 13-inch tool is taken off, and the full-size, convex,
grooved iron tool is put on; this has previously been fine-ground to the proper
curvature on the corresponding concave, tool. With this tool and washed flour
emery the diameter of the concave on the glass is increased to 23% or 23} inches.

The fine-grinding or smoothing of the concave is now done with the full-size
tool. The same grades of emery, the same lengths and speed of stroke, and the same
rules in regard to counterpoising are used as have already been described in the
case of fine-grinding the back of the glass (page 13). The length of stroke is
changed every eight or ten minutes, and the lateral displacement of the tool (given
by means of the transverse slide) is changed slightly at the end of every two or
three complete revolutions of the glass. The tool is taken off after each grade of
fine emery is used, and the tool and glass are carefully cleaned. With the assistance
of the counterpoise lever the removal of the tool is effected easily and safely, with-
out disconnecting it from the main arm of the machine ; this is well shown in Plate
rv, in which the grinding tool is shown hanging at one side of the glass.

The surface of the glass is examined with a microscope after each grade
of emery is used, to make sure that no pits from previous grades remain,

During all fine-grinding and machine-polishing a large sheet of heavy clean
paper or pasteboard is attached to the main arm in such a way that no particles of

16 THE MODERN REFLECTING TELESCOPE.

dust from the belts which control the slow rotation of the tools can fall upon
the glass.

The process of grinding larger concave surfaces without the use of full-size
tools is precisely similar to that described for a 24-inch mirror, up to the point of
substituting the 24-inch convex tool; from this point the grinding is carried on by
a continuation of the use of a half-size, convex grooved tool; this may be the same
iron tool which has been used for enlarging the excavation. When the diameter
of the excavation approaches that of the glass, the tool should be tested with the
spherometer for curvature, and, if necessary, ground in its corresponding concave
iron tool until its curvature is uniform and of exactly the desired radius. The
grinding of the glass is then continued with washed flour emery until the edge of
the excavation is within 1 inch of the edge. Experience in the previous use of the
half-size tool, in enlarging the excavation and in keeping the curvature of the glass
uniform and of the desired radius, will enable the optician to decide upon the vari-
ous lengths of stroke and the various settings of the transverse slide necessary in
this grinding and in the finer grinding to follow.

In fine-grinding a 30-inch concave with a 16-inch tool, strokes varying from 6
to 12 inches in length are used; for a 9-inch stroke the normal setting of the trans-
verse slide (7. é., one which would tend neither to lengthen nor shorten the radius of
curvature of the glass) would be such that the outer edge of the tool overhangs the
glass about 8 inches in the forward stroke, while the inner edge of the tool passes
about one inch on the other side of the center of the glass on the return stroke.

Throughout the entire process of fine-grinding with the half-size tool the length
of stroke is changed once every eight or ten minutes; at the end of every two or
three revolutions of the glass the setting of the transverse slide is changed, a little
at a time, for a considerable distance on either side of the normal setting; the set-
ting of the slide can be changed without difficulty, while the machine is running, by
merely turning a hand-wheel. By these means the formation of zones of unequal
focal length can be entirely avoided.

The same grades of emery are used, and the same rules in regard to counter-
poising observed, as with full-size tools. Notwithstanding the fact that the length
of stroke can be considerably greater than with full-size tools, each grade of emery
must be used for a longer time, on account of the smaller area of the grinding sur-
face. Glass and tool are thoroughly cleaned, and the surface of the former exam-
ined, after the use of each grade of emery, as before described.

Care must be taken during this work that the belts which rotate the turntable
are kept tight, so that no irregularity in the rotation of the turntable with refer-
ence to that of the crank-shaft can occur. It is absolutely necessary that all of the
fine work on large mirrors be done in rooms where no sudden changes of tempera-
ture can occur, and that nothing be allowed which might affect the temperature of
the glass locally.

If the concave mirror is intended for a paraboloidal one, the fine-ground surface

should be spherical, with its radius of curvature 2 F +i where F is the desired

THE MODERN REFLECTING TELESCOPE. 17

focal length of the finished paraboloid and R is the semi-diameter of the mirror:
the reason for this is fully explained later. I have never attempted to parabolize
while fine-grinding ; it is possible that it might be well todo this in the ease of
very large mirrors of short focus, but my practice has been to fine-grind and polish
to a spherical surface, free from zones, and then to parabolize by means of suitable
polishing tools.

CHAPTER X.
POLISHING.

Tue preparation of polishing tools has already been described. The polishing
rouge which I use is of the quality which is used in large quantities commercially
in polishing plate-glass. I prefer the powdered form always. This grade of rouge
is not expensive (it costs about 30 cents per pound), but, like all rouge which
I have seen, it contains hard, sharp particles which may cause scratches. It must
therefore be thoroughly washed in the following manner :

Ina clean, deep bowl C enough rouge is placed to fill it about one-third full ;
the bow] is then nearly filled with distilled water. The mass is very thoroughly
stirred with a clean wooden paddle, and allowed to settle for about twenty minutes.
The water above the rouge will now be perfectly clear; this water is siphoned off.
With a clean spoon the light and fine rouge constituting the upper one-third of the
precipitated mass is removed, and placed in a second clean bowl JY. The rouge
remaining in C’ may be again stirred up with an abundance of distilled water, and
allowed to settle as before, the water siphoned off, and the upper one-fourth of the
precipitated rouge removed and placed in J. The heavier rouge which remains
in Cis about half of the original quantity taken; this is usually reserved, and,
after further washing, is used for polishing the backs of mirrors, and for similar
work. Only the contents of the bow] J are used for fine work, and these are stirred
up again and again with distilled water during the process of polishing, and only
the fine, soft cream which remains on the top of the mass of rouge, when it settles
each time, is used for polishing.

The thin cream of rouge and distilled water is applied to the glass by means of
a wide brush consisting of a thin paddle of wood with clean cheese-cloth wrapped
and tied about one end. Brushes of the usual kind should not be used.

By taking these precautions, and by the use of the wax surface on the rosin
squares, scratches in polishing can be entirely avoided. It is true that the very
light, fine rouge polishes more slowly than the heavier and coarser rouge, but an
exquisitely fine polished surface is produced on the glass by its use. The wax
surface also polishes more slowly than a bare rosin one, but it has the very great
advantage that its action is more smooth and uniform than that of the rosin surface ;
the latter often tends to cling to the glass, and this unequally in different parts of
the stroke.

18 THE MODERN REFLECTING TELESCOPE.

The same question arises in regard to the size of polishing tools as in the case
of grinding tools,—whether they shall be full-size or smaller. In the writer’s opin-
ion fine plane and spherical surfaces up to about 36 inches in diameter are best
polished with full-size tools, which are moved by hand, by the optician and one or
two assistants, upon the surface of the slowly rotating glass. The upper parts of
the machine are, of course, removed during such polishing, which I shall call manual
polishing.

A 24-inch polishing tool, prepared as already described, with its wooden basis
23 inches thick, weighs about 25 pounds; this is not heavy enough for the best ac-
tion in polishing ; so about 50 % additional weight is put on in the form of 12 lead
blocks which are distributed uniformly and screwed to the back of the tool.
This gives a weight of about j4, pound for each square inch of area, which is found
to work well for all large tools. For tools 18 inches or less in diameter somewhat
greater pressure per square inch of area may be used. A 36-inch tool, with wooden
basis 33 or 4 inches thick, weighs 75 or 80 pounds, and needs no additional
weighting.

The work of polishing a 24-inch mirror with full-size tool will now be described.
Six strong knobs of oak wood are screwed to the back of the wooden basis, each
knob being at the center of weight of each sixty-degree sector of the tool. These
knobs serve for pushing, pulling, and lifting.

The polishing tool, which, with the glass, should have cooled over night after
the warm-pressing or rough-pressing previously described, is now to be cold-pressed.
Cold-pressing is absolutely necessary in all fine work on large optical surfaces.
In warm-pressing, both tool and glass are distorted by even slight warming, and
when they become cool a perfect fit cannot be expected. The glass is carefully
wiped with clean cheese-cloth, and an abundance of very thin mixture of rouge
and water is spread upon it. The tool is now placed upon the glass and allowed
to he for several hours, being moved about slightly every ten minutes to redis-
tribute the rouge and water, and to prevent the latter from drying around the
edges. The pressing may be assisted at first by means of a 20- or 30-pound weight,
the pressure of which must be distributed by some such means as three bars
laid upon the six knobs, and a triangle, carrying the weight, laid upon these. The
final cold-pressing must be done by the weight of the tool alone. The tool is
taken off and examined occasionally ; when it is sufficiently pressed the wax sur-
face appears uniformly smooth and bright. So perfect a fit is secured in this way
that there is no danger of injuring the form of the glass when polishing is begun.
This applies to all stages of polishing and figuring. A fresh supply of rouge and
water is now spread upon the glass.

The stroke of the 24-inch polishing tool is easily given by the optician and one
assistant, who sit on opposite sides of the machine; the glass slowly rotates with
the turntable, making about 2 revolutions per minute. The knobs on the back
of the tool are held in the hands, and the stroke is given by alternately pushing
and pulling; no vertical pressure whatever should be given by the hands. In
addition, a considerable side-throw is always given, first to one side, then to

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE — RITCHEY

PLATE VII

FIVE-FOOT MIRROR WITH FRONT AND BACK POLISHED APPROXIMATELY FLAT
LOOKING THROUGH THE GLASS

THE MODERN REFLECTING TELESCOPE. 19

the other; this greatly assists in preventing the formation of zones of unequal cur-
vature. Polishing may be begun with a stroke 6 inches in length, which of course
causes the tool to overhang the glass 3 inches at the ends of the stroke; between
20 and 25 double strokes per minute are given. The side-throw used with this
length of stroke is about 2 inches, ¢. ¢, the tool is made to overhang the glass
Bout 2 inches, first to the right, then to the left; the time occupied in passing from
the extreme right to the extreme left is about w Hate is required for 4 double strokes.
This stroke and side-throw are continued while the glass makes exactly 2 revolu-
tions ; the tool does not rotate with the glass, of course, while the stroke is being
given; the last stroke should end with the tool central upon the glass.

5

Tool and glass are now allowed to rotate together for 5 of a complete revolu-
tion, and each optician then grasps the pair of knobs next to that which he held
before, so that the stroke is now given along a diameter of the tool 60° from that
last used ; the length of stroke is now changed to 7 inches, and the side-throw
to 23 inches, and polishing is again carried on during exactly 2 revolutions of the
glass. Tool and glass are again allowed to rotate together for 5 of a revolution, and
polishing during 2 revolutions is now done with a stroke of 8 inehee and side- Hoe
of 3 inches. ates the next periods of polishing, each of 2 revolutions of the
glass, the stroke and side-throw are gradually shortened until a stroke of 4 inches
or less is reached; then the length of stroke is increased again.

When polishing has been carried on during 6 or 8 periods of 2 revolutions
each, it will be found necessary to supply more rouge. The only entirely satisfac-
tory method of doing this, when a full-size polishing tool is used, is to remove the
tool from the mirror, and quickly spread the thin cream of rouge and water upon
the glass as uniformly as possible with the cheese-cloth brush. The removal of
the tool is effected by the two opticians carefully sliding it off the mirror, and lift-
ing at the same time. The tool should be allowed to remain off the glass for only
as short a time as possible, so that the form of the latter shall not be altered as a
result of a change of temperature of the surface, caused by evaporation. For this
and other reasons, such as the prevention of dust, the air in the polishing room
should be kept moist by keeping the floor well sprinkled.

When the tool is replaced on the mirror it is lifted by both opticians so that
only a very small part of its weight remains on the glass, and is lightly moved about,
for 30 seconds or more, to distribute the rouge and water thoroughly before polish-
ing is continued. As before stated, the method just described is the only entirely
satisfactory one, known to the writer, of supplying rouge during the polishing with
a full-size tool. All methods of supplying rouge at the edge, or through holes in
the tool, are inadmissible when the greatest refinement of figure is required.

It is in order that they may be easily handled in the manner described that
full-size polishing tools should be made light. It would, of course, be possible to
devise mechanism by which tools of any size and ecient could be sufficiently
counterpoised, could be moved about upon the glass, and could be removed from
the latter for the purpose of supplying rouge. The simple and economical method
which I have described, however, works well for mirrors up to 36 or 40 inches in

20 THE MODERN REFLECTING TELESCOPE.

diameter. For larger mirrors it is more economical, in the opinion of the writer, to
use half-size tools for obtaining a fully polished spherical surface, and the same and
smaller tools for parabolizing. The method of using these will be described
later.

In general, it is much easier to prevent the formation of zones, and to eliminate
zones already present, with full-size polishing tools than with smaller ones. The
method of manual polishing just described, in which the length of stroke and the
amount of side-throw are very frequently changed, tends to give a spherical surface,
except for a zone around the edge of the mirror one-half an inch or less in width ;
this part of the surface will be of too great focal length, 7. ¢, will turn down or
back slightly, unless means are taken to prevent it. This tendency is most pro-
nounced when a long stroke is used to excess, or when the rosin squares are too
soft. It is entirely prevented by diminishing the area of the rosin squares around
the edge of the tool, by trimming their edges to such a form as is shown in Fig. 4,
page 28. The exact amount of trimming required depends upon the length of
stroke, hardness of rosin, and temperature of polishing room, and therefore can be
exactly determined only by experience.

A 24-inch mirror which has been properly fine-ground with emeries down
to 2-hour or 4-hour washed, is readily brought to a perfect polish with a full-size
tool in from 2 to 4 hours of actual polishing. If several broad zones of different
focal lengths have resulted from the fine-grinding, as frequently happens, these
zones can be gradually eliminated by a continuation of the use of the full-size
polisher as above described.

Attention must be given to the rosin squares, which gradually press down so
that their edges must be trimmed to keep the grooves of their original width and
of uniform width. When the bare rosin begins to show at the corners or edges of
the faces of the squares, which will occur after 6 or 8 hours’ use of the tool,
a new coat of wax must be applied, and the tool must again be thoroughly cold-
pressed. It must not be supposed, however, that cold-pressing is necessary only at
such times; in all fine work this pressing must be done whenever the tool has
remained off the glass for more than a few minutes ; after hanging face down during
the night the tool is always cold-pressed for about 2 hours before polishing is begun
in the morning,

Polishing with half-size or smaller tools is best done with the machine, instead
of by manual work. These tools do not have to be removed from the glass in order
to renew the supply of rouge; they are therefore connected to the machine and used
very much as half-size grinding tools are used; in my work they are made of such
weight that they need not be counterpoised. Very large or unusually heavy polish-
ing tools of this kind can, of course, be easily counterpoised when desired.

Great experience, constant attention to very frequent changing of the position
of the tool by means of the transverse slide, and frequent testing of the form of the
mirror surface are necessary in polishing with half-size or smaller tools, in order to
preserve the uniform curvature of the surface. This is greatly facilitated by trim-
ming the rosin squares at and near the edges of the tool, as in the case of full-size

THE MODERN REFLECTING TELESCOPE. 21

tools, but to a greater extent; the effect of the action of the edges of the tool
is thus softened or blended.

When a half-size or smaller tool has just been coated with wax, or is known
to be far from the exact form desired, it is first cold-pressed in the usual way on the
center of the glass. But the final cold-pressing of such tools should be done as fol-
lows: The entire surface of the glass is painted with rouge and water, and
the machine is set to give a “normal” stroke, 7. ¢, one by which the tool is made
to cover the entire surface of the mirror as uniformly as possible (without an
excess of action on any zone) as the glass revolves; the machine is run extremely
slowly, and the setting of the transverse slide is changed often; after pressing the
tool for an hour or two in this way, polishing or figuring is to be begun.

CHAPTER XI.
TESTING AND FIGURING SPHERICAL MIRRORS.

BrroreE describing the work of figuring concave mirrors, which is done with
polishing tools, it will be necessary to consider methods of testing. The principles
involved in testing concave mirrors at their center of curvature by Foucault’s
method. have been thoroughly explained and illustrated by Draper on pages 13-19
of his book, and by Dr. Common in his book On the Construction of a Five-Foot
Liquatorial Reflecting Telescope. Foucault’s original paper on this subject may be
found in Vol. V of the Annals of the Paris Observatory.

All mirrors, when being tested, are placed on edge, so that the axis of figure
is nearly horizontal, large mirrors being suspended in a wide, flexible steel band,
lined with soft paper or Brussels carpet; for glass mirrors larger than 30 inches in
diameter it is very desirable to have the grinding and polishing machine so con-
structed that the glass can be turned down on edge for testing, in the manner shown
on Plate u, without removing it from the machine. A 30-inch glass mirror 4
inches thick weighs about 260 pounds; mirrors larger than this are difficult
to handle without suitable mechanism.

A small, brilliant source of light, or “ artificial star” may be produced by pla-
cing in front of the flame of an oil lamp a thin metal plate in which a very small
pinhole has been bored. If the illuminated pinhole be placed about an inch to one
side of the principal axis of the mirror, and at a distance from the mirror equal to
its radius of curvature, a reflected image of the pinhole will be formed on the
other side of the axis, and at the same distance from it and from the mirror as
the corresponding distances of the pinhole itself. If the surface of the mirror
is perfectly spherical, and if there are no atmospheric disturbances in the course of
the rays, the reflected image, when examined with an eyepiece, will be found to be
a perfect reproduction of the pinhole, with the addition of one or more diffraction
rings around it, minute details of the edge of the pinhole appearing as exquisitely
sharp and distinct as when the pinhole itself is examined with an eyepiece. If the

99 THE MODERN REFLECTING TELESCOPE.

eyepiece be moved outside and inside of the focus, the expanded disk in both cases
appears perfectly round. Nothing can be more impressive than to see such a
reflected image produced by a fine spherical mirror having a radius of curva-
ture of 100 feet or more. Several such mirrors of 2 feet aperture have recently
been finished here.

The use of an eyepiece is interesting for such experiments as that just
described, and is important as a check upon the test with an opaque screen. The
latter test, however, which I shall call the knife-edge test, is used almost exclusively
for mirrors of all forms; it is far more serviceable than the eyepiece test in deter-
mining the nature and position of zonal irregularities, and is far more accurate in
determining the radius of curvature either of a mirror as a whole, or of any zones
of its surface.

If the eye be placed just behind the reflected image of the illuminated pinhole,
so that the entire reflected cone of light enters the pupil, the polished, unsilvered
mirror surface is seen as a brilliant disk of light. Let an opaque screen or knife-
edge be placed in the same plane through the axis as the pinhole, and be moved
across the reflected cone from the left, and just in front of the eye; if a dark shadow
is seen to advance across the mirror from the left, the pinhole and knife-edge are
inside of the best focus, and must be moved together away from the mirror;
if, however, with the knife-edge still moved across from the left, the shadow
advances across the mirror from the right, pinhole and knife-edge are outside of the
focus and must be moved toward the mirror. By repeated trials a position is found
from which the shadow does not appear to advance from either side, but the mirror
surface darkens more or less uniformly al] over: this is the position or plane of the
best focus, and it is with this position of the knife-edge that irregularities of the
surface, if any exist, are seen in most highly exaggerated relief; with this position of
the knife-edge, the mirror, if perfectly spherical, is seen to darken with absolute uni-
formity all over as the screen is moved across the focus, and the impression of a per-
fectly plane surface is given to the eye.

If, however, the mirror is not perfectly spherical, but contains several zones of
slightly different radii of curvature, a very common case, these zones will appear as
protuberant or depressed rings on an otherwise plane surface. The reason for this
is evident; the light from some parts of such zones is cut off by the knife-edge
before, from other parts after, the illumination from the general surface is cut off;
the surface is therefore seen in light and shade, 7. ¢., in enormously exaggerated
relief. The mirror must be regarded as being illuminated by light shining very
obliquely along the surface from the side opposite that from which the knife-edge
advances across the focus. The interpretation of lights and shades becomes easy
after a little experience; not only is the character of a zone—whether it is an ele-
vation or depression—readily seen, but its diameter and its width are readily
determined.

If the disk of glass is of sufficient thickness and of proper quality, and if
attention has been given to the uniform rotation of the turntable and to the pro-
tection of the glass from abnormal conditions of temperature during grinding and

bo
oo

THE MODERN REFLECTING TELESCOPE.

polishing, all irregularities of figure which occur are perfect zones or rings concen-
tric with the edge of the glass; that is, the surface is always a perfect surface
of revolution. If, however, these precautions have not been taken, or if the glass
has been improperly supported during grinding and polishing, or if it has been cut
out of thick rolled plate-glass, so that it is weak in the direction of one diameter, an
astigmatic mirror may be produced, in which the radius of curvature is slightly
different along two diameters at right angles to each other.

Astigmatism is easily recognized with either the knife-edge or the eyepiece
test. Let the plane of the apparent focus be determined with the knife-edge
advancing from the left, then from above, then from the right, then from a number
of directions between these three; if astigmatism exists the planes of the various
foci thus found will not coincide ; and the directions of greatest and least curvature
of the surface are readily determined. When the eyepiece test is used, an astig-
matic mirror does not give a sharp image even at the best focus; if the eyepiece be
moved outside and inside of this focus the expanded disk becomes elongated, and
is not uniformly illuminated ; the direction of elongation outside is at right angles
to that inside, and the distribution of light in the expanded disk is entirely different
outside and inside of the focus.

!

\

\ |

\i __LENS
hea wore Ae

DIAGONAL PRISM

PRINCIPAL

Axis

FIG, I.

ARRANGEMENT BY WHICH ARTIFICIAL STAR IS USED VERY CLOSE TO OPTICAL AXIS.

The general character of the tests having now been described, let us con-
sider some important matters of detail which are necessary for the greatest refine-
ment in testing all forms of mirrors.

By the use of a small lens and a diagonal prism, in the manner shown in Fig.
1, the lamp can be kept well out of the way, and the illuminated pinhole and its
reflected image brought very near to the axis of figure of the mirror. ‘This is of
much importance in testing mirrors of short focus or of great angular aperture, as
the danger of errors in testing due to working considerably out of the axis of figure is
avoided. As may be seen in the figure the pinhole is now placed at the surface of
the diagonal prism nearest to the mirror being tested. The arrangement should
be such that the cone of rays proceeding from the lens is considerably larger than
is needed to fill the concave mirror.

When being figured, mirrors are usually tested while unsilvered, since very
frequent tests are desirable. While the amount of light reflected from the polished

24 THE MODERN REFLECTING TELESCOPE.

unsilvered surface is surprisingly great, a much more brilliant “ artificial star” than
that given by the oil lamp is required for the greatest refinement and accuracy
with the knife-edge test, especially in the cases of plane, paraboloidal, and hyper-
boloidal mirrors, in which there are two reflections from the unsilvered surface. It
might be supposed that a larger pinhole could be used, and thus a more brilliant
illumination of the mirror surface secured ; but a large pinhole allows an apparent
diffusion of light over the mirror surface, which obliterates all the more delicate
contrasts of illumination due to minute irregularities of surface. With feeble
illumination of the surface the eye is entirely unable to detect slight contrasts,
which with brilliant illumination become strong and unmistakable. When the
knife-edge test is used with an extremely small pinhole of between 5}, and 31,
inch in diameter, illuminated by acetylene or (what is much better) oxy-hydrogen
or electric-are light, minute zonal irregularities are strongly and brilliantly shown,
which are entirely invisible with large pinhole or insufficient illumination. With
the arrangement of lens and diagonal prism (Fig. 1) either of the sources of light
named can be used without difficulty ; disturbances of the air from their heat
should be prevented by placing the light behind a partition with a window of thin
plate glass.

With the best conditions of apparatus just described, the degree of accuracy to
be attained with the knife-edge test is surprising. With a mirror of 2 feet aperture
and 50 feet radius of curvature, the plane of the center of curvature can be easily
located to within ;4, inch, and with care to within half of that amount. With the
dimensions eines a change of ;4, inch in the radius of curvature corresponds to
a change of — inch in the depth of the curve of the mirror surface. There can
be no doube that zonal irregularities of surface of half of this amount are readily
recognized.

We are now ready to consider the finishing of a spherical mirror, As before
stated, a continuation of the use of the full-size polishing tool tends toward the
gradual elimination of zonal irregularities. This work is often slow and laborious,
however, for when the mirror becomes nearly finished, so that any zones, when
seen with the knife-edge test, appear as extremely slight elevations or depressions,
the improvement becomes exceedingly slow. The work may be facilitated by the
local use of very small polishing tools upon protuberant zones. These tools are
usually from 2 to 4 inches in diameter, and consist of squares of rosin upon a basis
of brass; their faces are waxed and cold-pressed, and the squares around their
edges are trimmed in order to soften or blend the action of the edges; small local
tools with their surfaces trimmed as shown in Fig. 13 (in which the shaded parts
represent the rosin) are excellent for the purpose. These local tools are used as fol-
lows : the positions and width of any protuberant zones are carefully determined by
the knife-edge test, and the glass is replaced on the rotating turntable; stationary
pointers are clamped to the machine, and overhang the glass so as to indicate the
exact positions of the zones; the surface is painted all over with rouge and water,
and the optician works the small tools on the high zones by hand ; the rubbing
is done on each zone during several revolutions of the glass, the length and direc-

~~ ae

THE MODERN REFLECTING TELESCOPE. 25

tion of the stroke being changed after each complete revolution. Great care and
judgment must be used in this work, and the surface must be tested very often,
otherwise a wide zone will usually give place to several narrow ones. After the
protuberant zones have been softened down in this way the full-size polisher is
again used for finishing the surface.

A large and perfect spherical mirror is an indispensable part of the equipment
of an optical laboratory, as it affords what is in my opinion the most satisfactory
means of testing large plane mirrors. On account of the ease of rigorously testing
a concave spherical surface, this is the form which should be first attempted
by beginners in optical work.

CHAPTER XII.
GRINDING, FIGURING, AND TESTING PLANE MIRRORS.

Tue making of large plane mirrors of fine figure is usually regarded as much
more difficult than that of Jarge concave mirrors. The difficulty has been, in the
past, largely one of testing. With a satisfactory method of testing the large plane
surface as a whole, in a rigorous and direct manner, the problem is greatly simpli-
fied. So far as the writer is aware, no such test has hitherto been fully developed.
In Monthly Notices, Vol. 48, ». 105, Mr. Common suggests, very briefly, the testing
of plane mirrors in combination with a finished spherical mirror, and gives a
diagram in illustration; but no details in regard to the method are given. This
method has been developed and used for many years by the writer in testing plane
mirrors up to 80 inches in diameter. When this test is used, the difficulty of mak-
ing a 24inch plane mirror which shall not deviate from perfect flatness by an
amount greater than 55p.)9o inch is neither greater nor less than that of making a
good spherical mirror of 2 feet aperture and 50 feet radius of curvature, when
it is required that the radius of curvature shall not differ from 50 feet by a
quantity greater than ;1, inch.

SPHERICAL
MIRROR

TTELUCO 1-17 00 es a

PLANE MIRROR

Fic. 2. DIAGRAM ILLUSTRATING TESTING OF A PLANE MIRROR.

A spherical mirror A (Fig. 2), which should not be smaller in diameter than
the plane mirror B to be tested, is figured with the utmost accuracy, special care
being taken that no astigmatism, however slight, exists in it. The mirror A
is silvered; B is polished and unsilvered. The mirrors may be set up as shown 7?
plan in Fig. 2, the distance cm + mf being equal to the radius of curvature of A ;
both mirrors hang on edge in steel bands as already described. The light proceeding

26 THE MODERN REFLECTING TELESCOPE.

from the illuminated pinhole strikes 6, is reflected to A, thence back to B,
thence to a focus close beside the illuminated pinhole.

When using the knife-edge test the optician sees the mirror B brilliantly illu-
minated, and in elliptical outline, the horizontal diameter appearing foreshortened
by an amount depending upon the angle at which the mirror is viewed. With the
knife-edge test the surface of 4 is seen in relief, as a whole ; any zonal errors appear
enormously exaggerated, and their character and position are readily determined,
just as when a spherical mirror is tested at its center of curvature; these zonal
errors, of course, appear elliptical, on account of their foreshortening ; their effect is
doubled in intensity on account of the two reflections from L (assuming that the
illumination is as brilliant as the eye requires).

The test, as already described, is all that is necessary for the detection and loca-
tion of zonal errors. But something more is necessary in order to detect general
curvature, @. é, convexity or concavity, in B. Let us assume that the mirror, when
fine-ground and polished, is so nearly flat that no curvature can be detected with a
Brown and Sharpe steel straight-edge of the finest quality; and for convenience in
description let us also assume that the surface is free from zonal errors. Let the
knife-edge be moved across the reflected cone from the left; a focal point is found
at which the right and left sides of the mirror darken simultaneously ; this focal
point we will call f,. Now let the knife-edge be moved across the cone from
above, instead of from the left; a focal point will be found at which the upper and
lower parts of the mirror darken simultaneously ; this focal point we will call /9.
It is only when the mirror B is a perfect plane that 7, and f, coincide with each
other and with the point 7 (see figure). If B isslightly convex, f, and f, are out-
side of f(@. ¢, farther from the mirror than f) and f, is outside of f,. If Bis
slightly concave both f, and f, are inside of f, and 7, is inside of f,. In practice,
the exact position of fis not found (except incidentally when the plane mirror is
finished), for this would involve the very accurate measurement of the large dis-
tance cm + mf. The determination of the positions of f, and /, with reference to
each other is all that is needed.

That 7, and 7, do not coincide when B is convex or concave is due to the fact
that the curvation of & is apparently increased or exaggerated in the direction of
the horizontal diameter of the mirror, on account of its foreshortening in this diree-
tion, as seen from f; while the curvature in the direction of its vertical diameter is
not thus exaggerated. The effect is precisely as if the spherical mirror A were astig-
matic, the parts of the surface adjacent to the horizontal diameter having a different
radius of curvature from those adjacent to the vertical diameter. This effect is so
marked that an extremely small deviation of B from a true plane can be detected.
For example, if A and / are each two feet in diameter, the radius of curvature of
A being fifty feet as before, and if the angle which the line fm subtends with the
surface of B is 45°, a deviation from a true plane of 3999 inch in the surface of
B is readily detected. If the angle of the mirror B be changed to 30°, as shown
in Fig. 3, the accuracy of the test for general curvature is about doubled ; the latter
position, however, is not usually so convenient for determining the positions of zonal

THE MODERN REFLECTING TELESCOPE. 27
errors ; for the greatest refinement, therefore, the stand on which A and B are sup-

ported is so designed that the positions of the mirrors can be quickly changed so as
to give the greatest accuracy in each part of the test.

MIRROR

\ ILLUMINATED

PIN HOLE

PLANE MIRROR
Fic. 3. DIAGRAM ILLUSTRATING TESTING OF A PLANE MIRROR.

The use of an eyepiece in this test is important because it shows how fatal to
good definition is even a very slight convexity or concavity of a plane mirror when
used in oblique positions. If 7, and 7, coincide as closely as can be detected
with the knife-edge test (2 being free from zonal irregularities also) the re-
flected image of the pinhole, as seen in an eyepiece at f, is as exquisitely sharp
and perfect as if it were formed by the spherical mirror A alone. But if B is
slightly convex or concave the appearance of the eyepiece image is similar to that
which has already been described in connection with astigmatic concave mir-
rors; the image is not sharp even at the best focus; if B is convex, the image
becomes elongated in a vertical direction outside, and in a horizontal direction
inside, of the best focus; if B is concave the directions of elongation are the re-
verse of these.

The preparation of grinding tools for plane mirrors is similar to that of tools for
concave mirrors. Three full-size, flat iron tools are usually made, however, all of
which are grooved. These are ground together with carborundum of finer and finer
grades, until all appear flat when tested with a carefully kept Brown and Sharpe
steel straight-edge of best quality.

The plane mirror is fine-ground in the manner described for concave mirrors,
It is of course a rare occurrence to find a large plane mirror nearly optically flat
when it is first tested after grinding and polishing. My large mirrors almost in-
variably come out slightly convex when first polished ; this may be due in part to
the fact that the flat grinding tool becomes very slightly concave during the fine-
grinding of the glass, from being worked on top (see page 7). Slight convexity
of the mirror at this stage of the work is much better than slight concavity, for it
is much better and easier to remove a high center than a high edge, during the
process of figuring with polishing tools.

Manual polishing with full-size tools should be employed when the mirror is
not too large to allow this. The polishing is begun with the normal tool shown
in Fig. 4, in which the grooves are of uniform width throughout. After an hour's
polishing the mirror is tested; if it is found to be convex, polishing is continued
with the concaving tool shown in Fig. 5, in which all of the grooves are gradually
widened toward the edges of the tool, so that there is a progressive decrease of
action toward the edges of the glass; the amount of this widening must be

THE MODERN REFLECTING TELESCOPE.

THE MODERN REFLECTING TELESCOPE. 29

determined by experiment; it should be such that the convexity of the mirror is
slowly and uniformly decreased.

If the mirror, when first tested, is found to be coneave, the convexing tool
shown in Fig. 6 is used to continue the polishing.

The concaving and convexing tools often tend to introduce broad slight zonal
errors ; hence recourse must be had repeatedly to the normal tool. When all trace
of general curvature has disappeared, any remaining zonal errors are eliminated by
the use of the normal tool, and, if necessary, of the small local or figuring tools, (see
page 24),

JBABBBSB8B8B886p

BBSEE8s8 880886)

ee

(I2SBEBESBE88880E8D
JEEG8068888080

Fic. 6. CONVEXING POLISHING TOOL FOR FIGURING PLANE MIRROR,

If a finished plane mirror is available which is not smaller than the one being
figured, the work is very greatly facilitated by continually cold-pressing the polish-
ing tools on the finished mirror; every precaution must be taken, however, to pre-.
vent injury to the figure of the finished mirror by such cold-pressing.

In some of the writer’s early work, in which the thickness of mirrors was made
only one-twelfth of their diameter, it was found that a normal polishing tool, as de-
scribed above, tended to change the mirror very gradually toward a concave. This
was undoubtedly due to the fact that the friction of polishing warmed the surface
very slightly, thus expanding it and making it convex with reference to the polish-
ing tool; the tool did not follow this change of form readily, hence the central parts
of the glass were acted upon in excess, Furthermore, such thin mirrors, when un-
silvered, were so sensitive to slight changes of temperature that the presence of the

30 THE MODERN REFLECTING TELESCOPE.

optician’s body for a period of two or three minutes, at a distance of three feet
from a mirror which was set up for testing, would throw a previously plane mirror
convex by an amount many times greater than the smallest amount which can be
detected by the knife-edge test. When the thickness of mirrors is made equal to
about one-seventh of their diameter, their sensitiveness to all such temperature
effects is very greatly decreased. Furthermore, in the case of silvered glass mirrors
which are used for solar work, the writer has found that thick mirrors suffer very
much less change of figure from exposure to the sun’s heat than thin mirrors do.
Silvering affords a great protection from changes of temperature, since the silver
film furnishes an almost totally reflecting surface for heat radiations.

CHAPTER XIII.
TESTING AND FIGURING PARABOLOIDAL MIRRORS.

Tue work of changing a spherical mirror to a paraboloidal one is accomplished
entirely by the use of polishing tools, by shortening the radii of curvature of the
inner zones, instead of by increasing or lengthening those of the outer zones. The
methods of effecting this change of curvature will be described after the methods
of testing a paraboloid have been discussed.

Such testing can be done at the center of curvature, by determining there the
foci or the radii of curvature of successive zones of the mirror; it may be done at
the focus of the paraboloid, by the aid of a finished plane mirror which should be
at Jeast as large as the paraboloidal one; and it may be done directly on a star.
The first two methods named have the very great advantage that they may be
conducted without interruption, under the practically perfect atmospheric and
temperature conditions of the optical laboratory.

Lesting a Paraboloid at the Center of Curvature. A knowledge of the prop-
erties of the parabola enables the optician to compute the positions of the centers
of curvature of successive, definite, narrow zones of the mirror, and the surface
must be so figured that the radius of curvature of each zone agrees with the com-
puted value. In testing, each zone in succession is exposed by means of a suitable
diaphragm, all of the rest of the surface being covered. In practice, two entirely
different formulas may be used, depending upon the position of the illuminated
pinhole.

Let F be the focal length of a finished paraboloidal mirror, and R the semi-
diameter of any extremely narrow zone or ring of its surface, concentric with the
vertex or center of the mirror; the normals to this zone cross the axis at a point

2

whose distance from the vertex is 2 F+ FP? hence, if the illuminated pinhole be

2
placed very close to the axis, and at a distance of 2 F + in from the vertex, the

rays of light reflected from the narrow zone will form a focus or image in the same

.

THE MODERN REFLECTING TELESCOPE. 31

plane (at right angles to the axis) in which the pinhole itself lies. This is the
simplest formula which can be used, but it is not the most useful in practice.

In testing paraboloids at the center of curvature the writer has always used
the following method and formula: The illuminated pinhole remains fixed at the
center of curvature of the central parts of the mirror, 7.2, at a distance 2 F from
the vertex, where F is the focal length. The intervals, measured along the axis,
between the reflected foci of the various zones, are now twice as great as those
given by the method described in the preceding paragraph ; consequently these
foci can now be determined with twice the accuracy which ean be attained by that
method. Only the rays reflected from the parts of the paraboloid very near to
the vertex are now brought to a focus in the plane of the pinhole. If the para-
boloidal figure is perfect, the rays reflected from any very narrow zone whose semi-

R2 R4

diameter is R are now brought to a focus at a distance —“ + 6 F3 back of the plane

2F
R2

a ,
the pinhole, 2. ¢., at a distance 2 F + oF + — from the vertex of the paraboloid,

eae luSyh, ‘ ; . :
The quantity 16 Fs 8 8° small in the case of mirrors of moderate size and of ordinary

ratios of aperture to focal length that it can be neglected; even in testing the

outermost zones of the 5-foot mirror of 25 feet focal length, this quantity is less
ave :

than 0.002 inch, while the quantity 9 p amounts to 14 inches.

Now let us consider what is the best method of determining the planes of the
reflected foci. Draper, Common, and other workers used an eyepiece for this purpose ;
this serves well for mirrors of moderate angular aperture, but for mirrors in which the
ratio of aperture to focal length is as great as 1 to 5 or 1 to 6 this method presents
serious difficulties; if narrow zones are used the image in the eyepiece is blurred
and indistinct on account of the diffraction effect produced by the edges of the
zonal openings in the diaphragm, while if wide zones are used the difference of
focus of the inner and outer parts of a zone is so great that the image shows
evidence of marked aberration; with neither narrow nor wide zones can the
position of the focus be determined with very great accuracy. ;

In Publications of the A. S. P., vol. xiv., No. 87, Hussey gives a formula for
the position of the “circle of least confusion” when a zone of given width is used ;
if Hussey’s formula were employed and the pinhole were made very small and
round, with smooth edges, it is probable that much greater accuracy could be
attained than by the use of an eyepiece in the ordinary way.

The method of locating the reflected foci which is used by the writer is as
follows; it is capable of surprising accuracy when the optician has become experi-
enced in its use. The reflected focus of a zone is found with the knife-edge, pre-
cisely as the focus of a spherical mirror is found. The knife-edge is moved across
the reflected cone from the left; if the left side of the zone is seen to darken first,
the knife-edge is inside of the focus; if the right side darkens first, the knife-edge
is outside of the focus; when the right and left sides of the zone darken simul-

32 THE MODERN REFLECTING TELESCOPE,

taneously, the knife-edge is at the focus of the zone. One advantage of this
method is that it is independent of changes of focus of the eye itself; but the
great advantage is that very narrow zones or arcs can be used. Diaphragms
with zonal openings 1 of an inch wide serve admirably for mirrors of 10 or 15
feet focal length; indeed the width of the zones which are actually used is con-
siderably less than this; for, on account of diffraction, the edges of the openings
in the diaphragms always appear as brilliant lines, even while the illumination
near the center of the openings is being cut off by the knife-edge; it is therefore
only the illumination near the center that is used in making the comparison.

The diaphragms which I use in this method of testing do not expose entire
zones, but only pairs of ares on the right and left sides of the mirror. Fig. 7 shows
the diaphragm which was used in testing in this way the mirror of the two-foot
reflector of the Yerkes Observatory. The ares are cut in a long and narrow strip
of thin metal; this is attached to the inner edges of two wooden strips, a; these

x

I

WH

1:

Fic. 7. DIAPHRAGM USED IN TESTING A PARABOLOIDAL MIRROR AT I'S CENTER
OF CURVATURE,

|

edges are curved so that all parts of the thin metal diaphragm are nearly in
contact with the curved surface of the mirror. The edges of the openings are
bevelled so as to be extremely thin, and are finished dead-black. Twelve pairs
of ares were used, with mean radii of 1, 2,3, . . . 10, 11, and 11% inches.
The openings of these arcs are + inch in width. The foci of the successive zones
(except those near the center) can be readily determined by this means to within

THE MODERN REFLECTING TELESCOPE. 33

soo Inch along the axis, for a mirror of two feet aperture and of ten or fifteen feet
focal length.

Care must be taken when testing in this way that the entire mirror surface is
uniformly illuminated by the cone of light proceeding from the illuminated pin-
hole; this condition, once secured, is easily maintained, since the illuminated pinhole
remains immovable.

I have described at considerable length the methods of testing paraboloids at
the center of curvature, because of the importance of the subject, and because this
will probably continue to be a favorite method, especially among amateurs. But
when testing is done at the center of curvature, even with the extremely accurate
method just described, the making of a large paraboloidal mirror of great angular
aperture and really fine figure is an exceedingly difficult task. This is due in part to
the necessity of very frequent tests, in each of which the foci of a large number of
zones must be determined ; it is due far more to the uncertainty in determining the
exact nature of errors of surface (considering the surface as a whole) corresponding
to focal readings which do not agree with the computed values. In the case of
mirrors of small or moderate angular aperture, much important information can be
gained by viewing the surface as a whole, from the (mean) center of curvature, by
means of the knife-edge test; a finished paraboloid, when thus seen, appears to
stand out in relief, in strong light and shade, as a surface of revolution whose see-

tion is that shown in Fig. 8; knife-edge and pinhole are both at the center of curva-
ture of the zone a, the apparent curve of the surface should be a smooth one.
But in the case of a mirror of large angular aperture the change of curvature is so
rapid that only a narrow zone ean be seen well at one time, 7. ¢@., with a given focal
setting of the knife-edge.

Testing a Paraboloid at its Focus. ‘This method was briefly described by the
writer in the Astrophysical Journal, November, 1901. It is ineomparably more
simple, direct, and rigorous than the test at the center of curvature. A well-figured
plane mirror, which should not be smaller than the paraboloidal one, is necessary
in order that the testing may be done in the optical laboratory. In practice a small
diagonal plane mirror is also used, to avoid the necessity of a central hole through
the large plane mirror. Both of the plane mirrors are silvered. The arrangement
of mirrors is shown in Fig. 9. The diagonal prism is placed at f, with the illumin-
ated pinhole very near the axis; pinhole and knife-edge are in the same plane, at a
distance from the vertex equal to em + m/f, which is equal to the focal length of
the mirror. The paraboloid is now tested as a whole, without the use of zones,
precisely as a spherical mirror is tested at its center of curvature.

34 THE MODERN REFLECTING TELESCOPE.

PARABOLO!I DAL PLANE

ES

Fic. 9. TESTING A PARABOLOIDAL MIRROR AT ITS Focus.

If F be the desired focal length of the paraboloidal mirror whose semi-diameter
is R, then the spherical surface which is fine-ground and fully polished preparatory
- R? oy
to parabolizing should have a radius of curvature of 2F + ro This is because
parabolizing is done by shortening the radii of curvature of all the inner zones of a
mirror, leaving the outermost zone unchanged, as shown in Fig. 10; this is a far
easier and better method in practice than to leave the central parts of the mirror

SPHERE

PARABOLOID

FIG. 10, FIG. 11.

unchanged, and to lengthen the radii of curvature of all of the outer zones, as shown

in Fig. 11.
Let us now suppose that the concave mirror shown in Fig. 9 is a spherical one
rr 9

where R is the semi-diameter, and F is the

with radius of curvature 2F + es
distance cm + m/f, from the center of the mirror surface to the plane of the pin-
hole and knife-edge. If the spherical surface be now viewed from the point 7 with

the knife-edge test, it will appear to stand out in relief, in strong light and shade,

MI

FIG. 12,

as a surface of revolution whose section is that shown in Fig. 12, the height of the
protuberant center depending upon the angular aperture of the mirror. The reason
for this appearance is readily seen by reference to Fig. 10. To change the spherical
surface to a paraboloid, the protuberant center must be removed by the use of
suitable polishing tools, until the surface, as seen with the knife-edge test from the
point /, appears perfectly flat, 7. ¢., the illuminated surface darkens with perfect

nai

THE MODERN REFLECTING 'TELESCOPE. 35

uniformity all over. As the paraboloidal surface nears completion, an elevated or
depressed center, a “turned up” or “turned down” edge, or protuberant or de-
pressed zones, can be seen and their character and exact position determined, with
precisely the same ease and certainty with which similar irregularities are seen when
a spherical mirror is examined at its center of curvation with the knife-edge test.

It should be noticed that even when the pinhole and reflected image are very
near each other, as they should be, yet both may be far out of the axis of the para-
boloid, if the mirrors are not properly adjusted or collimated ; when this is the ease
the mirror surface, when seen with the knife-edge test, does not appear asa surface
of revolution, and cannot be properly tested. The mirrors may be collimated by
the following method, thus insuring that the pinhole and reflected image are both
extremely near the optical axis.

The mirrors are set up approximately right by measurement. A ring about an
inch in diameter, with two fine threads stretched diametrically across it, one verti-
eal, one horizontal, is set up near the plane of the illuminated pinhole, the intersec-
tion of the threads marking the desired position of the optical axis. A light,
stiff ring is made, which fits closely over the edge of the paraboloidal mirror, at the
front; this ring can be slipped on and taken off as required. Two very fine bright
wires are stretched diametrically across this ring, one vertical, one horizontal ; these
wires should be as close as possible to the face of the mirror; their intersection
marks the position of the center or vertex of the paraboloid. Two fine short lines,
one vertical, one horizontal, are scratched with a fiue needle-point at the center of
the silvered face of the small diagonal plane mirror. ‘The eye is now placed about
3 feet outside of the plane of the crossed threads, and an assistant changes the in-
clination of the small plane mirror, by means of three adjusting-screws at its back,
until the intersections of the threads, of the scratches, and of the wires are all
seen in exact coincidence. The assistant next changes the inclination of the para-
boloidal mirror (by means of three adjusting-screws at its back) until, with the
eye in the same position as before, the intersection of the threads, the intersection
of the wires, and the reflection of the intersection of the threads seen in the para-
boloidal mirror, all appear in exact coincidence; the position of the axis of the
paraboloid has now been defined. No attention is paid to the large plane mirror
in this part of the work. The illuminated pinhole is now placed in position, and
the large plane mirror is adjusted (by means of three adjusting-screws at its back)
until the reflected image falls in the right position with reference to the axis and
pinhole.

The frame which carries the paraboloidal mirror can easily be so designed that
this mirror can be removed and replaced repeatedly, while figuring it, without
sensibly disturbing the adjustments.

The difficulties of making short-focus paraboloidal mirrors of fine figure are
so greatly reduced when this method of testing is used that I believe that the gen-
eral adoption of this method by opticians would lead to such improvements in results
as to bring about a marked advance in the usefulness of reflecting telescopes. The
making of the large plane mirror which is necessary in this test becomes so simple

36 THE MODERN REFLECTING TELESCOPE.

and certain when the methods of testing and figuring described in the preceding
chapter are used, that I have no hesitation in saying that when a large paraboloidal
mirror of short focus and of the finest attainable figure is to be made, it is economi-
cal to make a plane mirror of the same size, with which to test it, if one is not
already available. The concave mirror is first figured spherical and is used thus for
testing the plane mirror while the latter is being figured; the plane mirror is then
used in testing the concave one during the parabolizing of the latter. Both the
plane and paraboloidal mirrors are then used in testing the small (convex)
hyperboloidal mirror while the latter is being figured.

Testing a Paraboloid on a Star. With this method the mirror surface, as seen
with the knife-edge test, presents the same general appearance as in testing in con-
junction with a large plane mirror; in the latter test, however, errors of surface are

Fic. 13. FULL-S1zE POLISHING TOOL FOR PARABOLIZING.

seen in greater relief, because the effect of such errors is doubled on account of the
two reflections from the paraboloid. In addition, it is impossible to overestimate the
advantage of being able to test as often as is desired, in the optical laboratory,
where atmospheric and temperature conditions can be controlled perfectly, and
where the mirror does not have to be removed from the polishing machine in order
to test it. In testing on a star it is seldom indeed that atmospheric conditions are
sufficiently fine to allow any except the larger errors of surface to be seen.
Changing a Spherical Surface to a Paraboloid. As before stated, this is ac-
complished by shortening the radii of curvature of all of the inner zones of the sur-

THE MODERN REFLECTING TELESCOPE. 37

face, leaving the outermost zone unchanged (see Fig. 10). There are two distinct
methods of accomplishing this: (1) by the use of full-size polishing tools, the rosin
surfaces of which are cut away in such a manner as to give a large excess of polish-
ing surface near the central parts of the tool; (2) by the use of small polishing or
figuring tools worked chiefly upon the central parts of the mirror, and Jess and less
upon the zones toward the edge.

(1) Parabolizing with Full-Size Tools. The rosin surface can be trimmed
in a variety of ways to give a great excess of action on the central parts of the
mirror. Fig. 13 shows one of the best forms of tool for this purpose, the shaded
parts representing the rosin surface, coated with wax. The form of the edges of
the rosin-covered areas can be altered as desired, and thus the amount of action on
any zone can be in some measure controlled. Length of stroke and amount of side-
throw are also very important factors in controlling the figure of the mirror. Tools
of this kind serve admirably in parabolizing mirrors up to 36 or 40 inches in diam-
eter, when the angular aperture is not very great.

(2) Parabolizing with One-Third-Size and Smaller Tools. In the case of
very large mirrors, when full-size tools are almost unmanageably heavy, and in the
ease of mirrors of great angular aperture, in which the departure from a spherical
surface is great and is effected with difficulty with full-size tools, one-third-size and
smaller figuring tools may be used. The machine should invariably be employed
in this work, the transverse slide being used to place the tool in succession upon
the various zones. In order to preserve the surface of revolution the setting of the
transverse slide should be changed only at the end of one or more complete revolu-
tions of the glass. The rosin squares of the small tools should be somewhat softer
than usual, so that the surfaces of the tools can accommodate themselves slowly to
the slightly different curvatures of the successive zones. The squares around the
edges of the tools should be trimmed, as before described, in order to soften the
action of the edges. The mirror should be tested very often, and the utmost care
taken to keep the apparent curve of the surface, as seen with the knife-edge test, a
smooth one, @. ¢., free from small zonal irregularities, at all stages of the parabol-
izing; this is not extremely difficult when the optician has become experienced in
the use of the transverse slide.

The mirror of the 2-foot reflector of the Yerkes Observatory, which has a focal
length of only 93 inches, was parabolized in this way by the writer. Two small
tools were used, of 6 and 8 inches diameter respectively. The actual difference of
depth, at the center or vertex of this mirror, between the paraboloid and the nearest
spherical surface is almost exactly 0.0004 inch. This difference is unusually large
in this case, on account of the exceptionally great ratio of aperture to focal length.
This difference varies, in different mirrors, as the fourth power of the diameter of
the mirrors, and inversely as the cube of the focal length. In the ease of Lord
Rosse’s great mirror, in which the aperture is 6 feet and the focal length 54 feet
(ratio 1 to 9) the corresponding difference at the center is only 0.0001 inch, very
nearly. In the case of the 5-foot mirror of the Yerkes Observatory, of 25 feet focal
length, the corresponding difference is about 0.0006 inch. This gives some idea of

38 THE MODERN REFLECTING TELESCOPE.

the actual amount of glass which must be removed by the figuring tools in
parabolizing.

CHAPTER XIV.
TESTING AND FIGURING CONVEX HYPERBOLOIDAL MIRRORS.

Tue methods of figuring and rigorously testing convex hyperboloidal mirrors
are now so thoroughly developed that the reflecting telescope can be regarded as a
universal photographie telescope of the highest class, capable of giving, at the
focus of the paraboloidal mirror of large angular aperture, the finest photographs
now attainable of large and excessively faint objects such as the nebule in general ;
while by the addition of a small convex mirror a great equivalent focal length
is obtained for the photography of bright celestial objects requiring large scale,
such as the moon, the planets, the dense globular star clusters, and the annular
and planetary nebule. The convex mirror of course serves as an amplifier, and
possesses the great advantages over a lens used for this purpose that the perfect
achromatism and the high photographie efficiency of the reflector are retained, and
that the mechanical arrangements are very compact and economical. In order
to give perfect definition the convex mirror must be an hyperboloidal one.

The writer has recently made two convex mirrors of different curvature, for
use with the 2-foot reflector. These give equivalent focal lengths of 27 and 38 feet
respectively.

Fig. 14 shows the arrangement of mirrors employed in the 2-foot reflector
when used as a Cassegrain; a small diagonal plane mirror is used at m, to avoid
the necessity of a hole through the center of the large concave mirror. / is the
paraboloidal mirror, with its focus at f; is the hyperboloidal mirror, the secondary

SS a eee.

So

a ae

FIG. 14.

focus or magnified image produced by the combination being at #’; the point ¢ is the
center of the hyperboloidal surface. Calling the distance fc = p and the distance

cm +m =p’, then P represents the amount of amplification introduced by the

convex mirror. The radius of curvature R of the spherical surface to which the
convex mirror is ground and polished preparatory to hyperbolizing is found with

2
sufficient accuracy for all practical purposes by the formula 5 — = R whence
Rae

pp

THE MODERN REFLECTING TELESCOPE. 39
For example, let the focal length of the paraboloidal mirror P, Fig. 14, be ten
feet; let jo= p = 2 ft. and em + mF = p'=8 ft. Here = 4; the image of

the moon or other celestial object produced at /’ is therefore four times larger in

‘ ; ; : 2 pp’
diameter than it would be at 7, the focus of the paraboloid ; and R = PP — 64

/
: p'—p
inches.
The method of testing the convex mirror while hyperbolizing it is shown

in Fig. 15. The illuminated pinhole is placed very near the axis at /. The diverg-

PARABOLO!I DAL PLANE

MIRROR F* Focus

Fic, 15. DIAGRAM ILLUSTRALING TESLING OF HYPERBOLOLDAL MIRROR.

ing cone of light strikes the small plane mirror, then the convex, then the large
paraboloid, whence if all of the mirrors are finished and are well adjusted or colli-
mated, the light is reflected in a parallel beam to the large plane ; returning, the rays
are brought to a focus very near the axis of figure and in the plane of the illuminated
pinhole. All of the mirrors except the convex one are silvered. The convex spheri-
eal surface with radius of curvature R, as above described, when viewed with the
knife-edge test from the point /, presents the same general appearance of a smoothly
curved surface of revolution, in strong light and shade, which a paraboloidal surface
presents when similarly viewed from its center of curvature (see Fig. 8, p. 33).
All that is necessary to produce the hyperboloidal surface is to soften down, with
suitable polishing tools, the apparent broad protuberant zone between the center
and edge, until the mirror, as seen from / appears perfectly flat; 7 @, until the
illuminated surface is seen to darken with absolute uniformity all over when the
knife-edge is moved across the focus. This hyperbolizing may be done with small
local or figuring tools, or with a full-size tool so trimmed as to give an excess
of action on the broad zone a, or (what is usually best) by a combination of
the use of both kinds of tools.

As in the case of the paraboloid, it is necessary in this test that all of the
mirrors be lined up or collimated with care; otherwise the surface of the convex
mirror will not appear as a surface of revolution, and cannot be properly tested.
The axes of the paraboloid and hyperboloid must coincide, and the face of the
large plane mirror must be at right angles to these axes. These adjustments are
made by means of an extension of the method of collimation described in the pre-
ceding chapter, p. 35. First the paraboloidal mirror is adjusted so that its axis
intersects the hyperboloid at its exact center or vertex; in making this adjustment
fine threads are stretched diametrically across the cell of the convex mirror, this

40 THE MODERN REFLECTING TELESCOPE.

mirror being removed during this part of the adjustment. Next, the small diagonal
plane is adjusted for inclination, care being taken that the intersection of the lines
scratched in its film is placed in the axis of the paraboloid. Then the convex
mirror is adjusted for inclination, by reflection. Finally, with the illuminated pin-
hole in place, the large plane mirror is adjusted, as previously described.

CHAPTER XV.
SILVERING,

Ir is not my purpose to discuss the various processes of silvering. Several
methods have been admirably described by Draper (see p. 2 of his book), by
Brashear, and by Common (see p. 159 of his paper On the Construction of a Five-
Foot Reflecting Telescope). 1 have used almost exclusively the formula published
by Brashear in 1884, in which sugar is the reducing agent. After experience with
this process, and when the grades of chemicals specified below are used, silver
films are invariably obtained which take a perfectly black polish, and which are so
thick as to be nearly opaque even tothe sun’s disk. Small mirrors are usually
silvered face down ; films which are satisfactory in all respects are obtained when
this is done.

In the case of large mirrors it is more economical of silver, as well-as safer and
more convenient in manipulation, to silver face up. Two difficulties occur, how-
ever, when this is done ; first, minute transparent spots are liable to occur in the
film; these are so small, however, that they can be seen only when looking
through the film at a bright object; second, the refuse silvering solutions must be
poured off the mirror, after the silver has been deposited, at exactly the right
stage of the reaction; if poured off too soon the film will be thin ; if too late, the
muddy-brown precipitate which settles upon the film will slightly tarnish the lat-
ter in such a manner that it will not take a perfect polish; it is only by ex-
perience that the optician is able to determine the right instant for pouring off the
refuse solutions. Mr. Common encountered similar difficulties in silvering face up,
and resorted to the use of solutions without caustic potash, and also to the use of
Draper’s method of reducing with Rochelle salt; these methods, while subject to
their own special difficulties, do not give the objectionable precipitate. The
writer has adhered to the use of a slight modification of Brashear’s formula already
mentioned, in part because no opportunity has occurred for comparing thoroughly
the merits of the various formule, and in part because the films obtained by this
method give entire satisfaction in use.

The Reducing Solution. This consists of distilled water, 200 parts; loaf-
sugar or pure rock-eandy, 20 parts; aleohol (pure) 20 parts; nitric acid (e. p.) 1
part. The proportions given are by weight. This solution is greatly improved by
keeping, a solution which has been made for several months working more surely
than one newly made. A gallon of this solution is usually made at one time.

THE MODERN REFLECTING TELESCOPE. 41

The operation of silvering a 2-foot mirror face up will now be described.
Tt will be assumed for the present that the back of the mirror is unsilvered. A
silvering table is used, which is a strong structure of oak wood having a tilting
frame carried on two trunnions, so that the mirron can be quickly turned from a
horizontal to a vertical position, for the purpose of pouring off the cleaning and
silvering solutions; a strong narrow edge-band of flexible steel prevents the mirror
from sliding off; the tilting frame is heavily weighted below so that it cannot
turn down accidentally. Thus all handling of the mirror while silvering is
avoided,

The old silver film, if one exists, is removed with strong nitric acid on a
bunch of absorbent cotton tied to a glass rod. The face and edge of the mirror
are then quickly washed with distilled water. A band of strong brown drawing
paper, which has been dipped in melted paraftin, is drawn around the edge of the
glass and tightly bound to it by means of a thin band of copper with tightening
screws; the paper should project about three inches above the glass; the joints
should all be made water-tight by means of more paraffin and a warm iron. A
dish about three inches deep is thus formed, with the mirror as its bottom,

A 10 per cent solution of pure caustic potash in distilled water is now used
for thoroughly washing the face of the glass and the inside of the paraffin band ;
this is done with a large bunch of absorbent cotton tied to a glass rod. This
‘solution is then poured out and the glass is similarly washed several times with
fresh supplies of distilled water, to get rid of all traces of potash. Enough dis-
tilled water is now poured on the glass to entirely cover it while the silvering
solutions are being mixed.

All of the vessels, graduates, ete., used for mixing the silvering solutions,
must be thoroughly washed, first with nitric acid, then with caustic potash, and
rinsed with distilled water, just as the mirror is cleaned.

For silvering the face of a 2-foot mirror, 2 ounces of silver nitrate
(Powers & Weightman) are dissolved in 20 ounces of distilled water. One and
one-third ounces of caustic potash, pure by alcohol (Merck), are dissolved in
20 ounces of water in a separate vessel, and the solution is cooled. Strong aqua
ammonia (pure) is added, drop by drop, to the nitrate solution, while the liquid is
thoroughly stirred; the mixture turns light-brown, then dark-brown; the am-
monia is slowly added until the liquid becomes clear. ‘The caustic potash solution
is now added slowly, with thorough stirring ; the mixture now becomes very dark-
brown or black. Ammonia is again added, with thorough stirring, until the
liquid again just clears. A solution of one-fourth ounce silver nitrate in 16 ounces
of distilled water having been prepared, this is added to the mixture, a few drops at
a time, with thorough stirring, until the entire solution has a decided straw color,
while remaining transparent. This straw color is the test for the condition of in-
stability which is absolutely necessary in order that the metallic silver shall be
thrown out of combination when the reducing solution is added later. The
solution is now thoroughly filtered through absorbent cotton.

A quantity of reducing solution is taken containing an amount of sugar equal

42 THE MODERN REFLECTING TELESCOPE.

in weight to one-half that of the entire amount of silver nitrate used; this is also
filtered. The silver solution and reducing solution are now both diluted with
distilled water, preparatory to mixing; the quantity of the diluted solutions,
together, should be sufficient to cover the glass about one inch deep.

An assistant pours off the water which has stood on the glass, while the op-
tician quickly mixes the dilute silver and reducing solutions in a large pitcher or
granite-ware bucket. The glass being horizontal, the mixed solution is immediately
poured on, and the mirror is rocked slightly by means of the tilting frame. The
liquid quickly changes to a transparent light-brown color, then dark brown, then
black, after which the silver immediately begins to deposit. The solution gradually
changes to a muddy-brown color, and in three or four minutes after the solutions
are poured on the glass, begins to clear; the light muddy-brown precipitate
settling upon the film. With the proportions given, the silver film should be
sufficiently thick in about five minutes after the solutions are poured on the glass
provided that the room, glass, and solutions are allat a temperature of sixty-eight
degrees or seventy degrees Fahrenheit. When first formed the brown pre-
cipitate is so light that it moves about with the rocking of the glass; but it very
soon deposits in large areas on the film. As soon as this begins to occur, the solu-
tion must be very quickly poured off the glass, an abundance of distilled water
poured on, and a large bunch of absorbent cotton, held in the fingers, instantly
used to displace all streaks of the precipitate which adhere to the film. The film
is now washed again and again with fresh distilled water and a soft bunch of cot-
ton; then an abundance of water is poured on and the film allowed to soak for an
hour. When this is poured off, the paper band is carefully removed, with the glass
horizontal so that no liquid from the edge can run upon the silver film; this must
be done quickly, before the latter has time to dry. A small amount of alcohol is
now flowed on the film; this is repeated several times to get rid of all water; the
glass is then turned on edge, and is quickly dried with a fan.

After standing for an hour or two in a dry room the film is to be burnished.
A soft pad as large as the hand is made of the softest chamois skin; this is used on
the film without rouge, with light circular strokes, to condense the silver. After
two hours of this work a little of the finest washed dry jeweler’s rouge is rubbed
into the chamois-skin with a piece of clean absorbent cotton; from thirty to sixty
minutes use of the pad with the same stroke as before should now bring the film to
a perfect polish, without scratches.

If the back of the mirror is already silvered, the face can be silvered by the
method just described, without injuring the film on the back; the mirror now rests
upon three curved and beveled blocks of soft wood which touch only the rounded
corner or edge of the back of the glass; extra precautions are now taken to prevent
any of the solution from touching the back. I regard this method as much better
in the case of large mirrors than to attempt to silver both back and face at the same
time in a deep tray ; in the latter method the difficulties of handling and properly
cleaning the mirror are almost insurmountable.

The back of the mirror does not usually need silvering oftener than once in

THE MODERN REFLECTING TELESCOPE. 43

three or four years. The face is usually silvered two or three times a year, to keep
it in the finest condition for photography, in which any yellowing of the film is
very objectionable.

CHAPTER XVI.
A SUPPORT-SYSTEM FOR LARGE MIRRORS.

Tue proper support of mirrors in their cells when in use in the telescope is a
matter of vital importance. Small mirrors can be made very thick and can be sup-
ported at their edges as a lens is supported; the cell must be so designed that no
sensible change of position of the mirror in its cell can occur, The necessity of
supporting large mirrors in such a manner as to prevent flexure from their own
weight, in all positions which can occur in use, has long been recognized, and elab-
orate support-systems for this purpose have been devised and used by Rosse, Grubb,
Common, and others. Comparatively little attention has been given, however, to
two additional requirements which are no less important; first, the position of the
mirrors in their cells should be defined with the greatest attainable stability, in
order to secure permanence of adjustment or collimation; second, the method of
support should be such that the silvered back of the mirror is exposed to the air as
freely as possible. It is assumed that a large mirror need never be turned farther
than ninety degrees from the position in which it lies horizontal upon its back.

In the Astrophysical Journal for February, 1897, the writer described a
method of supporting large mirrors which fulfills all of the requirements named in
the preceding paragraph. I have employed this method in the designs for the
support-system of the 5-foot mirror. These designs are described and illustrated
here.

I1.—The Back-Support.

Let us consider the mirror to be divided into twelve imaginary segments
of equal weight, as shown in Fig. 16, Plate vin. The back of the mirror rests, pri-
marily, upon three strong bronze plates, each ten inches in diameter, represented by
the double circles a Fig. 16 and at a Fig. 17, the center of each plate being exactly
behind the center of weight of the corresponding segments; these are called the
stationary plates. The upper surface of each plate is flat and is ground to fit the
flat back of the glass; the lower surface is spherical, and is ground to fit the large
spherical socket in which it rests. It will be noticed that these plates are near the
edge of the mirror, in the outer ring of segments; the base of stable support is
therefore large. It is evident that by properly designing these plates and their sup-
ports we can fix with very great stability the plane of the mirror which rests directly
upon them; there is no building out from the three primary points of support by
means of intermediate levers and triangles, as in the older systems.

The weight of the remaining nine segments of the mirror is just balanced by
means of nine weighted levers, each of which is entirely independent of every

44 THE MODERN REFLECTING TELESCOPE.

other, which lie in a plane parallel to the back of the mirror. One of these levers
is shown in elevation at c, Fig. 17, and in plan in Fig. 21. The positions of the
nine levers are indicated by dotted crosses in Figs. 16 and 18. These levers are
suspended between pivots screwed through lugs connected to the cell. The cone
bearings, shown in Fig, 21, are finely fitted, and are ground to reduce friction. The
long arms of the levers carry adjustable lead weights (d Figs. 17 and 21) which are
made in the form of plates, in order that they may occupy as little space as possible
perpendicular to the plane of the mirror ; the short arms of the levers are thus made
to press against the backs of the corresponding segments through the medium of
light plates of bronze represented by the single circles 4 Fig. 16 and at 6 Figs. 17
and 21.

The large mirror weighs very nearly 2000 pounds, so that each segment
weighs 1663 pounds. With the cell in a horizontal position the lead weight on
each arm is adjusted until it just balances a standard weight of 1662 pounds placed
upon the plate on the short arm. This adjustment being completed the mirror is
laid upon the support-system ; three-quarters of its weight is earried by the nine
levers, leaving one quarter to be divided equally between the three heavy plates a.
Thus each of the twelve segments is entirely supported at the back, independently
of all of the other segments. Now suppose that the edge-support, which will be
described below, be introduced, and the entire system, with the glass, inclined in
any direction and at any.angle; all of the levers and weights retain the same posi-
tion as before with reference to the glass, but they do not exert the same pressure,
on account of the inclination ; so far as the back-support is concerned there will
still be a perfect balance maintained in the case of each segment; this is true what-
ever point of the edge of the mirror becomes lowest—. ¢., in whatever direction
the levers lie with respect to the vertical plane through the axis of figure of
the mirror.

It should be noticed that in the case of each of the twelve 10-inch supporting
plates only a ring one inch wide around the edge is in contact with the glass; the
part of each plate inside of this ring consists of deep, narrow arms, which do not
touch the glass, and which allows free access of air to the latter.

For very large or thin mirrors a larger number of plates and levers can
of course be used. An incidental advantage which occurs when this is done
is that the base of stable support afforded by the three stationary plates is still
larger, compared with the size of the mirror, than when twelve plates are used.

I.—TVhe Edge-Support.

The relation between the back-support and edge-support is so intimate that any
inefficiency in the latter must injuriously affect the operation of the former, how-
ever perfect that may be in itself. In an equatorial reflecting telescope, different
parts of the edge of the mirror become successively lowest, as the position of the
telescope changes. With the flexible band and cushioned edge-support so much
used in the past, the heavy mirror necessarily changes its position, laterally, with

PLaTe VIII

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—RITCHEY

SUPPORT SYSTEM FOR LARGE MIRROR.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—RITCHEY PLaTE IX

Y SS 4 reer

SUPPORT SYSTEM FOR LARGE MIRROR.

THE MODERN REFLECTING TELESCOPE. 45

respect to its cell, in taking its position down against the edge-support; thus not
only is permanence of position lost, but this tendency to lateral shift must impair
the freedom of operation of the back-support system.

In the present plan four metal ares are used which rigorously define the posi-
tion of the mirror laterally. Two of these ares (¢ Figs. 16 and 17, and Fig. 19),
adjacent to each other, are bolted down to the cell, and their inner edges are scraped
to fit the ground edge of the glass; these are called the stationary ares; the other
two ares (7 Fig. 16 and Fig. 20), diametrically opposite the stationary ones, exert a
slight pressure against the edge of the mirror, by means of springs, for the purpose
of seating the mirror against the stationary ares and holding it there; this pressure
need amount to only a very small percentage of the mirror’s weight, for all
of the lateral pressure due to the weight of the mirror when the latter is inclined
is carried by a strong metal cownterpoising ring of T section (g Figs. 16 and 17) ;
this completely encircles the edge of the mirror, and fits it loosely, a band of leather
or thick felt paper being inserted between the ring and the glass. For convenience
in description, imagine this ring to be suspended from the tube above, by means of
three short wires, so that if the mirror were removed the ring could swing freely in
its own plane. The ring is pressed up against the edge of the mirror, when the
latter is inclined, by a system of twelve short weighted levers (A Figs. 16 and 17)
which hang perpendicular to the plane of the ring. ‘These levers are suspended
from the cell-plate behind the ring, by means of ball-and-socket joints, as shown
in Fig. 17, or preferably, to reduce friction, on pivoted universal or Hooke’s
joints. The ends of the short upper arms of these levers fit loosely into holes
in the ring; the long lower arms carry lead weights (¢ Figs. 16 and 17) which are
capable of slight adjustment.

Assuming that the counterpoising ring weighs 400 pounds, so that the com-
bined weight of ring and mirror is 2400 pounds, the adjustment of the edge-
support levers is effected by turning the entire cell to a vertical plane, with the
mirror and ring removed, and adjusting each of the twelve lead weights until it
just balances a standard weight of 200 pounds hung on the short arm of the lever
at the point where this is to touch the ring.

I regard the use of a support-system which will fulfill all of the conditions
mentioned at the beginning of this chapter as absolutely essential for large mirrors.
Only those who have tested large mirrors and combinations of mirrors in the opti-
cal shop, and those who have actually used large reflecting telescopes, can fully
appreciate the necessity of a support-system which will both support the mirrors
without constraint and flexure, and define their positions permanently with respect
to the tube and axes, in all positions of the telescope. These conditions can now
be attained easily and economically ; without them it is folly on the one hand to
expect good definition and successful photographs, or, on the other hand, to com-
plain that the reflecting telescope is subject to serious inherent difficulties which
eannot be overcome. In the case of large mirrors in which the ratio of thickness
to diameter is not less than as 1 to 9 or 1 to 10 the support-system just described
floats the mirror so perfectly in all positions which can occur in actual use that no

46 THE MODERN REFLECTING TELESCOPE.

flexure or distortion can be detected with the most sensitive optical tests. Fur-
thermore, with the method of edge-support described, and in the case of the 5-
foot mirror weighing a ton, no lateral shift amounting to 9455 inch can occur when
the mirror is turned in extreme oblique positions.

In Figs. 17 and 18 is shown the massive cell-plate of cast-iron which carries
the mirror and its support-system, and which is connected to the short cast-iron sec-
tion of the tube; this connection is made by means of strong adjusting screws, by
means of which the mirror and its support-system, as a whole, are adjusted for
collimating the mirror; these adjusting screws are shown at &, Fig. 18. Additional
screws are also shown at 7 in this figure; these are backed out of the way when
collimating is being done; when this is finished they are brought into position,
and assist in bolting the cell-plate rigidly to the tube. As is shown in Figs, 17
and 18, the central part of the cell-plate, a circle about 50 inches in diameter, con-
sists of open ribs or arms which allow free access of air to the silvered back of the
mirror.

When the face of the mirror is to be resilvered, the cell-plate, support-system,
and mirror are removed as a whole, and silvering is done in the manner described
in the preceding chapter, without taking the mirror from its supports or disturbing
the adjustments of the latter in any way. Furthermore, the mirror can be taken
out of the telescope in this way, silvered, and replaced, without sensibly disturbing
its collimation or the position of the focal plane. When the back of the mirror
must be resilvered, which need not be done oftener than once in three or four
years, the glass must of course be removed from its support-system.

This support-system, as described, may appear complicated and expensive; in
reality it is not so, for all of the levers, plates, etc., used for the back-support can
be exactly alike, as can also the levers used for edge-support ; even when a greater
number of levers than twelve are used the construction is simple and economical.

In Plate x is shown a 30-inch plane mirror supported at the back by twelve
plates and nine levers as described above; the mirror is shown unsilvered, so that
the plates are seen through 4 inches of glass. This is a part of the 30-inch ccelostat
recently constructed from the writer’s designs in the instrument and optical shops
of the Yerkes Observatory.

CHAPTER XVIL
A MOUNTING FOR A LARGE REFLECTING TELESCOPE.

In considering the requirements for a modern reflector mounting for photo-
graphic and spectroscopic work, the writer can probably not do better than to
describe the designs for the proposed mounting of the 5-foot reflector. These
designs are the result of experience both in optical work and in the use of the 2-
foot reflector and the 40-inch refractor of the Yerkes observatory in astronomical

photography.

‘SSVIO GUNGATISNN AHL HONOYHL NAAS ANV LYOddNSMOVA YOU SYAAAT AGNV SALW Id
“WOUMIN ANVId HONI-06 HLIM LVLSOTHOOD AOUVT

X 3LV1d A3HOLIY — SD037MONH OL SNOILNEINLNOD NVINOSHLING

THE MODERN REFLECTING TELESCOPE. 47

With the present great improvements in the materials and methods of machine
construction there is no longer any excuse for unstable and inconvenient mountings
for reflectors. The focal length of modern reflectors intended for photography is
short; the ratio of aperture to focal length generally used in such instruments will
probably be not greater than as 1 to 4, nor less than as 1 to 6; with such
ratios the mounting can be made extremely compact and rigid. By the addition of
a small convex mirror the equivalent focal length can be increased from three and
one-half to five times, and fine definition retained ; when this is done the actual
length of the tube is less than when the telescope is used at the primary focus.

The reflecting telescope defines well only at or near the optical axis; hence the
mirrors must remain in perfect adjustment with reference to each other and to the
eyepiece or photographic plate, in all positions of the telescope which ean occur in
use. Not only must the mirror supports be such as to define the position of the
mirrors rigorously always, as described in the preceding chapter, but the short
tube must be excessively strong and rigid so that no sensible flexure can occur,
This is especially necessary when the telescope is used as a Cassegrain, or as a coude ;
for when these forms are employed it is only when the axes of the paraboloid and
hyperboloid coincide that fine definition can be secured. When the necessity of
these conditions is fully realized by makers and users of reflectors, a marked ad-
vance in the usefulness of reflecting telescopes will result. It was the lack of such
rigidity and of such permanence of adjustments, fully as much as the lack of means
of rigorously testing the optical surfaces, which made the old Cassegrain reflectors,
including the great Melbourne instrument, such lamentable failures. I consider the
failure of the Melbourne reflector to have been one of the greatest calamities in the
history of instrumental astronomy ; for by destroying confidence in the usefulness
of great reflecting telescopes, it has hindered the development of this type of
instrument, so wonderfully efficient in photographic and spectroscopic work, for
nearly a third of a century.

When the telescope is to be used for photography, either direct or spectro-
scopic, it is indispensable that the mounting be so designed that reversal is not
necessary when passing the meridian; for it is frequently necessary to expose for
six or eight hours without reversal, on faint objects; and the best part of such an
exposure is that in which the celestial object is near the meridian. Several forms
of reflector mounting have been devised in which reversal is not necessary ; the well-
known English closed-fork mounting is one of them.

In designing the proposed mounting of the 5-foot reflector of the Yerkes
Observatory, of twenty-five feet focal length, the writer has adopted the form in
which a short open fork is used at the upper end of the polar axis. The tube
hangs between the arms of this fork, being carried on two massive trunnions; the

heavy lower end of the tube is so short that it can swing through, between the
arms of the fork, for motion in declination.

The fork mounting presents several marked advantages with respect to com-
pactness and stability, as well as convenience and economy, over all forms which
are modifications of the German equatorial mounting, in which the tube is carried

48 THE MODERN REFLECTING TELESCOPE.

out at one side of the equatorial head. The tube, carrying the great weight of the
mirror and its cell, is here supported at two opposite sides, instead of from one side
only, as inthe German forms; no heavy counterpoises are required; this form is
much better adapted for the coude arrangement of mirrors, so essential in work
with very large spectroscopes, only three reflections in all being necessary for this
arrangement ; furthermore, when the instrument is used at the primary focus, the
upper end of the tube is more easily accessible, in all positions of the instrument,
from an observing carriage attached to the inside of the dome.

The weight of the moving parts of the telescope will be about twenty tons.
On account of this great weight, and also of the overhang of the fork above the
bearings of the polar axis, an efficient anti-friction apparatus for the polar axis is
demanded, which will at the same time relieve the effect of the overhanging weight
of the upper end of the polar axis. The advantages afforded for this purpose by
mercury flotation, when this is properly applied, are so great, and the mechanical
details for such flotation work out so simply and economically, that this method
will undoubtedly be used.

The proposed mounting will now be briefly described in detail, and attention
will be called to many points which are indispensable to the success of a reflecting
telescope to be used for photography.

The equatorial head consists of three iron castings, the triangular base-plate m,
Plate x1, and the two posts 2 and 0, which carry the bearings for the polar axis,
Both posts are hollow, with walls 14 inch thick, and are bolted and pinned to the
base casting; the post ” contains the large driving clock.

The polar axis p is of hydraulic-forged steel, with a head or flange g, 48 inches
in diameter and 7 inches thick, forged upon it; the axis is 144 feet long over all, is
20 inches in diameter for a distance of 2 feet below the head, and is 16 inches in
diameter for the remaining 112 feet of its length; the axis is hollow, with walls
44 inches thick. The bearings of the polar axis are of hard Babbitt metal, and are
halved.

Attached to the lower surface of the 4-foot head of the polar axis is the large
hollow disk or float 7, 10 feet in diameter and 224 inches thick or deep; this is
constructed very strongly of angle steel covered with steel plates 3 inch thick; the
whole is finished smooth on the outside, and is turned true in a lathe. The cor-
responding trough s is of cast-iron and is turned true on the inside. The inner sur-
face of the trough is separated by 4 inch all around from the outer surface of the
float ; this space is filled with mercury. With the dimensions given the immersed
part of the float displaces about 45 cubic feet of mercury, which thus floats about
nineteen tons, or 95 per cent of the weight of the moving parts of the telescope. The
center of flotation is vertically below the center of weight of the moving parts.
Only three-quarters of a cubic foot of mercury is required to float nineteen tons in
this manner.

The importance in astronomical photography of the smoothness of motion
afforded by really efficient flotation of the moving parts cannot be overesumated.
The great size of the worm-wheel ¢ which rotates the polar axis, will materially

ital ad

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—RITCHEY PLATE >

Le] \ Jeg
Ey

a

DESIGN FOR MOUNTING OF FIVE-FOOT REFLECTING TELESCOPE.

PLATE XIil

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—RITCHEY

Fig 26

DESIGN FOR MOUNTING OF FIVE-FOOT REFLECTING TELESCOPE.

‘
*
‘
eS t
-
jm
~
J= 7
i
t
\,
. . =
* ~ =
ee _ > AP R
i : a . A UP ES bs =

THE MODERN REFLECTING TELESCOPE. 49

assist in giving smoothness and accuracy of driving; this worm-wheel is 10 feet in
diameter.

Attached to the upper surface of the 4-foot head of the polar axis, by means of
a circle of 2-inch bolts, is the large cast-iron fork w, different views of which are
shown in Plate x1 and Fig. 22, Plate xu. The extreme outside width of this fork
is 84 feet; it is of hollow or box section, with walls averaging 14 inches thick ; it
weighs about five tons.

Between the two arms of the fork hangs the short round cast-iron section w of
the tube; two 7-inch steel trunnions, having large heads or flanges, are bolted to
this casting, and turn in bronze bearings at the upper ends of the fork arms; this
part of the tube is 46 inches long; its inside diameter is 70 inches; its thickness is
1 ineh; it is reinforced at top and bottom by flanges. To the lower flange is con-
nected the cell-plate (described in the preceding chapter) which carries the large
mirror and its support-system.

To the upper flange of the short cast-iron section of the tube is bolted a strong
east-iron ring which forms the lower end of the main or permanent section of the
octagonal skeleton tube; this section is 13 feet 11 inches long, and 6 feet 8 inches
outside (diagonal) diameter. It is constructed of eight 4-inch steel tubes, connected
by strong rings designed to resist compression ; diagonal braces, which are connected
together at all intersections, greatly increase the rigidity of the structure. This
entire section is so rigid that it can be placed in a large lathe for facing the ends
parallel to each other, and for turning a slight recess in the ends for the purpose of
accurately centering the parts which are to be connected to them.

To the upper end of the permanent section of the skeleton tube can be attached
any one of three short extension tubes or frames, as desired ; two of these are shown
in Plate x1. The lower end of each extension is turned true, with a projecting ring
which fits into the turned recess in the upper end of the permanent section. With
this arrangement the various extensions can be removed and replaced without sen-
sibly affecting the adjustments of the mirrors and other apparatus which they carry,
with reference to the optical axis of the large mirror.

The extension which is shown in place on the telescope in Plate x1 and in Fig.
23, Plate xu, is the longest one; it is 6 feet 11 inches long; it is used for all work
at the primary focus of the telescope; it carries the diagonal plane mirror and its
supports, and the eyepiece and double-slide plate-carrier. This extension can be
rotated upon the turned end of the permanent section, so that the eyepiece or pho-
tographic apparatus can be brought to the side of the tube which is most con-
venient for observing or photographing a given object. The diagonal plane mirror
is of the finest optical glass, is elliptical in outline, is 15 x 22 inches in size, and is
34 inches thick; it is carried in a strong cast-iron cell, which is supported from the
skeleton tube by four thin steel plates, as shown in Plate x1. The diagonal plane
mirror is sufficiently large to fully illuminate a field 7 inches in diameter at the
primary focus. The double-slide plate-carrier is designed for 65x 84 inch photo-
graphic plates.

The other two extensions of the tube, which are only about 2 feet long, are

50 THE MODERN REFLECTING TELESCOPE.

employed when the telescope is used as a Cassegrain and as a coude respectively ;
each carries a convex mirror 19 inches in diameter and 31 inches thick, of the finest
optical glass, and of the proper curvature for the purpose desired.

Figs. 24 and 25, Plate xu, show the telescope used as a Cassegrain. In these
eases the amount of amplification introduced by the convex mirror is about 38}
diameters (see p. 38); the equivalent focal length is therefore about 874 feet,
and the ratio of aperture to focal length as 1 to 174. Fig. 24 shows the telescope
as used for direct photography with the double-slide plate-carrier at the secondary
focus. In Fig. 25 a spectrograph similar to the large Bruce spectrograph of the
Yerkes Observatory is shown attached to the north side of the short cast-iron sec-
tion of the tube; this affords a most stable base of support for the spectrograph,
at a point where it can be easily counterpoised.

Figs. 26 and 27, Plate xm, illustrate the use of the telescope as a coude; the
curvature of the convex mirror is now such that the equivalent focal length is
about 125 feet. The cone of rays from the convex mirror strikes a diagonal plane
mirror at the intersection of the polar and declination axes, and is by it reflected
in a constant direction, which can be toward either the north or south pole of the
heavens, as desired. This arrangement is almost indispensable when extremely
large and powerful spectroscopes and other kinds of physical apparatus are to be
used with the telescope; the focus is now in a constant position, so that such
instruments need not be attached to the telescope, but can be mounted on station-
ary piers, in constant temperature rooms, if desired.

A brief description of the mechanism for quick-motion and slow-motion in
right ascension and declination should be given. These are planned to be entirely

electrical, although hand-motions are added, to be used in case of an emergency. »

Quick-motion in right ascension, both east and west, is given by the reversible
motor w, this is connected by gearing to the large bevel-gear # through the
medium of an electric clutch y. The bevel-gear x is permanently fixed to the
polar axis. When the switch which starts the motor is thrown in, the electric
clutch y acts, and a motion of rotation is communicated to the polar axis; this
rotation is only at the rate of 45 degrees per minute; this is sufficient, since
reversal is never necessary ; hence very little power is required. The clutch is so
adjusted that it will slip when even slight undue resistance is encountered. When
the current is shut off from the motor the clutch is released automatically; the
polar axis is then free from the motor and gear-train.

Quick-motion in declination is given in a manner entirely similar to that in
right ascension, by a small reversible motor attached directly to the large cast-iron
fork; this motor drives, through the media of a gear-train and an electric clutch,
the toothed sector z, which is permanently fixed to the cast-iron section of the tube.

The driving-clock and 10-foot worm-wheel are “clamped in” to the polar axis,
when desired, by the electric clamps 7 which lock the 10-foot worm-wheel to the
bevel-gear «; the former is of course free to turn on the polar axis when not thus
clamped.

Slow-motion in right ascension is given by means of a small reversible motor

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE — RITCHEY PLaTe XIll

LARGE DOUBLE-SLIDE PLATE-CARRIER ATTACHED TO 40-INCH REFRACTOR ; YERKES OBSERVATORY.

THE MODERN REFLECTING TELESCOPE. 51

which acts on a set of differential gears in the shafting connecting the driving-clock
and the driving-worm. This device is used on the 2-foot reflector and on the
30-inch ecelostat, and is extremely simple and effective.

Slow-motion in declination is given by means of a small reversible motor
which acts on the long sector attached to the upper trunnion shown in Fig. 22,
Plate xa.

In concluding this necessarily brief and incomplete description of a modern
reflector mounting, attention should be called to an attachment which is absolutely
indispensable for the best results in direct photography of all celestial objects
requiring long exposure. I refer to the double-slide plate-carrier, by means of
which hand-guiding or correcting for the incessant small irregular movements of
the image, which are nearly always visible in large telescopes, can be done incom-
parably more accurately and quickly than by any other means now known. This
device is due to Dr. Common, who described it in Monthly Notices, Vol. 49, p. 297.
In 1900 the writer designed and constructed a small attachment of this kind for
use with the 40-inch refractor and the 2-foot reflector; this attachment and its use
are described in the Astrophysical Journal for December 1900, p. 355.

The photograph of the central parts of the Andromeda Nebula (Plate 1), was
made by the writer with this small plate-carrier attached to the 2-foot reflector.
The exposure time in this case was four hours. The images of the fainter stars on
the original negative are only 2 seconds of are in diameter; stars are shown which
are more than a magnitude fainter than the faintest stars which can be detected
visually with the 40-inch refractor; intricate structure and details are shown in
the nebulosity, which are entirely invisible with the 40-inch refractor and all other
visual instruments, and which have never been photographed before. When it is
remembered that the focal length of the 2-foot reflector is only 98 inches, and that
the aperture was in this case reduced to 18 inches, in order to secure a larger field
than is well covered when the full aperture is used, some idea can be gained of
the results which could now be obtained in celestial photography with a modern
reflecting telescope which would compare in size, cost, and refinement of workman-
ship with the great modern refractors.

In Plate xut is shown the large double-slide plate-carrier, taking 8 x 10 inch
plates, which was constructed from the writer’s designs in 1901, for use with the
40-inch refractor; the plate-carrier is here shown connected to the eye-end of the
great telescope. A description of this attachment, together with some photographs
obtained with it, will be found in the Publications of the Yerkes Observatory,
Vol. II, p. 389.

BY

Beach - BAAR WS

HAZARD PROFESSOR OF PHYSICS IN BROWN UNIVERSITY, PROVIDENCE, R. lL

(No. 1651)

CITY OF WASHINGTON
PUBLISHED BY THE SMITHSONIAN INSTITUTION

1905

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE

—$_—$_—_—— PART OF VOLUME XXXIV ——————

Hodgkins Fund

Peco NITINUOUS RECORD _ OF
VOsribiiiCc NUCLEATTON

BY

CAC BesgkaUES

HAZARD PROFESSOR OF PHYSICS IN BROWN UNIVERSITY, PROVIDENCE, R. I.

(No. 1651)

CITY OF WASHINGTON
PUBLISHED BY THE SMITHSONIAN INSTITUTION
1905

Commission to whom this memoir has been referred :
WILDER D. BANCROFT,
EDGAR F. SMITH.

ADVERTISEMENT.

In the present memoir, entitled ‘“‘A Continuous Record of Atmospheric
Nucleation,’ the author further discusses his researches on the nucleus, as
published in Experiments with Ionized Atr, Smithsonian Contributions to Know-
ledge, vol. XX1X, 1901, and in Structure of the Nucleus, issued as part of the
same volume in 1903. The investigation has been carried on with the aid of
a grant from the Hodgkins Fund of the Smithsonian Institution.

Doctor Barus describes the nucleus as a dust particle small enough to float
in the air, but larger than the order of molecular size. Such a particle pre-
cipitates condensation in an atmosphere saturated with water vapor in its
immediate vicinity. When these nuclei occur approximately of uniform size in
thousands and millions, they give rise to condensational phenomena of trans-
cendent beauty and importance. By far the greater number of nuclei are
initially ionized, or at least carry electric charge.

In addition to mechanical, thermal, and chemical processes, high potential
is shown to be a fruitful source of nuclei. Certain kinds of radiation, like ultra-
violet light, or the X-rays, or radioactive bodies, may also generate nuclei in the
dust-free air through which the radiation passes.

The term ‘“‘nucleation” is here used to denote the number of nuclei per
cubic centimeter regardless of their source or special properties.

The scope of the present memoir is summarized by the author in his preface.

In accordance with the rule adopted by the Smithsonian Institution, the
manuscript has been submitted for examination to a Committee consisting of
Professor Wilder D. Bancroft, of Cornell University, and Professor Edgar F.
Smith, of the University of Pennsylvania, and, having been recommended for
publication, it is herewith presented in the series of Contributions to Knowledge.

S. P. LANGLEY,
SECRETARY.

Smithsonian Institution,
Washington, May, 1905.

PREPACE:

“What is a nucleus?’’-asks my friend, smiling incredulously. The convic-
tion has become prevalent that only the ions induce condensation, and my own
consistent adhesion to the nucleus since my report to the U. S. Weather Bureau
(Bull. 12) in 1893 is often looked upon as heretical obstinacy. But this is quite
unjust, unjust even to those who have with such brilliancy maintained the
occurrence of condensation in ions. The nucleus has not left the field in dis-
comfiture. It has merely been forced reluctantly and under conditions of ex-
treme supersaturation to share its functions in this respect with the ubiquitous
and irrepressible ion.

But to reply: The nucleus is at the outset simply a dust particle, small
enough to float in the air, but much larger than the order of molecular size.
Such a particle precipitates condensation in an atmosphere supersaturated with
water vapor in its immediate vicinity, for the reasons long ago (1880) pointed
out by Lord Kelvin in his brief but epoch-making paper. The support of this
explanation was established experimentally by Coulier (1875), Kiessling (1884
et seq.), von Helmholtz (1886, 1887), and others, and, with particular ingenuity
and breadth of view, by John Aitken (1880, particularly 1888 et seq.)

A single nucleus, however, would be of but little interest. It is when the
nuclei occur approximately of uniform size in thousands and millions that they
give rise to condensational phenomena of transcendent beauty and importance.
To produce these legions of nuclei is not impossible by mechanical means, just
as we can, for instance, triturate a solid to a remarkable degree of fineness; but
the impalpable powders are perhaps best produced by chemical or at least by
very refined physical processes. Similarly, though a class of interesting nuclei
may be produced by vigorously shaking liquids, or, better, by mutually im-
pinging jets or by jets impinging on a solid obstacle, nuclei are more abun-
dantly produced by ignition or combustion. Such ignition, moreover, should
be unaccompanied by any kind of smoke, as the gross particles in this case are
an efficient means of absorbing nuclei. A clear non-luminous bunsen flame, a
red-hot metal or any other solid, like glass, for instance, is a powerful nucleator.
It is not even necessary that the solid be red-hot. Phosphorus is subject to a
peculiar kind of chemical reaction, whereby nuclei are produced at 13° or
a little below, and are then produced from the smokeless body in maximum

Vv

V1 PREFACE.

abundance. Fuming phosphorus is a relatively weaker nucleator. Many gaseous
sulphides, on mixture with air, become good nucleators. Even dust-free coal
gas and dust-free air, on commingling, set free nuclei.

Certain hygroscopic liquids like concentrated sulphuric acid are remarkable
nucleators. They probably make up a class by themselves. At least it is not
improbable that 1,000 or 1,000,000 molecules per cubic centimeter may escape
from such a body by ordinary evaporation, in spite of the low vapor pressure.
Since each such molecule is hygroscopic, stable nuclei may be formed in a
saturated atmosphere by condensation of the water vapor. Sulphide and
sulphur nuclei are in turn probably oxydized to sulphuric acid.

In addition to mechanical, thermal, and chemical processes, high potential
is a fruitful source of nuclei. A metal highly charged with electricity, or even
a glass insulator, or the nodal points in the metallic pathway of a stationary
electric wave are a source of nuclei. There is probably always an electric glow
present in such cases, though there need be no spark.

Finally (and here we reach debatable ground), certain kinds of radiation,
like ultra-violet light, or the X-rays, or radioactive bodies, generate nuclei in
the dust-free air through which the radiation passes.

Air originally made quite dust-free by filtration or otherwise, if exposed to
any of these sources becomes more or less filled with a freight of nuclei, fleeting
or persistent, and we may for brevity introduce the term nucleation to denote
the number of nuclei per cubic centimeter, regardless of kind or origin or other
properties possessed, and considered solely with respect to their tendency to
promote the condensation of water vapor in supersaturated moist air. If the
supersaturation is sufficiently pronounced the air molecules in successively
greater numbers as the supersaturation increases must themselves become
nuclei, probably beginning with the more complex systems. This, for instance,
occurs in the blues, opaques, and the succeeding browns and yellows of the first
order of the axial colors of the steam jet. The importance of experiments in
the spontaneous condensation of dust-free moist air was pointed out in my
report to the Weather Bureau in 1893, p. 48 et seq.; they were first carried out
in an independent manner and with exquisite finish by C. T. R. Wilson (1897).

The nucleus as an inert excessively small body, just transcending the order
of molecular dimensions, and occurring in immense numbers, has an interest of
its own; but this interest becomes much enhanced when it is found that by far
the greater number of nuclei are initially ionized, or at least carry electric charge.
The cases in which this does not occur are sufficiently exceptional to prove the
rule, though such nuclei need not for this reason be less efficient. They prob-
ably admit of a categorical classification, such as has been suggested above for
concentrated sulphuric acid and sulphides. Apart from these, all nuclei pro-
duced by ignition, by high potential, by the X-rays, or by radiation are power-
fully ionized. So marked is the quality that certain investigators (in particular
the younger von Helmholtz, 1887) have endeavored to find in the ionization a
sufficient cause for the condensation of supersaturated moist air, or at least an

PREFACE. vil

additional tendency to promote it. J. J. Thomson (1888) was the first to
adduce theoretical reasons.for the suspected condensational activity of the ion.
In the time since, so large in number and so important have been the researches
in which the precipitation of supersaturated water vapor on ions enters as an
argumentative premise, that insistence on the functions of the nucleus has
dwindled by comparison. I must nevertheless claim the right of an independent
investigator to interpret my work in a way which seems to me inevitable; and
I have therefore ventured to believe that, so far as experimental evidence goes,
the occurrence of electrical excitation is quite without influence in promoting
condensation of moisture in supersaturated air. However ionization may be
produced in the laboratory, whether by X-rays, or by ignition, or by high
potential, by chemical means, or even by excessively vigorous trituration as in
jets, it is always accompanied by nucleation. The average size of the nucleus
resulting depends for a given medium on the time of exposure to the exciting
cause and its intensity; or, in general, upon the number of nuclei produced per
cubic centimeter per second. Roughly speaking, if the conditions producing
ionization are sufficient and if they are maintained, there will be continuous
growth in the number and size of the nuclei. On this question I have
already expressed myself at some length in Nature (vol. LXIX, 1903, p. 103),
believing that ‘“‘out of all systems eventually issues a stable nucleation.”
“Why,” I ask, ‘“may one condense on a nucleus from which the soul has fled,
and still be permitted to call it anion? Why, indeed, does the nucleus persist
after the ionization has vanished; why does one not get back to dust-free air?”
I conclude that ‘‘electrification, if present simultaneously with nucleation, is an
incidental accompaniment with no immediate bearing on the condensation pro-
duced, and for this reason I have endeavored to account for the nucleus at the
outset, chemically.’ It is therefore merely necessary to summarize the point
of view in the following statement. Whenever ionization and nucleation are
associated in the outcome of any process, physical or chemical, the former is
generated proportionally to the latter, in such a way that each is produced at
its own rate, depending on incidental conditions. The subsequent life histories
of the nucleation and the ionization are distinct, nuclei being often surprisingly
persistent, ions by contrast characteristically fleeting. Hence it seems to me best
in keeping with all the data in hand, to regard the nucleation as the product
which owes its growth or origin to the expulsion (possibly also to the absorp-
tion) of the corpuscles representing the concomitant ionization. Moreover
ionization should be present only during the period in which the nucleation
varies, and a high order of nucleation may be associated with a very low or even
vanishing order of ionization. Many phenomena met with in the case of dust-
free air seem to be favorable to this view. Ignition and high potential nuclei,
X-ray and radiation nuclei in general, phosphorus and water nuclei, produced
throughout in dust-free air, all admit of this account of their occurrence and prop-
erties; and there is no observable case of a process producing ionization free from
nucleation, although there are many cases of nucleation free from ionization.

viii PREFACE.

What becomes of the ionization is a pertinent question: the ions probably
vanish by recombination, as they possess strong affinities for each other. Ejected
not by atomic but by molecular disintegration, we can scarcely attribute to
them phenomenal velocity. They may under favorable circumstances produce
fresh nuclei by absorption, by collision, but experiment does not show any ap-
preciable increase of nucleation during the period in which the ionization van-
ishes. If, however, the velocity of the ion is incremented by the presence of
an electric field, the production of fresh nuclei by collision may become per-
ceptible, and the result would then appear as if the nuclei themselves moved in
the electric field, whereas they are actually the inert residues left in the wake of
a fleeting electron.

Finally, it should be noticed that to produce condensation on X-ray nuclei
after long exposure, less than a double supersaturation is needed, whereas in
Wilson and Thomson’s case of condensation on ions, the supersaturation re-
quired is three- to four-fold. Thus the two views of condensation on nuclei
and condensation on ions would not in any case be mutually exclusive. Fur-
thermore, if initially (z. e., for short exposures, and nuclei in the extreme state
of fineness antedating growth) the nucleation is supposed to have ejected but
one electron per nucleus (an assumption which in one form or another must be
made in any other explanation), the present view is in no way incompatible
with J. J. Thomson’s method of measuring the charge of one electron.

If a nucleus like that of phosphorus, for instance, shows a continued ten-
dency to grow, until it finally appears as part of a visible smoke, there may be
continuous ejection of electrons within certain limits, as the growth matures.
In such a case, electric conduction through a gas freighted with these nuclei
would obey Ohm’s law, as is actually the case for phosphorus.'

To return from this digression to the present volume: the contents of the
first two chapters bear on my “Experiments with ionized air.”” The second
chapter originally carried the working hypothesis into further development, but
as I have not been able to supply the requisite numerical detail, I have retained
the experimental parts only. These chapters, like the earlier work, show, I
think, that whereas ionization vanishes with characteristic rapidity, the nuclea-
tion has a long lease of life by contrast. At the same time the ionization and
nucleation produced in any given process are proportional quantities. The

«In the time elapsed since these experiments were made, I have carried them much fur-
ther than stated in the text (cf. Science, xx1., 1905, pp. 275 and 563; American Journ. of
Sct. (4), 1905, XIX, pp. 175 and 349; Physical Review, July, 1905), showing among other
things that persistent X-ray nuclei pass into fleeting nuclei on removal of the X-ray tube to
greater distances from the outside of the fog chamber or on loss of intensity. Such fleeting
nuclei become persistent water nuclei on solution. For the case of radium in a sealed alumi-
num tube surrounded by a wall of lead 1 cm. in thickness, the nucleation is reduced by but
30 % of the value obtained when the lead envelope is absent. Hence the gamma rays which
produce but a few per cent. of the ionization are accountable for the greater part of the
nucleation.

PREFACE. ix

corresponding cases for nuclei produced by the X-rays are given in the earlier
volume and elsewhere.

The following chapters, III, IV, V, VI, have been written to draw a variety
of conclusions from the data in my work on the Structure of the Nucleus, which
escaped me in the earlier volume, as well as to correct certain errors relating to
the quantity of water precipitated under given conditions, and to the diffusion
of nuclei, to which I have already called attention elsewhere. Chapter VI,
relating to periodic distribution in the colors of successive coronas, shows under
what conditions the angular diameter of a ring of a given color may be used
for the estimation of the number of particles producing the observed diffraction
pattern. In Chapter VIII a definite practical application is made, for use in
the last chapter of the book.

Chapter VII shows a method by which fog particles, even of minutest size,
may be measured under the microscope, or microscopically photographed.
The peculiar difficulties encountered in the interpretation of these results, in
spite of the fact that fog particles are obtained in definite sizes and numbers,
are considered critically.

The chief results of the book, however, are given in the last chapter, which
is a record of over two years of observation of the number of nuclei present per
cubic centimeter of the atmosphere of the city of Providence. The nuclei are
abundantly represented, particularly in the winter months. Curiously enough,
the maxima and minima appear at about the time of the winter and summer
solstices respectively. The reason for this cannot be sought in the astronomical
circumstances involved, but must be atmospheric in character. I have supposed
that in addition to rain, light pressure, which must be more effective as the
days are longer, may have something to do with this. Under any circum-
stances a highly nucleated medium is an interesting medium. Since much of
the nucleation must be of local origin and referable to combustion, the question
arises, what has become of the ionization which was simultaneously generated ?
It has probably vanished as does the ionization in the experimental condensa-
tion chamber in the laboratory.

To reply to these questions systematically, observations have now for
nearly a year been taken at Providence and at Block Island simultaneously.
The latter station has many of the meteorological elements of Providence, but
Block Island lies sufficiently in the sea, and is in winter at least sufficiently
free from human habitation to present entirely different conditions as to nuclea-
tion. I have also in progress a continuous series of observations on the changes
in the lapse of time of the nucleation of dust-free (filtered) air, 7. e., of air free
from foreign nuclear ingredients. The results will be reported in due course

elsewhere.
Cart Barus.
Brown University, January 9, 1905.

~

TABLE OF CONTENTS.

Cuapters I—-I].—RE vaTInNG TO “EXPERIMENTS WITH IonIzED ArR,’’ SMITHSONIAN CONTRI-
BUTIONS TO KNOWLEDGE, VoL. xx1x, No. 1309.

CHAPTER I.
PAGE
The Relation of Ionization to Nucleation in Air after Contact with Phosphorus......... I
tr. Introductory. 2. Method proposed. 3. Water nuclei. 4. Comparison of the steam
jet and the condensation chamber. 5. Decay and absorption. 6. Apparatus. 7.
Manipulations. 8. Data for phosphorus ionization. g. Further data. tro. Effect of
different charges in the condenser. 11. Dried emanation. 12. Wet emanation. 13.
Residual ionization after one hour. 14. Nucleation partially precipitated. 15. Ioni-
zation of dry phosphorus nuclei. 16. Inferences. :
CHAPTER II.
The Relation of the Ionization and the Nucleation Associated with Water Nuclei,
rOcu@e dering Auteyeete, sorteetts as ois ius eealernere late «, hat Sheen CO GOCE eer 17

Ionization Produced by Shaking Solutions.

1. Introduction. 2. Apparatus. 3. Results.

Efficiency of Nuclei-Producing Jets.

4. Powerful methods of comminution. 5. Results.

The Ionization of Water Nuclei.
6. Introductory. 7. Apparatus. 8. Results. Initial charges. 9g. Comparison with
coronas. to. Evanescence of the charges of water nuclei. 11. Results with an Elliott
electrometer. 12. Further data. 13. Jets self-shattering or impinging on water. 14.
Summary of the relative degree of ionization and nucleation. 15. Spontaneous time loss
of nuclei. 16. Effect of condensation on ionization. 17. Effective condenser length.

Summary and Inferences.

18. Working hypothesis. 19. Charge and conduction.

20. Comparison of phosphorus
and water nuclei.

CuHapters [IIJ-VII,—RELATING TO ‘‘ STRUCTURE OF THE NUCLEUS,’ SMITHSONIAN CONTRI-

BUTIONS TO KNOWLEDGE, NO. 1373.

CHAPTER III.

Preliminary Survey of the Apertures of Coronas in Relation to the Number of Nuclei
and their Sizes

1. Introductory. 2. Apparatus and preliminary results. 3. Diameter of cloud particle.
4. Nucleation. 5. Cause of periodicity. 6. Effect of temperature. 7. Pressure decre-

ment. 8. Summary. 9g. Plate-glass apparatus.

xi

X11 CONTENTS.
CHAPTER IV.

The Number of Nuclei Produced by Shaking Different Liquids, and Allied Experiments.

1. Explanation. 2. Data. 3. Coronas in general. 4. Axial colors. 5. Carbon di-
sulphide.

CHAPTER V.

Diffusion of Vapor into Nucleated Air; a Correction®s...- eee ei ee eile

r. Apparatus and manipulation. 2. Equation. 3. Application and data. 4. Con-
clusions. 5. Diffusion from greater to less saturation. 6. Crucial experiment and
conclusion. 7. Nuclei produced by mixture of coal gas and air.

CHAPTER VI.

Periodic Color Distributions in Relation to the Coronas of Cloudy Condensation, with a
Revision of the Constants of Coronas#. =< -h24-e4 2 i ee eee taee:
Introduction.

1. Purpose and plan. 2. Apparatus. 3. Color distributions. 4. Apertures.

Data for the Welsbach Gas Burner.

5. Explanation of the tables and notation. 6. Charts. 7. Tables.

Method of Reduction.

8. Constants of the geometric sequence. g. Time losses. 10. Exhaustion losses. 1r.
Losses attributed to subsidence. 12. The optic constant. Diffraction methods. 13.
The optic constant. Subsidence methods. 14. Summary of optic constants. 15. Re-
sulting equations applied. 16. Remarks on the tables and graphs. 17. Diameter of
fog particles.

Data with Electric and Mono-Chromatic Light.
18. Tables. 19. Charts. 20. Diameters of cloud particles.

Miscellaneous Experiments with Deep Vessel.

21. Aperture of white disc. 22. Tablesfor the coronas. 23. Remarks on the table and
graphs. 24. Condensation chamber remote from the eye.

Long and Shallow Condensation Chambers.
25 Apparatus. 26. Reduction of data and tables. 6p = 17cm. 27. Smaller and
larger pressure differences. 28. Remarks on the tables. 29. Nucleation of the green
coronas,

Other Causes of Change in the Types of Coronas.
30. Thickness of cloud layer, 31. Obliquity of diffraction. 32. Effect of wave length.

Different Speeds of Exhaustion for a Given Pressure Difference.

33. Increased suddenness of condensation. 34. Results. 35. Growth of nuclei. 36.
Subsidence. 37. Exhaustion ratio. 38. Inferences. 39. Different exhaustion rates
for moderate nucleations. 40. Conclusion.

CHAPTER VII.

Direct Micro-metric and Micro-photographic Measurement of Fog Particles, with a
Summary of all Relevant Data... eee eee eee eee ere ene

I. Micrometry of Fog Particles.
Earlier Methods.

1. Introductory. 2. Apparatus. 3. Behavior of the precipitated droplets. 4. Pre-
liminary data.

PAGE

47

51

56

Io!

CONTENTS. Xili

Improved Method. aa
5. Number of droplets. 6. Diameters of droplets. 7. Graded particles.
II. Micro-photography of Fog Particles.
General Results.
8. Preliminary. 9g Apparatus and method. to. Incidental phenomena. Pitting.
tr. Dew. 12. Evaporation. 13. Moving and floating globules. 14. Graded particles.
Specific Results.
15. Photographic plates. 16. Tabulated results. 17. Remarks on the tables.
Inferences.
18. Precipitation per cubic centimeter. 19. Diameters and numbers. 20. Explanation
of discrepancies. 21. Summary.
III. Results from Subsidence.
22. Object and method. 23. Results. 24. Remarks on the tables. 25. Further
results.
IV. Summary of all the Results Obtained for the Relation between Nucleation and
Coronal Diameter.
26. Preparation of a table of reduction.
Cuaptrers VITI-IX.—App.icatTIons.
CHAPTER VIII.

The Coronal Method of Estimating Atmospheric Nucleation......................... 128
1. Introductory. 2. Apparatus. 3. Diffusion from two opposed surfaces. 4. Mis-
cellaneous tests. 5. Diffusion from a single surface. 6. Absorption and decay of nuclei.

7. Effect of pressure difference. 8. Precipitation per cubic centimeter. 9. Relation of
nucleation to aperture of corona. 10. Absence of electrification in cases of sudden con-
densation and of sudden evaporation. 11. Conclusion.
CHAPTER IX.
The Variation in the Nucleation of the Atmosphere of the City of Providence........ 139

Introduction.
1. Preliminary. 2. Apparatus. 3. Classification.

First Group of Observations.
4. Early observations. 5. Plate-glass apparatus. New apertures.

Second Group of Observations.
6. Plan of tables and graphs. 7-16. Successive monthly data for 1903. 17-25. Succes-
sive monthly data for 1904.

General Inferences.
26. Efficiency of Apparatus. 27. Variability. 28. Wind effect. 29. Rain effect.
30. Snow effect. 31. Cloud effect. 32. Solar effect absent. 33. Temperature

effect. Cold-wave effect. 34. Local effect.

Summary and Conclusions.
35. Mean daily nucleations. 36. Mean monthly nucleations. 37. Occurrence
of maxima and minima of nucleation during the winter and the summer solstices,
respectively. 38. Conclusion.

=

LIST OF FIGURES.

CHAPTER I.
PAGE
Ficure 1. Apparatus for comparing nucleation and ionization........ sare ey diayate: so tetaitien stele cad RoR 4
Ficure 2. Modification of preceding apparatus, with side influx tube and removable desiccator... . 5
CHAPTER II.
Figure 1. Apparatus for comparing nucleation and ionization...............00000 cece even seeees 17
BIGURE2=ioimplinedidetatliof coronalichamber: (Big. in, Arc. 4.0. > oc sac seliceee oeiee en scree alee 20
DISTRE Bo GS Otten ohana Ce ae ee een te Roser acinar ait adc nOn Grtone baon sb mneu notes r 20
FIGURES 4 to 6. Curves showing electrometer deflection (leakage), after consecutive half-minutes, for
ditterentycharges\and: potentialsiam: the condenser cus eee iis: nec Na ae eee ee eerie serait 24
Ficures 7 and 8. Curves showing radial currents (amperes) for different charges and potentials (volts)
LTTE NCONIC ENS EL ee tarevaysrsta charset ees olenck s te ravaicicud iseatedors te reievabnemiere crate Gen GR ES OO Eee 2
Figure g. Tube condenser with sliding and removable outer coating.................0.2--. eeeees 34

CHAPTER III.

Ficure 1. Chart giving curves showing relations of apertures computed from successive exhaustions
(1, 2), and the relation of diameter of fog particles computed from successive exhaustions (4,5) 41
FiGuRE 2. Chart giving curves showing ratios of nucleation computed from successive exhaustions and
from measured apertures (8, 9, 10, 11, 12), and corresponding diameters of fog particles in
METI UIMIELETSM (Ovary) Rapasretsyaacnyst se cine oan etegscziclee Ruleree © areete iota eaieyteslagenz ce vein atere Versi teamte corer creme era 42

EDGE RTD) LETSTOMN GHAMIDEL <(or2)ccc, 5d cs, nun cues oxoet ayers tebe ue on) syd este © cel sleieto a e)sapsracel opera) syehal ave alardia sce fate ee 52

Ficure 2. Chart showing the vapor pressures at different heights in the lapse of time when water vapor
TETRIS GSR UTNE AL IVa sne rst yoo 0eyas c=] ays ckacets fot stain os) abe retedogsvs Sicko’ ofoeel os atest ens eg eee eee aI 52

FIGURE 3. Same as 2, but for diffusion of air originally } saturated.....................020-0eeeee 52

BiGuReEnem Cmbical condensation Chamber. «access escile ove rece cre daar es + aces ee eerie Soe 58
Parone am SOU PA CONGENSA GION! CHAIN DER-c-iecivieicjereta leis sels wieivleniale Sie sveleiale sictalsieie sme esl eileen eats . 58
ICMiREEmm al yesfOri sudden CXHAUStION W/o craral=fe) ers ace ye Na lsuaeiehoes oye mace ey Saosin =n ieisiasye ee . 58
Ficure 4. Chart for Table 1, showing the relation of nucleation and of diameter of fog particle in

RERIMsOtknn ere per Lures | Ob Me! COLOMAS is, 1a rei. cree cis teh ck eusls cherie =leleis) coe fer ae) (5 aterete of Ore petencdel stem one oer 61
Ficure 5. Chart for Table 3, showing the relation of relative nucleation to the number or order of

Ewa EAS LO Ae pee Ae area ee Bhar cia oe er Cen ole censvckal dbaiayace ois cette, <1 eles haa eee te pens pe eee tuo) Ceee tetra: aia ee 79
Ficure 6. Chart for Table 8, showing the relation of nucleation and of diameter of fog particle in

MELINS OMe apes tres Olt De COLOMAS rs vis) «sie wld arco cleieivioie efdie)eiesets +] slere =< acaereyatebaiayetvereihie eas © 77
etme CAL bor L Able LO, SHOWIN SAME, rare ec reiers (eaaliejels eile os saicislshaneisiaie is + elld mul eiothietay ete everemieieteyers 79
Memos Ghartt Ore DIO T3y SOWIE) SAME vojeye cape aisyera oatarelesa’» (clela @ miei iaie a wlape “ole bisieelats Win)e/= ale /eysyelata 79
Ea tramcrm ce Oi tnt Ore hale x4) SHOW PYSATILE: (0, cie:0s facies «(a cietormiw «elie pivie sie Waicie/Slcuovels < Wisit y atereloie/oteie wisis 83
emMEEnce Chart toneha ble £4 ,, SHOWING SAMIE s:2).«.< c10/o1s sale, ere /arsl¢s/eiie pieteicla sss) «il sueleisiwial sieve svoretere aleleve aie 83

Xvi LIST OF FIGURES.

PAGE
FIGURE 11. Chart for/Table 15, showing Sameiin. «.-.«-1ely-nyelche tet ersioietelatenveieenere spate: Seneeeretae RRC go
Ficures 12 and 13. Charts for Table 18, showing the relation of the apertures of the coronas to the
number of exhaustionsy. <6. sisi s)2ie cis /e esse nus 6) sie stato oie )els mec tetenelelcherevenetet elle katate heter stead tetera 9°
Ficure 14. Chart showing the effects of different valves with nucleations increasing in the lapse of
ERIS sais aicoycy ie wires wae mem) eel =iielals ead ntfs! evaiad et = lol cPesteten felon = felis letete eet ets eet tee eee ee 83
CHAPTER VII.
FIGURE ©. Micrometric apparatus. cc c)c.c. cio ~ ie el ereiepete dee eneee ne eta eter ei ete ne eee ee 102
FicuRE 1a., The'same mounted in condensation) chambers quam cere trey ssi eaters eer 103
FiGuRE 2. Capsule. Plamiiccojoi.s05- toe) oresecsve steiesohars severe) eleneeateial eek feats eke terete teen eee eee ee 103
Ficure.3. Same. Sectional‘elevation.. 2. -).-/- scrsciset eee ee ise retreated teed tet tae eee 103
Figures 4-6. Diagrams showing the behavior of fogiparticles.. ijt s.r sede eens eee 105
FicuRE-7. Chart: giving.d.in terms Off's..5< 12 42 2 omteret-eeiars Berea reek ete ee eee 108
Bicure'8. Diagram of Crater’? «oie csieiei cote) oie lente) o ensreletor Sie reteset aT tele ege terate wat tone a et ea ete nea 113
FIGURE 9: Diagram.showing’ circulation: < 2.4 cac0-ciee ete seo eee ek ae ene eyes eee ene ee 113
FIGURE ro. Micro-photographsiof fog particles. sj. sis.) ee teehee eee eee 116
CHAPTER VIII.
FIGURE-1. Condensation chamber. ‘Sectional elevationy./—. 3. <\-).ic 0. en cic iene tee tee 129
Ficure ra. The same im fis. sre jo)s,<.s.0:s:5 sco vic 31 caohsus tore) eyehay ove a¥elle iy slo) evs fessjrals tote tele es = lot tetan eon delle eer 129
Figures 2-4. Graphs showing the progress of diffusion in the lapse of time....................-..-. 131
Ficure 5. Apparatus to detect ionization produced by sudden condensation.....................-. 137
CHAPTER IX.
Ficure 1. Diagram showing general disposition of apparatus................--2 eee erect eee 140
Cuarts 1-48. Daily observations of the atmospheric nucleation of the city of Providence........... 141
Cuarts 49 and 50. Mean daily nucleations in thousands per cubic centimeter..................-..- 223

CHART 51. Mean monthly nucleations from October, 1902, to October, 1904, in thousands per cubic
centimeten:-o-ccwes cide ee Wyn aga dk whew Dealers Re pfoleb epee, a IONE tas rtslca Seine NRC St AP eae pene eS vee a en 224

CONTINUOUS RECORD IO
MRVOSPHERIC NUCEE AT LON:

By Cart Barus,

HAZARD PROFESSOR OF PHYSICS AT BROWN UNIVERSITY.

CHAPTER I.

THE RELATION OF IONIZATION TO NUCLEATION IN AIR AFTER CONTACT WITH
PHOSPHORUS.

1. Introductory—The opinion was expressed in my earlier papers, that
wherever there is sufficiently intense ionization, there one may also expect to
find active nucleation; for it is hardly probable that a group of dissociated
molecules, neutral as a whole, can ultimately escape combination with each other
and the medium in which they are suspended. If these combinations occur in the
presence of water vapor, and particularly in a saturated atmosphere, the nuclei
due to solution may result. When the nuclei are produced from dilute solutions
by shaking, evaporation of the fog particle to the nuclear diameter might be
inferred; similarly, when the solute is produced by any kind of radiation or
emanation, each trace of solute may grow in bulk by absorbing water to the
nuclear stage. In case of an intense emanation like that from phosphorus, this
process may actually continue until a visible cloud is produced and the nuclei
attain the size of fog particles. But it is to be borne in mind that the nucleus,
dissolved or not, is present initially, and is in case of phosphorus producible in
dry air; whereas, in case of water nuclei produced by shaking or by jets, there
is no evidence of evaporation from the comminuted water particles, nor is it
certain that the nuclei here are mere water dust. One must keep in mind that
a nucleus may be the residue after the corpuscles representing the ionization
have been expelled.

2. Method proposed.—tlf the original emanation, highly ionized though
neutral as a whole, is put through the process of condensation, then, if the
negative ions are more efficient as condensation nuclei than the positive ions,
the nuclei after condensation, or even after remaining in a saturated atmosphere,

2 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

should become continually more positive,—assuming that a greater number of
negative ions are removed by condensation.

The investigation would therefore consist in testing the ionization imme-
diately coming from phosphorus as to its power in dissipating positive and
negative charges, and in comparing these results with the degree of ionization
after the emanation has produced condensation. In other words, it is to be
ascertained whether the. nuclei after a succession of condensations become con-
tinually more positive.

The results to be discussed in the following paragraphs have made this ap-
parently straightforward investigation of little avail: for after the emanation
has reached the nuclear stage, scarcely 3 per cent. of the original ionization is
left. The residue is then so small that a decision of a possible excess of positive
or negative ionization is difficult, because the whole is now of the same order as
the normal leakage of the electrometer.

In fact, the decision as to whether the positive or negative ionization is in
excess is now of very secondary interest, for the nuclei introduced into the
condensation chamber have already lost all but a trifle of their original charges.
The successive and even the initial condensations thus virtually proceed
without electrification.

The initial intense ionization nearly vanishes even in a moderately dry
atmosphere. Indeed, it is hard to understand how a neutral, intensely ionized
emanation can be produced from a body like phosphorus. It appears to me
that the emanation is a molecular body which is stable in the presence of an
excess of phosphorus, 7. e., at the surface, but which becomes unstable and
breaks to pieces in presence of an excess of air, on leaving the phosphorus. The
observed ionization is the accompaniment of this dissociation, and occurs on
the passage from the first environment to the second. If the ions were produced
by phosphorus directly, one would expect them to be either positive or nega-
tive, but not neutral.

3. Water nuclet.—After finishing the work with phosphorus, correlative
experiments with water nuclei were undertaken. It was found necessary,
however, to produce them in greater number than is possible by mere shaking,
to obtain marked effects. Hence jets were resorted to and studied in some
detail, as will be shown in succeeding chapters. The results obtained, though
closely resembling the phosphorus data in the main, differed from them inas-
much as the currents above a certain potential difference were constant, and
independent of the electromotive force of the condenser, while the ionization or
charge is not neutral as a whole. Nucleation again remained equally effective
after the ionization had all but vanished in the speedy way observed for phos-
phorus. Here, too, however, it may be plausibly argued that the nucleus is the
stable product after the corpuscles representing the ionization have been
extruded.

4. Comparison of the steam jet and the condensation chamber.—A digression
may here be made relative to the indications of the colors of the steam jet and

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 3

of the coronal phenomena, in relation to the number of nuclei concerned. The
usual and strong response of the steam jet is for axial blues and greens as far as
the purples of the second order. These are already too weak to be of effective
service for measurement. But at this stage of smallness of fog particle, the
coronal display has but begun. The strong blues of the axial colors correspond
to the diffuse gray fogs of the condensation chamber out of which the coronas
are gradually evoked when the number of particles ! has sufficiently diminished.
The two instruments are thus in a measure supplementary; the condensation
chamber gives intense evidence of the presence of nuclei long after the steam
jet would imply their absence. It is for this reason that ordinary smokes like
sal ammoniac do not affect the steam jet ? where a number of nuclei exceeding
a certain lower limit is necessary. The latter again 1s particularly active for
those intense and fresh nucleations which produce the browns and yellows of
the first order, implying conditions which it is impossible to produce in the
condensation chamber at all, until the lower limit of spontaneous condensation
of dust-free moist air has been exceeded.

5. Decay and absorption.—To account for the rapid diminution of the
number of nuclei in the phosphorus emanation in the lapse of time, two hy-
potheses are prominent. With finely divided and in so far highly potentialized
matter (possibly ionized positively and negatively), combinations of nuclei may
occur to the detriment of the number of independent nuclei. Such a decrease
would take place as the square of the number. On the other hand, it is equally
probable that the initial and very small nuclei are in rapid motion much like
molecules, and that the loss takes place by absorption or arrest at the walls of
the vessel. In my memoir on the subject, I included both hypotheses in the
computation; but finding that the phenomena could be adequately explained
by the latter, I ignored all spontaneous decay. Though this policy would not
be generally admissible, it is unlikely that nuclei can vanish initially at the
same rapid rate as the ionization. Indeed, evidence will show that it does not.
In case of water nuclei, which are much the more sparsely distributed, the
original number of nuclei can be proved by coronas to have varied but little in
the short time in which the ionization falls off to a few per cent. Hence the
nuclei must be regarded as parting with their charges more rapidly than they
are themselves absorbed in the lapse of time, and one will have to distinguish
between the velocity of the uncharged and of the charged nucleus in the electric
field, the latter being incremented by the electric forces. The cases will be
worked out in the chapters below.

6. Apparatus—The apparatus used in the present experiments was capa-
ble of a great number of variations. The essential purpose is to enable the
observer either to introduce phosphorus emanation at once into the electrical
condenser or else to introduce it after it has been saturated with water, suddenly

1 Phil. Mag. (6), iv, p. 24, 1902; cf. Structure of the Nucleus, Chapter III.

2 Thus smoke due to sal ammoniac if introduced into the steam tube will actually clear
the blue field produced by phosphorus nuclei, 7. e., will wipe out the condensation.

4 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

cooled, or otherwise treated. The following diagram, figure 1, will make the
adjustments clear.

FIGURE 1.—APPARATUS FOR COMPARING NUCLEATION AND IONIZATION. A, CORONAL CHAMBER; R,
EXHAusTION RESERVOIR; M, Mariotre Fiask; C, TUBULAR CONDENSER; P (TO BE INSERTED IN THE CON-
VEYANCE TUBE, e), PHospHORUS IONIZER; GanpD G’, Vacuum GauGEs; F, Cotton Fitter; B, STORAGE
Battery TERMINALS. TuBE p TO SucTION Pump, a TO ATMOSPHERE, ¢C AND ¢ FOR EXHAUSTION, ¢ FOR
CONVEYING NUCLEI INTO CONDENSER, d FOR FILTRATION, h FROM THE HyDRANT. SUPPORTS g, g, ARE
METALLIC, 2, 1, INSULATING.

The parts of the train of apparatus are the large copper Mariotte flask MV,
with a supply of water sufficient for aspiration, the condensation chamber A,
used both for producing coronas and for the aspiration and storage of air laden
with phosphorus nuclei, the exhaustion chamber, Rk, the phosphorus ionizer,
P, the tubular electrical condenser, C, and the electrometer, E. An accessory
desiccator, D, of the tower form may be inserted on the way, when dry air is
needed, as shown in figure 2.

R is connected through a stopcock with the suction pump at p, with the
atmosphere at a, and by wide tubing with the condensation chamber, A. It
carries a vacuum gauge, G.

A is connected with a stopcock with the cotton filter, /, with the ionizer
by b (where the tall desiccator may be inserted), with R by c, and also carries a
vacuum gauge, G’. A is further joined by stopcocks with the Mariotte flask,
M, for aspiration, and is graduated in liters on its side. It holds about 10
liters.

The ionizer, P, is a large U tube containing calcic chloride for desiccation,
kept in place by loose cotton plugs. One shank is nearly empty, and at g
carries thin pellets of phosphorus between strips of wire gauze.

The condenser is tubular, 2.10 and .64 cm. in diameter and 50 cm. long,
with the outer mantel permanently put to earth. The core is a brass rod, sup-

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 5

ported on hard rubber insulators, 7, 7, about 15 cm. long, and at a distance of
ro cm. from the end of the tube. This rod is highly charged from the storage
battery, B, and the leakage found from the electrometer, F, one pair of quad-
rants of which are in connection with r, 7, and the other pair put to earth. A
commutator enables the observer to use either the positive or the negative pole
of the storage battery for charging the system.

FiGURE 2.—MODIFICATION OF PRECEDING APPARATUS WITH SIDE INFLUX TUBE, k,f, m, FOR MAINTAIN-
ING CoNSTANT PRESSURE IN A, AND REMOVABLE DesrccatTor, D.

The electrometer was specially built for the present purposes, and the
needle was kept charged by a water battery of 48 volts, one pole of which was
earthed. The suspension is a silk bifilar moistened by a dilute solution of any
hygroscopic salt, and the battery charge is conveyed through the fibers. The
quadrants are supported on hard rubber insulators 10 cm. long. Difficulties
were encountered in using this apparatus as will appear below. In figure 2 a
side tube, f, between stopcocks k and n has been added, with the object of secur-
ing greater constancy of pressure’in A, the flow of water from M being via m j 1.

7. Manipulations—On raising the Mariotte flask, M, and opening ap-
propriate stopcocks, the emanation passes directly from the ionizer into the
condenser, and its ionization may be measured. A tower desiccator is here
to be inserted before P, to dry the air.

On lowering M, removing D, and reversing P, the emanation passes into
A. Here its nucleation may be tested by condensation, and it may thereafter
be introduced into C at once or after a number of condensations. The nucleation
may also be stored in a dry vessel between D and P reversed, and subsequently
transferred from the dry vessel into the condenser.

Finally, the phosphorus ionizer P and D may be quite removed and re-
placed with a pipe connection, while the tube 77 is adjusted for spraying. The

6 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

hydrant water passing under high pressure reaches the jet 7, placed suitably
within A. The water nuclei thus producible may be tested either by coronas
or electrically by passing them into the condenser, C. Compare Chapter II.

IONIZATION OF PHOSPHORUS NUCLEI COLLECTED OVER WATER.

8. Data.—The current through the condenser follows a law similar to
Ohm's. The constant appropriate for the comparison of the data may there-
fore be computed logarithmically. If C is the capacity of the apparatus in
parallel, E the potential of the condenser, z the radial electric current through it,
R its ohmic resistance for the given medium, s the deflection at the electrometer,

(dE /dt)/E = (ds/dt)/s=1/CR, or d(log s)/dt=.434/CR=a
where common logarithms are used and a is the constant sought. Thusa varies
directly with the conduction of the ionized medium traversing the condenser.

TABLE 1.—IONIZATION OF PHOSPHORUS NUCLEI.
dV/dt (IN LITERS PER MIN.) VARIABLE. NEGATIVE CHARGE.

Time. Deflection. a Remarks.
t m scm,
ro Tans .0073 Insulation (morning).
5 1225
10 1.5
DSi |) 40:5
Ils 06 9-3 .0097 Insulation (afternoon).
3 8.8
7 78
TLR 9-3 .013 Slow current of P nuclei (1L/min.) from aspirator (over
water).
I 8.9
2 8.7
3 8.5
DV. io, 8.4 .009 Faster current (same nuclei).
I 8.3
2 8.1
3 7-9
V. o 9-3 .0078 Do. (same nuclei).
4 8.5
8 7-9
12 7.5
WAG! 9.5 .O14 Very fast current of wet P nuclei. 10 L/min.
I 9.2
Ville 0, 9.7 .008 P nuclei enter by diffusion through middle tube of con-
denser.
I 9-5
2 9.3
3 9.2 :
VIII. o 9.0 007 P nuclei blown in with moist air very slowly.
I 8.7
2 8.6
3 8.5

‘ Nuclei put into receiver (A) by preliminary exhaustion.
2 In this and the other cases fresh nuclei were aspirated into A by passing room air over phosphorus.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. i

TABLE 1.—Continued.

Time. Deflection. a Remarks.
t m scm,
IX. o 12.8 321 P nuclei enter directly into condenser. Tubular ionizer.
I Sey
2 2.6
3 1.5
x © 9.8 .O14 Current stopped; observation within 5 ™.
I 9-4
2 9.1
3 8.9
xe io 7.0 203 Tubular ionizer. Very slow current (L/min.) into condenser
directly.
I 4.1
2 2.6
3 1.7
Eso II.10 .016 Nuclei from receiver. Very fast current.
I 10.70
SC 0 7-4 320 Slow air current through tubular ionizer into condenser.
+5 3-7
I 1.6
5 9
sec.
SVE 0 4.3 .054 Insulation immediately after air current stops.
10 4.0
20 3-9
30 3-7
4° 3-7
5° 3-7
60 3.6
70 3.6
2We ©) 9-7 .0085 Insulation.
16" 7p

‘In this and the other cases fresh nuclei were aspirated into A by passing room air over phosphorus.

TABLE 2.—SUMMARY OF THE PRECEDING. NUCLEI INTRODUCED INTO RE-
CEIVER BY ASPIRATING ROOM AIR OVER PHOSPHORUS.

Remarks.

ee lasulationsGnorning ier cls ates ai oleic rerio alrite okie atch tea
ie ne (GiitermgGn)) 7.07.5.) ain deere ee ene ee ne et ery eee eet
V. Slow current of phosphorus nuclei (liters per min.) in saturated moist air..
Nite Dowebuiast current (moliters sper taiit.) sensei ireieileranicietssl seeeieroiene ie

Wii Phosphorus) nuclei enter’condenser by diffusions .. 2... 42945.-4-+s-e es 8
Wile Lown! with) moist ain (Slowly,foc ose ee ile oe omic ein rae rea 7
IX. Phosphorus nuclei enter directly from tubular ionizer. Fast current of
TILA iy ATS eps eye Cee any seek Schon co acne aus eee say heen, le Toa mer usher acs teeeRer aha 320
Nem An Cucrent Stopped 3 Observation, within Sm ster. eee utetoe ae erie 14
XI. Slower moist air current through tubular ionizer directly into condenser. . 200
XTie Dor Slow moist air current, but) fresher phosphorus: -.- -2 22 .5.5.-....- 320
XIV. lonized air current stopped: immediately after........................ 54
POY Meme licr sta] yb1 OM house ste fe vane a Pes tade ovals oieraes naeniavas acaiay waa "s Bal eae o Biel becoish'sy oad eon ee RENe ome 8

Cases V, VI, tested for coronas, gave the usual intense and full series, beginning
with diffuse dense fogs.

8 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

Table 1 contains the direct observations, time in minutes, and deflections, s,
of the electrometer in centimeters. dV/dt, denoting the volume of nucleated
air put into the condenser per minute, is here estimated. The charge of the
core is negative.

Table 2 gives a summary of these results, among which may be mentioned
the following: The insulation is not above a=.o1 and usually lower. It is not
exceeded when phosphorus nuclei have access merely by diffusion (VII), nor
on being blown in from a wide vessel (VIII), the charge vanishing on the way.
The leakage is not exceeded when a slow current of highly nucleated air stored
over water is passed through the condenser (V), and but slightly for the case
of a fast current of such nucleated air (VI) taken out of the receiver A in the
figure.

By contrast, the excessive ionization (a=.2—.3), if the nucleation 1s at
once introduced into the condenser, is striking enough. Hence scarcely 3 per
cent. of the original ionization has survived after short storage in the receiver
in spite of the extreme density of nucleation which the coronas would cotem-
poraneously show. In fact, the ionization dies out almost at once in the con-
denser (X—XIV), even in the absence of water vapor.

g. Further data.—The experiments were now repeated as in table 3, with
the résumé of results shown in table 4. These are substantially like the above.
With fresh phosphorus the residual ionization of the nuclei-bearing air after short
storage over water is but a few per cent. of the original ionization. It was sup-
posed that on drying the nucleated air over phosphorus pentoxide, before
passing it into the condenser, the original ionization might be in part regained,
but the table shows not a trace of this.

TABLE 3.—IONIZATION OF PHOSPHORUS NUCLEI. dV/dt=ABOUT 2 LIT./MIN.
NEGATIVE CHARGE.

Time. Potential. a Remarks.
tm. Ss

o 15.8 .009 Room air (insulation).

2 15.2

4 14.6

6 13.9

° 20.5 024 Slow current of P nuclei in damp air.

I 19.4

2 18.5

3 17.6

° 18.9 .OIT Filtered moist air.

I 18.4

4 17-9

3 17.5

° Parait .O10 Room air.

I 20.7

2 20.2

3 19.7

"7.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 9

TABLE 3—Continued.

Time. Potential. a Remarks,
tm s
° 19.6 O21 Current of P nuclei in damp air passes over P,O,.
I 18.6
2 a
3 17.0
° 10.8 .OL7 P nuclei in damp air without P,O, (small deflection).
I 10.4
2 10.0
3 9.6
° 23.1 008 Room air.
I 22.6
2 222)
3 21.9
° 20.9 .o18 P nuclei in damp air (large deflection).
I 20.2
2 19.4
B 18.4
° 24.7 264 P nuclei directly from ionizer.
I 13.4
2 Fer
3 4.1
° 26.0 286 Do. Aspirator more constant.
I 14.4
2 6.9
3 3-9
TABLE 4.—RESUME.
a X 103 a X 103 corrected.
sated OMe (ROOMMAIT ecretscaye cies iets slashes mace = oe Sesame OSI 8-10 °
IPiilhwemexelGleinoy onehbey mas tates Ceee ae eee cme Sniapesc soe II °
feNiveleatedidamp ain (x)i(@resh P)io. 222. 2.222420. 02 cee 24 14
1) Nucleated damp air partially dried.................... 2I Ir
CUES Sita cedcrade Habe sae te 17 7
(G) Be ened ferrocene 18 8
, | hosphonuspnticleidinect: i)inajses cpa islet reece = 260 250
(@)e ashes Ge eaters eke 290 280

1 From condensation chamber.

2 From ionizer.

The charge in the condenser is negative as before. It should be more
rapidly dissipated if negative ions are precipitated more rapidly in the receiver,
than a positive charge.
aspirated over phosphorus into the receiver, from which it was then discharged
as expeditiously as possible, the time taken being from 5—ro minutes.

10. Effect of different charges in the condenser.—In the next experiments
the sign of the charge was varied. To find comparable results it was thus neces-
sary to maintain a definite current through the condenser, and about 2.5 liters
per minute was adopted compatibly with the dimensions of the apparatus.

To obtain dense nucleation, room air was again

1O A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

Nuclei were again aspirated into the receiver over phosphorus. The leakage was
apparently different for charges of opposed sign; but this was due to the in-
sulation of the condenser, which is greater for negative than for positive charges.
Deducting this, the values of the ionization, a, differ by quantities which lie
within the errors of observation.

As the receiver in the course of the efflux of nucleated air shows fogs of
continually increasing density, the spontaneous precipitation must have been
equally effective for positive and for negative nuclei.

TABLE 5—IONIZATION OF PHOSPHORUS NUCLEI. CHARGE IN CONDENSER
AT 10 VOLTS. dV/dt=2.5 LITERS/MIN.

Charge. Time. Potential. a Remarks.
m
~ ° 15.8 025 Nucleated damp air.
I 14.9
2 14.0
3 13.3
+ ° 16.0 .016 Insulation (air ionized ?).
I nce
2 14.8
3 14.3
aa ° E535 .022 Nucleated damp air.
I 14.5
2 13.8
3 13-3
+ fo) 15.3 .022 Nucleated damp air.
I 14.4
2 13.8
8 13.0
_ 17.9 OI Insulation.
17.3
16.9
16.6

TABLE 6.—RESULTS.

a X 103 (cor-

a X 103 rected.) Insulation a X 103 = 10-16.
fom 25 =O)
+ 22 +6
+ 22 +6
= 18 —§
ae 24 +8
= 17 aed

In tables 7, 8, similar results are given, but with the insulation tested after
each passage of nucleated air through the condenser. The results taken con-
secutively are shown in table 8.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. Tor

TABLE 7.—IONIZATION OF PHOSPHORUS NUCLEI. dV/di=2.5 LITERS/MIN.
CHARGE AT 20 VOLTS, 19.6 cm. DEFLECTION.

Charge. Time.

Potential.

aanda
corrected.

a

Insulation.

Remarks.

scale pts.
at °

WH HH OWN BH OW D FH OWN BH OW D HOW DH OW DH OW ND HOW DN

scale pts.
18.4
17.6
16.9
16.3
19.0
18.2
7
14.2
173
16.7
15-9
15.4
20.9
20.2
19.8
19.4
Lyn 7,
16.5
104
15.0
16.2
15.4
14.7
14.4
15-4
14.6
13.8
Teer
16.9
16.2
15-5
TAG
18.8
18.6
18.1

17-7

.o18
.009

-O14
-O12

.o18
.009

.O10
.009

.023
-OLL

.or8

.O17

SOLO

.009

-002

.009

-OOT

-O12

Nucleated damp air.

Insulation.

Charge +

+ 1+

a=.009

I2

oe
II

TABLE 8.—RESUME.

09 | Mean .o1o=a.

There is slight excess of leakage for negative charges; but as the insulation
was 10?Xa=g~12 for positive, and 1-2 for negative charges, these differences

are within the uncertainties of observation.

One may again note that if the

12 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

dense spontaneous fogs in the receiver were associated with selective condensa-
tion, then negative charges should be more rapidly dispelled.

11. Dried emanation.—In the following experiments (table 9) the phos-
phorus was previously dried over calcic chloride, and then introduced into the
receiver from the desiccator and the shortest possible connecting tube. Dry
ions thus suddenly came in contact with water vapor, and it was supposed that
an unequal reduction of positive and negative ionization might ensue. The
ions were stored less than 5 minutes in the receiver, the shortest time practi-
cable. Insulation of the electrometer and parts was determined before and
after each measurement with nucleated air.

TABLE 9.—IONIZATION OF PHOSPHORUS NUCLEI.* dV/dt=2.5 LIT./MIN.
TESTED AFTER 4-5 MIN.

Insulation a

Charge. Time. Deflection. a A Corrected.

s Before After

~

14.4 .O152 .0137 .O120 .0023
13.8
13.4
12.9
15.3 .0070 .0070 .0046 .0022
15.0
14.8
14.6
14.7 0.0135 .0096 -0092 .0O41
14.3
13.8
13.4

WwHH OWN HH OWN SO

* Phosphorus nuclei dried over calcic chloride and conveyed by a dry current of air, before storing over water

The residual ionization so obtained, a=.002—.004 for positive and .oo2 for
negative charges, is smaller than heretofore, but again practically neutral.
Thus very dry phosphorus nuclei seem to lose their ionization quicker than if
placed in ordinary air; but changes of the activity of the ionizer may account
for the difference.

12. Wet emanation.—For contrast, the nuclei were conveyed into the
receiver (table 10) in a wet current of air passing over phosphorus. A U-tube
was used, one leg of which contained wet sponges and the other the phosphorus
grid, the damp air from the former sweeping over the latter into the receiver.
The ionization found is distinctly greater! than the dry air data of the last
table, though it does not exceed the usual values for room air. The difficulty
of keeping the ionizing activity constant is again involved. According to the
table, positive charges are more rapidly discharged than negative charges,
which would indicate an excess of negative ions comparable to the case of
water nuclei in the next chapter. .

* American Fournal of Science (4), Xl, p. 327, 190t.

_— ie? ————<~ Geen wee ee

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 13

TABLE 1ro.—IONIZATION OF PHOSPHORUS NUCLEI. dV/dt=2.5 LIT./MIN.
TESTED AFTER 5-10 MIN.

: : Insulation a
Charge. Time. Deflection. a ee Gorectenk
t Ss Before After
° 8.9 .0207 .0098 -0098 -O109
(to cells) I 8.4
2 8.0
3 7:7
= ° 10.4 -O1I7 .0054 .0089 .0046
(to cells) I 10.2
2 9-9
3 9.6

1 Phosphorus nuclei conveyed in a damp current of air. Tested for coronas, the nucleated air of the
receiver shows the usual strength.

13. Residual ionization after one hour.—The storage of phosphorus nuclei
over water in the above experiments did not exceed 10-15 minutes. It was
thought that by giving the fog particles more time to subside, the sign of the
residual ionization might become apparent, supposing that more negative
nuclei are precipitated. The results in table rr are peculiar: whereas the
positive charges in the presence of the nucleation vanish more slowly than for
room air, the negative charge vanishes much faster. This would make an

TABLE 11.—RESIDUAL IONIZATION OF PHOSPHORUS NUCLEI AFTER ONE
HOUR. dV/dt=2.5 LIT./MIN. CHARGE AT ABOUT 20 VOLTS, OR LESS.

a

Charge. Time. Deflection. a Comacted! Remarks.
t Ss ,
fo) 23.9 .O121 .OT2t Room air.
(to cells) I 23.3
2 Doe
3 22.0
4 aed
5 20.9
+ ° 20.6 .O107 .OOT4 Damp nucleated air.
I 20.0
2 19.5
3 19.0
4 18.6
5 18.2 ‘
a ° 24.7 .0068 .0068 Damp nucleated air.
(7 cells) I 24.4
2 24.1
3 23.6
4 23.2 ;
a ° 23.0 .0036 + .0032 Room air.
I 227,
2 22.5
3 22.4
4 22.2
5 22.0

14 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

excess of a=.oo5 of residual positive nucleation over the negative nucleation,
and it would follow that the negative nuclei are precipitated faster. But with
a leakage in the electrometer of .o11 and .o04 respectively, without any ionized
medium, the result is not guaranteed, particularly as the positive leakage is large.

14. Nucleation partially precipitated —The nucleated air stored over water
in the receiver, A, was suddenly cooled and allowed to subside 5-10 minutes.
In this way greater chance was given for the differentiation of positive and
negative nuclei. The corrected values of a show that positive charge is removed
faster than negative charge, by a.=.oo1o and .oo4 respectively, implying excess
of negative nuclei. With the insulation varying from .0o3—.010 and .003-.013,
before and after some of the measurements, this result is again doubtful. It
should be noticed that it is the reverse of the preceding.

TABLE 12.—RESIDUAL IONIZATION AFTER PARTIAL PRECIPITATION.
dV /dt=2.5 LIT./MIN. CHARGE AT to VOLTS.

a

Charge. Time. Deflection. a Corected. Remarks.

t Ss

— ° 13.0 .013 .004 Nucleated air. Subsiding 4™ without ex-

(5 cells) I 12.9 haustion and 6" with exhaustion. Insu-

2 12.4 lation before .003 =a.
3 12.0

— I 11.0 .O15 ° Room air.
2 10.6
3 10.3
4 9-9

ao I 17.4 .022 .O14 Nucleated air. Subsiding 4™ without ex-
2 16.6 haustion and 8" with exhaustion. Insu-
3 15.6 lation before .co2 =a.
4 15.0 :
5 14.1

+ I 13.8 .013 Room air.
2 0327
3 13-4
4 13.0
5 12.4
6 1250

_ I 11.8 .O19 .O10 Nucleated air. Subsiding 4™ without ex-
2 I1.5 haustion and r2™ with exhaustion. Insu-
3 II. lation before .0079 =a.
4 10.5
5 9-9 bk

— I 9.6 .009 Room air.
2 9-3
3 9.2
4 9.0
5 8.8
6 8.5

15. Jonization of dry phosphorus nuclet.—In the present experiments a
dry vessel of 10 liters capacity was introduced between the tower desiccator
and the tube, P, figures 1 and 2. Phosphorus nuclei were now aspirated into this

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 15

vessel. They were discharged after 5 minutes into the condenser after removing
the phosphorus tube. Table 13 shows that almost all the ionization is lost by
this dry storage as the excess of leakage due to the discharge of the dry air.
through the condenser is a=.000 and .007, respectively. aes dry air shows
no preservative effect.

TABLE 13.—RESIDUAL IONIZATION AFTER DRY STORAGE OF PHOSPHORUS
NUCLEI! dV/dt=2.5 LIT./MIN. CHARGE AT to VOLTS.

Ns Deflection. a
Charge. Time. es a Gorectads

Remarks.

14.4 .006 .000 Phosphorus stored dry 25
14.1
13.9
Tete
7 .006 Room air.
11.6
11.4
Tn
a7, OL .007 Phosphorus stored dry 2-5"
13.4
13.0
12.6
ley .004 Room air.
11.6
Crs,
11.4

(5 cells)

WONHH OW HH OWN HH OWN HO

1 Intense antecoronal fogs obtained in the exhausted receiver, A, due to back motion of the nuclei into
the receiver through the aspirators.

In table 14, by a modification of the apparatus, a current of dry air passes
from the desiccator over phosphorus, and then into one end of a dry vessel of
ro liters capacity. At the other end of the vessel the air is continually dis-
charged into the condenser. But the usual negative result appears.

TABLE 14.—DRIED PHOSPHORUS IONS PASSED INTO A DRY VESSEL AND THEN
CONTINUOUSLY INTO THE CONDENSER. dV/dt=2.5 LIT./MIN. CHARGE
AT 1o AND 20 VOLTS.

Charge. a Remarks.
_ .O17 P ions.
(10 cells) — .000 Room air before and after.
(Gsicells)— .O12 P ions. —
.000 Room air before and after.

16. Inferences——The chief result of the investigation is the enormously
rapid reduction of the ionization of the phosphorus emanation in contrast with
the persistence of the nucleation. In other words, only a few per cent. of the

10 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

original ions are associated with the nuclei by which the dense fogs and coronal
sequences are produced, even at the outset. It was with a view to possibly
restoring some of this lost ionization that the great variety of experiments
detailed in the chapter were undertaken; but not a trace of restoration was
observable in any case.

Moreover, the whole of the original ionization vanishes symmetrically, for
the nuclei as a whole are neutral throughout. At least with so insignificant a
residue of the original ionization, the decision as to whether more positive or
more negative ions have vanished is a delicate one and of trifling interest in
this connection. For the phenomena are now all of the order of the leakage of
the electrometer and appurtenances. If when 100 ionized nuclei of the phos-
phorus emanation are suddenly introduced into an atmosphere saturated with
water vapor, the ionization of 96 has vanished without a record, while the re-
maining 4 are in equal number positive and negative, it is unlikely that negative
ions can have greater affinity for water vapor or be more remarkable in their
efficiency as condensation nuclei than positive ions.

Finally, it does not appear that the ionization lost so soon after the removal
of the emanation from the phosphorus surface can ever be restored, notwith-
standing the fact that in the condensation chamber nuclei may be made to
pass from the fog particle to the nuclear stage of size and density an indefinite
number of times. Hence it follows that it is the ante-nuclear stage by which
the ionization is introduced.

In one respect the experiments with phosphorus are unsatisfactory: it
takes some time before the condensation chamber can be adjusted for condensa-
tion on the phosphorus nucleus as it is necessary to introduce the nucleation
from without. Placed within the chamber over water, phosphorus emits a
dense filament of smoke, and is relatively inefficient. With water nuclei there
is no such difficulty; for here the nuclei are most efficiently produced in the
condensation chamber itself, while the ionization may be studied without loss
of time. The results are given in the succeeding chapter.

As a whole, the experiments agree well with the original hypothesis, that
nuclei and ions are distinct entities; that the former constitute the residual
product left after the corpuscles representing the ionization have been expelled.
Radio-activity in case of the relatively gentle breakdown of molecular structure
here in question can hardly be anticipated. If a nucleus, lke that of phos-
phorus, for instance, shows a tendency to grow continuously, until it finally
appears as part of a visible smoke, there may be continuous ejection of electrons.
In such a case electric conduction through the gas freighted with these nuclei
would follow Ohm’s law, as is actually the case for phosphorus.

CHAPTER II.

RELATION OF THE IONIZATION AND THE NUCLEATION ASSOCIATED WITH WATER
NUCLEI, PRODUCED IN AIR.

IONIZATION PRODUCED BY SHAKING SOLUTIONS.

1. Introduction.—In the present chapter there will be three subjects for
consideration. The endeavor must first be made to detect ionization in water
nuclei, produced as in the former memoir,! by shaking solutions. As the
ionization so recognized proves to be inadequate, the problem next in order will

Earth ,

FiGuRE 1.—APPARATUS FOR COMPARING NUCLEATION AND IONIZATION. A, CORONAL CHAMBER; R,
Exuaustion Reservorr; M, Marrotre Frask;C, TUBULAR CONDENSER; P, (TO BE INSERTED IN THE CON-
VEYANCE TUBE, e), PHosPHORUS IoNIzER; G,G!, Vacuum Gauces; F, Corton Fitter; Bb, STORAGE BATTERY
TERMINALS. TuBES, p TO SucTION PumP,a TO ATMOSPHERE, C,C!,FOR EXHAUSTION, @ FOR CONVEYING THE
NUCLEI INTO THE CONDENSER, d FOR FILTRATION, FROM THE HypDRANT. EXTREMELY FINE JETS OF
Water SHOOT OUT FROM THE NEEDLE-PRICKED LEAD PipPE,j7. Supports g, g, ARE METALLIC, 7,2, Insu-
LATING.

be the production of the maximum number of nuclei per cubic centimeter
possible, and will be considered in the second section of the chapter. The third
will then take up and examine the ionization produced along lines very similar

1 Structure of the Nucleus, Chap. V.
17

18 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

to those of the preceding chapter, and will lead to an examination of the infer-
ences already drawn for phosphorus nuclei.

2. Apparatus —To detect the electrification of nuclei produced by shaking,
a io per cent. solution of sodic sulphate was employed, as this had been shown
to produce the nuclei in largest numbers under otherwise like conditions. The
solution was used both to generate the nuclei by vigorously shaking the large
aspirator flask, A, figure 1, containing a small amount of the solution at the
bottom, w, as well as to discharge the nuclei into the condenser, C, by filling A
with the liquid coming from the large Mariotte flask, 17. Raising or lowering
the latter enabled the observer either to add liquid to A, or to withdraw it.
When a constant pressure was needed the device shown in the preceding chapter
(figure 2) was used, and the liquid passed along a side tube. The nucleated air
is discharged by the pipe, e s, into the tubular condenser, C, entering ats. The
core of the condenser is highly charged and connected with one pair of quadrants
of the electrometer, EF, the other pair being earthed and the needle kept charged
by a water battery, whose other pole is also earthed. The supports, 7 7, are of
hard rubber, those (g g) of the outer tube of the condenser are metallic, and
together with the base of the apparatus and the pipe, e s, are kept at zero
potential.

The operations are evident. Having charged the core, the rate of discharge
is found when the current of nuclei traverses the condenser, the volume supplied
per second being read off on a scale attached to A. Results for nucleated air
(shaking) are alternated with results for dust-free air obtained by aid of the
filter and stopcock at F.

3. Results —The results (two independent series) are given in the following
tables. The season was damp and unsuitable for electrostatic work, but the

TABLE 1.—ELECTRIC CHARGES OF NUCLEI PRODUCED BY SHAKING A DILUTE
SOLUTION OF Na,SO,. CHARGE IN CONDENSER, 40 VOLTS.
a=d(log s)/dt; dV /dt IN LIT. /MIN.

Condenser at dV /dt a Remarks.

+40 volts 7 -020 After vigorous shaking.
— 40 ae te 022 ae ae ee
+40 “ 7 .020 Without shaking.

—4o. * si .022 * fi

TABLE 2—Same as preceding.

—4o volts 3.0 .O10 Filtered air. Without shaking.
+40 3.0 -O10 i a i a
+40 3.0 .008 After vigorous shaking.

—40 3.0 -009 ts i x

1 Structure of the Nucleus, p. 121.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 19

data are sufficient to indicate the extreme smallness of the effect to be observed.
Ohm’s law being assumed as usual, as a rough approximation to the truth, the
readings in scale parts, s, are read off, at the successive times, t, of observation.
The constant then follows as a=d(log s)/dt and dV/dt denotes the number of
liters of air, nucleated or not as stated, put through the condenser. In table 1
the leakage, a, is larger for negative than for positive charges, but the effect is
equally large no matter whether the influx is dust-free and filtered or whether
the nuclei produced by shaking traverse the condenser. Shaking is thus without
an electrical effect, so far as can here be discerned. What has been observed
in both cases is a continuous drift of the needle.

In table 2, for another adjustment, the positive and negative charges leak
out equally fast when dust-free air constitutes the medium of the condenser.
When shaken nuclei circulate through it, the negative charge leaks out more
rapidly than the positive charge, implying positively charged nuclei. But as
the result is no larger than for dust-free air, it is again probable that the mere
drift of the needle is being observed.

The results from both tables are therefore negative, showing that the excess
of leakage for a charge of one sign over that of another must be of the order of
.oor when referred to minutes, if the nuclei in question are produced by shaking.
This result, however, is not unexpected; for it is not more than about a thousand
nuclei that are here available, and with the necessarily small charge residing on
each, a detection of the effect will not easily be accomplished in connection with
moist air. The following pages, moreover, will show how quickly the charge
vanishes, and I am not sure that the experiments were made expeditiously
enough. Hence I waived the experiments temporarily, to be resumed with a
more efficient method of producing water nuclei and during the dryer atmos-
pheric conditions of winter.

THE EFFICIENCY OF NUCLEI-PRODUCING JETS.

4. Powerful methods of comminution.—It appears from the preceding
section and elsewhere that the number of nuclei produced by shaking is rela-
tively very small, and the coronas correspondingly simple. To obtain more
nuclei a much more violent method of comminution must be resorted to, such as
is given if fine jets of water, generated under high pressure, are shattered either
against a solid obstacle or against each other. Fortunately, ordinary hydrant
water contains enough solute to answer the requirements, and the construction
of the jet is thus a straightforward problem. It will be found that for each jet
there is a maximum of productivity, and one is thus able to make a series of
jets, each corresponding to a definite number of nuclei under like conditions.
Each jet has a definite saturation number, and while the maximum saturation
producible in this way is naturally far below the efficiency of phosphorus and
other chemical ionizers, so far as nuclei are concerned, the aggregate ionization
is not very different.

20 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

In table 3 the results obtained with jets of a variety of patterns and num-
bers are given in detail. They were all fed with hydrant water at 60-70 pounds
pressure. In these experiments the jet, 7, replaced the phosphorus, the nuclei
being actually produced from the spray in the vessel, A, figure 1. Details of a
simplified form are given in figure 2, where 7 is the jet to be tested, screwed to a
brass pipe, 7, joined by gas couplings (unions at U, etc.) to the pipe h from the
hydrant. In many of the jets the spray is broken against the sides of the
vessel, this being the most efficient mode of comminution. The excess of water
is carried off by the cock, k, for which there is a side branch, p, with a special
stopcock. When k is closed, the jet may often be used to discharge its own

FIGURE 2.—DETAIL (SIMPLIFIED) or A, Fic. 1 FIGURE 3.—ForMs OF IMPINGING JETS.

nuclei into the condenser without the intervention of the Mariotte flask, M,
figure 1.

The types of jets used are shown in figure 3 in cross section, and the num-
bers in the table correspond to those of the figure. In No. 1, two parallel eighth-
inch lead pipes emit jets from their sides in such a way as to impinge on each
other nearly at right angles. It was not possible to completely or continually
shatter both jets mutually in this way. In No. 2, radial jets from a quarter-
inch lead pipe impinge on the walls of the receiver; in No. 8 the pipe has been
thinned. In No. 3 two capillary adjutages produce jets which impinge on each
other, and the same is the case, with evident modifications, in Nos. 4 and 7. In
No. 6 an oblique lead buffer has been added, while in No. g the jets issue from a
finely perforated copper plate, and impinge turbulently on the pool of water
below. In addition to these, ordinary lava tips (No. 5) and other steatite jets
were used.

‘The holes were usually pricked with fine cambric needles. Of all the jets,

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 21

Nos. 2 and 8 were not only the simplest but the most efficient relative to the
quantity of water consumed. Adjutages introducing resistances gave inferior
results, but are needed for special purposes. To change the efficiency at will,
the number of holes may be varied from two to fifty.

5. Results.—In table 3 the description of the jet, the corona produced
and its serial number, together with the number of nuclei per cubic centimeter
corresponding, are given in each case. Certain data of the aperture, s, are some-
times added for identification. For convenience in comparisons, a brief table
of coronas is subjoined to the next table.

TABLE 3—NUMBER OF NUCLEI PRODUCED BY DIFFERENT JETS. WATER
PRESSURE, 60-70 Ibs.

Kind of Jet. Corona. Ss n Remarks.

Nomi One jet. Simple eG 6000

“1. Two impinging jets. i Ane 30000

eee Ts : . x 4.2 29000 | After 5™ spraying.

ames ee is 4.5 40000 yy aKeye

pene aweniuy: ge’ br 6.0 80000

~ 2. Fifty radial jets. g’ br too0oo | Jets .o4 cm. diam.

2 LOO ; - gbr 80000 | (Lead walls too thick.)

“3. Short adjutages. gy ob 80000 | Jets .1r cm. diam.

“ 3. Do. Mouths flat. weg 50000

~ 3. Finer holes. w rg 4.2 30000 | Jets .o5 cm. diam.

per" . w rig Zxoroeey ||| a? doe

eee Six4 ets: g’|blr 40000 | Punctured sheet lead.

meOo Chree <“ w rig 30000

eee woo + w r|g 4.0 25000 | Jets .o5 cm. diam.

eo mee DIT hye: wpbegr I00000 >) Oe saan

It is frequently difficult to place the coronas, and for this reason, several
cases, one of which is given in table 4, were investigated by successive exhaus-
tions, as explained in the preceding memoir. The identification of the coronas
is then more easily possible, beginning with the green corona. The table also
contains direct measurements of the successive apertures, with the number of
particles computed therefrom. The results show a peculiar periodic discrep-
ancy, the nature of which will be treated at length elsewhere.' For the present,
the apertures serve merely for the identification of coronas.

The results obtained for jets will conveniently be discussed in the next
section in connection with the electrical data there set forth. Here I need only
point out that the maximum nucleation obtainable with jets is obviously de-
pendent on their pressure, number, fineness, etc., and probably on the degree
with which the air current simultaneously generated has been removed. The
presence of this current eventually sweeps out the nuclei as fast as they are
generated, and for this reason it makes no difference whether one begins with

‘ Chapter VI.

22 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

filtered or with room air. A jet free from air currents should be most efficient,
cet. par. Curiously, the maximum nucleation obtainable with jets lies just

TABLE 4.—EFFECTS OF SUCCESSIVE EXHAUSTION. WATER NUCLEI. ESTI-
MATED NUCLEATION.

No. Corona. s n X 1073.
79 | wpbr — 120
13 | wre = 75
14 | w clg — 65
15 | wibjr . — 45
16 w cig 3-9 25
17 w|b|r 23 12
18 | corona 2.8 8
19 3 Bak 6
20 | a 2.2 4

Nucleations corresponding to successive coronas:

No. 5 olive | — No. 13 w,rg 75000
6 wy | 200000 14 welg 65000
7 wo 180000 15 w p cor 55000
8 wr 160000 16 g’ bp 45000
9 we | 140000 17 wrg 35000
10 wp 120000 18 weg 25000
II gbp 100000 19 coronas 20000
12 g’o | goo000

below the maximum nucleation of atmospheric air as found in the winter observa-
tions. Probably this is a mere coincidence.

THE IONIZATION OF WATER NUCLEI.

6. Introductory—In my report to the Smithsonian Institution (August,
1902) and elsewhere,! I pointed out the desirability of further investigations on
the Lenard ? effect. While my work in this direction was in progress, a paper
due to J. J. Thomson’ appeared covering similar ground. Nevertheless, I
shall venture to publish the following results, since the subject is looked at from
a somewhat different point of view, obtained from coronal and other measure-
ments. My chief purpose, however, was to find in what degree the theory given
in my experiment with Ionized Air * was to be modified to meet the conditions

' Science, XVI, p. 633, 1902.

? Lenard, Wied. Ann., xivI, p. 584, 1892.

* J. J. Thomson, “Experiments with Induced Radioactivity of Air, and on the Electrical
Conduction Produced in Gases when they Pass through Water,” Phil. Mag. (6), Iv, p. 352,
September, 1902.

* Smithsonian Contributions to Knowledge, 1309, 1901.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 23

for water nuclei. I have not, however, with the data now in hand been able to
complete this to my satisfaction, and have for this reason confined the present
chapter to experimental work, to the exclusion of theoretical considerations of
the kind given tentatively elsewhere.1

7. Apparatus.—If in the receiver, or condensation chamber, A, the
metallic pipe, e, joining at b, leads directly to the tubular condenser, C (radii
1.05 and .34 cm., length, 50 cm.), the apparatus, figures 1, 2, takes the form
adapted for measuring the initial ionization of the nuclei. If the cocks k and d
(filter) are closed and the fine radial jets are put in action by opening the water
faucet h, the charged air is gradually expelled through b as the water level in A
rises. When an efficient jet is used the rate is usually about 2 liters/minute.
This velocity may be increased or diminished by aid of the flask, M, figure 1,
attached at k.

Since the jets, 7, impinge on the walls of the vessel, this is kept uniformly
moist or better coated with water, and therefore continually put to earth by
the hydrant connection. Similarly the pipe, e, leading to the outer coating of
the condenser is with this continually put to earth. The core of the condenser,
insulated by long hard rubber supports, retains charge well even at high poten-
tials, and in spite of the damp gases, because of the remoteness of the supports.

8. Results. Initial charges—The following table gives the data obtained,
when the nuclei generated by the spray are at once passed into the tubular con-
denser, whose inner surface is charged as stated, the outer being put to earth.

TABLE 5.—IONIZATION OF WATER NUCLEI. dV/dt=2 LIT./MIN. TOTAL
CAPACITY OF ELECTROMETER AND CONDENSER, 72 cm.

Condenser charge. | Insulation before. | a Insulation after. | ds/dt. cm./min.
| =
+20 volts a=.013 .338 a@=.010 | -79
: .490 ae
—20 volts a=.020 .072 a=.002 | 32
.OQT 38
.108 | -40
124 | -40
+20 volts a=.022 2123 aos -79
| 335 | 78
.498 71
.872

The method consisted in testing the insulation of the condenser, immedi-
ately before and after the introduction of water nuclei. The table gives the de-
flection in centimeters, after intervals of 1 min., } min., and 1 min., in each of the
cases, respectively. The conduction, a=6 (log s)/ét, is computed by assuming
Ohm’s law; but in case of the medium of water nuclei, it is seen at once that
Ohm’s law does not apply, and that the conduction, a, increases enormously as

1 American Journal of Science (4), XV, 1903, p. 105; tbid., p. 217.

24 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

the charge on the condenser vanishes. The average value of the conduction,
moreover, is quite of the order of values found for phosphorus nuclei in the above
tables, under similar circumstances. As the data in the table show a for succes-
sive minutes, its variation in the last series is from .223 to .872 in 3 minutes.

Since 2.3 aC=1/R, where C=8/10" farads is the capacity of the condenser
and appurtenances, the initial and final resistance would be

R=25X10° and R=6X 1o° ohms.

It follows then that if the equation of the current be taken, or

ion(U+V)e (E/l) in the usual notation, the number of charged nuclei, 1,

20 volts.

0
onun. 7 R 3 0 7 R 3

FIGURES 4, 5, 6.—Curves SHOwING ELECTROMETER DEFLECTION (LEAKAGE) AFTER CONSECUTIVE
Harr Minutes, FOR DIFFERENT CHARGES AND POTENTIALS IN THE CONDENSER.

increases as the potential difference, E, diminishes. The same is true for the
negative current with a smaller coefficient.

The results for the conduction, a, become more interesting if the electro-
meter deflections are charted graphically in relation to time as in the annexed
figures 4, 5, 6. It is thus seen that the current is surprisingly constant, while
the initial potential difference of about 20 volts gradually quite vanishes.

9g. Comparison with coronas.—The number of ions which may be com-
puted from the given currents is excessive when compared with the number of

ti

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 25

nuclei, particularly when considered in parallel with the corresponding case of
phosphorus. The contrast would be even more striking if the water nuclei
could be tested immediately after production. Several inferences are thus
suggested: either the charge of each nucleus is many electrons, or nuclei are
lost at the outset at a rapid rate (this is disproved by experiment), or each
nucleus emits many electrons.

to. Evanescence of the charges of water nuclei—The same remarkable con-
trast between the initial charges and the subsequent charges on the nuclei that
has been already pointed out for phosphorus will now be observed, if only a
little time is allowed to intervene. Table 6 refers to nuclei produced in the
receiver, A, figure 1, by allowing the accumulating water to run off by the
cock, k. They were then conveyed to the condenser, C, about 5 or 10 minutes
later, by aid of the Mariotte flask, M. The conduction is enormously reduced,
though for positive charges in the condenser it is greater than for negative
charges, showing that an excess of negative nuclei has persisted.

TABLE 6.—IONIZATION OF WATER NUCLEI AFTER 5 MIN. a=6 (log s) /8t.

Condenser at Time f. Deflection s. Observed a. Insulation a. Corrected a.
volts. min. cm. |
+30 fe) 9 O19 — .009 .OLO0
I oD
2 73 |
3 16.6 |
=30 ° 18.9 -005 + .002 | .007
I a7
3 oy)

In table 7 the data refer to nuclei which were left in the vessel A for about
one hour after they were produced. The original ionization has all but

TABLE 7.—IONIZATION OF WATER NUCLEI AFTER +t HOUR. a=6 (logs) /8t.

Condenser at Time t. Deflection s. Observed a. Insulation a. | Corrected a.
=
volts. min. cm. |
+20 ° 20.0 -O13 -009 .004
I 19.3 |
2 18.8 |
| 3 18.3
: 2 =
—20 ° 26.1 .003 — .ooI .002
I 26.1
2 25.8
3 25-7

26 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

vanished; nevertheless there is still an excess of negative nuclei, as shown by the
greater leakage of positive charges in the condenser. IH negative nuclei had been
more rapidly precipitated in the intervening hour in A, the reverse should have
been the case; there should have been an excess of positive nuclei, and negative
charges in the condenser should vanish more rapidly.

Tested for coronas, even after about one hour, 50,000 were left, or over
1 of the original 10° to 2X10°, a result quite out of proportion with the loss
of ionization. The electrical and the condensational phenomena are thus
distinctly separated.

11. Results with an Elliott electrometer —For reasons which need not be
stated, the electrometer of modern type in which the charge is imparted to the
needle through the suspension, notwithstanding its sensitiveness and low capa-
city, was not adapted for further experiments. Accordingly, the data of the
following table were obtained with an ordinary electrometer, with the quad-
rants permanently charged with a water battery. The core of the tubular
condenser communicated with the needle. This adjustment was chosen be-
cause the leakage here was relatively smaller, though the high capacity of needle,
jar, and condenser is unfavorable to sensitiveness. The table contains results in
which the potential of the core of the condenser was altered in steps of one half.
Care was taken to determine the insulation immediately before and after each
measurement with the nucleated medium. The leakage is seen to be always
greater at the beginning than at the end, which is the usual phenomenon of
absorption and release of charge in the insulators. If any trace of radioactivity
occurred it would be obscured by this phenomenon.

The results of this table may be summarized.

TABLE 8._CHARGES OF WATER NUCLEI. dV/dt=2LIT./MIN. CAPACITY 409 cm.
DEFLECTION OF THE ELECTROMETER, s.

Leakage ds/dt per 2™.
: aoe ] j Current
Electrometer Before. 18 | After. C (dE /dt) X 10".
ara en nucleation.
charge : a cm. amperes.
| Gi |
+at 81 volts 13 -64 | .09 Tee
—at 81 volts 7 39 .06 67
+at 40 volts 10 67 | .06 1.47
—at 40 volts is 34 .05 64
+at 20 volts 08 | 63 03 T.45
—at 20 volts .06 2 | .02 57
|

The ionization here is somewhat greater than the preceding, but the differ- _

ence is at once referable to the gradually increasing size of the holes in the lead
jet as the result of long spraying. Slight changes in V lit./min. are now of
importance because of the rapid loss of the charges in the influx tube of the
condenser.

ti

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 27

Whereas the current for positive charges decreases with the potential, the
current for negative charges increases; but this, too, is due to incidental reasons
of the kind mentioned. Seen in the light of the results preceding and following
it, the general evidence of the table is rather to the effect that the current in
the condenser is constant, independent of the electromotive force when the
gradient exceeds about 20 volts per radial centimeter.

These results are given in the chart, figure 7.

2X70" AMD.

{ |

02 _
ovolt” 70 Z0nasOm 40. +5089 COs On) acOmmamnOD

Ficures 7, 8—Curves SHowinc RapiaL CURRENTS (AMPERES) FOR DIFFERENT CHARGES AND
PorTeENTIALS (VOLTS) IN THE CONDENSER.

12. Further data.—A series of results quite similar to the last, but with a
more sensitive electrometer, is given in the next table. As a rule, positive
charges were taken in succession, though a number of incidental data accom-
pany the table. The insulation of the electrometer was found before and after
each measurement with nuclei. Having been taken on different days and not
in a single sweep, the results cannot be quite coincident, because of jet differ-
ences, water pressures, etc., as already stated.

These results are summarized.

The currents are given in the chart marked figure 8. They show that

above 10 volts the currents are practically constant, remembering that any
change in V due to water pressure, etc., will convey the nuclei more rapidly into
the condenser, and from their exceedingly rapid decay at the outset the currents
will necessarily be variable. Below 10 volts the current decreases with the
potential, but remains quite appreciable even when the potential is zero with
the absence of charge in the condenser. The two observations made for the
negative charge indicate similar relations, when taken in connection with the

preceding results.

28 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 9—CHARGES OF WATER NUCLEI. dV/dt=2LIT./MIN. CAPACITY 409 cm.
ELECTROMETER DEFLECTIONS, s.

Leakage ds/dt per 2™.
C (dE/dt)
Electrometer | Before After. Xx 10m.
charge. | cm cm. amperes.
|
|
+at 81 volts 22 17 2.17
—at 80 volts 23 -19 95
+at 81 volts aan 22 1.74
+at 60 volts a2 50g) 1.62
+at 40 volts | onE -O4 1.48
+at 20 volts 05 00 1.54
—at 20 volts a0 105 59
+at 20 volts | .08 02 1.51
+at 20 volts .08 -05 1.66
+at ro volts 02 .00 1.51
+at 2 volts -O1 .00 61
-+-no charge .00 .00 .13
+at 100 volts Ore 2 1.59

Corona on immediate condensation: white, crimson, green, being No. 9 with about
150,000 nuclei per cm}.

* Smaller electrometer factor.

The average number of ions in those cases where positive and negative
charges were observed are again found to be slightly larger than the preceding,
due to further enlargement of the holes of the jet, whereby fresher nuclei are
put into the condenser.

The table states that the most advanced corona obtainable did not exceed
the middle green-blue-purple type of my series,! throughout the whole of the
work. It makes little difference whether the corona is taken instantly or a
few minutes after the jet is shut off. The number of nuclei therefore is constant
throughout the experiments, being about 10°.

13. Jets selj-shattering or impinging on water—To make sure preliminarily
that no induced radioactivity is demonstrable within the limiting potentials to
be employed, the experiments of the following table 10 were devised. Here the
large vertical jet (No. 9, with 18 needle holes, discharging about 8 liters per
minute into the water below and violently churning it) was put in action, and
the air above the water in the aspirator discharged into the condenser by the
rise of level due to the jet. The table shows the insulation before and after the
passage of nucleated air, for different potentials in the condenser. The data
are given in centimeters of deflection per minute (ds/dt). Hence the currents
are 1=ds/dtX2.4X107** amperes.

1 Phil. Mag. (6), 11, pp. 80-91, 1902, corrected in Am. J., xv1, 1903, p. 325, and in
Boltzmann’s Jubelband, p. 204, 1904. Cf. Chap. VI.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 29

TABLE 10. —-ABSENCE OF APPRECIABLE RADIOACTIVITY. VERTICAL JET
NO. 9 IMPINGING ON WATER WITH VIOLENT CHURNING.

Leakage ' before Condenser charge while nuclei Leakage ' after Leakage ' after
expt. are passing. _ expt. expt
positive charge. negative charge.
.07 0 volts 14 23
.00 a ule) 05 -06
.06 aT ON .00 OL
03 qr AQ). “02 .12
.06 120) ae .08 -02
.10 = OO: .07 20
.20 =O Olt 18 .09
ans OOM at .1O si
18 —80 “ a2 -07

1 Condenser at 80 volts in first row. Positive and negative charges follow charges of same sign during
passage of nuclei.

As it is the purpose of the present section to find the nucleation and the
ionization of jets when shattering against a solid obstacle is avoided, the ex-
periments were all made with jets either discharging vertically down into water
or with jets impinging upon each other either vertically or horizontally.

The table shows that the insulation is, as a rule, better (smaller leakage)
after than before the passage of nuclei, for potentials of the same sign, so that no
induced radioactivity can be detected in comparison with the absorption and
release of charges by the apparatus. The absorption phenomenon is strongly
marked when the sign of the charge is changed, as in the two experiments after
nucleation.

TABLE 11.—IONIZATION AND NUCLEATION OF WATER NUCLEI.

Jet. BV fa ©) OORT |i a nee Ue dea aecieaetae
lit./min. volts, amperes. amperes.
No. 9, with 18 holes .o5 cm.| 8 +81 2.2 55 30000
diam. in copper plate. Vio-| 8 — 81 1.0 25
lenitichtumminers - 2 shscs 2. |
No. 9, with 8 finer holes in lead) 2 +—-20O | 12 2 corona
plate. Slight churning... . small
INGER ET Re ech. oakegess 3 +80 1.0 66 30000
4 +80 E.2 .60

INO MGR oie Ger scaieveste thas 3 +80 8 .50 80000
6 +80 2.4 80

INDE EP sessed aco dle aex's: « 2 +81 = i239 120000
+81 — Qtr
+81 — 1.74

The ionization of the air from this turbulently discharging jet was now tested
as in table 11. The coronas obtained did not go further than to correspond

30 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

to about 30,000 nuclei, whereas those of the radial jet, No. 2, correspond
to 105 or more. Hence weak ionization is to be expected, if the two occurrences
go in parallel. The table (part 1) shows the positive and negative currents
when 8 liters of nucleated air pass through the condenser per minute. Hence
for 2 lit./min.

at + 80 volts, t=.55 X10 ™ amperes,

at — 81 volts, \= DK TO e
giving a mean current 7=.40 X10 "', as compared with 1.6 X10 ™* amperes for
the radial jet. Thus the mean current is about 4 times smaller, and the mean
nucleation also about 4 times smaller, as nearly as can be ascertained. There-
fore, the reduction of nucleation and of ionization run in parallel.

For the flat-bottomed lead jet with 8 very fine needle holes, the ionization
was almost inappreciable. The current was about 10°‘? amperes, not much
exceeding the ordinary leakage of the electrometer. The corona was corre-
spondingly small.

The next experiments, made with two capillary threads of water impinging
on each other (jet No. 7), are given in the third part of the table. The different
data for the mean current in case of a positive charge in the condenser and a
supply of 2 lit./min. of nucleated air passing through it show that 7=.6X10°"
amperes. The coronas correspond to about 20,000 or 40,000 nuclei, evidencing
a relatively large ionization.

The last experiments of the table were made with the large oblique jet
No. 3, and the coronas here obtained are just inferior to those of the radial jet,
corresponding to about 80,000 nuclei per cubic centimeter. The positive cur-
rents are about 7=.7 X10°"' amperes. When dV /dt= 3, almost half the ions are
lost in transfer.

Experiments were made at somewhat greater length with two oblique
capillary threads of water shattering each other, as in jet No. 3, above. Special
care was taken to prevent the jet from striking the walls of the vessel. As the
self-shattering was very complete, the spray reached the water with but little
churning. In spite of the small amount of water used, however, relatively
many nuclei were produced, the number estimated from coronas being 40,000
per cubic centimeter.

Moreover, to keep the conditions more uniform, the water level in A, figure
1, was kept constant (efflux from k (Fig. 1) being just as large as the water com-
ing from the spray), and the nuclei were removed by a current of air flowing
through A into the condenser, C, at the rate of 2 liters per minute. The currents
so obtained are relatively large as compared with the number of nuclei, which is
due to the condition that the space in which the nuclei are produced is less than
4-as large as usual above, and that therefore nuclei are fresher on entering the
condenser. The mean positive and negative currents were

at +8o volts, 4=1.1 X to ™ amperes,
at— 8o volts, t= .3 X 10° amperes,
the mean, .7 X10 “' amperes, being about half as large as the currents for radial

jets, whereas the nucleation is deficient.

($ 14).

exceptionally small relatively to the positive currents.
No charge was imparted to the condenser by the spray, the current vanish-
ing with the potential; but this also occurs at times with the radial jet.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 31

The reason will presently appear
The currents for negative charges in the condenser, moreover, are

TABLE 12:.—SPRAY FOUND IN SMALL SPACE AND REMOVED BY AUXILIARY
AIR CURRENT. JET NO. 3, CAPILLARY THREADS, SELF-SHATTERING.

| |
|
Condenser 4X 10% | Meanz X 10 zt X ro per ;
charged to ie observed. corrected. ds/dt 2 lit. /min. | Nucleation.
= — | =
volts. lit./min. amperes. amperes. | | amperes.
+80 2 1.03 I.14 —= 40000
+80 2 Ina at |
+80 2 1.19
+80 2 1.00 | -
+80 2 1.16 | -—
—8o 2 35 30 |
—8o 2 |

TABLE 13—SPRAY FORMED IN GRADUALLY DIMINISHING SPACE AND RE-

MOVED BY RISING SURFACE OF WATER.
| ( .38
+80 B85 1.19 1.04 -44 .62 40000

( -53
37
+80 B55 -99 42
“45

—8o0 325 = aBi5) oo -20

+80 2.5 aaa 72 al 58

+80 1.0 = .29 | — 58

14. Summary of the relative degree of ionization and nucleation.—The
number of nuclei, and the electric conduction of the nucleated air, are quantities
which increase and decrease together. Nuclei and ionization are produced
whether the jet is shattered by a solid obstacle, by two impinging jets, or by
jets impinging on a surface of water, but the efficiency of the spraying arrange-
ment depends on the degree of comminution produced, and in this respect the
jets shattered at the highest velocity by a solid obstacle are preferable. When
the jet is shattered on itself or on a surface of water, the electrical current
vanishes with the potential difference in the condenser, so far as can be seen;
or at least is less than 5 per cent. of the constant positive current. When the

32 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

jet is shattered against a solid, the charging current, for the condenser at zero
potential, is about 8 per cent. of the constant current; but it may also be
absent.

The ratio of the ionization to the nucleation does not always appear as a
fixed quantity; from which it follows that the mean charge per nucleus depends
on incidental conditions of freshness, the nature of the jet, its impact, etc.
Similarly, the ratio of positive to negative ionization does not seem to be a
fixed quantity, but to vary under the same conditions. Nuclei generated in a
small space are more highly charged because they can be more swiftly trans-
ferred to the condenser.

Finally, the maximum nucleation for any jet is reached when as many
nuclei are produced per second as are lost in the same time. Unquestionably
the air current accompanying the action of a violent jet contributes to this loss,
by washing the air against the sides of the vessel and the surface of water.
Hence jets with a strong single direction, even if made up of filimentary jets,
produce few nuclei.

Finally, the reason for the unique efficiency in the capillary oblique jet was
specially verified. Supposing that the high ionization relatively to the nuclea-
tion in this case is due to keeping the water level near the jet and expelling the
nuclei by an auxiliary air current from a small volume, I made the following
experiments in which the nuclei were discharged by a rising surface of water by
aid of the Mariotte flask. Table 13 shows that on successive half minutes the
currents ds/dt increase rapidly, as was supposed. Moreover, when referred to
an efflux of 2 liters per minute, the amperes are now actually of the low order
corresponding to the nucleation of the jet.

The effect of different volumes is also seen from the table, which proves that
proportionality is roughly admissible. This also follows necessarily from the
equation of the phenomenon given elsewhere (cf. § 18), and has been carefully
verified for phosphorus.

15. Spontaneous time loss of nuclet.—The following table (14) shows the
spontaneous loss of nuclei in the lapse of time. The nuclei were produced in the
receiver in the usual way, and their number was then determined by the con-
densation produced after a stated interval. The approximate number or order
of the corona in my series is nevertheless somewhat difficult to determine, and
the number of nuclei estimated therefrom not quite definite. As this number
is an exponent, arithmetical progression indicates geometric progression in the
number of nuclei.

The radial jet, No. 2, shattering itself against the sides of the vessel is
strongest as a nuclei producer, and the large oblique jet, No. 3, considerably
below it in efficiency. The capillary oblique jet, No. 3, is remarkably efficient
relatively to the quantity of water used. The vertical large copper jet, No. 9,
used in tables ro and 11, is a very poor producer of nuclei, though using about
8 liters of water per minute and in spite of the turbulent churning of the pool
below.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 33

TABLE 14.—LOSS OF WATER NUCLEI IN LAPSE OF TIME. ESTIMATED.

Jet. Lapse of time. Corona. Nucleation.
min. | |
No. 2. Radial. of | w bryyb g | 120000
a8 wre 140000
6.0 | gbrbr T00000
10.0 wr|g 75000
a3 w br bg 120000

No. 3. Large.! Eo | gbrbr 100000
R32 gy brbr go000

6.0 Vacs 70000

a3 gy brb goo000

10.0 | wee 65000

Ba wrbg 75000

3 | y’tg 80000

No. 3. Capillary. Eg gyrbr | 50000
a3 gy brbr 45000

10.0 gy brbr 45000

|
| |

1 Note that the second jet gradually loses efficiency.

16. Effect of condensation on itonization.—The following table shows the
effect of precipitation on the ionization of the water nuclei. The original current
(without condensation) is given both for positive and for negative charges.
The condensation was produced by exhaustion immediately after the jet was
shut off and but a few minutes allowed for subsidence of the fog. Only a small
number of nuclei relatively to the total number therefore can have been re-
moved. On the other hand, however, the original ionization has vanished as
the result of condensation, for the residual currents are below those correspond-
ing to the normal leakage of the condenser and electrometer, and are thus mere
errors. One may note that both the positive and the negative ionization 1s com-
pletely removed by condensation, even though subsidence of fog particles has
been all but excluded.

TABLE 15._IONIZATION OF WATER NUCLEI AFTER PARTIAL PRECIPITATION.

Condenser charge at dV /dt No: oF ere i X 10" corrected. Gasaey Sane
volts. lit./min. amperes. amperes.
+80 2 ° 1.34 aie
+80 eS) I 03 07
—8o0 2 ° 77 03
—8o0 I I .06 .06

—8o0 2 I 03 02

34 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

Another peculiarity may here be referred to: when the radial jet is very
active it is capable of charging the neutral condenser. This, however, is not
usually the case with a jet striking water. The following table, in which s
denotes the galvanometer deflection, shows that the radial jet, when shattered
on a rigid obstacle, does not always convey charge.

TABLE 16—CONDENSER! CHARGED BY RADIAL JET. dV/dt=2 lit./min.

Time, = oO | ym | 2m au | 4™ Su | 6m
Charge, s= — | .0O | .O1 +.00 | +.02 .O4 | —
Leakage, s= .O1 — — — |

- .00 | .00
dV /dt=4.5 lit./min.

Charge, s= —
Leakage, s= | .00

-0O

= | = .0O

1 Combined capacity 409 cm.; 5 volts per scale part.

In these experiments non-symmetrical charge was not detected when the
nuclei from the radial jet passed through the uncharged and insulated condenser.
This was even true when the air current carrying the nuclei was increased to
nearly 5 liters per minute. Reasons for this diversity of behavior have yet to
be sought.

17. Effective condenser length—It is finally desirable to ascertain whether
the charges of water nuclei are actually lost to a few per cent. in the first few
centimeters of the condenser, very near the influx tube. A condenser was
therefore constructed the length of which could be varied by placing earthed
tubes, 27,=2.1 cm. in diameter and of different lengths, /=60, 30, and 15 cms.,
around a fixed charged insulated core, 2r,=.64 cm., concentrically, with the
usual precautions.

RN SSS NN TO

FIGURE 9.—TuBE CONDENSER WITH SLIDING AND REMOVABLE OUTER COATING.

Figure 9 shows the apparatus where r is the inner and C the outer coating
of the condenser, the latter held in the sleeves, ll. The insulators and the

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 35

metallic supports are capable of sliding to and fro in the base plate and of being
clamped in any position. Influx of nuclei occurs at s. To change C, the sleeves
ll are loosened and the rod rr removed, after which the tube C may be slid off
and another inserted. Set screws and clamp screws complete the adjustment
as shown in the figure.

Table 17 shows the results in which the insulation of the condenser was
determined before and after each measurement with the nucleated medium.
The condenser lengths, 60 and 15 cms., are inserted as a sufficient contrast.

TABLE 17._EFFECT OF LENGTH OF TUBULAR CONDENSER. dV/dt=1.9 LIT./
MIN. A=3.5 VOLTS/em. C=409/9 X10" FARADS.

Leakage.

Condenser | Corrected
Length. charge at ds/dt, Observed. BSC eas

Beforc . During.* After.

cm. | volts. em./min. | em. /min. cm./min. amperes.

60 +80 .02 39 .00 1.07
40
“43

= “45
15 +8o .05 oot -00 1.12
42
-47
52
15 +80 .00 -44 .02 1.20
“45
“47
‘ 58
15 — 80 03 25 -00 62
24
20
30

* During the passage of nucleated air through condenser.

It is seen that the currents are certainly quite as large, cet. par., when the
length of the tube condenser is 15 as when it is 60 cm. It is actually larger at
15 cms., due to the gradual enlargement of the needle holes! in the lead jet,
whereby fresher nuclei are conveyed into the condenser. The currents for
positive and for negative charges have the usual relation to each other.

The table shows another interesting fact, already pointed out above, that
the current, ds/dt (per min.), increases as the water level in the receiver rises,
or as the discharge into the condenser is fresher. One naturally inquires what
the maximum charge of each nucleus would be if there were no conveyance tube.

1 The fine holes clog with lead hydrate when the jet is left standing in a damp atmosphere,
and the obstruction is gradually removed by the friction of the water. Old jets long unused
therefore show small electrical currents as compared with new jets.

36 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

The successive values of the table (one minute apart) correspond on the average
to about 16 per cent. per minute. In the present installation this jet was
unable to charge the condenser, the charging current being less than 107%
amperes, about of the same order as the leakage.

_ One may conclude, therefore, that the loss of charge per minute, 7. e., the
electrical current radially traversing the condenser, is practically independent of
its length if the latter exceeds a few centimeters, for the air current and width
given. All but a few per cent. of the charge are lost in the first few centimeters
ahead of the influx tube of the condenser. The experiments are thus in keeping
with the surmise of the preceding paragraphs.

SUMMARY AND INFERENCES.

18. Working hypothests.—In conclusion a brief summary of the working hy-
pothesis from which most of my work has proceeded may be added for reference.

Let the ions be regarded as charged nuclei, and let there be an average of
q electrons per nucleus. Let the loss of ions be due merely to absorption of the
charges at the boundary of the region. This is virtually stating that the loss is
as the first power of the number 1, per cubic centimeter. Whether the charge
travels with the nucleus, or whether it travels from nucleus to nucleus along a
highway of nuclei, as it were, is left open, but the charges are lost at the bound-
ary at a more rapid rate than the nuclei.

To fix the ideas, let a tube condenser of radii r, > 7,, and length, l, be given,
and let v (cm./sec.) be the velocity of the air current bearing charged nuclei
longitudinally through the condenser. If V is the volume of this air in liter/min.
entering the condenser at one end, 2 (7;—17;) v=10.7 V.

The loss of nuclear charges is then due to two causes: (1) These charges
have a specific velocity, k (absorption velocity in a given cardinal direction) in
the absence of the electric field. Charges are lost in pairs by this non-directed
motion without producing current. (2) The nuclei have a second velocity,
U’, in the same direction per electron carried and per volt/cm. of the field.
Hence the number of nuclei, 7, at the section / cm. from the influx end, where
n=Nn,, is given per unit of length by

Na ges ale, = Fea)

where K=k + gEU’/(r,—17,) and #=.377 IK (r, + 1,)/V.

The radial current at the same section, if the potential difference between
the surfaces 7, and 7, of the tube condenser is E, will not depend on k, but on
US so that

—di=2n(r,+r,) neU'@ (E/(r,—1,)) dl
or eventually
16.7 Vn,eq?(l— =)
where C is the capacity of the condenser and e the charge of one electron, while
q such charges travel per nucleus.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 37

Experiment shows that the currents are of about the same order when
charged water nuclei from an intense high-pressure jet and when charged
phosphorus nuclei are passing longitudinally through the condenser. But as
the number of water nuclei as tested by coronas are, even in the coadensation
chamber, not above ro° per cubic centimeter, while the number of phosphorus
nuclei may reach ro’, the charge q in electrons per nucleus is large for water
nuclei and small for phosphorus. Similarly one may expect the water nucleus
derived by a mechanical process to be larger than the initial phosphorus nu-
cleus derived chemically, so that k is larger in the latter case.

Hence it is assumed that in case of water nuclei, k is negligible in com-
parison with gEU’/(r,—r,) and equation (2) becomes, if qU’=U

—C dE/dt=16.7 Vn, (eq) (1—@— 377 (2 +) UE/V (12—1)),
This equation, which fits the phenomena very well, predicts saturation as the
exponent is essentially dependent on FE.

On the other hand, in case of phosphorus nuclei, k is large in comparison
with gEU’/(r,—r,), for here a single electron travels with many nuclei. The
exponential term in (2) vanishes or

— C dE/dt=16.7 Vn,eq?EU'/k (r,.—1,)

which is virtually Ohm’s law.

An endeavor has thus been made to explain the two types of conduction in
question, the charged water nucleus type and the phosphorus nucleus type, by a
simple self-contained hypothesis. I have not, however, been able to complete
the numerical details to my satisfaction, and will therefore leave the subject
here without further comment.

19. Charge and conduction.—The data have shown that positive as well as
negative charges are dissipated by water nuclei, immediately after they have
been produced, and that the ionization, if it may be so called, is quite of the
order of that of phosphorus, while the nucleation is much smaller. After being
stored but a few minutes, the nucleation loses all but a few per cent. of this
property of conduction, behaving in this respect again like phosphorus nuclei.
The number of nuclei does not appreciably vary in the same time. The char-
acter of the ionization (whether positive or negative nuclei are in excess) remains
intact so long as it can be observed. Hence the large initial and the eventual
very small conduction (a few per cent. of the original value) may be regarded
as two successive phases of a single continuous phenomenon, either of charge or
ionization or conduction. It seems to me therefore that it is not necessary to
distinguish the initial charges from the initial ionization. The experiment as a
whole shows an attenuation of the Lenard effect, continuously through infinite
time. One is at liberty to refer the conduction either to charged nuclei or to
ionized nuclei unless some distinctive definition is adopted. Both occurrences
are similarly reduced. The present case of river water is one in which there is
an excess of negative over positive nuclei. In other cases (pure water) the
reverse may be the case, or, again, there may be an absence of an excess of

38 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

either sign. If the nuclei were without charges, however, the medium would
not conduct. In’a condenser, positive or negative charges are sooner dissi-
pated according as there is excess of negative or positive nuclei in the medium,
respectively.

20. Comparison of phosphorus and water nuclet.—Between phosphorus and
water nuclei there is in the first place the essential difference that whereas the
current in the first case obeys Ohm’s law, roughly, it does not do so’ in the
second, being more and more independent of the electromotive force as E
increases above about 15 volts per cm. Similarly, the coronas for water nuclei
usually terminate with the middle g-b-p type, whereas in case of phosphorus
they go to indefinitely higher orders, beyond the first in the series. Parallel
to this there may run a difference in the size of nuclei. The inference is war-
ranted that phosphorus nuclei are small as compared with water nuclei,
inasmuch as the latter owe their origin to mechanical conditions, while the phos-
phorus nuclei arise under molecular conditions and molecular dimensions. As
the observed electric currents are about of the same order in both cases, it follows
that the charges per nucleus are very much larger for water nuclei than for
phosphorus. If water nuclei could be examined immediately after production,
7. é., in the same degree of freshness as is customary for the phosphorus nuclei,
the contrast would be enormous.

In both cases, however, whenever ionization and nucleation are associated
phenomena, the number of ions generated varies directly with the concomitant
number of nuclei.

In other respects there is great similarity in the behavior of the two types
of nuclei. The enormous charges of ionizations at the beginning vanish to a
residuum of a few per cent. in a few minutes if confined by a receptacle, while
the nuclei are not affected either as to number or condensational properties by
the presence or absence of the primitive charge. It is not unreasonable to
suspect, therefore, that the water nucleus, like the phosphorus nucleus, may be
the permanent residue produced by the expulsion of the electrons representing
the ionization: for whenever nucleation and ionization arise in a common
source, any increment of the former is accompanied by a corresponding incre-
ment of the latter.

CHAPTER IIT.

PRELIMINARY SURVEY OF THE APERTURES OF CORONAS, IN RELATION TO THE
NUMBER OF NUCLEI AND THEIR SIZES.

1. Introductory—Throughout my earlier work with coronas, I have relied
chiefly upon the color sequences, and have taken the data for numbers and sizes
of cloud particles (a fixed degree of supersaturation presupposed) from the
tables given elsewhere.'. When apertures were measured this was done chiefly
for the identification of the series to which the corona belongs. There is no
doubt, however, that an expression for the diameters of particles in terms of the
aperture of the coronas would be a great and immediate convenience, par-
ticularly as facility in using the color sequences is apt to be lost, unless one is at
work with them continually. Apart from this, the colors represent steps of
progress, while the apertures should be continuously, even if irregularly, variable.
The purpose is then to find under what conditions the discrepancies of aperture
may be reduced to a minimum.

If the supersaturation is constant throughout, the diameters of cloud
particles and their distance apart will in general be proportional quantities.
Let m be the grammes of water precipitated, » the number of particles per
cubic centimeter, D=n ~’ their distance apart, d the diameter of each, s the
aperture of the corona. If, therefore, for normal coronas d=a/s, where a is a
constant found by purely optical experiments,

n= (6m/ za) s3= (6m/7)/d?=1/D*, and d=D(x/6m)”*.

But it is doubtful if these equations are true even for normal coronas;
they must certainly be a very crude approximation for coronas of the higher
orders, where d and D are possibly both implicated in producing coronal effects.
If one builds up a system of glass plates each sprinkled with lycopodium par-
ticles, the diffraction pattern, which is finely multi-annular for a single plate,
is a mere blur for ro plates placed within a linear foot, for instance, without
changing the aperture appreciably. If the source of light and the eye are both
distant, the coronas gradually lose sharpness and soon cease to be measurable
as the number of plates increases. This indicates that greater uniformity of
distribution and equality of diameter must be met with in case of cloud par-
ticles, but it leaves the question open whether the distance apart of particles is
not from the outset a consideration.

1Am. Fourn. of Science (4), XUl, p. 81, 1902; Phil. Mag. (6), 1v, p. 26, 1902; cf.
Structure of the Nucleus, Smithsonian Contributions, 1903, Chapter III.
39

40 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

What is further menacing is the distortion produced by spherical and
cylindrical vessels, the surfaces of which are rarely quite concentric. In my
work with globes, I assumed that if the annuli showed no distortion and were
small in aperture as compared with the aperture of the globe, distortion could
be neglected. It is questionable, however, if this observation is vouched for,
since the apertures of coronas are peculiarly sensitive to refraction, particularly
when the distances of eye and source from the receiver are purposely chosen
large.

Again, the quantity m is dependent on temperature. It is necessary
therefore to refer coronas to a standard temperature as well as to a given degree
of supersaturation, and the correction is important if the coronas are to be used
in estimating the number of particles.

Finally, the ratio of densities before and after exhaustion is a seriously
difficult datum to determine, for it depends on the degree to which adiabatic
conditions have been attained. It is here that the work is lable to be dis-
crepant. Hence a determination of apertures has an ulterior value, for it is not
improbable that the two series of results will mutually interpret each other. The
present chapter bears out this surmise, though it is merely to be regarded as a
rough test of my earlier results (/. c.). An independent survey is made in
Chapter VI with plate-glass vessels.

2. Apparatus and preliminary results —The following charts contain a
preliminary survey of the sequence of coronas, their apertures, and the number
of particles of specified diameter encountered. The data for diameter and
number, d and n, are taken from my work on successive exhaustion (I. c.), where
the experiments are largely non-optical, and they are compared with the corre-
sponding data d’ and n’ which follow from measurements of aperture. The eye
and source of light are distant 1 and 3 meters, respectively, from the condensa-
tion chamber between them. This was here a long cylindrical vessel of as clear
glass as possible, 50 cm. long and 13 cm. in diameter. The observations were
made parallel to the axis, absence of distortion being assumed for the axial plane,
an assumption which was justified by trial comparisons with plate-glass ap-
paratus, though the latter was not quite large enough for the complete survey.
The method of work was otherwise the same as that described in the earlier
papers.

The results of the work may be given without tables in the accompanying
charts, figures 1, 2, 4, 5, 8, 9, in which the old results for d and m (computed from
successive exhaustions) are laid off horizontally, the new results d’ and 1’,
computed (as stated) from the observed aperture, vertically. The discrepancy
of the two sets of data is enormous, and the curves all show sustained period-
icity. All measurements of aperture, s, are made to the inner edge of the
red ring, and show the diameter of the central disc.

3. Diameter of cloud particle.—The variations of d and d’ are on the average
6d=1.46d’, from curve 4, and 6d=1.6éd’ from curve 5. In other words, the
diameters obtained for coronas by computation from the conditions of suc-

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. AI

cessive exhaustion are about 1.5 times larger than the same data estimated
directly from the apertures of the coronas.

Moreover, the new values of diameter d’ show a curious periodicity which
must be peculiar to them, since the old values from the manner in which they

a t

70cm

70

CHART 1.—CURVES I AND 2, RELATIONS OF APERTURES COMPUTED FROM SUCCESSIVE EXHAUSTIONS
(s), AND DirectLy Measurep (s’). Lonc CyrinpricaL RECEIVER, 13 CM.IN DIAMETER. CURVE 3, THE
SaME FoR PLatTE-GLAss APPARATUS 20 CM. DEEP. CORONAS HERE DIFFICULT TO PLACE.

CuRVES 4 AND 5, RELATION OF D1aMETER OF FoG ParTICLE COMPUTED FROM SUCCESSIVE EXHAUSTIONS
(d), AND FROM MEASUREMENTS OF APERTURE (d’), BOTH GIVEN IN CENTIMETERS. LONG CYLINDRICAL
RECEIVER. THE Types oF CORONAS ARE MARKED gr (GREEN CENTERED), cr (WHITE-CRIMSON CENTERED).

CuRVE 5 DROPPED .o002 cM. SMALL Dots REFER TO A SPECIAL SERIES.

42 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

were obtained (geometric progression) cannot be periodic. There is accelerated
increase of diameter toward the crimson types, Nos. 9 and 14, and a falling off
which may even be a retrogression toward the green types, Nos. 4-5, 11—12,
1s-16, these being respectively the crests and troughs of the wave. The

|

!

720

‘00

80

20

6

6 8 70 th 14. 1é i

6 8 70 7R /4- 76

CuHart 2.—CurRVEs 8, 9, 10, Ratios oF NUCLEATION CoMPUTED FROM SUCCESSIVE EXHAUSTIONS (”
PARTICLES PER CUB. CM.), AND FROM MEASURED APERTURES (n’). LonG CYLINDRICAL RECEIVER.

CURVES II AND 12, THE SAME FOR ANOTHER CYLINDRICAL RECEIVER, 20 CM. DEEP. CURVE II REFERS
to WATER NUCLEI, CuRVE 12 TO PHosPHORUS NUCLEI.

CURVES 6 AND 7, CORRESPONDING DIAMETERS OF FoG PARTICLES IN CENTIMETERS.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 43

undulation continues even beyond this, but it is then difficult to identify it, as the
annuli become crowded into the normal coronas. The results are similar in all
the curves. In other data obtained during experiments with jets in Chapter IT,
S$ 4, 5, 17, and given in the present figures 6, 7, 11, 12, the same undulatory line
is encountered both for phosphorus and for water nuclei, with maxima at the
gth and r4th coronas. Here a different vessel (aspirator 32 cm high) 228 cme
in diameter) was used, and the ratio, 6d=1.36d’, is distinctly smaller for this
case, showing the marked influence of distortion due to the vessel. The same
ratio (1.3) will be adduced below in connection with the preliminary experi-
ments with plate-glass apparatus.

4. Nucleation.—Since nd? is constant, remarks of the same general char-
acter may be made for the nucleation, u, except that the discrepancy will be
reciprocal in character and enormously exaggerated. If on the average d=1.5d’,
n= 3.4n; if d=1.3d’, n’=2.2n, but the undulations have now become so sweep-
ing that a ratio can only be inferred for the small coronas.

5. Cause of pertodicity.—If one inquires into the cause of the periodic dis-
crepancies, it appears that the crimson coronas are too small or else the green
coronas too large, for the data computed from exhaustions cannot be periodic.
The former being white-centered with a diffuse red margin, it is impossible to
mistake the outside edge of the first rig for the inside edge. The blue-green
coronas, however, show a uniformly colored disc, and here the first ring may
be of the same color as the disc, and the corona would then be measured to the
outside margin of the first ring. From this point of view only the crimson
coronas are adapted for measurement, and both curves would then give d=1.3d,
and n’=2.2n. Since the curves actually give evidence of diminishing aper-
ture while the droplets certainly decrease in size, this explanation is plausible,
though it does not agree well with the evidence trom normal coronas. The red
and crimson coronas are the only ones which retain the white center, and the
phenomenon may in so far be regarded as similar to the case of normal coronas.

6. Effect of temperature.—The explanation of the discrepancy between
d and d’ (computed from exhaustions and measured from apertures, respectively)
reduces in the most favorable case to d=1.3d’, and for this two explanations
must be examined. Supposing that one does not inadvertently measure into
a ring, the value of m which enters into the computation of d is very variable
with temperature. For 6p=17 cm., for instance,

aitatOn, m= 3.7 X10° grams per cub. cm.
20°, 4.6 X 107° seid ese,
° —6 “cc
3°, 5-7 X10

Since d varies as m3, for the same nucleation the values of d at 10°, 20°,
30°, will be in the ratio of 56, 60, 64, respectively, and the coronas will be in
the same degree smaller. Per degree between 20° and 30° this amounts to
about .8 per cent. of the value at 20°. Hence to bring the values of d computed
from successive exhaustions into coincidence with the data computed from

44 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

apertures would require a temperature excess of nearly 30°, which is out of the
question.

7. Pressure decrement.—As none of the explanations are satisfactory, light
of a different character may be thrown upon the discrepancy by computing by
the approximate method of the earlier memoirs the density ratio, y’, of the gas
after and before exhaustion, corresponding to the observed values, s’. Since if
n is the nucleation, z the order of the corona in a geometric series, b the co-
efficient of time loss, t the time interval between exhaustions,

log n=z (1+ Ot) log y,
the equation corresponding to a different exhaustion ratio would be
log n’=z (1+ bt) log y’
if the same corona, 2, and time interval, ¢, is implied.
Hence log n/log n’=log y/log y’, while n= (6m/za’) s* = As*. Therefore,
log y’/ log y= (log A+ 3 log s’)/(og A +3 log s).
The computed values s=alnz/6m are given and in the chart, figures
1 and 2. From the latter for s=5.0, s’=8.0 to 9.0 cm. From the earlier
memoir,! the value computed for y was .819. Hence
s'=8, y’ = .807,
==; y’ = 804,

whereas y=.819 was the value computed in my work on coronas for the ex-
haustion 76—58 cm.

Since, roughly, y= (p/p.)”", where p=76 and y=1.4, the following
values of 6p obtain:

ocm., 0p=18.0 cm.
.O 19.1
° 19.4

Thus if the pressure decrement on exhaustion had been taken 1 cm. higher
than the observed value, the apertures computed from successive exhaustions
in the former memoir would agree with the average apertures directly measured
in the present paper. Observationally this is out of the question, but it is
nevertheless difficult to know just what pressure is effective in the adiabatically
cooled receiver (cf. Structure of the Nucleus pp. 35, 38), since neither the
isothermal nor the adiabatic conditions will rigorously suffice. The memoir
shows that isothermally y=.764; adiabatically y=.825; adiabatically with
allowance for condensed water y=.819, as already specified. The aperture
data demand y=.805, which is even nearer to the isothermal y than the value
taken.

Incidentally one may note the precision with which y must be entered or
the pressure difference determined, if the observations are to be sufficiently
close to admit of a computation of dand n. In other words, it is probable that
the ratio y may be determined with greater accuracy from the successive aper-

’ Structure of the Nucleus, Chapters III and IV.

Pyle imo aay

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 45

tures as a whole, notwithstanding their periodic character, than by direct meas-
urement. This is what I meant by stating that the two sets of observations
would probably sustain each other, for nobody would be justified in using the
apertures of abnormal coronas, unless such use was suggested and guided by
independent evidence. The subject will be resumed in Chapter VI, and treated
from a point of view different from the present, which is merely tentative.

8. Summary.—tThe result of this paper is then favorable to the use of the
apertures of coronas in place of the colors of the annuli, for estimating the
number of particles corresponding to a given degree of supersaturation at a given
temperature. Full allowance must, however, be made for the occurrence of
periodic variations of aperture in relation to the diameter of the fog particles;
in other words, a given aperture is only of value when qualified by the type of
corona (whether of the crimson or green order) to which the aperture belongs.
Thus it will not in any case be possible to dispense completely with the color
pattern. It was with the object of finding these corrections systematically that I
began a series of experiments (Chapter VI) with new forms of plate-glass appa-
ratus, and I shall there refer to other developments. Homogeneous light,
though in many respects desirable, gives effects so faint as to be useless in
practice.

With the above data I am able to make an independent estimate of the
number of particles in the saturated phosphorus emanation. The number
found for the first fog of the series was (Phil. Mag. (6), Iv, pp. 25-26, 1902)
nN=6X 83,000; since n’=2.2n, n’=6 X 183,000 particles per cub. cm. Now the
density ratio before and after exhaustion is y, so that 1~y is the volume of
saturated emanation added. As this has passed directly and slowly over
excess of phosphorus, it must be very nearly saturated, becoming diluted on
mixture with the dust-free air of the receiver. Hence, if m, particles per cub.
cm. correspond to saturation, (1-y) = ,=6X183,000; or n,=10"X6.
There must therefore be at least 6 million nuclei! per cub. cm. of the air in
contact with a surface of phosphorus. The value following from my electro-
meter work was n,=2X10°. The two methods are absolutely distinct, but
lead to data of the same order. It is because of the general reasonableness of
the data which have followed from my simple hypothesis throughout a very
wide territory of observation that 1 have felt bound to adhere to it.

PLATE-GLASS APPARATUS.

9. Description —To test the results just adduced, the apparatus shown in
Chapter VI, figure 1, and in Chapter VII, figure 1a, was constructed. The
frame, 20 cm. deep, 35 cm. long, 27 cm. high, was of wood, nicely joined, and
covered within and without with a mixture of burgundy pitch and beeswax
while hot. The front and rear faces are of 1/4-inch plate-glass, cemented on by
the same resinous mixture, and further held in place by the wooden clamps,

1 The factor 6 is introduced in conformity with the work of Chapter VI.

46 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

secured by the brass bolts at their ends (see Chapter VII). F is the filtering
attachment with a cock, E the exhausting attachment, P the nucleator. Ther-
mometers show the temperature both of the air within the chamber and of the
water at its bottom. The goniometer is in front and the source of light in the
rear of the apparatus, and the exhaustions are made in the way frequently
described in these memoirs.

In the preliminary results, 6p=4.5 cm. was the pressure decrement on ex-
haustion, t= 23°, the temperature of the saturated air and water. The eye and
the light are at distances 85 and 235 cm. from the central plane of the apparatus.
Since s, for lycopodium is by experiment .75 cm. and d,=.0032 cm., a=d,s,=
.0024. Hence at 20° d=.0024/s and n=6ms?/za’.

The results, sand s’, are constructed in the chart, figure 3, and show the same
general character as the results already discussed. Moreover, since s’=1.3S5,
the two sets of data are more nearly in correspondence here than was the case
with the cylinder above. Definite results of this character for higher values of
op will presently be given (Chapter VI), after a few incidental questions have
been disposed of.

CHAPTER IV.

ON THE NUMBERS OF NUCLEI PRODUCED BY SHAKING DIFFERENT LIQUIDS AND
ON ALLIED DATA.

1. Explanation.—In my report on the nucleus,' I showed that the number
produced in a given mode of comminution was least in pure water, greater in
dilute organic solutions, and still greater in dilute inorganic solutions, all of the
same strength. Results were also given for other solvents than water, in par-
ticular for benzol; but I was unable to reduce the data to the same scale as for
aqueous solvents, as the data needed for the reductions were not at hand. I
have since found that the method of Wilson and Thomson? lends itself to
benzol, and have therefore computed the data over again, as shown in table r.

TABLE 1.—NUMBERS OF NUCLEI PRODUCED BY VIGOROUSLY SHAKING DIF-
FERENT SOLUTIONS IN THE SAME MANNER. CONCENTRATION 1 %.

Solvent. Solute. Number of nuclei per cub. cm.

Water (Pure water) 130
% Sucrose |
Glucose |
Glycerin \. 630
Urea |
Tartaric Acid |
= Na.SO, |
e K,SO,
| Alum

” CaCl ReCl

ie NaCl, HCl {

* Ca2NO, |

H,N NO,

' Al3NO,

c Fe3NO,

- | Na,PO,
Benzol | Naphthalene 3500
Benzol Paraffine 5000

- 1300

2. Data.—The pressure reduction used to effect the condensations was
throughout 6p=16 cm. Hence at about 20° the adiabatic fall of temperature
in case of a benzol-air medium should be as far as — 10.2°, the rise of temperature

1 Smithsonian Contributions to Knowledge, No. 1373, Chap. V, 1903.
> Phil. Mag. (5), XLVI, p. 538, 1898.
47

48 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

thereafter (due to condensed liquid) to 11.3°, and consequently the liquid
benzol precipitated per cubic centimeter m=30.4X10 ° grams. The gonio-
meter factor was a=.0031 =ds, being the product of the diameter d of the fog
particle and the aperture s of the corona. Hence the number of nuclei per
cubic centimeter is finally ~=1.95 (zos)*, all the coronas in question being
normal, excessively intense and brilliant.

This may be compared with water. The corresponding temperature reduc-
tion of the water-air medium is to —7.6°, the rise of temperature due to the
ensuing condensation as far as 9.5°, so that m=4.5X10 ° grams per cubic
centimeter almost 7 times smaller than the corresponding datum for benzol.
When the same goniometer as above is used, therefore, 7=.29 (10s)°.

The curious result thus appears that the number of nuclei produced by a
definite amount of shaking is least for water, about 5 times greater for dilute
organic solutions in water, about ro times greater for dilute inorganic solutions
in water, and about 30 to 4o times greater for dilute solutions of non-conductors |
like naphthalene and paraffine in benzol. It is difficult to even conjecture a
reason for this behavior.

3. Coronas tn general.—The coronas in benzol for the above pressure
differences, 6p, are all normal, even if nucleation from sulphur, phosphorus, etc.,
is introduced. From the slow diffusion of the vapor they soon become distorted
during successive exhaustions unless the vessel is shaken between them. It is
interesting to show, however, that in spite of the normal coronas the high
initial nucleation is fully accounted for. To do this I shall select a series of
observations for coronas in benzol vapor at random (I. c., p. 56). Sulphur
nuclei were used and the vessel shaken between observations. The table gives
the results.

TABLE 2—CORONAS IN BENZOL VAPOR. SULPHUR NUCLEI. oOp=18 cm.
n=6m/nd?. m=33Xt10-°g. Per cub.cm. d=.00144/s.

Exhaustion No. Observed d X 10%. Computed d X 10?. Computed 7.

6800000
3200000
1400000
610000
270000
120000
52000
23000
T0000
4400
1900
850

Fog

mI AnfW NH O

°
ve OR H DOANE

o
SWN Se
N ~I Don O 6
BwWN HAA

4
i

Computed exponentially the initial nucleation would run up into the
millions. The observations are not, however, in keeping with such a locus, and

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 49

conform more closely to 1=d (1/d,—z0/a) or s=s,—oz and ds=a. For
present purposes this is near enough. I shall therefore lay off the aperture, s,
as a linear function of the number of the exhaustion, z, for which the observa-
tions show per unit of z, in case of sulphur nuclei, 6s=.28, and in case of punk
nuclei, ds=.19. The initial aperture computed herefrom as the mean of six
series, in each of which the nucleation was introduced independently, is for
sulphur, s,=3.4 and for punk, s,=2.2. Hence u,=840,000 in the former case
and 1,= 230,000 in the latter.

Since the pressure ratio was in each case 1.36, the nuclei in the influx air
passing over burning sulphur or glowing punk must have been 3.8 times more
numerous. Thus there were nearly 3,000,000 sulphur nuclei and nearly 900,000
punk nuclei per cubic centimeter in the laden air currents entering the con-
densation chamber.

I shall show in Chapter VI that the equation applicable to the present
experiments is

ees
Neto => MET (p—S/s*),
Zi

where 7 is the initial nucleation, y the volume ratio on exhaustion, z the
number of the exhaustion, and S an appropriate subsidence constant. The
function 7 is a product of the terms (1—S/sz) (1—S/sz1,)... (1—S/s2_,),
so that Z is the number of the exhaustion in which the first corona is seen and
1I=1. When the particles are as large as is the case for benzol the subsidence
function is of prevailing importance and masks the exponential function as all
the observations for benzol show. I have carried this method out for water
vapor, obtaining consistent results throughout. The present observations for
benzol are scarcely systematic enough to make it worth while to compute S,
and the experiments should be such in which the diffusion and homogeneity of
vapor is insured by continued rotation of the vessel rather than by shaking.
But there can be no doubt that, with proper precautions in this respect, the
number of nuclei furnished per cubic centimeter by any given nucleator can be
determined with benzol vapor as the coronas are all normal, even for large
values of 6p, with certainty.

4. Axial colors.—It is because of the relatively great number of relatively
large particles in case of benzol and similar hydrocarbon vapors, that the axial
colors are seen, and may be traced into much higher orders than is the case with
water vapor. The yellows, browns, etc., of the first order may be easily ob-
tained with the steam jet, though they cannot be produced in the condensation
chamber by any means except by pressure differences causing intense spon-
taneous condensation in moist air. The subsequent violets, blues, etc., how-
ever, are here distinctly seen as far as the orange red of the second order, after
which the admixture of white light makes recognition of color more and more
difficult. With hydrocarbon liquids like gasolene, benzine, etc., the axial colors
are seen much farther along the series even through a short column, and they
are intense in the drum. The difficulty encountered in observation is due to

5° A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

the slow diffusion and consequent absence of homogeneous vapor. I hope,
however, by keeping the drum in rotation around the axis of vision, as already
suggested, to counteract this discrepancy, and correspondingly to prolong the
series.

5. Carbon disulphide-—The vapor of this reagent is another in which
coarse normal coronas usually appear. The endeavor to produce the higher
coronas with sulphur, punk, or air nuclei fails if the pressure differences are of
the same order as those used for water. Particles of the fog are usually about
d=.oo1 cm. in diameter for strong nucleation, and the strong coronas produced
on shaking showed diameters of the order of d=.0015 under the given conditions
of exhaustion. Relatively large coronas were obtained with nuclei which
apparently rise from this reagent spontaneously. Thus after about 2 hours
d=.oo2, after 6 to 15 hours d=.oo12 cm. was observed. The fact that the
coronas increase in size in the lapse of time suggests other explanations than the
slow diffusion of vapor or the difficulty in keeping it uniformly saturated when
successive exhaustions are made. For in this case coronas would decrease and
the size of particles increase, whereas the reverse is observed.

The computation of the number of nuclei per cubic centimeter for carbon
disulphide is more precarious in view of the high vapor pressures and the de-
ficiency of data applying throughout the range of temperatures involved. For
the case of a pressure decrement of 6p=18 cm., from 76 cm., and at 20°, the
adiabatic fall of temperature would be as far as — 34°, the rise thereafter due to
condensed liquid as far as 5°. This implies 53X10 °° grams of moisture per
cubic centimeter, whence with the above goniometer the number of nuclei per
cubic centimeter would be = 34 (10s)?=.10/(10d)*.

The coronas obtained by spontaneous nucleation thus correspond to
N=13,000 after 3 hours and ~=50,000 after 6 hours or more. Finally, punk
nuclei after two or three exhaustions with shaking were still present to the
number of 775,000 per cubic centimeter.

eh eae

CHAPTER V.
THE DIFFUSION OF VAPOR INTO NUCLEATED AIR; A CORRECTION.

1. Apparatus and mantpulation.—The apparatus with which experiments
of the present kind are made is conveniently described by aid of the accompany-
ing diagram. The appurtenances necessary in practice are given in my report
on the “Structure of the Nucleus” (Smithsonian Contributions, No. 1373, 1903),
to which reference has frequently been made. A, Figure 1, is a tall glass vessel
about one meter high, either cylindrical or rectangular in section, in the latter
case with opposed plate-glass sides. The liquid, L, whose vapors are to be
tested, is placed in the bottom. The wide tube, c, is used for sudden exhaustion,
while a vacuum gauge, g, registers the pressure differences. The tubes a and
b to the top and the bottom of A serve for the admission either of filtered
air or of nucleated air. They are used together, one for influx and the other
for efflux, in connection with the suction of an aspirator.

When the diffusion of the necessarily heavy vapors from L is to be meas-
ured, the air in A is first cleansed of vapor by a current of nucleated air from
ato b. Thereafter the stopcocks are closed at a stated time. If now at a
subsequent time a sudden exhaustion is made in A through c, for a stated
pressure difference, op’, shown at g, the progress of the diffusion may be com-
puted from the height of the fog-bank after an allowance is made for the rise
due to the exhaustion.

On the other hand, if the aspirating current is of filtered air and moves in
the direction from b to a, over the surface of the volatile liquid, the receiver, A,
should become uniformly saturated to a high degree throughout. If nuclei are
added at a stated time below near the surface of the liquid, the corresponding
height of the fog-bank seen on exhaustion at a later time should indicate the
rate at which the nuclei diffuse, if they diffuse more slowly than the residual
concentration of vapor. This method for nuclei, which I pursued with entire
confidence, leads, however, to erroneous results, as the present paper will show ;
for the diffusion of the nuclei is a much more rapid process than the accompany-
ing complications of vapor diffusion.

2. Equation.—To state the case specifically, let p be the vapor pressure
relative to the saturation pressure at the temperature $,, at a time ¢ after
diffusion of vapor commences and at a height x above the surface of the liquid
in the receiver, A. Then from well-known principles it may be shown that

(1) (DS fee dq

where k is the coefficient of pressure diffusion.
51

52 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

If the exhaustion at the time, ¢, is made from air pressure, pé to p’, cor-
responding to the temperatures $, and $’, the relation is approximately 9’/9,
=(p’/pl)*-”*® where a correction for precipitated liquid, etc., is needed.

The vapor pressure corresponding to the reduced temperature, 8’, so ob-
tained after division by the saturation pressure at 9,, is, then, the value of p
in equation (1), which therefore, like « and ¢, is known, so that k may be
computed.

3. Application and data.—In order to have an example for use in the
discussion below, I computed the case for water vapor, which though unsuitable
from its lightness for experiment, is convenient for comparison with other
vapors, almost all of which are heavier than air. The well-known expansion of (1),

p=1—j—{x/2(kt)°— 23/3 X 8(kt)** +25/3 X55 X 32 (ki) *—— +
judiciously manipulated is sufficient for the purpose, though I afterwards
availed myself of the tables in Dienger’s Method of Least Squares in the absence
of larger tables.

The results for water vapor were given in a table, with the time, ¢, in min-
utes and the height of the fog-bank, x, in centimeters.

The table also contained a second series of data, for the case in which the
diffusion takes place into a vapor 4 saturated, to which reference will be made
below. The results of the table may be constructed graphically, showing re-
spectively the advance of diffusion at a given height and at a given time. The

FIGURE 1.—DIFFUSION CHAMBER.

FiGurE 2.—CHART SHOWING THE VAPOR PRESSURES, P, AT DIFFERENT HEIGHTS, %, IN THE LAPSE OF
TIME, WHEN WATER VAPOR DIFFUSES INTO AIR.

FiGuRE 3.—THE SAME, FOR DIFFUSION INTO AIR ORIGINALLY 4 SATURATED.

latter are exhibited in figure 2, in connection with figure 1. From either set of
curves the parabolas which show the rise of a given vapor pressure in the lapse
of time may be obtained by graphic interpolation.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 53

If the exhaustion chosen is such as to reduce the vapor pressure to 4
(this corresponds roughly to a pressure difference of 6p’=17), the intersection
of the vertical line of the figure with the successive curves will show the heights
of the fog-banks on condensation. Thus after 20, 40, or 60 minutes the fog-
banks, having attained heights of, roughly, 22, 33, and 4o centimeters, will be
in good position for observation.

For any other vapor than water, the times will increase inversely as the
coefficients of diffusion. Thus for benzol the time intervals should be increased
about 24 times.

4. Conclustons.—The striking feature of these curves is the extreme slow-
ness of diffusion even for water vapor. At but 20 centimeters above the liquid
surface it takes half an hour to reach semi-saturation. The case is accentuated
for other liquids where the coefficients are smaller, as, for instance, for the follow-
ing liquids at about 20°:

Vapor, H.O CH,O C,H,.O C,H,O (Cal: CS,
ke 123 .16 .12 07 .09 I

If, therefore, the fog particles are relatively numerous, large, and subside
rapidly, the air will soon become highly de-saturated. In other words, if the
air in the receiver A is cleaned of nuclei by condensation, there is no vapor
available to replace the moisture lost.

In case of water vapor the fog particles are small and subside slowly while
the vapor is lighter than air. Hence the latter is liable to be reheated from
the rapid radiation of gases assisted by convection, as stated, before much
de-saturation takes place, unless the vessel is very long and the sides dry. Pre-
cisely the opposite is the case for the hydrocarbon vapors in spite of their vola-
tility, since the fog particles for the same nucleation are larger and fall rapidly,
and where the vapors are heavier than air. After successive precipitations at a
given pressure difference the vapor may be so far de-saturated that it nearly
ceases to condense even if nuclei are present. That it can quite cease to re-
spond is impossible, for some vapor must return to the air after condensation
almost instantaneously; but it is not improbable that a vapor exhausted to a
slightly higher pressure difference will fail to respond thereafter at the original
pressure difference. Thus there is considerable chance for error, and what is
taken for the diffusion of nuclei added near the surface of the liquid may actu-
ally be the diffusion of the liquid itself. This will even be the case if an aspira-
tion current, as in § 1, falls sufficiently short of saturation, supposing always
that the velocity of nuclei is relatively large. True, in the experiments which I
made, the two sets of results for nuclei and for vapors differ radically in order
of values, in distribution among different vapors, while for carbon disulphide
the gradually increasing apertures of the coronas is certain evidence of greater
concentration of nuclei. But these and other occurrences may each in their
turn be explained away.

5. Diffusion from greater to less saturation.—To facilitate the discrimination

54 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

in question, the diffusion of vapor into partially saturated vapor may be com-
puted, as has already been done since

/ 2VRt

p=1+%(p.— 1) f,e-dg

and the initial saturation is p.=} at t=o. The results are constructed in
figure 3, and from them the parabolas showing the rise of the levels of succes-
sively increasing saturation may be derived.

An inspection of figure 3 shows that if the exhaustion were carried some-
what further than corresponds to the lower limit p=4, the fog-banks would
be capped at a definite height, and that the latter would be enormously influ-
enced by slight changes of pressure decrement on exhaustion. Experiment
bears this out. Even for fixed pressure differences (6p) the condensation
must progress with a sweep from the bottom upward, and if the very small
particles last formed evaporate fast enough, an upper demarcation of the fog-
bank will again show itself which would easily be mistaken as a true case of
the diffusion of nuclei. In this way the diffusion of about semi-saturation
(p=.5) into benzol vapor initially about 4 saturated would fully account for
the apparent diffusion of nuclei into benzol vapor shown in the memoir cited.

6. Crucial experiment and conclusion.—Special experiments must therefore
be made to decide whether, when nuclei are added at the bottom of a homo-
geneous column of nearly saturated vapor, the observed diffusion is that of
nuclei through the vapor, or of a greater concentration of vapor through homo-
geneous nucleation. For this purpose it is sufficient to add the nuclei in suc-
cessive experiments at the top and at the bottom of the receiver, A, figure 1.
The nuclei in such a case must diffuse alternately downward and upward, while
the vapor diffuses upward only. Such experiments since made with care
showed that the addition of nuclei above or below the column of vapor is without
effect on the observed diffusion. Hence it follows not only that the diffusion of
the vapor and not of the nuclei has been observed, but that the nuclei must
diffuse much more rapidly than the vapor. Indeed, in the time in which the
nuclei travel from top to bottom of the tall vessel nearly 1 meter high, the
vapor has scarcely risen, and the fog-bank seen on exhaustion lies close to
the surface of the liquid.

An attempt to measure this rapid diffusion of the nucleus in benzol vapor
by the present direct method failed, chiefly because all attempts to rigorously
saturate the air in the receiver with the heavy vapor in a reasonable time were
seriously hampered by convection. The results merely showed that the ve-
locity of the nucleus in benzol vapor must be quite of the same order as in water
vapor, but sharp data could not be obtained.

A curious observation, obtained particularly in the case of coronas from
alcoholic fog particles, deserves mention. Here the tendency to irregular
coronas decreases as the number of nuclei becomes smaller. The final coronas
are generally regular, though small. It follows from this that the diminished

S
i.
,

OO ————— sO Et

mb open

Ea

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 55

saturation in the upper parts of the vessel, due to the precipitation of fog
particles with relatively slow diffusion of vapor, eventually becomes more and
more negligible. Hence the small region from which each nucleus draws its
liquid charge is apparently limited by the low rate of diffusion relatively to the
rate of subsidence of the nucleus. In large coronas, subsidence is slow, and the
region from which vapor reaches the nucleus is correspondingly large. In
small coronas, subsidence is rapid, and the region from which vapor is received
dwindles in local extent.

7. Nuclet produced by the mixture of coal gas and air—Some time ago I
noticed that if coal gas is examined by the steam jet or color tube, as described
elsewhere,' a faint pink flush is seen in the field of the tube. This indicates the
presence of nuclei to the extent of many thousands per cubic centimeter in the
gas. Inasmuch as such nuclei could not be retained in the gas pipes (they
would soon be lost either by subsidence or diffusion), an explanation of the
phenomenon was difficult to suggest. Recently I examined the question by
the aid of the present method of coronas. Coal gas stored over water and
suddenly cooled shows no condensation. It is therefore free from nuclei, as
would be anticipated. Filtered air under the same conditions behaves in the
same way. If, however, coal gas and filtered air are mixed and then examined,
nuclei are abundantly present, to the extent of several thousand per cubic
centimeter, showing that chemical reaction (attributable to the presence of
sulphide gas as an impurity) has taken place.

If the air is introduced above the lighter coal gas, the nuclei are obtained
at once as a result of the mixture, by convection. If the coal gas is introduced
above the air, nuclei are not at first in evidence, but they appear later as the
result of diffusion at the surface of contact. The case of the steam tube is now
obvious, seeing that the gas is here necessarily introduced in contact with air.
These nuclei are not ionized, as special experiments with a condenser showed.
Very probably the product of the oxidation is sulphuric acid.

* See Experiments with Ionized Air, Smithsonian Contributions to Knowledge, No. 1309, 190r.

CHAPTER VI.

PERIODIC COLOR DISTRIBUTIONS IN RELATION TO THE CORONAS OF CLOUDY
CONDENSATION, WITH A REVISION OF THE CONSTANTS OF CORONAS.

INTRODUCTION.

1. Purpose and plan.—The growing importance of costric dust ! in rela-
tion to geophysic phenomena suggested the need of developing a method by
which the atmospheric dust contents could be speedily and systematically
determined. An appropriate method for this purpose was tested in a number
of my earlier papers * which gave promise of being in a measure independent of
merely local or accidental dust distributions. It is based on the measurement
of the angular apertures of the coronas produced on suddenly cooling moist
atmospheric air under definite conditions. Observations of atmospheric nu-
cleation made in this way for about two years show results of considerable
interest.

There is some difficulty, however, in reducing these data to absolute values
(number of nuclei per cubic centimeter), inasmuch as the coronas obtained
with lamp light very frequently pass beyond the ordinary white centered normal
type into the more complex forms corresponding to very small particles. I
have therefore been obliged to make an extended study of coronas. The
method pursued consisted in highly nucleating the air stored within a given
receiver over water (with adequate provision for continued saturation), and
then withdrawing definite amounts of it by successive partial exhaustions. If
the nucleated air is replaced by filtered air free from nuclei, the residual number
of nuclei in the receiver must decrease in geometric progression with the number
of partial exhaustions. The latter, moreover, produce the sudden cooling by
which the coronas are obtained. Let m be the moisture precipitated per cubic
centimeter, in any exhaustion, the number of cloud particles contained, d the

diameter of each: then n=6m/zd*. Since for the successive partial exhaus-
~ tions m is constant, n follows from d, and vice versa.

Two methods are available for the absolute measurement of d. One may

* The pioneering work of Aitken is well known and cited in my earlier papers.
> Science, XVI, p. 948, 1902; Physical Review, xvi, p. 193, 1902; ibid., XVII, p. 233, 1903.
’ Phil. Mag. (6), tv, p. 24, 1902; American Journ. of Science (4), x1, p. 81, 1902;
ibid., XV, Pp. 335, 1903; Physical Review, 1. c.; Smithsonian Contributions to Knowledge, No.
1373, XXIX, pp. 1-176, 1903.
56

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 7

OL

deterniine the apertures of the coronas (so long as these are normal) by a suit-
able goniometer, or one may find the rate of subsidence of the cloud particles.
Both are approximate and limited in scope, as they fail in the cases of the higher
transient coronas. Two methods, furthermore, are available for measuring the
nucleation, 7, or at least relations of m. Aitken’s direct dust counter or a
similar apparatus may be applied (work! with this end in view is given in Chaps.
VII, VIII), or the values of » may be made to decrease geometrically in the
way just specified until normal coronas are obtained, for which d follows from
aperture. For the last of these methods I have already published data; but
in the course of over a year’s additional experimentation a number of new
developments have shown themselves which it is my purpose here to elucidate.
In the first place the method formerly used for determining m gave results
much too small. These are corrected in the present paper. In the second
place, the coronas were supposed to be observed under adiabatic conditions of
temperature; direct experiments in this paper show that the air temperatures
during which the coronas are observed are nearly isothermal. Moreover, the
new results prove that in addition to the systematic loss of nuclei by exhaustion,
as thus fully computed, there is an additional loss which has hitherto escaped me.
Each exhaustion, in fact, is accompanied by a definite loss of nuclei for which
reasons must be investigated (§ 10).

Finally, I have in this chapter used both electric- and mono-chromatic
light as a source, as well as the Welsbach mantel employed for practical pur-
poses. Naturally from the introduction of intense violets the coronas become
more complicated, but it is only in this way that their true nature may be
detected.

2. Apparatus.—The apparatus in which the present experiments were
made differs from the earlier forms merely in the employment of plate-glass
condensation chambers. A variety of forms were used, some bulky and nearly
cubical, like figure r (20 cm. deep, 25 cm. high, 35 cm. long), others (figure 2)
long and narrow (15 cm. deep, 11 cm. high, 55 cm. long). Practically an appa-
ratus 25 cm. deep, ro cm. high, or less, and 60 cm. long would be most generally
suitable. They were all lined, except on the opposed plate-glass faces, with a
double layer of cotton on a copper frame. The chamber is to be mounted on
trunnions, E, t, so as to admit of easy rotation around a horizontal axis at
right angles to the line of vision. Holes, A, A’, must be provided so that the
plate-glass may be cleaned within, with a probang. The chamber carries a
stopcock, F, leading to a cotton filter, and another, P, leading to the nucleator
(preferably phosphorus). The trunnions are wide and hollow, and exhaustion
is made through one of them, F, while the other, ft, is either closed or may
serve for the admission of a thermometer. To produce a definite amount of

1 Aitken’s dust counter may be dispensed with, and the intensity of the nucleator deter-
mined by condensation in benzol vapor, in which the coronas are allnormal. See Smithsonian

Contributions, 1. c., p. 55 et seq.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

FicureE 2.—LoncG CONDENSATION CHAMBER.

FIGURE 3.—VALVE FOR SUDDEN EXHAUSTION.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 59

exhaustion an external vacuum chamber provided with a mercury gauge and
a wide stopcock suffices. A jet pump is suitable for evacuation.

The condensation chamber is placed between the goniometer and the
source of light, nearer the former if large coronas are to be obtained. In my
experiments the distances were usually 84 and 250 cm., or about as 1/3. The
eye at the goniometer is focussed unconsciously on the distant source (334
em).

The exhaustions must be made systematically in connection with a seconds
clock, to admit of allowance for the time losses. The time during which the
fog remains suspended is particularly important, and must be uniform and as
short as possible.!

3. Color distributions.—In classifying the coronas, a statement of the
colors of the first two or three annuli, counted from the center, will usually
suffice. For the case of the electric light the central patch remains white, or
at least opalescent or bluish. With the Welsbach lamp a central disc of vivid
green or green-yellow, or even yellow, is frequently observed; but the use of the
electric light in parallel series shows this to be due to the absence of strong
complimentary blues and violet.

For convenience in specifying color, the following abbreviations will be
used throughout: w, white; p, purple; c, crimson; r, orange-red; br, brown;
0, orange; y, yellow; g, green; b, blue; v, violet. Mixed colors are written
together; thus bg is blue-green, rv red-violet. An accent denotes an approxi-
mation to the color; thus b’is bluish, which has been otherwise indeterminable.
A dot or capital denotes a deep or dark color; thus b or B is dark blue.
A mere line denotes a color ring too narrow or dark to be recognized. This is
the frequent transition from red to green, marked w rg.

Beginning with the most intense nucleation obtainable, 7. e., with particles
of the least size producible, the following coronas appear in succession, at first
filmy and fleeting, but eventually brilliant and dense. The numerals attached
to the series are arbitrary.

iE. On
Il. wveg’; b’ br’; w'gv; wy v bg’; wyove’; wcygv’

There is thus an obvious tendency for the colors succeeding white to follow each
other in the order of wave length, as the particles continually increase in di-
ameter. All intermediate gradations are represented. The second cycle is
nearly complete, the first (?) cannot be obtained except in the opalescent orange
tint, unless the steam jet is employed. The second annulus of any corona
is apt to vary in width so as to be unequally important.

The next series (I11) for successively larger particles is a contraction of the
preceding. There is obviously much overlapping. The following types of
coronas may be cited. The colors are very brilliant. The second “‘green”’

1 Smithsonian Contributions to Knowledge, |. c.

60 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.!

corona is particularly characteristic, consisting of three broad color bands.
The disc is green with the Welsbach lamp.

Ill. wvp b rs we b Pp; w yo (b) gbr; w r (b) igure

The next series (IV) is a variation of w r’ b g r, approaching the steady
normal coronas of the next cycle. The colors are very closely packed together,
so that it is difficult to produce definite types of them at will. Very small
differences of diameter of cloud particle materially change the details of the
color scheme. Incidentally, however, the “‘green’’ corona, wg’ b pis obtained
particularly with the Welsbach lamp; the red of the first ring changes from
y’ to br’. wr |g is frequent.

In succeeding coronas the normal type is practically permanent and the
observable variation is merely in diameter.

4. Apertures.—For the measurement of the relative apertures (s) of the
coronas the inner edge of the first ring or the diameter of the white patch is
unsuitable, because this demarcation is usually vague. On the other hand, the
demarcation between the first and second color rings is usually very sharp
and the colors in contrast. Most of the measurements have therefore been
made to the outer edge of the first ring. Naturally there will be periodi-
city from the fluctuation of wave length specified, but this periodicity per-
sists when homogeneous light is employed. Unfortunately, the coronas are
usually so faint that the simple means for homogeneous light are not available
and electric or sunlight must be employed. For practical purposes, colored
annuli are thus inevitable.

The opalescent colors of the series marked II above soon fade. Evapora-
tion takes place while the partially exhausted air is regaining its original tem-
perature. Particles become irregular with no markedly preponderating size.
Initial coronas are fleeting, final coronas washed. The evaporation effect is
much less evident in Series III and the succeeding series.

DATA OBTAINED WITH THE WELSBACH BURNER.

5. Explanation oj tables—To correlate the present with my earlier in-
vestigations I will give a series of results found by using a small circular part of
the Welsbach mantel as a source of light. Coronas in this case are more easily
identified because of the simplified color scheme, to the practical advantages of
which I have already referred. ;

In table 1, z denotes the number of the partial exhaustions each of volume
ratio, y, and made in succession, ft, the current time in minutes (the interval
being about 3 minutes to allow for adjustments and for diffusion), s the chord
of the angular radius, gy, at radius R, so that s/R=2singy. The eye and source
of light were at distances 85 and 250 cm. from the intervening condensation
chamber, and the former was focussed for long distances. The pressure and
temperature of the atmosphere were P and 6, and the fixed pressure decrement

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 61

on exhaustion uniformly 6p=17 cm., nearly, so that the precipitate per cub.
cm. is m=4.7X10 °. Measurements were made to the outer edge of the first
ring. In the column marked “coronas,” the color of the annuli is specified
from within outward, using the abbreviations stated above. The nucleation is
marked n’ if computed from the aperture s standardized with lycopodium, n’”’
if computed from s standardized by subsidence measurements, N if computed
relatively as a geometric progression, m when the latter is reduced as shown
below, and the absolute values corrected for time and exhaustion losses, etc.
The initial nucleation is shown under n,, and corresponds to s=4. The other
coefficients, 4, referring to time losses, a referring to exhaustion losses, S refer-
ring to subsidence losses, will be presently explained. Though m is measured
for the partially exhausted receiver, a final correction (1/y) need not be added,
for the influx of filtered air leaves the nucleation undisturbed. The ratio n’/N
=275 s3/10°'*” constructed in the charts shows the wide departure from
the constancy which would be anticipated. Diameters of the fog particles are
given under d, the accents referring to the method of computation. All data
will be fully discussed below.

6. Charts ——The charts show the relations of important quantities in the
tables. Thus r=n’/N may be laid off in relation to the number of the exhaus-
tion, 2, as in figure 5; but, generally, the nucleations, m, and finally the di-
ameters, d, of the fog particles, are given in terms of the apertures, s, where the
angular radius P=s/60.

7. Tables.—The data investigated for the Welsbach burner follow.

TABLE 1—CONSTANTS OF CORONAS. WELSBACH LAMP. CONDENSATION
CHAMBER, 20 cm. BROAD, 25 cm. HIGH, 35 cm. LONG; DISTANCES OF EYE
AND SOURCE OF LIGHT FROM CHAMBER, 85 cm. and 250 cm., RESPECT-
IVELY; 9=22°; BAROM., 75.34cm.; dp=16.9; v=.77; @=.064; B=0; S=2.65;
a!’ =.0029 (SUBSIDENCE); a’ =.0032 (LYCOPODIUM); n,= 209000; PHOSPHORUS
NUCLEI. d/’=.0029/s; m=4.7X10°° MEASUREMENT OF s TO OUTER EDGE
OF FIRST RING.

Bi t Ss Corona. | n’=275 s3.| n” =370 s3.) NIZ (a2) No n aa ie
o | min. cm. <ton < ror? (2.84) Kio. 600000 cm.
I 17 wi be Ee (2.19) 460000 .000270
2 20 — |b’r’ — —- (1.69) 355000 290
3 2 — |g’ — -- (1.30) 273000 320
4 27, | 10.3)|.y 0b 300 403 1.000 210000 350
5 30 8.3| weg 157 212 751 158000 385
6 34 6.8! wpb g’ On. | |] aero) 25157 117000 425
MlecmiGoleb p | 59-4 | 79-9 -403 84600 475
8 41 5-8| wre 53-0 | De .289 60700 530
9 44 4.6| w br b'|p | 26.7 36.0 .204 42800 595
10 48 4.2| Wrg 20.4 27.4 3037 200 28800 680
II 51 3-7| corona 13.9 18.8 .0gO 208 19000 780
12 54 ane Pl 9.0 Te -055 220 11500 920
13 57 2.6 4.8 6.5 .032 205 6700 -OOITIO
14 60 2.0 22 3.0 .O15 199 3100 1430
15 63 Te) fi 6 8 .004 220 800 2250

62 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1.—SECOND SERIES. n,=190000; a=.o61.

|
Fe zt | s | Corona. | n’=275 s3.| n”=370s3.| NII G2) No n te
| al
° — — —_ — (2.84) 540000 .000256
I 4 — _- | — (2.19) 416000 278
2 7 — |b’b -- (1.69) 321000 304
3 ro | — | ¢ — | «(@230) 247000 331
4 13 | 10.6] yob 327 440 1.000 190000 362
5 16 | 8.3] weg? G7 212 -751 143000 398
6 | 19 | 60|/gbr 59-4 70:04) mieSSi 106000 440
7 22 5-6] yg o bg 48.4 OF lees oo 75600 492
8 25 5-3| weg 41.0 5 5er -280 53200 555
9 | 28 | 4.4] g’lb'|p 24.2 32.6 .196 37200 623
To 31 4.0| wcg 17.6 23.7 -130 182 24700 715
II 34) 3-6| wbrb‘g’r 12.8 7-3 .083 207 15800 830
12 37 2.9 | corona 6.7 9.0 -050 181 9500 980
13 40 | 2.4 ‘: 3.8 5-1 .027 189 5100 -001 200
14 43 1.9 “sh 1.9 2G .OIL 230 2100 1630
{5 46 5 7%: 9 ne} -002 = 400 2850
16 49 1.0 - aa 4 == a ° a=

1 Gap in coronal sequences probably due to accidental delay. Similar difficulties at end of series.
Timed fog intervals desirable.

METHOD OF REDUCTION.

8. Constants of the geometric progression—To determine whether the
factor of the geometric progression of successive nucleations, z, was to be com-
puted isothermally or adiabatically, a series of direct temperature measurements
was deemed necessary. These were made by aid of a thermocouple of ex-
tremely thin wires (.0o7 cm. in diameter), of copper and german silver. The
junction within the receiver was not soldered, but the flexible copper wire
looped once around the other. In this way the variation of the instantaneous
air temperature in the receiver could be closely followed. It was necessary to
use a sensitive astatic galvanometer, and the measurements are thus subject to
the fluctuations of the earth field. As it is the immediate purpose of these data
to determine about how soon the isothermal conditions are re-established by
radiation from without, the irregularities are of little consequence.

Table 2 contains three series of results, the upper end of each row corre-
sponding to the period of exhaustion, the lower to that of (slow) refilling.
Readings were taken in intervals of half a minute. The table shows that after
the lapse of one minute following the sudden exhaustion the temperature has
been regained to within a degree. As the coronas can hardly be observed and
measured within this time, the exhaustion ratio may be computed isothermally.
I have therefore computed the density ratio of nucleation p/p’=n/n’, before
and after exhaustion, as follows.

Since p=Rp$ in the usual notation of Boyle’s law, and p=P-—p’ where P

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 63

is the reduced reading of the mercury gauge and p’ the vapor pressure of water
vapor,

pip = (R—=p) (2—68/3)/(P’—p’).

The correction 68/9, being by the table .7/293=.0024 or about } per
cent., may be neglected. Hence y= p’/p=1—06p/(P—p’), where sp is the pres-
sure difference selected. Thus in table 1, N, not corrected for time losses, etc.,
would be

N=y'=10°'? where y=.77 and N=10° 73,

In this way the auxiliary ratios n’/N in the tables were constructed. This is
the first departure from the method of my earlier work (Smithsonian Contribu-
itons, No. 1373).

TABLE 2.—AIR TEMPERATURE! IN CONDENSATION CHAMBER DURING AND
AFTER SUDDEN EXHAUSTION AND DURING INFLUX.

Time after exhaustion. 1st Series °C. 2d Series °C. 3d Series °C.
On TON ys = 107° = TOs Sa
a5 mney] eS Sigh
I.0 ee eS a
T.5 ae —- 7 — 4
2.0 — al es
235 stp dies a ee
3-0 1.4 =F Tes : aed:
3-5 7 2 ak
4.0 4 1.1 I
4-5 al aS i
5-0 3 4
5:5 “3
6.0 I

1 Irregular march after exhaustion due to unsteady earth field.
2 Throw on sudden exhaustion.
3 March on filling slowly with filtered air.

9. Time losses—Nuclei apparently decay spontaneously in the lapse of
time, #, and a correction is to be added to N. Since this loss is relatively small
in view of the short time intervals occurring above, n=n,10%"") may be
assumed for convenience. Hence if 2 be the nucleation due to a given corona
seen at low pressure or the identical nucleation at atmospheric pressure after
filtered air has been added, the next corona after z exhaustions and ¢ minutes
will correspond to n,=n 107! ¥*®“"™ whence

B=(x/(t—t,) ) log (n,/n 9’).

04 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

Thus it is merely necessary to know the relative values of m or the nucleation
ratios, and these may be obtained by any method of smoothing, by plotting
n’ or N in terms of s.

In table 1, V has been found for each aperture, s, and 8 computed as stated.
The values of # so found are of the mean order of .oo2. They apparently de-
crease toward the end of each series as the number of nuclei grows smaller; but
this conclusion is not vouched for, as 1t will presently be shown that a specific
loss of nuclei accompanies each exhaustion. In other words, n=Nr, (1— a2),
where 7, is a constant, or the numbers diminish faster than is attributable to
time loss alone. The effect will be particularly marked for the smaller initial
time intervals of the original nucleations.

Accordingly I have made another computation, writing approximately

n/n’ =(N/N')(1 +a (2’—2))

and finding # from this by the equation already given. These values of # are
inserted in the last column of the table. They are irregular in distribution and
of the mean value 6=.0016.

The chief result of this paragraph is the relatively small value of the co-
efficient of time loss of nuclei. Its effect on the results may therefore be neg-
lected (tested), particularly as the effect 10°°*°=1.004 for z=1 is in part
compensated by the temperature factor 68/8=.0024 of the preceding para-
graph. Direct experiments below ($$ 27, 28), where identical results are obtained
for different time intervals (2 min. and 4 min.), vouch for this conclusion. Meas-
urements of coronal aperture of the present kind, hampered as they are by the
difficulties of observation, of reduction, of subsidence, etc., cannot offer more
than an estimate of the conditions sought.

to. Exhaustion losses—I shall next consider the independent destruction
of nuclei which accompanies each exhaustion, and for which it is difficult at the
outset to assign a reason. It will be shown that it is due to the subsidence of
fog particles. This loss did not appear in my original investigations, probably
because the spherical receiver used was not lined with wet cloth.

From what has preceded, the relative number of nuclei after z exhaustions
is 10°'**, whereas in the region of normal coronas the absolute number is
certainly very nearly n’=Cs*, where Cisaconstant. This will be independently
verified presently by special experiments. Hence the ratio r=Cs*/10°'*?
should be constant, whereas experiment and the charts, figure 5, show (roughly)
that r=r, (1— a2), a being the coefficient of exhaustion loss and r,=Cs,° the
arbitrary initial ratio for =o, preceding the first corona observed. Thus

ae— (1—(s/Ss)3/aor 2):

If r=o for z=2’, 1/2’=a, so that a appears as the reciprocal of the abscissa
underlying the positive ordinates in the first quadrant, and may be taken at
once from the graphs. These values are given at the head of each table.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

65

TABLE 3.—CONSTANTS OF CORONAS, COMPUTED BY SUPPOSING a =€ONST.
DISTANCES OF EYE AND SOURCE 85 AND 250 cm. FROM RECEIVER, RE-
SPECTIVELY. @=22°;

a=.062; B=.002.
N,=750000; M=4.7 XIO-

6

P=75.34 cm.;

P IONIZER.

Op=16.0; Y=-750; Ww—N,105-> (a—az);
MEASURED TO OUTER RED, INNER BLUE OR GREEN.

; @=.0032; 6m/nmd? = 2758.

First SERIES.
3 t s Corona. n! — 27583 |N (1 — az) No i x Sai eae
min. cm.

I 1G] nucleation. —

2 20 b’ r’ 504 460000 ==
3 23 ee 355 323000 aa
4 27 10.3 | yob 300000 240 227000 300

5 30 8.3 | wicg 157000 a0713 157000 204
6 34 6.8 | wp’ Beg’ 86300 .120 ITOgo0O 146
7 37 6.0 | g Bp 59400 .082 74700 130
8 41 5-5 | wrg 53600 055 50400 152
9 44 4.6 | w br B\p 26700 .037 33500 99
10 48 4.2 | worg 20400 .024 8500000 21800 98
11 51 3.7 | corona 13900 O15 g 100000 13900 87
12 54 Bee x5 gooo .009 9700000 8500 74-5
13 57 2.6 ie 4800 .005 goo0000 4900 51.0
14 60 2.0 - 2200 .003 8000000 2500 30.0
1S 63 Tes As 600 -OO1 5400000 1000 10.7

mean, 9100000
SECOND SERIES.
Zz t s Corona, n! = 27553 | N (1-@z) Ng Pal i ee i
min. | cm.

I 4 nucleation.

2 7 b’B 504 378000 —
gi 10 g’ 355 266000 —

4 13 10.6 | yob 327000 -249 186000 327

5 16 8.3 | weg 157000 a3 130000 204

6 19 6.0 |gBr 59400 | .120 goooo 100

7 22 5-6 | ygo bg 48400 | .082 | 61600 106

8 25 5-3 | weg 41000 055 41500 II7

9 28 4.4 | g’ Bip 24200 .037 27600 89.5
10 ile ALON | awe Cie 17600 -024 740000 18000 85.0
II 34 3.6 | w br B g’r 12800 O15 840000 11500 79-5
12 37 2.9 | corona 6700 .009 720000 8400 Binag
mye) 40 2.4 ‘S 3800 .005 7 10000 4000 40.0
14 43 1.9 * 1900 .003 680000 2100 26.0
15 40 Das a 930 -OO1 830000 825 16.6
16 49 1.0 270 .000 — — 6.3

Nore:—In N = ro--* the constant is corrected for time loss.

The constant in n’ is obtained from earlier measurements with lycopodium.

has been used in the charts.

This practice was later abandoned. (@ 9).

The auxiliary column n’/N

66 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

The occurrence of different slopes in the curves is an obvious result of the
unavoidable differences of initial nucleation. The effect is a shifting of the z
values. Thus in two series for the same corona similarly observed, but with
independent initial nucleations, r=7r,(1—az)=71',(1—a’2’), where 2’
whence a’=a/(1—az’,), the constant 2’, being the index of the difference of
initial nucleations.

A computation on the assumption of constancy of a was made for com-
parison, and is given in the preceding table, 3, corresponding to table 1 above,
and with the same notation.

11. Exhaustion loss attributable to subsidence—The misleading feature of
a is its apparent constancy for a given receiver and a given scheme of observa-
tions. It will now be shown that this result is a mere approximation and that
the phenomenon may be fully explained in terms of subsidence. In this case,
10* R=g pv, where KF is the radius of the water particle and v its rate of sub-
sidence. Since 2 R=d=.0032/s, approximately, v= (1.78)’/s’, or if v’ refers to
minutes, v’ = 190/s?.

The relative loss, 1, per minute, is for a vessel of height h and nucleation
n, l=v/h=190/hs?. If, as in the above condensation chamber, the height is
h=26.5 cm., l=7.2/s’, or, in tabular form,

= 3 —
e

of -
S107)

s= I 2 3 4 5 em.
l= 7.2 1.8 .80 45 .36

numbers which are astonishingly large, but must be near the truth.

Let the time consumed in observation be 1/2 min. and 1/4 min., respect-
ively; then the ratio r in table 1 may be corrected in the region of normal
coronas as follows:

TABLE 4.— CORRECTION OF r=27553/107'*” IN TABLE 3, SHOWING THE EFFECT
OF SUBSIDENCE.

: | rx103 | 7 X10-3 ||Mean r X 10-3
8 os . | 2 corrected. | ‘ corrected. corrected.
| |
10 98.0 4.2 | 20 98.0 | .10 98.0 98.0
II 87.0 257 .26 108.0 | SLs 97.0 103.0
12 74-5 3.2 | 235 118.5 18 95-5 107.0
13 51.0 2.6 | oS 121.0 28 85.4 103.0
14 30.0 2.0 | go 126.0 | 45 79.0 103.0
15 10.7 3 130.0 | 75 . Re 102.0

The first of these values of 7 is under-corrected, while the second r is over-
corrected. The mean of these corresponding to a time of observation of 3
minute is constant. Now as the time during which the fog is left undispelled
after exhaustion for measurement is actually of this order, there can be no
question but that the error is due to subsidence. The equation should therefore

read:
N=CS—Mro. (5/528) Sy ee)

Sh

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 67

as will be more fully explained below. Here S is an appropriate subsidence con-
stant depending on the mean time of observation or on the time within which
nuclei are suspended in the fog particles. Conformably with this view a or S
depends essentially on the height of the vessel in which the condensation is pro-
duced, being larger, cet. par., in shallow vessels, seeing that v/h is larger. ($25).

12. The optic constant.—The proportionality of diameter with the inverse
aperture may be assumed for normal coronas. The occurrence of periodicity in
the higher coronas, even if merely a question of color were involved, would
modify these simple conditions for these cases. It is well known that for a
single particle, the masterly work of Lommel' has given a complete treatment
of the diffractions in terms of Bessel functions.

In meteorological work for a particle of diameter d and for uniformly nor-
mal coronas, the equation sing=1.22A/d is usually assumed, if the angular
radius of the corona is y and the wave length in question A. Since in my goni-
ometer 2 sing=s/R, where R= 30 cm., ds=a=73.2A. ‘Hence, for the succes-
sive spectrum colors, the following values obtain for the constant a:

color, c if oO y g b Vv
a= .0O51 .0046 .0044 .0042 .0039 .0034 .0029

In view of the theoretical uncertainty of these values in the case of the
distribution of particles met with in the above experiments, I have usually
relied on the results of direct comparisons with the corona of Lycopodium spores
where d,=.0032 cm. Placed in the position of the near plate of the coronal
chamber, the corresponding aperture for lycopodium spores was s,=1.05 cm.,
at the far plate, s,=1.03. Hence a=d.s,=.0034 for measurements to the
outer edge of the first ring. This corresponds to the preceding value for blue,
though these apertures were measured through ruby glass. In the above
tables, where merely relative results were in question, a=.0032, a datum of
earlier measurements, was inserted. Reduction was deemed needless. For
the case of measurements to the inner edge of the first ring or to the edge of the
white disc, s,=.7 to .8, the demarcation being more vague. Assuming the
latter, a=.0026 for such measurements. Still more troublesome is the measure-
ment of a when the condensation chamber is remote (250 cm.) from the goni-
ometer and near (85 cm.) the source. The datum was a=.oo125 from s,=.39.

13. The optic constant. Diameters from subsidence.—This is an independ-
ent method of standardizing s. In my earlier work the condensation chamber
was not cloth-lined, and the subsidence data quite untrustworthy, showing
rapid retardation of abnormally high initial values due to evaporation. In the
present cloth-lined receiver kept wet on all sides, subsidence data are reasonably
satisfactory. The coronas, however, change character during subsidence, and
in case of the initial opalescent coronas (Series IT above) all coronas vanish into
a mere fog before subsidence is even appreciable. Finally, the upper plane

‘ Lommel, Abhandl. der kin. Bayerischen Akad. der Wissensch., xv, 1886.

68 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

boundary of the fog, which at the outset appears as a sharp horizontal line about
50 cm. long, even after 1 or 2 min. becomes more and more vague. Subsidence
is here accelerated. Hence it is chiefly for the normal coronas that subsidence
data are available, and, fortunately, it is precisely here that they are wanted.
In other cases the occurrence of periodicity and the rapid change of coronas
makes the interpretation tedious and difficult.

The following table contains the results for subsidence. Stokes’s well-known
formula reduces to R=g X10 *)'v where R is the radius of the fog particle
sought. Temperature is denoted by 6 and current time by t, depth of the fog
line by x, and rate of subsidence by v. The diameter computed from aperture
is inserted for comparison. 2Rs=a is the new optical constant sought. The
braces denote the observations of x taken to compute v whenever the corona
changes character.

It appears from the earlier parts of this table that although the coronas
may shrink and change appearance, this is not in the same measure the case
with the rate of subsidence. In the latter parts of a given series of observations,
therefore, the diameters from aperture are always in excess of those from sub-
sidence. With the growth of droplets, the two phenomena refer to different
particles, and hence only the original observations, 7. e., those within the first
minute, are of any value.

TABLE s5.—SUBSIDENCE OF FOG PARTICLES. PHOSPHORUS NUCLEI. VIS-
COSITY OF AIR, 7=.00019.

First SERIES.

6§ t Corona. G | x v 2R X 104 10, x 2R.s =a
— =
oC. h. m. cm. cm. cm./sec. em. cm.
GE Cu | LEN Ny lees Wate es 9.1 Onn) .030 3 Bay .0028
2 2 Ki
3 | shrunken | |) ears
4 | and foggy 8.5
z | Sec
6 |gbp 6.4 ; { .047 4.0 5.0 .0026
7 |
8 | wbrb|p ae 542
/ a 5-5
9 |wob 5.8 | -047 4.0 5.5 (.0023)
8.5 {
13, | corona | |) ato [* #2092 54s, jeenO3 .0027
14 5 oh S35
15 | 11.0
16 | -x560
| ==]
° |
21 | corona 2.8 f 2.4 8.8 10.0 .0025
2035 )
22
C7

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 69
SECOND SERIES.
4 t Corona. S x v 2R X tos 104d 2R.s =a
ae yt
mig? |) sete igs || tee ogo) 6.8 ° 025° 2.9 3-4-7 .0020
44 -| wit’ TG
———— 4 —_
fo
48 | wog 9-7 025 2.9 353 .0032
49 | foggy 1.5
50 | wr’ 5.5 042 3.8 5.8 (.co2rt)
4.0
54 | wpb 6.8 ° .042 3.8 4.8 .0025
55 (6.7) 2.5 (.046)
56 | w’ oO’ 6.5 Bay .050
(foggy)
57 | wo’ 5.2 9.0 058 4.1 6.2 (.c021)
12 I | wog 6.0 ° -O71 4-9 5°55 -0029
2 4.0
3 5-5 8.5
4 13.0 075 4.9 6.0 (.0027)
13. | web g’r’ 4.2 oO 158 752 7.8 .0030
13-5 4.1 4
14 | wo’ 4.0 9-5
14.5| w’ 0’ 3-9 I4. 167 7.4 8.0 (.0030)
THIRD SERIES—AIR NUCLEI.
4 t Corona. v 2R X tot | @ X 10¢ | 2k. si=a
he) om.) |
I5°| 12 20 |wbp ory 5-9 7.3 .0026
20.5
21 ;
21.5! w br b|p ar 5.9 Ess (.0025)
22
26 | corona B12 ° 22 8.5 10.0 .0028
26.5 6
27 15

7° A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

FourtH SERIES.

Corona. Ss 2R X 104 | d X 104 |2Rs =a
w br bg’ p 4.7 5.8 6.8 .0027
«
corona 3-4 8.5 9-3 -0029
whbrbg’r Geil ee 6.3 .0026
7
10.0
wog 5-9 ° .078 5.0 5-4 .0020
2
4
7
9
wobg 4.6 ° “133 6.6 7.0 .0030
3-5
7°5
12.0
corona Bor ° 300 9.9 10.3 .0031
6
17
25
weg 5-5 ° .083 5.2 5-7 -0029
2.5
5-0
7-5
wrl|g 4-5 ° au58 7.0 Fis 0031
3-7
9.0
1.3
fe) corona 3:0 ° +300 9-9 10.5 .0030
25 6 ,
ro) 17
5 260

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 71

It is nevertheless interesting to note that the values of a obtained are of a
reasonable order of values even for the higher coronas, in which periodicity
supervenes. The corresponding graph (not shown) brings this out clearly. If
the values of a which correspond to normal coronas be selected, the following
summary is obtained:

TABLE 6.—VALUES OF a FROM APERTURE AND SUBSIDENCE.

Series. Corona. Ss a Mean.

I normal cen .0027

I s 2.8 25

2 ye 4.2 30

2 wbp 4.5 26

3 normal 3.2 28

4 pe 4.7 27

4 3-4 29 |

4 ae 5.1 20

4 wobg 4.6 30 + 00291
4 normal Bai 20

4 Ww 2B {y 4-5 31

4 normal 3.0 30 J

mean 00283

The mean of all the series is a=.00283; the mean of the fourth series,
which is more uniform, a=.00291. The latter datum will be taken in the
following computations.

14. Summary of optic constants.—The following series of values of a=ds
has been obtained when the measurements of aperture are made to the outer
edge of the first ring.

Optically (blue), a= .00344
From lycopodium (d,=.0032), .00336
From subsidence, .00291

The latter datum is decidedly the smaller, corresponding closely to optical
puce-violet (.00293). If, in place of the above expression, the elementary
optical equation 2 sing=s/R=\/d or a= 30) had been taken instead of a= 73.2A,
even the extreme red would show but a”=.0023.

The datum for subsidence being simplest in character is apparently the
most trustworthy. Since n=(6m/za’) s°, if the method of Wilson and Thom-
son ! be used for the computation of m the following values in grams per cubic
centimeter are applicable at the temperatures stated, for the pressure difference
op=17 cm.:

0= 10° 20° 30°

m= B47 6 LO! ALO XTON - 5:7 X10°

6

' Cf. J. J. Thomson, Phil. Mag. (5), Xtv1, p. 538, 1898.

72 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

For the pressure difference é6p=22 cm. which will occur below,
5.5 X10. 67 ><T0>"

“6

m= Al2x1 0m

‘For 6p=8.5 cm.
m= 2 OR 210 X LOs * 22O)< roma
The effect of temperature on latent heat is not considered, since the data are
not fully known. Its effect may be 1 to 2 per cent. More important is the value
to be taken for the heat ratio y=1.41.

15. Resulting equations applied—From what has been stated, it follows
that the first quantity to be found is the initial nucleation, 7., 2. e., the nucleation
which obtains when z=Z. This depends on incidental conditions, such as the
intensity of the ionizer, the first corona seen (Z), etc., and is therefore quite
arbitrary. In table 1, for instance, u,=,. Hence

n,=Nn,10% 9 "8% (1 — S/s?) (1—S/s?) (1—S/s’) .. .(1— S/Sé-»);
which will be abbreviated
n,=n,10% 9 "89 TT (1—S/s*).
4

This equation affords in the first place a means of computing S. For in
the region of normal coronas u is given by the apertures of the coronas. If, for

brevity, 7,=1,/10°'8”,
ec om ro G@=S/si_ Jia —Sfs2),

from which S is determinable in terms of pairs of values of r and s; or S may
be even more simply found from two successive normal coronas. The follow-
ing table shows the values found for the two series in table 1.

TABLE 7.—VALUES OF S FROM 124.:/f2-1=1—S (t/Se2x+1/82) +S*/522 183.

YX Io-3

SERIES I. 100

TABLE I. 132
68.4
117 ¢
40.5
Io0o
144
68.4

SERIES 2. 39-4 .139 .0048 2.73 |
TABLE I. 114 |
53-7 .196 92 2.88 ee
397 204 .0208 2.27 J
27.8 -453 486 1.84

26.3 Bae +1230 2.02
34-9

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 73

As both series were made in the same way, the mean value S=2.65 will be
inserted. With this value of S the data of the column Ni (1—.S/s’) may be
computed throughout. Then in the region of normal coronas the fundamental
constant of the reduction follows as

n= 3705°/NI (1— S/s?).

With this constant, the true value of the nucleation (number of particles per
cub. cm.) is computed for all coronas as

n=n.NII (1—S/s’).

All these data are found in table tr. It should be noticed that the coefficient
370 is obtained from subsidence.

x10?

.
[2

TAB. 1

8 70

1 2

FiGuRE 4.—CHART FOR TABLE 1, SHOWING THE RELATION OF NUCLEATION (7) AND OF DIAMETER (d)
or Foc Particle IN TERMS OF THE APERTURES (Ss) OF THE CORONAS.

20
Pies X10;
@
=
3

S
o-

8

16. Remarks on the tables and graphs.—The graphs, figure 4, for tables 1 and
3, show four independent series of observations of diameter, d, and nucleation,
n (particles per cub. cm.), in terms of the relative aperture, s=60 sing, where

74 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

pis the angular radius. The partial exhaustion is to 17 cm. and the standard-
ization is by subsidence ($ 14). If standardized by diffraction, the n-data
would be about .6 smaller, or the upper ‘“‘green”’ corona, for instance, showing
n=98,000, would then show u=60,000 nuclei. The d effect is much smaller,
being .2 larger.

Towards the end of any series the numbers diminish more slowly than the
formula requires, but this is naturally the result of the evaporation seen, for
instance, in the shrinkage of the coronas. On the other hand, when subsidence
is exceptionally rapid, the time during which subsidence takes place, and which
can in no case be sharply given (it was not directly timed), is seriously
large. On the whole, the agreement is better than was anticipated, and
certainly trustworthy. The results coincide in a general way, moreover, with
the data found by assuming a constant as was done in §§ 5, 9, and the latter
equations may be regarded as approximations by expansion of those in this
section.

The graphs, ” in terms of s, give evidence in the first series of three cycles,
the lower two being merged. In the second series there are apparently four
cycles, the two lower being distinct. The horizontal position of the cusps is as
closely in accord as the measurements justify. The vertical position suffers
from the shift and difficulty surrounding the absolute evaluation of n. Through-
out their extent, however, the fundamental similarity of the graphs is unmis-
takable, as will be further shown in the corresponding curves for ruby light
below.

Since n’=6m/2d?=(6m/za’) s*=23 (s/10*A)*, approximately, the fluctua-
tion of m with A is obvious; but the feature of the phenomenon is none the less
the occurrence of cyclic variations in the color of the innermost ring. The correc-
tion implied in the last equation would be more than sufficient. The violet

coronas are to be depressed as regards 1 and the red coronas raised in their 1 |

values, showing that in the former the measurement referred to red surpassing
the last violet, and in the latter to violet beyond the red. It is expedient to
state these data in relation to the diameters of the fog particles under observation.

17. Diameter of fog particles. Having determined the true values of n,
the diameters of fog particles may be computed for each aperture, since

d= 6m/an=.021n-”°.

The results are given in the tables and are plotted in the corresponding graphs.
Each of these (d as a function of s) shows the three cycles already determined,
and the cusps lie at d=.0007 to .co008 cm. and d=.0005 to .o0055 cm., or that
the intermediate and particularly luminous cycle covers a range corresponding
to about ten times the wave lengths (.c0o004 to .oo008 cm.) of the visible spec-
trum. But two of the cusps are unmistakably marked, while in other respects
the graphs retain the hyperbolic contour, ds=const.

Since n~*/3 is the cubical volume which contains one fog particle, d/n-*/3

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 75

is the ratio of the diameter of particles to the distance between particles, con-
stant throughout. The distance between centers is thus about 48 times the
diameter of particles for the temperature and pressure conditions prevailing
during the exhaustions.

One may note that the diameters found are independent of m; for in the
above notation let z and D refer to the normal corona virtually used for stand-
ardization. Then, as the series stands,

s-1
Ne—n,10" 9°81 1—\S/s*),
4
or after reduction, since the same equation also holds for Z,
c=)
t/a —(s,/a)*10° ~ °§2 1(a— S/s2)
Z-1

where s, is the aperture of the normal corona numbered Z. Thus d depends on
a, y, and S, and does not therefore differ much from my earlier values except in
so far as a and y were differently determined and S not observed.

Finally, since nd*=6m/2=const., the relation of m and d are reciprocal,
and maxima in 7 thus correspond to minima ind. If d is determined too large,
n will be too small. The curves bear this out. The periods indicated by the
cusps in the d curves are more appropriately referred to below. They may be
placed as follows: in the first curve they lie at d=.00069, .00053, .00039; in
the second at about d=.00079, .00055, .ooo40 cm. In conformity with the
work below their mean position may be rated at d=.00072, .00054, .00036, or
in the ratio of 4, 3, and 2. In other words, they are, roughly, multiples of the
cycle datum .o0018 cm., and throughout large as compared with wave length.

MONO-CHROMATIC LIGHT.

18. Tables——The coronas are too faint for effective observation with
mono-chromatic light obtained from simple sources like the salt flame. I
therefore used the electric arc as a point source and obtained sufficient limita-
tion for the present with a double thickness of ruby glass. This arrangement
has an ulterior advantage as it is thus possible to observe the colors of the
annuli as well as their red diameter on interposing the colored glass. This
greatly facilitates the reductions. The observations made with the large
cubical condensation chamber are given in the following table arranged on the
same plan as the preceding. The constant a” is as before definitely determined
from subsidence, while a’ refers to preliminary standardization with lycopodium.
The time loss # is neglected. The subsidence loss has been separately
computed and differs slightly from the above, showing more expeditious
work.

76 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 8—_CONSTANTS OF CORONAS. ELECTRIC AND RUBY LIGHT. CONDENSA-
TION CHAMBER AS IN TABLE 1. DISTANCES AS IN TABLE 1. 0=15.5°;
BAROM., =75:8 cm.; Op=16.9 cm.; y=.773 @= 00325 \fi—o- |S —exco am —anon
10-° GRAMS; a’’=.0029; a’=.0032; 1,=188000; PHOSPHORUS NUCLEI; d=

.0029/s. MEASUREMENT TO OUTER EDGE OF FIRST RING. 2 SIN g=s/R.
First SERIES.

2 t s Corona. zu Tey | at oes NO(1-8) oe n <a ae
o | min.| cm. | :

I 41) — --- —— | 2) 412000 | .000269
2 44 || —— |v" of x — ne) 318000 293
3 48 | — | wyeg'v’ -- Te3)) 244000 320
4 51 9.0] wopg 182 | 244 1.000 188000 349
5 | 55 | 83| weye 145 | 195 -746 140000 385
6 58 6.5|wgbp HK) || 94.1 554 104000 426
i 62 | 5-7) wyg.b 47-5 63.6 402 75600 473
8 65 5-4| wrg 39.2 52.6 .286 53800 531
9 68 4.7| weg bp 26.0 34.8 201 37800 595
10 71 4.2; wrbg 19.2 25.7 nay 188 25800 676
II 74 3.7, wy bg 12 7ieee 17.0 og! 186 17200 Tes
12 78 3-1] wobg 7.8 10.5 .056 186 10500 913
13 81 2.7) corona 4.9 6.6 .033 200 6200 .OOTOgO
14 84 20 iy 1.9 | 2 O17 188 3100 1370
15 87 |) erg! = aT | 1.0 006 180 1000 1960

SECOND SERIES.—n°= 200000; @=.0610.
I — — — — — 222 440000 .000264
2 54 | — | wg’ — — 7) 340000 287
3 Ry Nae || ye ean — — re 260000 314
4 61 9-5|wyg 214 287 T.000 200000 342
5 64 8.8) weg 170 228 Oy) | 149000 378
6 68 4.1] wgbr 89.5 120 593 | 112000 415
7 71 6.0] wgvr 54.0 | 72.3 SAS ian at 81200 462
8 74 5-7| wog 40229 62.0 352 58600 514
9 77 5.2| we br 35-2 amet eA ye2 2971 41600 576
10 80 | 4.3| wobg 19-9 | 26.6 .208 183 29000 652
II 83 | 4.1| wobg’r’ Eis 2 ee 2a a 161 238 ?19400 744
12 87 3.3| corona G-4e 12.6 -121 203 12400 866
13 go | 2.8 sf SS al 7-4 -095 197 7500 | .001020
14 93 202 ‘ 2.6 | Be5 .073 181 3900 1270
15) ||| 96: | x66 cA Len 5 .056 210 1400 1770
16 99 1.2 -4 6 .043 135 800 2130
19. Graphs for nucleation.—The results for mono-chromatic light are

constructed in the chart, figure 6, and show the three prominent cycles with
suggestions of others crowded into the region of normal coronas; but the latter
The two curves are more closely in agreement both in their
All four curves
agree, however, in their essential points, as closely as may be expected. The

are uncertain.
horizontal and vertical dimensions than was the case above.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. den

region of lower cusps is between = 35,000 and 45,000, and of the wpper cusps
between 7 =95,000 and 105,000, or at s=4.5 and s=6.0 in the two cases. These
cusps mark the region of “‘green”’ coronas. The angular radius is p=4° 18’
and #=5° 44’, respectively. Other remarks already made apply here.

80

= 60
Ss
3S
~<

S 0
8
><
=

0

Ruby Light
eee TAB.8
aeenez0 — 5 5a SEG

FicurE 6.—CHART FOR TABLE 8, SHOWING THE RELATION OF NUCLEATION (”) AND OF DIAMETER OF
Foc Particie (d) in TERMS OF THE APERTURES (S) OF THE CORONAS.

20. Graphs for diameter.—There is a slightly different constant involved
here, as the temperature is lower. Since 6=15.5°,

d=6m/2n=.020n “°.

Hence the edge of the cube containing one fog particle is just 50 times the
diameter of the particle for the uniform conditions of condensation selected.

The curves for d” (smooth hyperbola since d’s=const.) and d are drawn
in relation to s as usual, and the latter give evidence of three or more cusps
with intervening minima. These in their maximum and minimum relations
are reciprocal with the corresponding curves for 1. Both curves agree with
each other and with the preceding set for d.

Horizontal positions are naturally better than vertical positions because of

78 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

the shifting attending the standardization of m andd. If the horizontal position
of the cusps be taken from the figure as s=2.7, 4.2, 5.6, 8.3 in curve 1 on the
left, and s=2.8, 4.1, 5.6, 8.3 in curve 2 on the right, the mean ratios are 2,
3,-4, 6, so that s=1.4 or p=1° 18’ repeats itself. Since d=.0029/s nearly, the d
ratios should be as 1/2, 1/3, 1/4, 1/6. The figure shows roughly in curve 1
10°d= 1100, 710, 540, 400, and in curve 2, 10°d=1050, 740, 550, 350, the means
of which values are in the ratio of 6, 4, 3, 2, as nearly as can be expected, the —
data being multiples of .ooo18 cm., here as above. The presence of a period
between 4 and 6 is not in evidence.

MISCELLANEOUS EXPERIMENTS WITH THE DEEP VESSEL.

21. Measurement to the edge of the white disc——Before citing data obtained
from measurements of the aperture of the disc, it is desirable to reduce the
preceding data, s, (aperture to the outside edge of the first ring) to s; (aper-
ture to the inside edge of the first ring) by direct experiments. This is done in
the next table (9). The relation within the normal region is practically linear
and s;/s,=-76.

Careful inspection of the coronas seen with electric light shows that even
when normal the edge of the white disc shades off through y, 0, br, etc. There
is no real demarcation of the white disc, but the r’ edge is rather the beginning
of a diffraction cycle in which (counting from the outside inward) r, 0, y, merging
into white, appear, and when the higher coronas are reached the missing g, b, v
become evident in their turn.

TABLE 9.—OUTSIDE AND INSIDE DIAMETERS OF THE RED RING. DISC, ww’ y’
br’ vp’ b g y’r; EDGE, a w y HAZE. MEAN RATIO, s,/s,=.77.

Corona. Si so Si/So Corona. Si Se 5i/So
wybrbgy’r 2.1 3.0 42 w|rbg Bar 4.2 4
2.9 4.2
Do. 2.4 Be 78 wobg 2.8 3.5 81
3.0
Do. 2.2 2.9 ats
3.0
wyob|r 4.0 Bar .80 Do. 2.7 3-4 79
é 4.8 3-4
wgblp 3.6 4.6 78 N. corona 2.2 2.9 76
change to wyg 4.5 .
wobg’ 4.0 4-9 | 83 < 1.8 2.5 74
4.6
wy’rlg 3.2 4-3 | -73 se Hee 1.9 69
4-4

Measurements of s, before and after s5;.
Table 9 shows that if ds,=.o0291 (subsidence), then

a;=ds;=.00291 X.77 =.00223 and n=(6m/7a}) (10s)3=810s°.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 79

Direct measurements of the a, by comparison with lycopodium gave S;=.78
and .84 for the near and far faces of the chamber whence a;=.0026 (diffraction)
and m=515s*°. This is in the usual way smaller, and n/s* larger than the
subsidence datum (a;=.00223), but the latter will be taken here as above.
Both values are compared in the tables, the ratio being .64=n’'/n".

22. Apertures of the discs of the coronas.—Table 10 contains two inde-
pendent series of results arranged as in the preceding tables. The source of
hight is the Welsbach mantel. The data are further shown in figure 7.

-24g0 7-4 -* = Ray
264 230

ve als ashe
[t. |
emer oe tas GB 0 es

FIGURE 5.—CHART SHOWING THE RELATION OF RELATIVE NUCLEATION (r) TO THE NUMBER (z) OR
ORDER OF THE EXHAUSTION FOR TABLE 3.

FiGurRe 7.—CHART SHOWING THE RELATION OF THE ACTUAL NUCLEATION (”) TO THE NUMBER (2), OR
ORDER OF THE EXHAUSTION, FOR TABLE 10. AtLso THE DIAMETER d (IN 10-5 CM.) OF THE FoG PARTICLES
IN TERMS OF 2.

Figure 8.—Do. ror TaBLE,13.

80 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 10.—CONSTANTS OF CORONAS. CHAMBER AND DISTANCES AS ABOVE.
@=21°; BAROM., 75.3 cm.; @€p—16.9 cm; y=.775 | @—.062) Si—o8 Si — 1c eae
.0026; a’’=.00223; mM=4.7 X10~°® grams; ”,=200000; s MEASURED TO INNER
EDGE OF FIRST RING (APERTURE OF DISC); PHOSPHORUS NUCLEI;
WELSBACH LAMP.

" 0 ‘ie S | d=.oar

Zz t Corona. Ss n' =515s3 | n” =S810s3 NI(1-3) n=n, NII | Bay

min. em. X 103 X 103 | cm.
I 9 -= ~ 2.8 560000 -000255
2 12 pega! —- : 2.2 440000 277
B 15 g’ b’ aaa ef 340000 300
4 18 ygo 9.2 401 631 1.32 264000 328
5 22 y obg 7.4 208 328 1.000 200000 360
6 25 w clg 5a8 76.7 121 oem 150000 395
hoe 2 g’ bp Gay 95-3 149 Stig 110000 440
8 32 g’ bp ee} 76,7 [21 .408 81600 485
9 | 36 wrg 4.3 41.0 64.4 -299 59800 | 537
10 39 w br blr Bey 26.1 41.0 215 43000 | 600
II 43 w olg 3.4 hee 33-3 -149 29800 680
L2 40 N. corona 2.9 13.2 20.8 .103 20600 705
13 50 a Del 8.0 12.6 066 13100 890
14 53 2am Bal 8.0 -O4I 8200 -OO01040
15 Oy 1.6 2:3 3.6 -022 4500 1270
16 60 ; D338 Te 1.8 -009 1800 1730

SECOND SERIES.—@=.061; N,=175000.

X 103 X 103
I — —_— = = 2.2 385000 -000288
2 56 —_ — = = Tey 300000 314
3 59 gy II. 13 230000 343
4 63 y obg 6.9 173 272 1.000 175000 375
5 66 wecbg’ 5-0 66.4 104 -749 131000 414
6 69 eg’ bp Gee 72.6 1I4 -548 96000 400
7 73 y olg 4.6 SS 81.4 -401 70200 510
8 76 w eclg 4.0 33.0 51.8 .290 50800 570
9 79 glb p 3.6 24.0 37-7 .206 34000 650
10 82 v’cbg 2.8 11.9 18.7 -142 24800 720
Il 86 y’ trbhg 2.6 9.1 14.2 .0924 16200 830
12 890 corona 213 6.3 9.9 561 9800 980
gee 93 a 1.9 3.8 6.0 333 5800 | .oorr7o
14 | 96 . Tes 1.9 3.0 169 2960 1460
I5 | ror 1.0 6 TO .0060 TO5O 2060

The subsidence constant must be smaller in this case, since S,1,=s;
(1—s3,,/ys3), and it is clear that S and s? must change in the same ratio
nearly. It was shown that s;=.775,, or s?=.59s2, and hence S;=2.6%.59
=1.56. An actual computation gave S;=1.3, the difference being easily
referable to incidental discrepancies.

23. Remarks on the table and graphs.—The green corona now shows the
nucleations (110,000 +82,000)/2 and independently 96,000, in the two series,
agreeing reasonably well with the above 85,000, 106,000, 104,000, 96,000, the

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 81

standardization being throughout by subsidence. This agreement is further
substantiated below. The difference of values here is largely an actual differ-
ence in the positions of the coronas.

The curve r=515s*/to°'"” if constructed in full shows the eventual
linear march already instanced. As above a=.062 nearly.

The curves of nucleation, 7, in terms of the relative aperture, s, indicate the
occurrence of four cycles, though the lower is uncertain. Both curves run in
parallel as nearly as the observations allow, and they show actual cases of over-
lapping or reéntrance.

This reéntrance is particularly evident in the d curves in which four distinct
branches overlapping at their edges may be made out. The positions of the
cusps, as indicated by the green coronas, may be placed at 10°d = 450, 650, 950, cm.
or about in the ratio of 3, 4, and 6 times .ooo15 cm. The corresponding number
found above was larger than this in the ratio of 18/15, while the ratio of the red
and blue wave lengths involved would be 1.4.

24. Condensation chamber remote from the eye-—The distances are in-
verted in the following experiments, being 250 cm. from the eye to the con-
densation chamber and 85 cm. from this to the source. All coronas therefore
are relatively small and the measurement of s by the same method much more
difficult, particularly in relation to standardization. Measurement of s was
naturally made to the outside of the first ring. Results with Lycopodium seen
through ruby glass and a source of electric light gave (d,=.0032), s,=.344 cm.
Hence a=d,s,=.oo11 and n=6750s*. If the value of a be reduced to the
conditions holding for subsidence, a” =.oo11 (291/336) =.953 X10 *, whence n=
10,4008. Table 11 computed with this value follows. Unfortunately, three
important data were lost, and the series are deficient in this respect.

TABLE 11.—CONSTANTS OF CORONAS. CHAMBER AS ABOVE. DISTANCES FROM
EYE AND SOURCE, 250 cm. AND 85 cm. RESPECTIVELY. @=21°; BAROM.,
75.3 cm.; Op=16.9 cm.; y=.77; B=0; S=.36; 2,=260000; a’=.0011; a’ =.00095;
m=4.7X10~6 grams; s MEASURED TO OUTER EDGE OF FIRST RING; PHOS-
PHORUS NUCLEI; WELSBACH MANTEL.

z t Corona. s n’ =675083 |n” =1040083| NII (x -3) n=n. NIT dsnee
min. em. xX 103 X 103 | cm.
I 21 — — — = D2 570000 .000254
2 24 Wt’ 3.0 315 486 D7 440000 276
3 | 28 B.D: 3-7 356 549 1.3 338000 302
4 32 y obg Ieee 205 409 | 1.000 260000 330
5 36 wegy erro 182 281 747 194000 363
6 40 wpbeg’r 242 76.9 119 556 144000 400
ale ea gbp 2.3 82.3 127 .404 105000 446
8 48 wre. 2.0 58.0 89.4 .290 75400 496
9 52 w br b g’r Te a8 S10 205 53300 560
10 56 wyobg’ — —— Saad -I4I 36700 633
Il 60 coronas Te mee 20.4 | .094 24400 724
12 64 os Tea 10.3 15.9 -059 15300 850
13 69 Ss I. 6.7 10.4 -035 g1oo .0O01000

82 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

SECOND SERIES.—n” = 280000.

X 103 X 103 |
I 24 — —- —— aa 2.2 620000 -000246
2 27 w’ r’ 4.0 432 666 a7 480000 268
3 30 e biips 3.6 328 505 1.3 | 360000 205
4 34 y o bg — + — 1.000 280000 321
5 39 wrg | 3.1 201 310 747 209000 354
6 43 weg’ | 2.4 99-2 153 -556 156000 390
7 47 gbp | 2:2 FPG III -404 113000 435
8 51 wrog 222 aS III -290 81200 485
9] 55 | welglr’ 1.8 42.5 65.5 .205 57400 544
IO 59 g’|bip — -- — 141 39500 616
Ir 2 w clg 7’ Teg 16.7 25.4 094 26300 705
12 65 wobg Co 10.3 15.9 059 16500 | + 824
13 70 corona 1.0 G27) 10.4 035 9800 980

The values in their relation to s work out similarly to the above data.
The two cusps are in their usual location. The green coronas correspond to
Nm=105,000 and 113,000 nuclei, respectively, somewhat larger than the above
order of values, but not more so than the difficulties of standardization would
lead one to anticipate.

The d values are difficult to interpret, as the curve is fairly uniform but for
the break introduced in the w g b p region. Both curves are in close agree-
ment, and under the circumstances it is impossible to locate the cusps and their
relations.

LONG AND SHALLOW CONDENSATION CHAMBERS.

25. Apparatus—The purpose of the present section is primarily to give
greater scope for the observation of the higher coronas, 7. e., those of large
aperture. Incidentally it affords a means of testing the validity of the correc-
tion for subsidence, S.

To obviate fragility of apparatus and inconvenience in the exhaustions, the
volume of the long vessel is suitably decreased by lessening the height. This
does not interfere with the work, since only a diametral section of the color
rings is needed while the apparatus becomes more manageable.

The vessel is shown in the diagram, figure 2 (above, p. 58), and is (in the
clear) 55 cm. long, 11 cm. high, and 16 cm. deep in the line of vision. Less
height and even greater depth and length would have been preferable, but in
such a case the correction for subsidence becomes unwieldy. As usual, there isa
central brace and a lining of wet cloth, and the front and rear faces are of plate-
glass ¢ inch thick. Exhaustion is effected through the wide hollow trunnion, E,
the other being closed and holding a thermometer, or other accessories. JF is the
filter and P the nucleator, both attached with a cock. Rubber corks, A, A’, close
the wide holes necessary for cleaning. The external vacuum chamber was such
that the pressure differences were reduced about one half on suddenly opening
the wide cock into the condensation chamber. There was no leakage, and the

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 83

easy means of wetting the whole inside by revolution of the apparatus around
the trunnions was a great convenience. But in spite of this, color distortion of
the coronas was an almost invariable occurrence, particularly when the interval
between observations was short. Since in comparison with the case of the taller

FIGuRES 9, 10.—CHARTS FOR TABLES 13, 14, SHOWING THE NUCLEATION (7) AND THE DIAMETERS (d)
or Foc ParricLes FoR DIFFERENT APERTURES (S) OF THE CORONAS.

Ficure 14.—CHART SHOWING THE Errect oF DIFFERENT VALVES WITH NUCLEATIONS INCREASING IN
THE LAPSE OF TIME, FOR TABLE 19.

vessel above, which showed no appreciable distortion, vapor diffusion is here
much enhanced, the discrepancy must be referred to the diffusion of nuclet.
This would have to take place unassisted by the convection of vapor.

26. Reduction of data.—For the pressure difference 6p=17 cm. and at 20.7°,

84 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

m=4.7 X 10 °grams, and therefore, as above, n= 365s°. The constancy of S is
less evident than above, being in the respective series

I II Ill IV Mean
S=6.8 4.8 6. G5 5.8
7.3 8.5 6.5 6.4
4.2 2.1 3-9
Der

the variability being, however, largely due to the accidental delays inevitably
encountered in observing. Nevertheless some other cause is at work simul-
taneously toward the end of the series. Here the nuclei are not removed
apparently.

The mean value S=5.8 will be accepted for the body of the observations,
but toward the end of the series the special values computed for S will be in-
serted as well as the alternative mean values. The ratio of heights of the
condensation chambers here and above is 11/26=.42, while the inverse ratio
of the subsidence constants is 2.6/5.8=.45, showing close consistency for ex-
periments of the present kind. Table 13 for 6p=17 cm. will now be intelligible.

Since n=6m/zd’, it follows that d=.o21n~“3, so that the last column
follows from the preceding.

The charts, figures 8 and g, corresponding to these tables, show the cycles
already mapped out above, but less distinctly, from the difficulty of maintaining
homogeneous nucleation. Naturally, the large coronas, being vague in outline,
are difficult to measure, and the middle cycle thus comes out clearest.

TABLE 13.—CONSTANTS OF CORONAS. CONDENSATION CHAMBER 55 cm. LONG,
11 cm. HIGH, 16 cm. DEEP; DISTANCES FROM EYE AND SOURCE, 85 cm.
AND 250 cm. RESPECTIVELY; @=21°; BAROM., 75.6 cm.; 0p=16.9 cm.;y=.77;
B=0; S=5:8; a’=.0032; a= 0020; mM=a4 7X 105-; %.— 173000; Ss eM AS UNE
TO OUTER EDGE OF FIRST RING; PHOSPHORUS NUCLEI; ELECTRIC ARC.

First SERIES.—RuBy LIGHT.

|
Zz | t | Corona. Ss n’ =27553 | n!’ =36553 | NIZ (1 -3) n=n NIT eon
min. | cm. X 103 X 103
I 2 | wrg am = — ay] 294000 .000315
25a | ver a om — tg 225000 345
3 6 | wypg 9-7 251 333 1.000 173000 378
4 8 | wrog 8.8 IgI 253 -722 125000 420
Bi xo webr’ 6.8 86.3 II5 -515 89100 470
6 2 wogbp 5-9 58.0 77.0 347 60000 540
7 14 wobrg ait 36.6 48.5 .222 38400 625
8 16 | corona 4.0 17.6 23.4 DLSs 23000 740
9 18 | ‘ 3.0 7.8 10.4 .059 10200? 970
IO 20 | = 1.6 T.2 1.6 -O10 1700? | .001760

1 Streaks of color and horizontal bands visible. S for normal coronas specially computed.
2 Alternative, 11200, 3300.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 85

SECOND SERIES.—n,=327000. WuitEe (Evecrric) Licut.

X 103 X 103

I fo) — -- — = =

2 2 wpg 12.2 499 664 | 1.000 327000 -000305
3 4 wep 11.8 458 606 | -739 242000 337)
4 6 Wayavess 10.4 309 409 } 545 178000 373
5 8 v ro g’ 8.2 151 201 so 130000 415
6 Io webr’ 6.3 68.8 o.3 | 27.0) i) 9 1200 467
4 12 w br g|p 6.1 64.1 | 85.0 | 184 60200 536
8 14 w br b g’p 4.7 20-4 || a SOne nN) sekOL alee OOOO 620
9 16 N. coronas 4.0 17.6 23.4 | .072 23600! 730
10 18 . 2.8 6.4 8.5 | 026 | 8700! | .cor020
ch 20 “a 2.0 2.4 sei | .OTO 3300! 1410

Ale
1 Alternative, 22200, 2800, o.
FourTH SERIES.?—n,=546000. WHITE LIGHT.
* 103 X 103

° 2 w’ 1’ 13.2 630 840 1.000 5460000 .000257
I 4 Ww pv g’ [1.2 390 518 Tae 406000 285
2 6 wep 11.6 434 561 +544 297000 315
3 8 Wy Dv sg 9-7 251 333 395 216000 35°
4 10 wlog 8.8 187 248 288 157000 390
5 12 w br b 6.7 82.7 IIo .206 113000 435
6 14 wyg 6.1 62.4 82.9 .138 75400 496
7 16 wbrbg Gea 36.6 48.5 .089 48600 575
8 18 wpg es 21.9 29.0 .052 287005 685
9 20 corona B35 11.8 TG a7 .029 15800 837
10 22 e 2.5 4.3 Bel .O10 5730° | .oo1170
It 24 - 1.6 1.2 1.6 003 | 16405 1780
12 26 rs lost

2 Cycles nearly absent. Third series omitted.
3 Alternative, 25300, 10400, 4900, o.

The green coronas reproduce the orders of nucleation already attributed
to them, showing 1=g0,000, 110,000, 90,000, 90,000, a favorable result in view
of the large differences of subsidence encountered.

27. Smaller and larger pressure differences—To further elucidate the
effect of the subsidence error and to exhibit the sequence of coronas more
closely, experiments were now made with lower and higher exhaustions than
the above. In view of the smaller steps of pressure, it was also thought that a
more definite location of the cusps of the cycles would be possible. This, how-
ever, was not the case, for if the interval between observations is too small, the
coronas are never free from distortion. Table 14 in particular, in which the
interval is but one minute, gives evidence of this, but it is even present in
the succeeding tables where the intervals are two minutes or more.

To reduce the data, the values of m must first be computed. If the latent

86 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

heat of steam is considered constant (Regnault’s value, 606 at o° C.), and the
ratio of specific heats is 1.41, the following values may be derived:

For dp=22; 0=10 M—=h.2 XO? For 6p=8.5; @=10 m=21X107—6
20 Gas 20 26
30 6.7 30 28

The effect of a rise of temperature on latent heat would be an increment of
the order of .o5 per cent. per degree. The specific heat ratio is also variable,
and for lack of data applying to the region of low temperature in question these
subsidiary variations must be disregarded.

If the data for 6p=17 above be included, a table of double entry may be
drawn up adapted for all pressure differences between 8 and 22 cm. and for
temperatures between 1o° and 30°.

The optic constant in the two cases becomes (since 1 = 6ms*/7a°)

For 06p=8.5 cm., n=1225° (diffraction) and m= 202s? (subsidence).
For 6p=22 cm., n= 284s? (diffraction) and m= 470s? (subsidence).

Finally the volume ratio, y, is at 6p=8.5,
y= (57-7 —8.5)/ (75-6 — 27.9) =-853;
and at dp=22 cm.,

y= (73.7 —22.2)/(76.0— 2.3) =.700,

with which values a practical table is appropriately drawn up, time losses being
ignored as above.

The subsidence constant computed for the different series shows the same
falling off of value just before the coronas vanish, to which attention has already
been called. The following cases may be instanced.

TABLE 12——VALUES OF S FOR SMALL CORONAS.

Low pressure, 6p =8.5 cm. Higher pressure, 6p =22 cm.

Ss S s Ss S Ss S:
—— |
|
4.1=3.3\+| (6-4) |) 5.4405 Ae5—3).0) ||) 6:0) |) -Any— 326) S55
3.32.6) || 426 | 4.5—3°0 B.0—2.7) Sate 3.0—25O lecaG
2.0—2:0 | 3.2 | 346—2.8 Oe TOM ese Oi esO— ce Om eso
2.0—141 | 3.4 |\-228—oem Tea SOM e2sS T-S—T-On|| eae:
Probable S= 4.8 4.5 5.6
ést= 7 2m ae

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 87

The means of the best values are in the five cases S=4.8, 6.7, 6. 3, and
S=4.5, 5.6, respectively, the mean of all of which is S=5.6. This value is
practically the same as in the preceding table for 6p=17 cm. The effect of
waiting 2 or 4 minutes between observations does not appear in the individual
results. The decay-effect of waiting 1 or 2 minutes is apparent, but this is due
to the necessarily much more expeditious observation of coronas (time of fog sus-
pension) in the former case. Curiously enough, at 6p=8.5 cm. S=6.5 is larger
than at 6p=22 cm. where S=35.0.

The constants for computing diameters of fog particles (d= 6m/zn) are at
Sa5em, 10 Xd—1.71n “2 and at 22 cm., 10°7Xd=2.27n “3. Thus at ép=8.5
cm. the ratio of distance between centers of cloud particles, n~'’%, and their di-
ameters is 58.4, and at dp=22 cm. the corresponding ratio is 44.0. It follows
then that for the same corona, the distances apart are materially different in the
two cases. Hence if the interstices enter into the character of the diffraction
pattern and distribution of axial colors, these should be different for the two
pressure differences in question.

TABLE 14.—CONSTANTS OF CORONAS. ARC LIGHT, CONDENSATION CHAMBER,
AND DISTANCES AS IN TABLE 13. 9=21°; BAROMETER, 75.6 cm.; 0p=8.5 cm.;
J—.o55 P=o; S=4.8; a’ =.0034; a’ =.0029; m=2.6 X10 °; n.=825000; s MEAS-
URED TO OUTER EDGE OF FIRST RING; PHOSPHORUS NUCLEI; st=1".

|
z t Corona. s n'’=200s3 | NIT Gs) n=n,NIL Bee
|
min. em X 103
al 3 wy 13.0 440 He L7 965000 .000173
° 4 wo Tar | 354 | 1.000 825000 182
I 5 we Ter 354 | .830 685000 194
2 6 wp 9-3 | u6m | .690 569000 206
B 7 wvg 3) |) oz 554 457000 222
4 8 w bv g = | .448 369000 239
5 9 wvb 9.8 188 300 297000 256
6 IO wep 9.0 146 .292 241000 275
a If wyg 8.8 ners O -234 216000 285
8 12 wobg 8.7 132 188 155000 318
9 13 wog 8.0 102 =L50 124000 343
10 14 wrg 7.3 79.4 -T1Q 98200 370
IL 15 weg 6.6 57-4 .092 75900 404
12 16 webr’ Bay 38.0 .070 57500 443
13 7 webr’ 5-7 38.0 O51 41900 493
14 18 w brg 5-3 29.8 -037 30500 550
15 19 wrbp 4.5 18.2 .026 21600 617
16 20 corona 4.1 14.2 O17 14000 710
17 21 a 3.3 75 .009 7600 ' 868
18 22 i 2.6 | Bar .005 3700 ! .OOTTOO
19 23 . 2.0 | a7 .002 1730! 1420
20 24 a Lee a3 -000 200 ' 2720

1Alternative 8580, 4130, II50, O.

88 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 15.—CONSTANTS OF CORONAS. CONDENSATION CHAMBER AND DIS-
TANCES UNCHANGED. ARC LIGHT; PHOSPHORUS NUCLEI; s TO OUTER
EDGE OF FIRST RING; 0=21°; BAROM., 75.6 cm.; 6p=8.5cm.; y=.85; =o;

S=6.7; a’ =.0034; a’ =.0029; N,=110000; m=2.6XtI0 °.

First SERIES.—dét=2™.

Zz t Corona. Ss n” =20083 | NII (1 -3) n=no NIT Tou Y
min. cm. X 103 cm.

I ° ae ae am aaa

2 2 w oy bg 10.5 235 1.000 110000 -000358
3 4 wog 9.0 148 -799 87900 385
4 6 weg 8.1 106 .627 69000 418
5 8 webr’ 6.3 50.0 -478 52600 455
6 IO wo b! 5-6 36.0 -340 37400 510
7 12 wipg 5-4 31.4 -229 25200 585
8 14 wbrbg 4.5 18.8 arG x 16600 670
9 16 corona 3.6 9-7 -087 9530? 807
10 18 x 2.8 4.4 -039 4340? .0OI050
II 20 2A 1.9 .O17 1870? 1390

1 Seen very obliquely this changes to w g.
2 Alternative 9530, 4040, 500.

SECOND SERIES.—1,= 680000; S=6.3; dt=2™.

cm. X103

I 2 wr’ a

2 4 wt’ - fog —

3 6 wr’ ~=

4 8 w 0’ — -

5 10 we Dsk2 460 1.000 680000 -OO0O195
6 12 w bvg 12.4 381 .824 560000 208
7 14 wvbg 11.8 329 675 459000 222
8 16 wbg 11.0 266 -549 373000 238
9 18 wep 10.7 245 -446 303000 255
10 20 Ww yg vp 10.2 212 .360 245000 273
II 22 w oylg 10.1 206 290 197000 204
12 24 wlog 9.6 180 .232 158000 316
13 26 wrgy 8.6 127 185 126000 341
14 28 weg’ 7.8 95-0 145 98600 370
15 30 wegebp 6.4 52.4 aor 75500 404
16 32 wygbp 5-9 42.2 .080 54400 451
17 34 wrog | 5.8 39-0 .056 37900 509 ,
18 36 w br big a 26.6 .038 26200 575
19 38 w y bglr 4-4 17.0 .025 17100 663
20 40 corona 3.6 9-7 -O14 9790 800
21 42 Ss 2.8 | 4-4 -006 4350 -OO1050

TABLE 16.—CONSTANTS OF CORONAS.

A CONTINUOUS RECORD OF

ATMOSPHERIC NUCLEATION.

89

CONDENSATION CHAMBER, DISTANCES,

MEASUREMENT OF s UNCHANGED; ARC LAMP; @=24.5°; BAROM., 76.0

con, Op—22'cm.; y=.7o; B=o; S=4.s; 1,= 360000; m=6.1X10°%; a’=.0032;
a’’=.c00290; PHOSPHORUS NUCLEI; 6t=2".
z t Corona. Ss n’ =360s3 | n” =470s3 | NI (1 = n=noNIT Geecores
Se rape
min. cm. X 103 X 103 cm.
I 2 w 0’ — = = = be Bae
2 4 wc’ — —_ a = =
3 6 w e’|p’ 1362 828 To8o0 1.000 360000 .000320
4 8 w y|vb 10.5 421 550 683 246000 362
5 10 weg 8.5 221 289 -456 164000 415
6 13 wg blp 6.0 79.6 104 207 107000 478
7 15 wrog 5-9 73.8 96.3 .183 65900 562
8 17 wyob 4.5 33-9 44.3 III 40000 664
9 19 w brb 3.6 16.8 21.9 o6r 22000! 810
10 21 corona Ba) Tia. 9-3 026 9290' | .oo1080
II 23 Ey 1.8 2.3 3.0 008 2920! 1590
12 26 = 9 a a oor 360! 3200
1 Alternative 22000, 10500, 2660, o.
SECOND SERIES.—”,= 410000; S=5.6; dt=4".
min. cm. X 103 X 103 cm.
I 4 wv Teo 7 738 "963 1.000 410000 .000305
2 8 w giv 10.1 371 484 675 2777000 350
3 12 wrg 8.8 245 320 -444 182000 400
4 16 w g blp 6.3 90.0 117 .287 118000 462
5 20 wrg 5.8 70.2 91.7 73 70900 548
6 24 wybg 4.7 37-4 48.9 .IOI 41400 656
7 28 corona 3.6 16.8 21.9 053 21600 814
8 32 = 2.6 6.7 8.7 O21 86907 | .co1100
9 36 - 1.8 2a 2.7 .007 27907 1610
10 40 1.0 4 a5 .OOI 490? 2900
2 Alternative 8690, 1230, 0.
28. Remarks on the tables —Considering the graph, figure 10, correspond-

ing to table 14, it is seen that the curve as a whole lies above the locus n’ = 200s".
This might be rectified by using a larger value of S than the small one (S=4.8,
while the mean value is 5.6) specially computed. The divergence of the S-
effect is enhanced in view of the large number of exhaustions.

The graph n=n,NI/ brings out for the first time the cycle of the first color
series, w y to w p, which in the shorter apparatus above was not fully attain-
able. The successive colors of the first annuli vary in the order of wave length,
though nothing was observed above the first w y. The next color cycle (en-
larged 10-fold in the adjoining curve) is very complete, though irregular from
the inevitable color distortion, the interval between observations being but one
minute. The following color cycles II] and IV, which include the vivid annul1,

go A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

manifest the same relations already instanced above. The small value n=
70,000 which belongs to the green corona, and the correspondingly small values
for the other coronas, are in accord with the small pressure difference, dp=8.5,
chosen; but it will presently appear that this value is even smaller than would
be predicted. The crests of the n-cycles may be located at = 25,000, 65,000,
and 550,000, from which the corresponding d-values would be dX10°=585,
430, 210 cm., indicating diameters of fog particles in the ratio of 3:2:1.

FIGURE 11.—CHARTS FOR TABLE 15, SHOWING THE NUCLEATIONS (7) AND DIAMETERS (d) OF THE FoG
PARTICLES, IN TERMS OF THE APERTURES (Ss) OF THE CORONAS.

FIGURES 12, 13.—CHARTS FOR TABLE 18, SHOWING THE RELATION OF THE APERTURES, S, OF THE
CoRONAS TO THE NUMBER, 2, OF EXHAUSTIONS.

The d-values are much more difficult to correlate, although the main cycles
are apparent. The appearance here is that of four independent loci which
appear united as the result of color distortion. The view which ascribes to
these curves different parameters is in many respects plausible. There are
three green coronas discernible corresponding to 10°Xd=275, 443, 617, about
in the ratio of 3:5:7. But this result is again merely tentative.

Table 15 contains two series of results in which larger times intervene
between the exhaustions to insure more thorough diffusion and less distortion
of color. The data are mapped out in the same chart, figure 11, one above the
other. In the m-curves four cycles may be made out, the lower less distinctly,

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. QI

as this is concealed in a measure by the specially computed S-values. There is
reasonable agreement with the graph 200s’, though the upper cycle lies above
it. In the d-curves the agreement with .oo29/s is throughout as close as the
difficulty of observation permits. Three sweeping marches from violet to deep
red may be distinguished.

In the last table (16) high pressure differences 6p=22 cm. are brought to
bear, and the time intervals é6t=2 and 4 minutes intervene between the ex-
haustions. A definite effect of the latter does not appear, since the nucleation of
the green coronas, for instance, is 7= 107,000 and 118,000, with more nuclei indi-
cated for the larger time interval. The reverse would be the case if there were
marked time loss. The corresponding charts may be drawn to show the relations
in detail. The curve 470s% lies above the higher coronas, whereas in the preced-
ing cases it lay below them, indicating the difficulty encountered in computing
sufficiently correct values for the subsidence constant, S. While the n-values
show the usual relations, the d-values are more difficult to interpret; but three
cycles may be made out, with the middle one unusually bulging and contracted.

29. Nucleation of the green coronas.—These values, though difficult to
obtain and suffering from cumulative errors, are nevertheless consistent; and
the nucleation to produce the green corona seems to depend on the distance
between particles. Put d/n~"3=D. Then the above results show that
n=107(D—.o0114). Hence if the distance of particles is 87 times their diameter
the green coronas should appear for vanishing nucleations. This implies a limit
of values of 2 by which green coronas may be evoked. Furthermore, since the
apertures of this type of coronas remain within a range of values nearly the
same for all conditions of nucleation, the value of m may be found below which
green coronas do not occur. For the diameter of fog particle is here about
d=.00046 cm., and therefore n=15,300. One may argue, therefore, that at
least 15,000 nuclei per cubic centimeter must be present if coronas of the middle
green type are to be possible.

TABLE 17.—_NUCLEATION OF THE GREEN CORONAS.' 1 (computed)=107 (d|Y¥n

—.OIT4).
Table. n X 10-3 ot Table. n X 1073 ot Table. | n X 1073 | bt
min. min. | min.
14 49.7 ee 13 89.1 2 16 107 2
15 52.6 2 13 110.0 2 16 118 4
15 5 2 13 gi.2 2 |
— 94.2 2
Mean 57-0 = 96.1 | 113
10'd= 45 cm. (obs.) 46 cm. 47 cm.
Op= 8.5 cm. 17 cm. 22 cm.
djy/n= O17 021 023
107 3n= 57 (comp.) 96 113

1 Earlier values, 6p = 17 cm. and cubical vessel, 107° =85, 106, 104, 96, and 96, 96, the mean being =
97000.

Q2 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

While these inferences are essentially uncertain and might even seem
attributable to the time coefficient, 8, neglected, the observations do not war-
rant such exception. Thus at 6p=22 cm. and 6¢=4 minutes, the total time
consumed was 36 minutes, whereas in the case of 6p=8.5 cm. it was only 32
minutes for 6t=2 minutes, and only 21 minutes for 6f=1 minute. The total
amount of time loss must therefore actually have been smaller in the latter cases.

OTHER CAUSES OF CHANGE IN THE TYPES OF CORONAS.

30. Thickness of cloud layer.—In passing from the cubical to the long
apparatus, the thickness of the cloud layer decreased from 20 to 16 cm. On
tipping the former vessel on its trunnions, and looking through the chamber
diagonally, the thickness of cloud layer could be increased to 30 cm. In all
these cases, the type of corona and its diameter showed no appreciable change.
If symmetry about the center is maintained, the effect of the thickness of the
cloud layer seems thus to be absent.

31. Obliquity of diffraction—When the direction of the diffracted ray
differs by a small angle from the direction of the normal ray of light, the type
of corona does not change as to color of annuli. If, however, the angle of de-
viation due to diffraction is very large, the corona may change in character.
Thus for great obliquity the orange-red corona was found to change into the
preceding green corona. In other words, a larger particle on very oblique
diffraction may produce the same corona as a smaller particle on less oblique
diffraction.

32. Effect of wave length—Since at 6p=17 cm., a=73A, and therefore
n=6ms*/7a3=6 ms3/72\3(73)3= 2355/1075, the equation for different colors
would be (denoting the colors by subscripts) 1=67s3=97s3}=10753=118s$
= 15453= 2215$=360s3, etc., while for other pressure differences the same
relations would be preserved.

Testing this by the very full series in the second part of table 15, and re-
membering that the measurement is made to beyond the first ring, the fol-
lowing data may be deduced:

S=1I030 w oy (p)g n (observed) = 197000 n (computed) = 186000
636 wre’ 126000 126000
262 wgbp 75500 75200

If the w r g’ corona is taken as correct, the w g b p corona will also be, as well
as the w oy (p) g corona. But in the latter case the narrow purplish ring which
intervenes before the green and the general difficulty of defining the mixed
colors of the second ring makes inferences of the present kind precarious.
Nevertheless it is probable that the rather sudden transition of green to blue in
the colors of the second ring is associated with the underlying cause of peri-
odicity. Examples of this kind might be multiplied.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 93

DIFFERENT SPEEDS OF EXHAUSTION FOR THE GIVEN PRESSURE DIFFERENCE.

33. Increased suddenness of condensation.—As a final test of the trust-
worthiness of the above sequences, it was necessary to repeat the results with
some form of valve more nearly instantaneous in its action. In the case of
air nuclei, special comparisons instanced below (Chap. IX, $ 3 ) showed that
for a reasonable relation between the sizes of the condensation chamber and the
vacuum chamber, an ordinary plug stopcock was quite as serviceable as an
instantaneous valve, the coronas observed with the air nuclei being in both
cases the same. With phosphorus nuclei and at the outset of the experiment,
however, this is not quite the case, at least when the nuclei are very numerous
and very small.

The design of the new valve was very simple. In figure 3, p. 58, V leads to
the vacuum chamber (large aspirator flask), and C to the condensation chamber.

TABLE 18—THREE GEOMETRICAL SEQUENCES OF CORONAS, FOR INSTAN-
TANEOUS VALVE. CHAMBER, 20X26 X35 cm.3; DISTANCES, 85 AND 250 cm.;
Goce BAR OM. 74405, 0p —17 cm:; J=.707> P=o; S—2.1- S’— o>: a (SUBSIE
DENCE) =.0029; m=4.8 X10-°g; n{=308000; n,=340000; PHOSPHORUS NUCLEI;
WELSBACH LAMP, AND s MEASURED TO OUTER EDGE OF RED RING.

First SERIES.

Zz t Ss Corona. mn’ =376s3 | N IL (1-3) Goa n’ N Il’
) min. cm. Nucleation. X 103
I I fog
2 4 (13.0) wr’
3 7 (11.0) olive
4 10 Pare
iF 13 10.6 y obg 534 1.000 .00030 308000 1.000
6 17 9.1 wrg’ 288 iS 33 232000 754
7 20| Teil weg eg 562 30 174000 -504
8 23 | 6.3 gbp 94.0 416 40 129000 -419
9 26) 6.1 gy br b’ 85.3 302 45 939090 -305
10 29 | Gs! wrog 85-3 .218 50 68100 221
II 32| 5.0 | w br cor 47-0 .158 56 49600 161
12 35) 4.6 Ww 0 cor 36.6 -IUI 62 35100, 104
13 38 4.0 corona 24.1 .077 71 24600 .080
14 41 3.4 | 14.8 -O51 81 16300! 053
15 44! 2.9 Sah 9:7 .032 04 10500 .034
16 47| 2.6 Mi 7,0 .O1g -OO1I2 6200 .0120
a .OTO .OOT4O 3400 OIL
ne a ‘i ae Note: n= |
32I X 108
303° |
| 278
| | 285
| 345
! | | 313
i | Mean:} 308000
|

94 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

R is a soft rubber cork which can be raised or suddenly lowered (in the latter
case by the blow of a swivelled hammer not shown in the figure) by actuating
the knob or handle at the upper end of the rigid rod r._ There is a stuffing box
at S. The end of the tube p has been turned off smoothly and serves as a seat
for the plug, R. Stiff glycerin is used as a lubricant. All passageways and
pipes are wide, the latter at least one inch in diameter. The valve has retained
its efficiency after countless experiments, but it must be left open when not in use.

34. Results—The results are given in table 18, on a plan similar to
the above tables, and the calculations are made in the same way. All
operations were strictly timed, and it was thought best to compute S from
subsidence as

S!=s7 (D5 Giie/ Se):

The corresponding n, is written 1. The result is S’=1.95. Computed from
the observations themselves, S= 2.1, 1.7, for instance, in the first two parts of
the table, a difference which can only be explained on the ground of obser-
vational error. Usually 3 minutes were consumed by the operations between
the exhaustions, while the fog was dissipated (by the influx) within 15 sec.
after the exhaustion. In the last series but 2 minutes are allowed between
the observations, but there seems to be no appreciable difference in the data so
far as this cause is concerned.

SECOND SERIES.—O=24°; barom.=75.08 cm.; oO¢=3"; y=.767; u,.=212000; 1,/=

305000; m=5.0 X10 °g/cm3; S=1.7; S’=1.95.

z t Ss Corona. n’ = 376s3 Ni (1-2) x Baa n’ N Il’
tll| 43 fog 390 X3=
to 3n’
2 45 (13) w r’ —
3 48 (12) b’ B —
GO Fest g B =
5 54 ist y’o 624 1.000 .0003I 305000 1.000
6 | 57 9.9 Wie 378 -758 34 231000 -756
7 60 8.3 weg 223 571 38 173000 -568
8 3 6.2 gbp 92.8 .427 42 129000 -423
9 6 6.4 gy|b’ 102.2 Baus 40 93900 308
IO 9 6.0 wrg 82.4 .230 51 68600 Rae
oe 12 ree w pcor 55.0 .168 57 49700 -163
12 15 4.5 Ww yo cor 36.8 shore 64 35400 -116
13 18 4.0 w br cor 25.9 .085 WE 24700 -o81
14 21 B55 corona 16.7 .059 83 16500 -054
15 2 3.0 re II. .038 96 10500 -035
16 2 27 * 5-5 .024 .OOITS 6400 .021
Note: 1,/ =
317 X10?
320
310
Bu5
265
Mean: 305000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 95

THIRD SERIES.—@= 24°; barom.=75.08 cm.; 6f=2".

Z t Ss Corona. n' =376s3 | N IT (1-2) aes “5 n’ N Il’
I 40 fog
2 42 wr’
Sule 44 (13) Wp
4 46 (12) olive
5 | 48 (11) oat
6 50 Ce wyo 540 1.000 .00030 346000 1.000
7 De 9-4 wns 324 -759 33 255000 739
8 54 7.8 wpg 185 558 36 196000 506
9 56 6.5 gbp 107 .408 40 145000 .420
IO 58 6.6 yo 112 .293 44 106000 | .307
II 60 6.2 weg 95-2 .210 49 77800 a2
12 62 5a w br 58.1 .I50 55 56700 164
13 64 4.6 wo cor 38.1 Ten 61 40500 suns
I4 66 4.3 corona Bin. .070 69 28400 .082
15 68 3.6 oe 18.2 .046 78 19400 .056
16 7o Ber Ss TESS .028 go 12600 .0364
17 iD 2.8 7 8.54 .0153 .OO107 7700 .022
18 74 2.4 os 5-38 .0035 127 4500 .031

Note: 1,’ =

323 X10

380)

320) NU

320 |

383 J

(420)

Mean: 346000

n’ and S’ computed from subsidence and time.

To compare the new observations with the older ones above, it is con-
venient to lay off the aperture, s, in terms of the order of the exhaustion, z, as
has been done in the charts, figures 12 and 13. The following peculiarities are
observed: Below the g-b-p coronas (s=6 to 6.5 cm.) the general slope of all the
curves is nearly the same. Above these coronas, the older results obtained with
the plug valve correspond to a decidedly steeper slope than the new results.
Curiously enough, therefore, a greater number of exhaustions are required in
the case of the instantaneous valve to pass from a given corona to a given
succeeding corona than in the case of the stopcock. The instantaneous valve
thus removes a smaller relative number of nuclei per exhaustion than does the
stopcock, so long as fresh nuclei or great numbers are in question. Subsidence
of fog for the very fine particles here in question need not be considered in
explanation. In fact, the recent work is even more rapid than the old.

35. Growth of nuclei.—The simplest way of accounting for this result is to
assume that there is a continual growth of nuclei in the interval between the
observations, whereby those of extreme smallness come first within the range
of action of the instantaneous valve. The excess of available nuclei obtained

96 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

in this way would gradually decrease toward the end of the experiment so that
the slopes of all curves compatibly with the observations would gradually be
the same. Below the middle g-b-p corona, the curves if placed together would
nearly coincide. The difficulty with this hypothesis is the absence of any
obvious effect when the time between observations is varied. The curves are
about the same for 6f=2 min., 3 min., or even 12 min., so far as observed.

36. Subsidence.—The effect of errors in the subsidence constant may be
estimated. Writing the equation n,=n,10° ?'*?77, in the approximate form
nN, =Nn,10% (1 —S (e + 3 Wea 3 ,,+)), whence if N= 10 @-2) bey,
én, =Nn,N 6S = (S).

It is preferable to use the inverse method, putting u,=n,/N (1—S2(2)),
whence 6n,=¥2(=) OS.

Putting ,=10500, n,=212000, N=.0703, =(<)=.265,

6n,= 40000 6S, nearly.

Thus if 6S=.2, 6n,=8000, or relatively 8000/212000=.038. It follows
that by the error of .2 made in the estimate of S, , will not be affected more
than about 4 per cent., which in no way accounts for the observed discrepancy
of the new and the old results.

37. Exhaustion ratio.—Again if the above value of m, be taken, the effect
of an error in the exhaustion ratio y will be

— 6n,/n,= aaa Z)6y.

For 2,=212000, s—-Z=15—5, y=.767, an average case in the preceding
table, 6n,/n,=136y, or for

Oy —0n,/n,=130 %
.05 65 %
.O1 13 %

showing that great care is needed in relation to y.

38. Inferences.—To reach an opinion as to the cause of the observed
initial diversity of rates one may note that for equal pressure differences, smaller
nuclei are necessarily caught by the instantaneous valve than by the stopcock.
More nuclei are thus within reach in the former case: but apart from this,
since all other conditions are the same for both cases, the rate of denucleation
should be the same, even if the absolute number of efficient nuclei removed by
exhaustion is greater for the instantaneous valve. If therefore one begins with
the same corona which implies identity of diameter of particle and may be as-
sumed to imply an identical number of available or effective nuclei in both cases,
the two curves should be identical. In figures 12 and 13, considered first be-
tween the upper w-y-o and the middle g-b-p corona, the new and the old curves
cross at the large w-y-o corona. If under identical exhaustions the number of
available nuclei are therefore successively greater for the instantaneous valve

ee eS a

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 97

than for the stopcock, there must be an accession of nuclei between the ex-
haustions in the former case or a loss of nuclei for the case of the plug.

Between the middle g-b-p corona and dust-free air, nuclei are removed at
the same rate in the two cases, as the curves here have practically the same
slope and would be brought to coincidence throughout if the upper coronas
were to coincide. Inasmuch as in the case of atmospheric air the g-b-p corona
is not exceeded, it is thus immaterial which form of valve is used, and the direct
experiments of Chapter IX bear this out.

The alternative that nuclei are generated by very sudden expansion is
without correlative evidence. At least all my experiments to detect ionization
produced by sudden expansion have failed. It is equally difficult to account
for an abnormal loss of nuclei in the earlier stages of the experiment of table 18.
True, it is here that the ionization of phosphorus vanishes, though this evan-
escence is enormously rapid by comparison.

The nucleations of the g-b-p corona are by table 18, 7=129,000, 129,000,
and 145,000, respectively, values decidedly larger than were found above, and
due to the greater geometrical remove of the normal coronas from the g-b-p
corona. As about one more exhaustion is required on the average in the new
data, the corresponding results for the same z would be, 7=94,000, 94,000,
106,000, which agrees with the order of values above.

39. Dufferent rates of exhaustion for moderate nucleations.—In this place it
is interesting to insert a series of direct comparisons on the number of nuclei
within the reach of a half-inch exhaust pipe with an ordinary plug stopcock,
and the number caught when the valve is as in figure 3 with all pipes over one
inch in diameter and less than one foot long. These are given in table 19, the

TABLE 19.—DIRECT COMPARISON OF RESULTS FOR STOPCOCK .5’”” DIAMETER,
AND INSTANTANEOUS VALVE 1” DIAMETER. LARGE CON-
DENSATION CHAMBER, AIR NUCLEI. 6Op=17 cm.

Experiment No. Remarks. Corona. Ss 1 =37553
em

I Stopcock weg’ Cas 62000

2 ree gbp 5-9 77000
3 ae ce g’ b p ceo 73000
4 aie ip: gbp 5-7 69000
5 Inst. valve wp 6.9 124000
6 - s wp 6.9 124000
7 = oe wp 6.9 124000
8 Stopcock wp 6.4 g8000
9 Zs i wp 6.7 112000
10 “ - Wp 6.8 rt 18000
Ir Inst. valve weg a3 149000
12 is i Ww cg 7 137000
a3 i - weg 7-2 140000
S14 Stopeock weg 7.2 140000
ee | € ‘< weg 7.3 149000

98 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

data for the stopcock alternating with data for the instantaneous valve. Un-
fortunately, the time intervals between observations were not quite equal, as
the necessary adjustments consumed varying amounts of time.

These data are shown graphically on the chart, figure 14, p. 83, the abscissa
being merely distributive. If the mean data of each set be taken the results are

for the stopcock: s=6.57, S= 284,
for the inst. valve: s=7.04, S=3 40,

or the ratios of s* are about .81. Since, however, in view of pericdicity, the
g-b-p coronas show abnormally small values of s, it is better to compare the
red coronas only. In this case the mean values are

for the stopcock: s=7.0, S= 343,
for the inst. valve: s=7.2, S= 373,

or the ratios of s* are .g2. Hence the two valves differ as to their data by less
than 10 per cent. If this value be ascribed by the value of m assumed it would
not suffice for the values obtained by photography.

In table 18 the ratios of s* for the large crimson coronas under like occur-
rence were on the average .42, widely different from the present, showing that the
discrepancy in table 18 enters with the very small nuclei of the very large
coronas.

One may note that the nucleation of the large room is gradually increasing,
due to a single small gas burner (source of light), although the air was all pumped
into the condensation chamber from the floor of the room.

40. Conclusion.—As none of the explanations given are satisfactory, it
seems well to restate the case in conclusion with additional remarks on a possible
explanation. No matter which group of sizes of nuclei are efficient in producing
coronas, the effect of successive identical partial exhaustions must be to reduce
the numbers by the same relative amount. If a given corona is obtained for
the slower and the faster exhaustion for the same pressure difference, etc., it
may be assumed for argument that the same number of efficient nuclei must
occur in both cases, though they may not be the same nuclei as to size. It
follows that the same corona should occur in the two cases in each of the suc-
cessive exhaustions. This is only true below the middle g-b-p corona, 7. ¢., for
relatively fewer nuclei (nucleation below about 100,000 per cub. cm.). In
this region of distribution, both valves eventually remove all nuclei, and they
remove them at the same rate. :

For nucleations above 100,000, however, proportionately more nuclei are
apparently removed by the slower exhaustion, cet. par., than by the faster
exhaustion, so far as coronal evidence is in question, precisely as if the faster
exhaustion were itself productive of nuclei (as, for instance, by breaking up
coarser into finer aggregates). One may note, moreover, that in this stage of

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 99

the work, the phosphorus nuclei are at first ionized. Subsidence is without
effect.

It is just here that another important question is suggested which may
perhaps offer evidence in explanation. It was shown above (§ 8, table 2) that
the moist air after exhaustion very nearly regains its original temperature after
the lapse of even half a minute, while the coronas persist throughout this in-
terval and much longer without appreciable change of character. It is difficult
to understand why the fog particles do not evaporate proportionately to the
rapid rise of temperature, unless there is rapid evaporation and diffusion from
the relatively warm inner surface of the walls of the condensation chamber, im-
mediately after exhaustion. In any case, the method above used for computing
m, the moisture precipitated per cub. cm., will give a result too large, for it
takes no account of the rapid increase of temperature in question after the
fog particles are produced. The swifter the exhaustion the larger this dis-
crepancy (which is probably indeterminable) is liable to be. Thus in the above
case for the pressure difference 6p=17 cm., at 20°, the cooling ideally as far as
—9g.6°, rises to 8.8°, in consequence of condensation of fog particles, but within
4 minute the temperature is nearly 20° again. Hence the precipitated
4.6 X 10 °g/cm$ at 8.8° is to remain undisturbed while the moisture content of
saturated air rises from about 8.7 X 107° at 8.8° to about 17.2 X 10 °(g/cms) at
about 20°, leaving an actual deficit of about 4 X 107°g/cm*. This would be out
of the question unless moisture evaporated immediately from the damp and
warmer walls and the pool of water in the bottom of the apparatus to supply
the deficiency. But these conditions are vague, for this moisture may actually
be precipitated on the colder fog particles. To some extent, therefore, a degree
of uncertainty is left in the determination of m, the moisture precipitated per
cub. cm., inasmuch as the actual temperature at which the fog particles persist,
and to which they have accommodated themselves, is left in doubt.

It is well to observe, however, that as there will presumably be more evap-
oration from the fog particles while the normal air temperature is being
regained in case of very rapid than for the slow exhaustion because lower tem-
peratures are in general associated with the former case, the discrepancies
of plates 12 and 13 above the g-b-p corona may possibly be explained in this
way. Compatibly with observation the effect would be less marked as the fog
particles are larger. Finally, very large coronas are always more fleeting in
character.

Experimentally and with a bearing on Chapter IX the question is easy of
decision. So long as the pressure difference corresponding to the lower limit
of spontaneous condensation of moisture from dust-free air is not approached
(it will usually be reached at about %p> 20 cm. for the above types of appa-

1 Recent experiments have shown that the very small nuclei associated with larger nuclei evaporate
their loads of water after condensation in such a way as to form water nuclei. As the latter must be
larger than the original nuclei now held in solution, a reason for the excess of nucleation detected by the
instantaneous valve is suggested.

100 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

ratus), and so long as the g-b-p corona is not exceeded (which again is the
actual case), the degree of rapidity of condensation on the usually occurring
atmospheric nuclei is a matter of little consequence. Thus the same coronas
appear with non-filtered atmospheric air, cet. par., both for the $-nch plug
valve and for the 1-inch instantaneous valve, even for large variation in the
size of the condensation chamber.

CHAPTER VII.
I. MICROMETRIC MEASUREMENT OF FOG PARTICLES.
Earlier Methods.

1. Before using the data computed for the dimensions of fog particles in the
reductions of my observations of atmospheric nucleation, it seemed expedient
to endeavor to obtain corroborative values by some straightforward method.
Aitken’s dust counter had naturally suggested itself early in the course of my
work; but the results so obtained are essentially indirect, as the fog particles
are not themselves observed. It was necessary, therefore, to devise apparatus
by which the identical fog particles of a given corona could be directly en-
trapped and held for examination. This was eventually accomplished in a way
admitting, apparently, both of the measurement of the diameters of the par-
ticles and of counting the number precipitated under known conditions. More-
over, the particles caught, however fine (even less than .0003 cm. in diameter),
can often be kept in place for observation for some time, so that microscopic
photography would appear*to be at once applicable not only for the purpose of
obtaining size but number.

Many investigations are thus suggested, as, for instance, a repetition of
Thomson’s method for determining the charge of an ion. Again, while the
corona gives merely the average size of the cloud particles, the microscope is
particularly available for indicating variations of diameter for the particles of
the same corona. In fact, the water particles of the coronas as caught on the
plate are not of a size; they are graded, and hence the nuclei are probably also
graded in size.

2. Apparatus.—Aitken’s beautiful and highly ingenious instrument is well
adapted for the purposes for which it was designed. Apart from this, it will
furnish only an estimate of the dust contents sought. The droplets evaporate
too rapidly, and are often too numerous for exact counting. The need of
mixing atmospheric air with dust-free air with shaking and stirring is an inter-
ference with the nucleation. In fact, water nuclei may even be generated in
this way, possibly by the friction of air passing across damp surfaces. There is
the tendency of the plate after long use to fog permanently or to collect droplets
on its own account. Finally, it would be very difficult to remove the contents
of the coronal chamber to the dust-counter without reducing the nucleation

during the transfer.
Ior

102 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

I therefore endeavored to ascertain whether the particles might not be
made visible directly. The chances of success seemed small indeed, particularly
as Assmann had failed to see the particles with magnifications of even 400
diameters. But after long trial, the result was eventually accomplished in a
way that now seems surprisingly simple.

The compound microscope, M, magnifying about 100 diameters, is pro-
vided with a filar ocular micrometer, ~. The objective and the whole lower part
of the microscope is submerged in the condensation chamber, being suspended
for this purpose from the wide rubber cork, C. All lenses below C are her-
metically sealed with wax. The microscope originally carried a rigid stem, 7,
to which were attached the plate, s, to be examined, the mirror, m, and the
metallic disc or shield, p. Afterwards the more flexible adjustment shown in
the figure and described below was adopted. The lower side of p, which is flush
with the objective, and the upper side of s are covered with wet blotting-paper,
the latter being perforated to admit light into the microscope through the thin
cover-glass placed at s and held sharply in focus by a suitable clip. The field
within which drops are to be counted is bounded at pleasure by the wires of the
micrometer.

FIGURE 1.—MICROMETRIC APPARATUS.

This apparatus was totally unsuccessful. Drops were but rarely seen to
fall on exhaustion, while the dew soon gathered on the plate, s, in such a way
as to be easily mistaken for droplets; for the dew evaporates like the latter
when the microscope is removed, and the regularity of the pattern on the plate
is the only distinguishing feature.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 103

Various modifications of this apparatus were then used, among them cap-
sule forms, figures 2 and 3, similar to Aitken’s, but containing a very thin plate
of glass or mica or celluloid slightly raised above the base on pellets of wax.
It was supposed that this would counteract the tendency of the drops to vanish
by evaporation from the warmer glass surface. Capillary metallic tubes led to
the curl aneroid, a, the filter, f, and to the.cock for influx of air, the only large

FIGURE ta.—THE SAME MOUNTED IN CONDENSATION CHAMBER.

tube being the exhaust pipe, e. Condensation again occurred as a microscop-
ically granular deposit spontaneously on the raised surface, under all circum-
stances, and the experiments were failures. After oiling the filmy mica surface,
p, however, droplets were often seen to fall and either to stick fast or to float.
These could at times be counted (2000-5000 per cub. cm.); but the rapid evan-

FIGURE 2.—CAPSULE. PLAN. FIGURE 3.—SAME. SECTIONAL ELEVATION.

escence of precipitated droplets and the failure of all attempts to reach systematic
results induced me to abandon the capsule.

I therefore returned to the apparatus in figure 1, using at s a plate of thin
microscopic glass covered with a thin film of oil and exposed in the capacious

104 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

vacuum chamber. The experiments were now phenomenally successful. Thus
for the aperture s = 5 the mean results were m = 150000, and fors = 4.6 (wg bp),
n = 140000. The precipitation of globules was clearly seen, and they persisted
even after the exhaustion was removed. The numbers being excessive and
referable to globules swept in by lateral air currents, an improvement was now
added by increasing the diameter of the disc p to about 5 cm. The improved
apparatus gave no results whatever, and the mere addition of the wider disc
wiped out all precipitation. But this capricious behavior is characteristic, for
next day drops were seen to fall as follows:

Si sAen wgbp n= 6.5 X tot
4.6 wgbp 4.7
6.3 wegebp 1333

after which no precipitation could be caught in the 6 subsequent exhaustions
by the identical method. The same unaccountable irregularity was noted in
the afternoon. Next day the first experiments showed

Ss = 6.0 weg n= 7.3 X 104.
6.4 Wgvp 12.0
etc. :

after which further precipitation did not occur.

The apparatus was then again modified to the final form shown in figure 1,
by inserting a thin brass tube laterally through the stopper, C, and firmly solder-
ing this tube above and below at e to the body of the microscope. A rod snugly
fitting this tube thus provided an eccentric focussing device, abcd, with a stuffing
box at b, and an external handle at a. The latter is adjustable by aid of a set
screw so that the plate s may be kept in focus during rotation of the rod. To
catch the droplets, the plate s is rotated into the position s’ quite beyond the
shield, for a definite time (15-30 seconds), and then returned to s for examina-
tion. In this way the definite results were obtained, in a manner to be further
detailed below, with the apparatus free from capricious behavior. It is of par-
ticular interest that the particles caught on the oiled surface persist as brilliant
round globules for a long time (sometimes 10 min. or more) ina saturated atmos-
phere. They very gradually vanish as a rule on the readmission of air into the
condensation chamber.

To remove the globules for the next experiment, the influx of air is thus
not always sufficient. It is necessary to withdraw the microscope from the
condensation chamber bodily and to wave it about a few times in dry air. On
returning it to the chamber the plate is then again clear and white.

At first the plate was oiled by a small flat piece of blotting-paper saturated
with oil and held on a stem, care being taken to remove all excess. Clean
machine oil or ordinary illurinating oil or a mixture of the two subserved the
purpose about equally well. Probably the best method of oiling consists in
dipping the plate rotated outward to s’ in very hot melted vaseline (to drive

:
?

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION, T05

away moisture), removing the excess while hot by filter paper and when cold
submerging the plate in petroleum for transparency. With solution of vaseline
in benzine, etc., I have been much less successful. Damar varnish and turpen-
tine was much used in the final work. When drops are to be counted by the
method given below, the oil film must be practically solid; otherwise the capil-
lary forces produce an immediate and often startling redistribution of the pre-
cipitated granules, though they but seldom coalesce.

3. Behavior of the precipitated droplets —In case of a petroleum film on the
plate, the water droplets were sometimes seen to fall and float on the film, which
is positive evidence against spurious droplets. They are usually black and cir-
cular in outline, but when the light is intense and axial, they are bright. Fixed
globules are apt to be larger and more irregular. Particles may sometimes be
seen to coalesce on collision, but this is rare.

On tipping the microscope so that the light does not penetrate the vividly
colored drops axially, they seem to cast shadows in opposed directions for
symmetrical inclinations on both sides; but in view of the aplanatic properties
of spheres, the phenomenon is probably a case of refraction, with the shadow
beginning at the edge of a caustic. Similarly, on moving the lamp horizontally
to either side from the position corresponding to axial illumination, the globules
become opaqte, and look like round shining steel beads. The diameter of the
beads has but little effect. If the lamp 1s moved until the field is dark, the
plate looks like the starry heavens. These stars seem to be above the drops.

=0- © -O=
® ® @

oO, © «Do
(6)

FIGURES 4, 5, 6. DIAGRAMS SHOWING THE BEHAVIOR OF FoG PARTICLEs.

After remaining in the plate for some minutes the fixed droplets often
become rosette-shaped (apparently), at first showing a mere black spot in the
center of the color disc, which gradually enlarges to a ring-shaped appearance
slowly moving radially outwards. As a rule, the color! is eventually the same
on the inside and the outside of the enlarged ring, the ring itself appearing red
with black demarcations in the surrounding green field, as shown in the figure.
On influx of air the structure becomes washed. This ring-shape may be merely
apparent, but the small globules when at first deposited never show the same
color within and without, the former being uniformly red and the latter white.

1 The colors observed were afterwards found to be due to chromatic aberration of the

microscope.

106 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

As the rings are not easily produced with a very viscous oil, it is probable that
the droplet has here penetrated to the glass and that the oil film is drawn over
it by the capillary forces at the common edge of the three media.

4. Preliminary data.—Before adopting the eccentric focussing device,
many experiments were made to ascertain the cause of the uncertainty in
catching the drops on the plate when kept in place, seeing that sometimes the
precipitation was abundant, while at other times under the same apparent con-
ditions drops did not fall. Failures occurred both for high and for low nuclea-
tion. From the outset it was improbable that radiation from the outside could
affect the result. It was eventually ascertained that on tipping the condensa-
tion chamber after the fog had formed, so that the subsidence would reach the
plate obliquely, a precipitate would usually appear. Again, an oblique current
within the chamber and passing across the plate usually produced a deposit.
Hence the drops actually exist within the fog, and success in bringing them down
upon the plate is probably conditioned by very close isothermal adjustment of
the plate to its surroundings, added to the advantages gained from incidental
and favorable air currents. Hence a little time must always elapse before the
drops persist at the plate, and hence the droplets from a shallow capsule do
not appear. Using a film of mica as a plate the result was the same, and it is
useless to attempt to enumerate the drops by this method. Those which fall
are carried in by grazing air currents, while no drops are obtained from the
fiducial space under the objective.

Nevertheless, the measurement of the diameters, d, of the drops obtained
by the above method without modification is an excellent test of the results
obtained elsewhere by computation. The factor of the ocular micrometer
described above was .oo2 cm. per turn of the screw, or .cooo4 cm. per scale part
of the drum divided into 50. The extent of plate covered by the breadth of the
spider lines was about .0003 cm. The finest particles are of about this diameter,
so that such measurerents must at best be much inferior to photography with
a scale attachment. The results are given in the following table, in which only
those results among many are inserted for which the observations were clear
and satisfactory. The coronal color with its diameter, s (chord of a radius of
30 cm.), are as observed when the eye and the source of light (Welsbach mantle
seen through a small circular hole) were at distances 85 cm. and 250 cm., re-
spectively, from the center of the condensation chamber. The exhaustion was
usually to a pressure difference of 17 cm., but this is of no significance when
diameters are alone to be observed. The particles were collected by tipping the
chamber, sometimes in large numbers, but at other times sparsely distributed
without apparent cause. Nuclei were conveniently obtained from burning char-
coal. Both floating and fixed globules were examined with strong microscopic
illumination. It was difficult to retain a clear image without frequently re-
moving the plate, as the adjustment for focussing the plate within the chamber
had not yet been adopted. A table showing the results from coronal measure-
ments under the same circumstances is added.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 107

TABLE 1.—DIAMETERS OF CLOUD PARTICLES. PRELIMINARY RESULTS,
FIXED PLATE. :

s Corona. Computed d. Measured d. Remarks.
cm . cm. cm
6.5 wbp .00046 .00060 Fixed
6.5 w|b|p 46 084 ‘a

= olive 2 050 Floating
—— olive 32 040 Fixed
6.4 g’|blp 46 084 a
9.1 wr bg 37 052 Floating.
3.0 corona 82 096 Fixed.
3.8 oe 76 116 os
5.8 weg’ 50 088
4.8 corona 60 088
2.4 i? 120 222

TABLE 2—DIAMETERS OF CLOUD PARTICLES. PLATE ADJUSTABLE.

s Corona Computed d. Measured d.
em. cm. cm.
Phosphorus — olive .00030 .00042
nuclei, 10.0 w o bg 35 oa
8.2 weg 39 060
7EO wpbg 44 050
5.6 wrs 52 060
Bn3 corona go II7

Floating globules were often observed in the act of coalescing; but this is
much rarer than the passage of a floating droplet over a fixed one without
interference. A distinct central bright area shading off into darkness was seen
even in the floating globules when axially illuminated by intense light. The
larger drops were often metallically green. The colors vanish after long stand-
ing, and they are particularly vivid immediately after falling. It was not even
now possible, in spite of all precautions in tipping the vessel, to obtain an abun-
dant crop of drops at pleasure.

The results are given graphically on the chart, figure 7, in comparison with
the computed data of my earlier experiments (upper curve) as well as with my
later experiments (lower curve). The observed results lie below the former
where adiabatic conditions were assumed, and above the more recent experi-
ments where the effect of the successive expansions was computed isothermally
In other words, the observed diameters are intermediate, but nearer the older
results. I will pass over these preliminary data, as they are probably too high
because of the difficulty in focussing among others, to resume the subject in con-
nection with the results obtained under much more favorable conditions below.

108 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

see © ae
se BOG [eee

4 ° 6 0 al
ia

a
8 bxtoin
$s
0

0 KR 4 6 8 70 7k “4

Figure 7.—Cuarr GivinG d In TERMS OF s.

Improved Method.

s. Number of droplets—The following results were obtained with the
definite form of apparatus shown in figure 1. A method of estimating the
nucleation from a direct count made under the microscope is obtained as follows:
Let the plate be so rotated eccentrically as to catch the descending fog particles
for a definite interval of time, t. If v be their velocity of subsidence, all par-
ticles within a height, , will be caught, if

h=vut (1)
and J— TOW) aon (2)

where the usual value of the viscosity of air has been inserted. Furthermore,
m grams are precipitated per cub. cm. by the given exhaustion, and if 7 be the

nucleation
n = 6m/7d5. (3)
Finally, if c is the area of the field seen in the microscope and n’ the number of
particles falling into this field
n’ = nhe. (4)
From these equations is obtained by eliminating v as

S| ze te ate
n= —@——
tem?/3106.-

The values of the constants usually adopted were t = 30 sec., ¢=144 X 10° ° Sqe
cm., m/3= 2.8 X 1074, whence yJn=1.75 Xn’. The experiments to test this

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 109g

method often gave serviceable results, some of which are inserted in the follow-
ing tables; but at times the u-values are out of proportion. The reason of this
is threefold: In the first place 7 is found from "3 with the usual difficulty.
Again, in a simple arrangement like the above, air currents cannot be quite
excluded. They may arise incidentally in the apparatus or the motion of the
plate even if parallel to itself may stir the air unless some form of guard ring
attachment is added. Particles are thus swept down upon the disc before and
after the exposure, as was actually observed. The difficulty may be removed
by adding a capsule above the plate or simply by decreasing the distance be-
tween the shield and the plate to a millimeter or less. Finally, if the oil film is
semi-fluid and not quite fixed, if there is slight creeping, as was usually the case,
the particles are redistributed after falling along stream lines where they cohere
in strings and bunches, but without coalescing. This was also observed, and in
fact the capillary forces involved are apt to be strong enough to counteract
viscosity.

I have not thus far spent much time in correcting these defects, chiefly
because the new results for the diameters of fog particles are more immediately
interesting. The data are given in table 3.

TABLE 3— OBSERVED DIAMETERS AND NUMBERS (per cub. cm.) OF CLOUD PAR-

TICLES. m=4.7 X10~g per cub.cm.; If c= 144 X10 °sq.cm., and t= 3osec.; Py
=1.75 n’. Generally p/n = 2.11 n’/tc m*/3 10°. Micrometer factor, .coco4 cm.
~ ; , Observed Computed Observed Computed

& Corona. |c X 10° ¢ n i ; a d

cm. cm.? sec. em. cm

aad olive 140 30 207 271000 | 250000 —— —

6.0 wrg 140 30 27. | 105000 goo0o0o — _
Phosphorus 10.2 wobg’| 70 30 17 | 210000 | 210000 | .00036 | .00036
nuclei. 7.4 wpbg] 7o 15 6 —_— ~- 48 4
ia w’ bp — — — = 56 2
4.7 | corona cal al a aa So 61

Particles as small as .o003 cm. present throughout.
Phosphorus 6.5 gbp T40 15 10 43000 | 100000 | .00064 | .00046
nuclei. 6.5 | g’ bp = = = = 56 40
6.1 | wrlg ad a aaa an 2 48
Particles as small as .o003 cm. always present.
|
Air nuclei. 4.5 cor 140 15 9 30000 40000 | .00064 | .00064
Cr nen eae es = =a a 52 46
8.1 wc bg’ Ou aS 14 120000 | 150000 48 39
8.2 | we bg’ 7o 15 16 | 180000 | 150000 | — 39
, |

Particles graded as usual.

I1O A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

6. Diameters of fog particles—If the diameters measured are plotted in
a chart together with the results computed from successive exhaustions in
the older and in the more recent experiments, the present values again lie be-
tween the two curves, but now much nearer the lower (recent) curve than before.
I shall not pause to interpret the differences which remain, but only to remark
that the capillary forces at the area of contact of the droplet even with the liquid
oil film may transform it to an oblate spheroid, and that diffraction at the
circular edges of the drops is not excluded. If the nucleation, ,, obtained
from successive isothermal exhaustions and subsidence measurement, be ac-
cepted as correct (lower curve), the ratios of the nucleation found from the
different methods tested will then be

From subsidence, a=.0029; d/d,=1.0 n/n,=1.0

From lycopodium (d=.003 cm.), @=.0034; “ =1.2 <= 01
From diffraction (blue), a=.0034; BO =D “« = 61
From micrometer measurement, @=.0037; “ =1.3 = = 7-48
Old results (adiabatic conditions assumed), ‘‘ =1.6 a 24

Since 1 is obtained from the cube of d, large differences of this kind are as yet
inevitable, particularly as the particles measured in these different cases are not
the same.

7. Sizes of particles graded.—The point of particular interest which comes
out on using the eccentric plate to catch the subsidence during 15 or 30 seconds,
and at once examining the deposit, is the result that particles of all sizes are
present. By far the greater number, however, have the maximum diameter.
These particles are caught from the fog without interference, and it is not prob-
able that coalescence or evaporation have been appreciably operative, so long
as the corona remains the same throughout the micrometer measurement. The
probable explanation is this: while the pressure decrement is growing from zero
to the maximum 6p, condensation is taking place on the greater number of
particles throughout the whole of this interval. In other words, although the
nuclei are graded in size, the greater number exceed a certain dimension and
require almost no pressure decrement to induce condensation. These are the
particles (diameter exceeding a certain inferior limit) which give character to
the persistent corona. A minority of the graded particles are below the di-
mension in question, and upon these condensation does not take place until the
higher values of the pressure difference are reached; some may even require
the full decrement, 6p. Thus it is that in the deposit of fog particles, one finds
those of diameter .oor cm. intermixed with others of smaller diameter, even as
far as .ooo2 cn. or less, all shining like beads. When fresh phosphorus nuclei
are first introduced into the condensation chamber the result is a gray fog, but
a relatively s‘rall white reddish corona is nevertheless discernible. Accordingly,
the crop of droplets seen under the microscope contains not only surprisingly
small but also relatively large droplets, with all intermediate diameters. Hence
the indefinite fog and the small corona. The large olive (g b p) corona and

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. IIt

other of the early coronas are very apt to fade into a coarse white reddish corona.
This is the evaporation of the smaller particles into the larger, which accounts,
moreover, for the loss of nuclei during the first precipitation, to be caught in
subsequent exhaustions. The successive coronas in a series gradually become
sharper and the larger particles more uniform, but extremely fine particles are
still present even when one approaches the normal coronas. The fine particles,
however, belong to coronas so large and diffuse that their coronal effect scarcely
modifies the strong coronas of the large particles even before the former vanish
by evaporation.

When I first observed these different sizes of drops caught on a single plate,
it seemed not improbable that a difference of the condensational effect of the
negative and the positive ions might here be actually in evidence; but as all
intermediate sizes are present at the outset, and particularly as large and small
droplets still appear together long after all electrification has certainly vanished,
this conclusion is not warranted. What continually favors uniformity is sub-
sidence of fog. As the phosphorus nuclei are graded, it is probable that the
very fine droplets are due to the initial or primitive nuclei from which the larger
nuclei have grown by coalescence; or the fine droplets may be due to air nuclei
associated with the phosphorus nuclei. All this will appear in the more minute
photographic study of the subject detailed in the next section, and it will be
further interesting to decide whether the nuclei generated by the X-rays are not
also graded below a certain usually much smaller maximum diameter. That
this maximum diar eter will increase with the lapse of time allowed for coherence
may be inferred.

The coarse and washed type of coronas obtained with nuclei produced by
the X-rays is evidence of graded size, while the fog particles, so far as I have yet
caught them, are of varied dimensions. In these cases the X-rays reached the
inside of the condensation chamber through its waxed wood walls lined with wet
cloth. To obtain a fairly strong and large corona an exposure to the rays lasting 5
to 10 minutes was needed, as the radiation was not very intense. In this interval
the original extremely small nuclei are probably undergoing continuous growth,
for instance, by cohering, so that on exhaustion particles of all sizes are revealed.
In addition to the ragged coronas there is copious rain. Under these circum-
stances it seems reasonable that the time loss of nuclei must at the outset be
proportional to the square but finally to the first power of the number, assuming
that eventually the large nuclei do most of the catching.

Il. MICRO-PHOTOGRAPHY OF FOG PARTICLES.

8. Preliminary.—In the preceding section ! I described a series of experi-
ments in which the diameters of fog particles were microscopically measured,
directly. In the present section these particles are micro-photographed, and
the negatives subsequently measured. The results, though interesting as a whole,

1 See also American Fourn. (4), XVII, p. 160, 1904.

TieTe A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

are not as immediately available for quantitative discussion as was hoped.
The number of nuclei per cubic centimeter and the diameters of the fog par-
ticles are respectively below and above the computed nucleation (Chap. V1),
and the globules photographed are rarely of the same size for a given corona.

Very curious results, apparently capillary in character, were obtained,
showing permanent pitting effects of the subsidence of fog on a film of viscous
oil and persistent motion of globules in liquid oil.

9. Apparatus and method.—The apparatus needed is a modification of that
described in § 2 of this chapter for the micrometry of fog particles, with the addi-
tion of a camera above the microscope. The usual form of camera attached to
a substantial eccentric axis so as easily to be rotated into place or removed
therefrom for inspection is satisfactory. The revoluble condensation chamber
must be clamped in place. The camera may be focussed by the stage focussing
screw of the microscope, so that a lens is apparently not necessary in the camera;
but it is essential that the magnified fog particles be seen very clearly with the
eye at the microscope, before photographing them, and this procedure is there-
fore not quick enough. It was thus found necessary to add a good lens to the
camera adjusted for parallel rays and to adapt the eye for the same infinite
focus by concave glasses. In this case, when the particles had been caught and
put in place under the objective by the eccentric stage device described above
($ 2), the camera could be at once swung into position for photography. The
endeavor to adapt a small kodak for the work was not very successful.

Magnification may be secured either at the objective, or at the ocular.
Some space between the objective and the plate is desirable for safe manipula-
tion, and therefore a half-inch objective will serve the present purposes better
than a quarter-inch lens. The illumination must be axial to avoid astigmatism,
but the use of condensers has not as yet been tried, and would probably pro-
mote evaporation of the fog particles. In general, reasonably small magnifica-
tion, much light, and rapid photography are best conducive to success. In this
way the time of exposure was gradually reduced from 20 seconds to 2 seconds
or less. The positive ocular seen in the plates was used merely in the absence
of a suitable negative ocular, inasmuch as the former was provided with a filar
ocular micrometer. But a negative ocular containing a plate ruled in square
millimeters would be far preferable.

The plate in these experiments was covered with an even coating of Damar
varnish, neither too moist nor too dry. A clean microscope cover-plate of glass
is dipped in the varnish, and the excess removed by placing it on edge. The
film, which must be smooth, clear, and even, will be ready for use in a few hours.
If too dry it should be soaked in turpentine, otherwise the particles adhere
broadly to the plate and rapidly evaporate. If the plate is too moist, the
particles float and cannot be photographed. No precise rules can be given.
Naturally, the conditions surrounding the plate should be as nearly as possible
isothermal, which implies a capacious air chamber for condensation.

10. Incidental phenomena. Pitting—A curious phenomenon sometimes

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 113

accompanies the deposition of the fog particles on the film of viscous varnish,
inasmuch as they leave a permanent impression. In other words, the plate be-
comes more or less permanently pitted after the fog particles are gone, appearing
washed or dull to the naked eye, and not regaining the clear state until after the
lapse of 12 hours or more. The effect occurs only in the case of fog deposition
and is never present on the plate in the absence of a precipitate of fog. Hence
it cannot be an air bubble effect.

The cause is probably to be associated with surface tension since the weight
of the particle is negligible. If the particles were to break through and reach
the bottom of the film of varnish, there is no obvious reason why the pitting
should vanish like a viscous phenomenon in the course of time.

It is conceivable (Fig. 8) that the surface tension of the varnish is locally
lessened by slight admixture, or at least the proximity of water, and that an
alveolar structure of the surface is the result. In certain slides (No. 22) the
presence of these “‘craters,’”’ as they may be called, seems to be clear in the pho-
tograph. At other times droplets shrinking in their cavity by evaporation
were actually observed. But the phenomenon is rare and observation there-
fore uncertain.

t1. Dew.—When the plate is dry, the beginnings of the formation of dew
on its surface are apparent after long exposure. The dew particles are very
fine even as compared with fog particles, the former as observed lying within
.ooor centim. Their number is enormous, aggregating to fully 2 million or
more per square centim. of the surface (slide No. 27). They do not further
seem to interfere with the deposition of fog particles than by promoting adhe-
sion, but this is naturally objectionable.

pam eel

FiGurE 8.—DIAGRAM OF ‘‘CRATER.” FIGURE 9.—DIAGRAM SHOWING CIRCULATION

12. Evaporation—The droplets, originally sharp in outline, become
vague and washed on evaporation, doubtless because their curvature decreases,
while the area of adhesion remains the same, to the detriment of the nearly
spherical curvature at the beginning. Floating globules are usually much more
uniform in size and remain more uniform, because the differences due to adhe-
sion are absent (see $$ 9,13). It has been stated that when there is evaporation
the photograph fails and a blank plate results. It may be assumed that a
successful photograph implies as little evaporation of fog particle before the
taking of the picture as during this interval.

13. Floating and moving globules.—A final very interesting phenomenon
is met with in the case of fog particles floating in a liquid film of oil. It fre-
quently happens under these circumstances that there is a sharp line of de-
markation in the field of the microscope, probably an edge of contact of the
semi-fluid matter on the plate. In all such cases there is apt to be continuous

114 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

motion of the submerged or partially submerged fog particles, to and from the
edge on one side. They approach and leave in great armies, at first, gradually
dwindling in number as they pass out of the field of view, eventually to be lost
by evaporation. Sometires both advancing and retreating fog particles are in
sharp focus at once. At other times the advancing set is obviously above or
below the other, to the extent of one or more tenths of a millimeter. Cases also
occur in which there are two edges or a sort of geographical strait in the field of
view, in which case particles are frequently seen moving from edge to edge
until they vanish in number, probably from evaporation. Vortical or involved
orbits of fog particles also occur.

In explanation of these phenomena it is necessary to bear in mind that the
edge of retrogression mentioned is always nearly fixed, under the microscope.
Hence it is not probable that that motion of the mixed oil (varnish-turpentine)
due to concentration on evaporation can be the cause; for in this case the liquid

would visibly gather itself up into a drop showing a shifting edge in the field. |

The possible motion of a liquid on its surface skin or similar capillary phenom-
enon is equally hard to reconcile with the stationary edge. The most probable
explanation, it seems to me, is given by the annexed diagram, figure 9. If the
film on the plate of glass is microscopically uneven and slightly inclined, the
motion of a relatively liquid layer over a fixed layer may enclose a shallow
region of liquid, in which eddying is kept up, as shown in the diagram, remem-
bering, of course, that all motion is observed under the microscope. ‘The
particles indicate the motion of the skin circulation.

14. Graded particles —Suggestion may finally be made as to the cause of
the observed gradation of particles, where such gradation appears simultaneously
with clear-cut coronas. In case of the X-rays the coronas are vague and washed,
accompanied with copious rain. What is seen is a coarse red-rimmed fog.
Here gradation is obviously due to the corresponding gradation in the sizes of
the nuclei, as explained above. Something similar shows itself in the initial
phosphorus or sulphur fogs.

The grading in question cannot be due to coalescence, not only because
such coalescence is but very rarely observed, but because the volume increase
is as the third power of the diameter. Sizes as 1 to 3 being very common, this
would mean an equally frequent coalescence of 27 droplets, which could not
escape detection. Moreover, the coronas in air retain a nearly fixed diameter
until they are lost by subsidence.

Some difference of size must be due to a difference of surface adhesion;
but if the oil stratum is of the same character throughout, this is not liable to be
large without being detected in the picture.

The final cause for gradation, apart from original differences of nuclei, is
evaporation. The occurrence of marked evaporation is at times beyond ques-
tion, and it is then impossible, or nearly so, to obtain a photograph. In general,
however, evaporation is obscure, and one may argue, as already suggested,
that in the time (about 60 seconds) needed for adjustment and photography,

i

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. Im5

the evaporation should be no less than during the anterior 15 or 30 seconds of
subsidence during which the fog particles were caught. This would seem to be
especially true where the particles persist for several minutes after photography,
as is usually the case. Nevertheless, it is not impossible that the first precipitate
prepares the plate (by evaporating partially) for the subsequent precipitation.
The subject will be resumed in connection with definite results below.

Results.

15. Photographic plates—Unfortunately, it is impossible to reproduce the
photographic negatives obtained, except in certain cases; for not only are the
fog particles frequently too small, but slides suitable for the measurement of
number are indistinct in relation to diameter. To obtain a plate in which the
particles actually stand out is naturally a matter of chance, as this will depend
on many conditions (light, evaporation, focus, etc.) beyond the observer’s con-
trol in expeditious work. Curiously enough, the plates obtained for fog par-
ticles condensed on the persistent nuclei produced by the X-rays in dust-free
air are the best of my series, and are therefore reproduced.

To measure the sizes of the particles on the plates a filar micrometer of
low power was used. To count them, the slide was divided off into fields of
convenient size, andthe number then enumerated under a lens, taking all the
fields in clear focus in succession.

16. Tabulated results —The following tables show the general results for
about 50 photo-micrographs. The kind of nuclei used are given in the first
column, the coronas and their angular diameters (s/30, nearly) in the second
and third. The mean diameter, d, of fog particle measured from the photo-
graph and the limits of d observed, follow. The next two columns show the
number of nuclei, 7, obtained from the photograph by the equation of this
chapter, § 5, and the observed limits of ” for the different fields counted. The
last columns are explanatory and usually show the film of oil or varnish used to
catch the precipitate and the character of the plate.

17. Remarks on the tables—The use of rough oil films (tallow, paraffine,
wax, etc.) is naturally unsatisfactory, but the particles are easily recognized
(No. 3). Plates Nos. 4, 6, 22, though very perfect as photographs (large camera
magnification) showed a permanent impression which lasted 6-12 hours, as
already stated. In No. 8 only a few particles were caught. Nos. 9 and 10 were
taken with a 14-inch objective, and though good in themselves were hard to
obtain, and showed the advisability of the weaker objective (4-inch), favorable
to shorter times of exposure. No. 11, and particularly No. 22, gave evidence of
the occurrence of “‘craters”’ (Fig. 8) in the varnish film left after the evapora-
tion of the fog globule. As a whole, the photographs on the X-ray nuclei
are the most successful, and in No. 23 in particular the particles stand out on
the photograph. In No. 31 particles were deposited on the plate during the
vortical motion occurring on influx of air while in No. 32 the motion of the
particles (film too liquid) shows streaks on the plate.

No. 2) ae X-1
nuclei. Se
part, .0030 ¢
Fall 15 sec. D

in negative,

No. 30.—X-ray
nuclei. Scale
part, .0o30 cm

Fall 45 sec

). 23. — XN

No. ro.—Air nu- nuclei. Seal
clei. Scale part, part, .0030 ¢
oor2 cm Fall Fall 30 sec,

No. 9.—P nuc
Scale part, .o
cm. Fall 4

sec.

No. 22.—Scale
part, .0o30 cm
“Craters.’”’ Fall
30 sec. X-ray
nuclei.

No. 4.—P nuclei.
Scale part, .0044
cm. Fall 20sec.

FIGURE I0.—MICRO-PHOTOGRAPHS OF FoG PARTICLES, PRECIPITATED ON PERSISTENT NUCLEI USUALLY

PRODUCED BY THE X-RAYS IN DuST-FREE AIR
116

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. Ly

TABLE 4.—FIRST SERIES OF NEGATIVES MEASURED FORd AND n. m= 4.7 X 10~.

p/w = 2.11 n'/tcm*310°. t=20%to30°. Op=17 cm.

| |
& | ‘6 a
Bie aac fae
Oe TEM eer ral oer: Se
Nuclei.| ¢ |No.] Corona. s x S 8 - psi 5 e s Remarks. Film
3 . on ee m
Sales =
& e a
Le
|
SEC: em. |cm.| cm.
iP 2 |Lycopod. 312 -
Pp 3 |weg’ 8.9] 55 | 4-7 -- aaa Rough. Tallow.
ie 20°] 4 |wcg’ 8.6] 76 | 5-10] 250 —7000| 160] .6 | Persistent. Damar.
Py = \20 5 | olive” -|>12 | 57 | 4-9 5h 36-220) |sanollic. — Pine: “
Air 30f| 6 |weg’ 8.4] 69 | 3-10] 33 —270 | 151|5. | Persistent. o
Air 30 | 7 |corona 5-6|109 |ro-13| 18 —1500| 56]/3. |Good but in-
(stale) jured.
iP — | 8) |\fine fog —"|| sol! Evaporated.
Pe a5 | o|iwap 7.3| 9O |-5—14| 75 —1400] 130/1.7 | Clear.
30?]| ro |corona 5.6] 84 | 7—-10|?250 |175—-2600] 56] .2 | Clear.
een | LIN liane 7:6, — | — | — — — | — | Failed.
Craters.
ener sles wart >12 | 54 |-4-7 ?4 — | — | Vague and
faint.
Air “ |13 |corona AO 77 -\5—- LO 8 I-125 45 |6.
(stale)
ve Aen ee ig sa 4.6| 96 | 7-11 ” —170 26. . |\Clear.
nem ie \|(eT'S * 4.6] 97 |.g-10] 13] 12=330 2/3. |Clear but not
all focussed.
iP EO liyaoren g.2| 62 | 4-10]? 16 7-230 | 180/11. | Clear but dark.
ie SE TE wicre? 8.3] 82 | 5-12} 61] 3-1300] 150]/2.5 | Good.
iB Seale On Oza 6.5| 62°] 3-10 6 —340 | 105 | — | Evaporated.
iF “119 |we g’ 8.3] 55 | 3-21] 34] 2-970 | 150|4.4 |Clear but un-
certain.

The results with fog particles condensed on nuclei of atmospheric air (table
6) and photographed by a small kodak were not very successful, due to second-
ary causes. In the small vessel used, the tendency to evaporation was accent-
uated, and the fog particles had in many instances evaporated before the
photograph could be taken. Hence d was not measured.

In table 7 the chief purpose was a comparison of the precipitates obtained
when using an ordinary stopcock to effect the exhaustion and on using the in-
stantaneous valve described above.

Inferences.

18. Precipitation per cubic centimeter—The precipitation, m, computed
from the plates Nos. 4, 5, and 6, for instance, would be 57 X 107°, 4.8 X 107°,
5.7 X 10 ° grams per cubic centimeter, respectively. In the first case the deposit
is excessive, probably due to eddy currents, in the second nearly correct, in the
third too large. It is preferable, however, to compare the values of d and n
obtained from the photographs directly with the data from coronas.

118 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION

TABLE 5.—NUCLEI DUE TO X-RAYS OR TO FILTERED P EMANATION; DATA
FOR d AND n. m=4.7 X10; [Yn = 2.11 n'/tcm’? 10°; t= 15—-45° ;
op =17cm.; FILM, SOFT DAMAR VARNISH.

Sala | 2 |eso| gs
Nuclei. | Fall. | No. | Corona. Ss x Bx x = E x = Remarks,
3 de | = a
Ss.
X-ray | 30 | 20 | wp’ 6.9| tor | 6-11 | — Fogged plate.
a 30 | 2t | wr’ 5.7| oof} — | — Do.
Sa Z0. | 22) | wep: 5-7| 133 | 7-16 4.3| 60] 14 | Pitted. Shows crater.
ve 30 | 23 | corona 2.8} 129 |10-20 1.3 8) 6 | Good.
i 30 | 24 | wp’ 5.3! — — — | — | — |Lost.
30 | 25 | wp’ 523i) L250 | Sa 2h e244 |ieenot || e2e2 (Goode
30 | 26 |wr’ 3.6] 86 | 5—ro 3-1] 18] 5.8} Too late.
z rh | 27 | wr Z| 3 Onl 72 4-1] 11] 2.7|/Good. Shows dew.
3 15 | 28 | wr’ 4.4] 105 | 8-15 | 19.7} 36] 1.8
BO) W520) | Wake 2.8] 210 |16—-24 | ?.17 8|?47 | Evaporated?
is 45 | 30 | wr’ 4.6] 139 | g-20 1.9] 42 |22 Good.
P Sait | syestes >12 | 65 | 5s—to-|-—— | — On influx.
iB 30 | 32 | wy’ 10.4) — — — | — Streaks.
P (stale) | 30 | 33 | wp 5-5) S510 5% il 27-9155] ae
iE 30 | 34 | wbr 4.5| 86] 7-12 4.9] 40] 8 | Good.
P 30. || 35) | wit’ >12 | 65 | 4-11 |?13.8] 250] 18 !Good (mixed small and
large)
iB 20 | 36 | corona 2:6] 140 | 6-20} 5.8! 4 | 1.2|)\Goods
ie 30 | 37 | wp 5-3| 95 | 8-11 |?13.8] 50] 3.6] Mixed small and large.
P (stale) | 30 | 38 | wr’ >12] 39) 2-6 |? 3.1] 250/ — | Out of focus.
iP 30 | 39 | we'’g 7.8} 87] 8-10 |? .10} 140] — | Out of focus.

TABLE 6.—AIR NUCLEI; DATA FOR n, m=4.7 X10%g/em?; {/m=2.11n'/tem™? 10°;
t= 307-157; 0P=17 cm.; c=5 X 26 X (.00304)?.

| From From

No. Corona. Ss eb photograph] coronas Ratio. Remarks.
n X 1073. | n X 1073,

40 y’ og’ 6.8 I05 30 10.8 II7 10

41 w’ og’ 6.8 87 30 6.7 117 07,

42 wog 6.8 Failed — 117 — Accident.

43 w br cor 4.8 1O2)'936 7.4 45 6.1 Good.

44 w br cor 4.8 EGnmeis 9.0 45 5.0

45 w’|b|r’ 4.6 5 aneLs, 12.3 42 3-4 Very good.

46 wrg rai} 67 15 6.1 58 9-5 Vague.

47 wog 6.8 Failed = = =

48 wrg 5.8 Failed — a =

49 gbp 6.4 7S aeerTs Be 100 Bn3 Good.

50 gbp 6.4 TOG) ors 85.8 100 12 Good.

19. Diameters and numbers.—The d-values given by the photographs are
almost without exception larger than the corresponding diameters computed

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. I1g

TABLE 7-—MISCELLANEOUS EXPERIMENTS. ROOM AIR FROM FLOOR. LARGE
VESSEL. ,/n=6.237. EXHAUST THROUGH STOPCOCK.

Na. (Garon 5 on i By photograph! From coronas | cot

WW XA10ms-) ||| n X 1073. obs

53 wp cor Ts 144 30> || 26 127 4.9

54 wo g’ 9.8 168 20), | 42 200 4.7

55 wrog 8.8 147 30 28 170 6.1
56 wo g’ | 9.8 73 30 2.4 200 | ?

EXHAUST THROUGH INSTANTANEOUS VALVE.

Eva wog 9.8 138 30° 24 200 8.4
55 oo 4.6 Q2 30 7.2 2 5.8
59 pace. | 8.3 130 20)) |] f1g ESI i
Eee la S| on ae
g | ie 170 30 45 140 Ber

62 weg’ 7.9 30 18 145 8.1
| | Mean, 6.3

from coronas. No doubt this is in part due to adhesion; but the chief reason
is the occurrence of so many large particles with the small, so that the average
value of d given by the photographs is within certain limits an arbitrary quan-
tity, depending on the distribution of particles selected for measurement. If
the values of d be plotted graphically in terms of s, they do not make a smooth
curve, and these values are from 1.5 to 2 times larger than the coronal values
of d. For this reason the measurement of d in the photographs was abandoned.
The data are much inferior to the above micrometer measurements of floating
globules.

The n-values in the photographs, being obtained independently and not
conditioned by so perfect a reproduction of form, etc., as are imperative in
measuring diameter, may be considered more trustworthy. Nevertheless, they
also fail to suggest a smooth curve, though they are, as a whole, very much
smaller than the coronal n-values. If observations obviously in error be
omitted, the following are the mean values of the ratio m (coronal) to n
(photographic) obtained from the successive tables. Only in two or three in-
stances are these ratios less than one.

Table 4. Ratio=4.1,

“cc Re “c = 3.8,
“c 6. “cc =4.4,
“i 7. “cc = 0),

or on the average the coronal n-values are about 4} times as large as the n-
values obtained from the photographs.

120 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

20. Explanation of discrepancies.—The reason for this result would at
first sight be obviously given by the evaporation of the fog particles on the
plate, before and during the photographic exposure. But as the d-values were
from 1.5 to 2 times too large, it is necessary to guard against accepting this
explanation too hastily: for inasmuch as the d-values and the n-values are in-
dependently given by the photograph, their relation must be

nd} = 6m/z.

Now it is rather curious that while the m-values are 4.5 times too small, the in-
dependent d*-values should be 3 to 6 times too large. In other words, on
using the average d- and the average u-values given by the photographs, an
approximately correct value for m, the precipitation per cub. centim., follows,
even though d and u are found quite independently. Add to this the fact that
if evaporation occurs the photograph usually fails entirely. It is quite im-
probable that on the average a definite number, about 78 per cent., of the fog
particles should incidentally evaporate.

Before proceeding further it will contribute to clearness if some of the
better data of table 4 be summarized. The d-values show the marked occurrence
of larger particles on the photograph, whereas the smallest particles more nearly
correspond in size.

TABLE 8--SUMMARY.OF CERTAIN DATA FROM TABLE 4.

Diameter | Diameter | Number Nam bear
7 - | Timeot 5 : from from from umibemtom
No. Nuclei. 2 Ss Corona. B é coronas
subsidence. photograph] coronas | photograph moss
d X 104. d X 104. | m X 10-3. BOTS
sec. cm. cm. cm.
4 iB 20 8.6 we g’ 5-10 3.8 250 160
5 ee 20 > 12 olive 4-9 B52 Gu 2.50
6 Air 30 8.4 weg’ 3-10 3-8 33 160
Ose 45 7-3 wp 5-14 4.1 75 130
Io P 30 5.6 cor 7-10 Ga2 ?250 60
23 X-ray 30 2.8 cor 5-20 10.6 13 8
25 7 30 5e8 Ww 1’ S20 555 24 50
27 a 15 Bal i De Pat 4 II
30 as 30 4.6 a g—20 6.2 2 42

If the micrometer data ($$4-6) be used for the d-values, the ratios of
coronal and measured diameters are

sS= 4 6 8 10 12
Ratio = Tage Te T.2 T.2 T.2
1.6 1.6 aoe cal Te

or the d3-values are 1.7 to 4.1 times too large. Now in these instances the fog
particles were often measured while floating, so that adhesion in these instances
must have been a negligible factor. Nevertheless the d-values are not all in-
compatible with the n-values.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 2m

21. Summary.—The curious state of the case mentioned is not reassuring,
Briefly, the photographic m-values and the photographic values for diameter,
d, do not make an incompatible system, though both differ materially from the
coronal values, the former (d) being larger, the (7) smaller. If one ascribes the
large visible diameters to adhesion, and the small numbers counted to evaporation,
the compatibility of the two sets of independent data (d and 7) is not explained.

I have therefore concluded that the results in question suggest that the
smaller particles evaporate rapidly into the larger. Hence while the mass of
water precipitated per cubic centim. remains constant, the diameter of the fog
particles soon increases while their number decreases (by evanescence of the
smaller) in such a way that 1 X d* remains very nearly constant throughout.
In place of evaporation, capillary coalescence of the initial very minute droplets
is an even greater probability; and such-coalescence I have often seen under the
microscope on observing dew droplets! very fine and close together. It is rather
interesting that the evidence in favor of this complicated behavior is so strong.

That such evaporation occurs appreciably in the coronas is not probable, be-
cause ina good apparatus they retain their character during the whole of the subsid-
ence. The plate of glass being slightly warmer contributes to the effect observed.

Hence neither the micrometric nor the photometric method can be relied
upon for undistorted results. Whenever graded particles are present the evi-
dence must be sought in the washed and blurred coronas. If the coronas are
clear and multi-annular, the gradation observed under the microscope must be
a secondary effect not present in the coronas themselves.

III. RESULTS FROM SUBSIDENCE.

22. Object and method.—The difficulties mentioned in the last sections in-
duced me to look for corroborative data in the evidence obtainable from special
experiments with subsiding fog particles. These are easily made in the
apparatus for measuring atmospheric nucleation as described in Chapter
VIII, Fig. 1. The observer, provided with a stop-watch, simply determines
the time during which the straight horizontal line of fog descends a given dis-
tance, say 5 centimeters, from the top. In other cases there were two marks
between which subsidence was noted.

23. Kesults—In the experiments given in the following tables, the watch
was started simultaneously with the exhaustion and stopped when the given
fall of fog line had been reached. The equations are then successively

to‘d = 18/9, ds = @, n = 9/(d? X 10°) = 1550/v*/?,

where 1 is the number of particles per cubic centim., d the diameter of each,
v the rate of subsidence.

‘ If a microscope plate is dipped into thin, rapidly drying methyl alcohol varnish, a milky
deposit of dew is often seen on the film when solidifying. This film behaves under the micro-
scope as stated in the text.

122 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 9—VALUES OF a ANDn. AIR NUCLEI. 1 = 6m/z2d?=9/10'd?; 10'd = 18/0;

asia.

Corona. s 103 XU 10! Xd a Io-3 Xn Dae

cm. cm./sec. cm. cm./sec.

wolg 5.8 77 5-0 .0029 72 70

wrg 5-4 84 5-4 34 56 80

gy o 6.5 72 ng 34 62 60

wrog 6.2 89 5-5 34 53 65

wog 6.2 89 5°55 34 53 65

wrg 5-9 85 5-4 32 56 70

g’|b|p 4-5 180 7.6 34 20 123

Do. 4.5 116 6.1 28 39 123
Do. 4.5 130 6.5 29 33 125 a

w|b|p 4.6 122 6.3 29 36 115

wrg 5-9 107 3-9 35 44 7°

wrg 5-7 102 a7 33 48 75

gbp 4.6 128 6.4 30 34 115

y’ bg 6.2 77 5-3 33 59 65

wrg 6.2 79 5-4 33 59 65

g|b|p 4.6 156 7-1 33 25 115

Wgi|r 4.5 192 7.9 35 18 123

wrg 4.5 122 6.3 28 36 123

wre 5.6 94 5-5 31 53 75

wrg 525 104 5.8 32 40 80

vy Og 5.8 107 5-9 34 44 7°

Wree 5-7 125 6.4 36 35 75

weg 5-4 86 teal 30 55 80

yo b’ 6.2 76 ces 33 60 65

g|b|p 4.8 140 6.7 32 29 105

gbp 6.0 72 5-3 Re 60 65

Mean a in decades .00322 |
304 |
327 - @=.00320
|
329
TABLE 1o.—FURTHER RESULTS FOR SUBSIDENCE. n=6m/z2d?=9/10°d?; 10d =
13/0; ds — id:

Corona. to? Xv roi X d Ss a eee
y br bg 4.2 79 Sar 6.5 .0033 58
wog 4.1 74 4.9 6.3 31 62
w br cor Buk 135 6.6 4.8 32 105
w|b|p 2.9 114 6.1 4-5 27 122
w y cor 3.0 125 6.4 4.6 29 117
w|b|r 3.2 106 5-9 4.9 29 100
Ww y cor 2.3 167 7.4 ans 20 215
weg 2.7 143 6.8 4.1 28 150
w br cor 2) 200 8.1 3-4 28 230
wog 2.8 125 6.4 4.3 27 135
w rg 2.7 140 6.7 4.2 28 145

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 123

Corona. $19.5 103 Xv 104 Xd Sa fi > ee
cor 2.2 157 7.1 3.4 24 25000 235
wp cor 3-3 94 Beg pan 28 54000 go
w|b|p 3-0 102 Gi 4.6 26 49000 117
cor 2.3 2.30 8.6 3.6 31 14000 210
cor 2.9 ?109 5-9 4.5 27 44000 123
wbp 2.8 III 6.0 4.3 26 42000 135
wog 2.8 135 6.6 4.3 28 24000 135
w|b|r 3.0 98 5.6 4.6 26 51000 117
gbp 3.0 102 Ca 4.6 20 49000 Tey,
wre 3.8 81 Bed 5-9 30 68000 68
gbp 3.0 105 5.8 4.6 27 45000 rae)
wrg 2.9 II5 6.1 4.5 27 39000 1233
Wp cor 3-4 QI 5-4 feo 28 57000 go
wog 3:0 122 6.3 4.6 2 36000 117
w rg 2.8 140 6.7 4.3 29 30000 135
wog 2.8 10g 5-9 4.4 20 44000 130
gbp 2.9 122 6.3 4.5 28 36000 123
g|b|p 3.1 105 ceo 4.8 28 46000 105
wp cor Bia 100 er 4.8 28 48000 100
gbp 4.1 68 4-7 6.3 30 86000 62
gbp 2.9 100 Bea 4.5 26 48000 123
w br cor 3.0 100 Se 4.6 26 48000 Tela
w brr 2.9 105 5-8 4.5 26 46000 123
Ww oO cor 2.9 120 6.2 4.5 28 38000 123
wog 2.8 140 6.7 4-3 29 30000 135
wre 2.8 143 6.8 4.3 2 28700 135
wp cor 3.0 era 6.0 4.6 28 41000 117
w rg 2.9 143 6.8 4.5 30 28700 T2333
gbr 3.0 109 5-9 4.7 28 44000 110
w br cor 3.0 128 6.4 4-7 30 34000 Ito
gbp 4.0 60 4.4 6.2 27 106000 64
gpb 4.2 64 4.5 6.5 2 99000 58
gbp 4.0 69 4.7 6.2 29 87000 64
w rg 4.0 85 Gee 6.2 33 60000 64
wyg 4.2 79 Te 6.5 33 68000 58
w y cor 2.9 128 6.4 4.5 29 34000 123
w 0 cor 2.9 152 7.0 4.5 31 26000 123
Wrog 4.0 76 5.0 6.2 20 72000 64
WwW pcor Bre 98 5.6 4.9 27 51000 100
w br cor 3.0 120 6.2 4.6 29 38000 117
cor 2.9 120 6.2 4.5 28 38000 120
Ww p cor 3-3 gt 5-4 5.2 28 57000 88
wbp 3.0 96 5.6 4.6 26 51000 117
Wels; 2.9 132 6.5 4.5 2 32000 123
Wy ey Bet 140 6.7 4.2 28 30000 135
Ww p cor eS 98 5.6 5.1 2 51000 go
vaetin 2.9 T3'2 6.5 4.5 29 32000 123
cor 3.0 116 6.1 4.6 28 40000 reanG)
Wp cor 3.6 gr 5-4 5-5 30 57000 80
ivi Tan 4.0 78 5.0 6.2 31 72000 65
gbp 3.0 105 5.8 4.6 27 46000 117
wg cor 1.8 215 8.3 2.8 24 15800 330
wr cor 2.0 200 8.0 Be 25 17300 280
wr cor 2.0 200 8.0 aii 25 17300 280
cor T.9 210 8.2 2.9 24 16400 320
cor 2.4 156 sil Bal 20 25000 195
cor 26 205 8.2 3n2 26 16700 205

|

124 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

24. Remarks on the tables —Constructed graphically, the results of table 9
are irregularly below the coronal results. The results of table ro are somewhat
smoother and in the main agree with the coronal data. Thus for the green
coronas (see Chapter VI, § 29, table 17)

coronal datum, g b p, #=96,000,
from table 10, g b p, #7= 106,000, 99,000, 87,000, 86,000; mean, 7 =95,000.

or the mean values are about the same. In case of low nucleations and small
coronas the subsidence results are much too high as compared with the coronal
results, implying too small a value of the v of subsidence.

The only explanation which suggests itself for this unexpected behavior is
the occurrence of acceleration in the motion of the fog particles. Accordingly
the following experiments with very small coronas (large particles) were tried

as a test:
Normal corona with s = 2.9.

Fall, if = 2 cm. time = 15 15 13 16 16
4 25 21 24 25 28
6 32 27 30 30 —_—
Thus VF = 1.0 Mean t = 15 Ratio = 15
I.4 25 17
sey 30 18

indicating that the uniformly varied motion is only very gradually retarded by
the air resistance encountered. The following table contains similar data for
larger coronas, in none of which the evidence of acceleration is absent.

TABLE 11—MEAN RATES FOR SUCCESSIVE DISTANCES OF 2 cm. EACH.

Time for successive falls of 2 cm. each.

Corona. Sso t sec. t sec. t sec.
w rg 527. 18 23 22
w|b|p 4.6 20 18 16
wrog 4.6 20 15 12
wrg 4.4 17 12 10
wp cor 3-4 13 8 5
w br cor 4.6 — h = —
wbp 4.7 19 14 —
wre 4-7 20 14 12
wreg 4-5 15 14 II
wbp 4.6 23 15 17
cor 1.9 13 8 6

These results rob the subsidence method of much of its trustworthiness
except when the particles are very small, when other difficulties step in: for.
large coronas are not persistent in character.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 125

The case may be stated by computation as follows: The velocity at which
force is annulled by the resistance is

—1¢LO-0 RS) —-3 a> LO°

from which the following table has been computed.

TABLE 12.
Si) 1.5 em. d = .oo180 cm. Ui = 1100 (Cie SEC:
2 143 -63
3 098 a2)
4 072 -161
5 055 093
6 047 .068
7

042 .053

If the results of this table are laid off graphically in a chart, the column
marked ‘‘constant ro} X v’’ in tables 9 and 10 above may be obtained from it.
It will then be seen that in most of the instances of small coronas in table ro, the
observed velocity is less than the limiting velocity. Hence these data are to be
rejected, and it follows that the subsidence method is scarcely applicable until
the middle g-b-p corona (s = 6.2—-6.5) has been reached. The mean value of a
computed from the admissible data of table ro is in successive decades, .00284,
.00290, .00293, .00290, giving a mean value of a=.00289. This agrees closely
with the datum (.0029) accepted in the above coronal tables, but is much below
the value in table 9. Hence the following experiments were made with large
coronas in the cubical apparatus.

25. Further results.—The data of the following table 13 were obtained by
observing the time of fall for three successive distances, of 2 cm. each (usually),
with three stop-watches. The results taken were those cases only where the
three intervals observed are nearly the same. The fog was usually precipitated
in the long vessel for observing atmospheric nucleation on air nuclei. If the fog
line became billowy before the last observation was reached, this was discarded.
The mean results are all given at a temperature of about 20° centigrade.

The new results like the above fail to suggest a smooth curve; though it
would be difficult to obtain them under more generally trustworthy conditions
than in this large vessel, showing a fog line half a meter long. The slightest
variation of temperature, etc., gives rise to an undulatory fog line or produces a
washed upper surface to the fog-bank. After a lapse of time of about 1 min.
the demarcation is rarely available. In the very large coronas above g-b-p, the
color and the character of the coronas is fleeting. In fact, a total change is hable
to occur within a fall of one centimeter. Such results have no meaning.

If these data be compared graphically with the results of tables 1, 8, and 13,
Chapter VI, they will be found to agree with them as well as these results

126 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.
agree with each other. In fact, the subsidence data fall very nearly on the
curve corresponding to the fourth part of table 13.

TABLE 13—DIAMETER AND NUMBER OF FOG PARTICLES FROM SUBSIDENCE
MEASUREMENTS. d X 10* = 18,/v; n=9/@ X 10°; 6p = 17 cm.

Corona. s Mean v X 103.| Meand X 1o*.| Meana X 1o*.| Mean ” X 1078.
cm. cm/sec. cm. per cm*.

wrg 5.8 86 5.3 31 62
wrg 5-9 82 5.2 30 66
wog 6.2 74 4.9 30 76
wo bg 6.6 68 4:4 31 86
wo bg 6.8 69 4.7 32 85
wgebp 12. 34 Ba — 250
gbp 6.3 62 4.5 28 99
gbp 6.3 62 4-5 28 100
yg br bg 6.6 66 4.6 31 gi
wrg 5-9 80 ai 30 68
wrg Sey 81 Go 29 66
y o bg 6.2 72 4.8 30 80

Do. 6.2 fu 4.8 30 82
weg 5.8 83 5.2 30 5 64
W p cor 52 97 5-6 20 5!
wo bg 9.2 37 355 32 220
y obg 11.6 31 3.2 37 280
wo bg 9.4 34 B33 31 240
wo bg 9-4 34 Bias ai 240
wo bg 9.9 36 3-4 34 225
weg’ 8.0 43 3.7 30 170
w yojbg 6.2 74 4.9 30 76

The nucleation for the g b p corona (m = 100,000) agrees closely with the
mean results above (1 = 96,000), table 17.

Beyond the g b p corona (7 = 100,000), however, the new data for m are with
few exceptions much larger than the coronal values, at least with the interval
s=6to1ocm. From the difficulty encountered in observing this very slow
subsidence and the variable coronas, the discrepancies are much more liable to
rest with the subsidence data than with the other (coronal) group.

IV. SUMMARY OF RESULTS FOR NUCLEATION.

26. Preparation of a table for deducing the nucleation from the observed
coronal aperture.—In conclusion, it will be useful to collect all the data obtained
for the nucleation on a single sheet. Accordingly a chart was drawn up (too
large for convenient reproduction here) containing the results of Chapter VI,
table 1, series 1 and 2, table 8, series 1 and 2, and table 13, series 1, 2, and 4,
as well as the best subsidence data of the last section. The latter are to be dis-
tinguished by a special symbol. The curves are found to lie closely together
until the lower g b p corona (s = 4.3-4.5) is reached, where the first periodicity
occurs. They then pass with wider divergence through the middle g b p corona

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 127

(s = 6-6.5), finally with still wider divergence to the upper g b p corona (s = 12).
The nucleations corresponding to these double inflections are about as 1, 2, 3
(1 = 4,000 to 5,000, 10,000, 300,000). The mean curve as far as the middle g bp
corona 1s to be separately constructed with the ordinates (nucleations n, while
the abscissas are apertures, s) enlarged ro times. The lower periodicity must be
ignored from the crowded condition of coronas here; the upper periodicity
preferably presented as a discontinuity of the curve.

The graph so obtained was used for all the reductions of the nucleation of
Chapter IX and elsewhere. From it table 14 was constructed. The nucleation
n may finally be reduced to air at normal pressure and temperature.

TABLE 14.—FOR THE REDUCTION OF NUCLEATIONS (# nuclei per cub. cm. of the
partially exhausted fog chamber). FROM APERTURES s AND S. PRESSURE

DIFFERENCE, 17 cm. of Hg.

s, chord in cm., when the goniometer arms are 19.5 cm. long.

‘ cc

Sy a * 30
iss Sy WeryGLOns. Corona | Ss oS nm X ro-3. Corona.
|

em. em | cm. cm

1.0 italy Teh normal 2.8 4.3 28

re 7 Te | 2.9 4.5 Bi

a2 1.8 2.0 3.0 4.6 35 wr

Tin 2.0 2.5 Sail 4.8 38

1.4 ane 3.0 Bu 4.9 2

1.5 2.3 4-0 3-3 5-1 45 g’ Bp

1.6 2.5 5-0 3-4 5-2 49

ey, 2.6 6.0 3.5 5-4 re w P cor

1.8 2.8 7.0 3.6 5e5 57

1.9 2.9 8.5 a7 oy 61

2.0 Bait 10.0 | 3.8 5.8 66 we

ai 3-2 Ta 3-9 6.0 70

2.2 3.4 a5) 4.0 6.2 75 wr

23 Bas 15.5 4.1 6.3 80

2.4 Bei Tei 4.2 6.5 85

Bas 3.8 20 we 4.4 6.8 95 ygbog

% : 22 |

ae os A | 4b 6.5 100 gBp
uF zis | 4.5 6.9 120 | w P cor

Note.—The photographs of §$15-21 were made from fog particles condensed (as stated) on persistent
nuclei. No attempt was made to photograph the fog particles condensed on ions or fleeting nuclei, because
the high exhaustions needed would have endangered the large condensation chamber. I shall therefore
return to the subject elsewhere with a modification of apparatus and with this specific point in view.

CHAPTER VIII.
THE CORONAL METHOD OF ESTIMATING ATMOSPHERIC NUCLEATION.

1. Introductory—To produce coronas the nuclei must be very closely of
the same size, for in a large trough a rigorous uniformity of diameter of fog
particle and possibly of distribution is implied, if the corona is to be sharp and
brilliant. Particles of even slightly different sizes would give a blurred effect or a
mere fog. Therefore, as I understand it, the effect of ordinary dust to some
degree vanishes from the corona, and the nucleation observed is probably
something more definite. It-is for this reason that in spite of very discouraging
drawbacks my interest in the subject has not waned, though I am well aware
that the effect of chemical products of combustion in winter, such as sulphuric
acid, or ionized matter in general, has not been eliminated. One may note, for
instance, that the distribution of atmospheric electrical potential is a maximum
in winter and falls off in its yearly period in a way similar to the observed nuclea-
tion; that there is frequent occurrence of day minima in both cases; that
maximum nucleation occurs, as shown in Chapter [X, during the winter months,
when one must certainly anticipate the maximum of dust contents during the
summer.

The subject of atmospheric nucleation, as a whole, has received enhanced
interest in view of its bearing on Arrhenius’s theory of the geophysical import-
ance of cosmical, and in particular of solar dust. Some limitation has been put
on the light-pressure theory of Schwarzschild, but this has rather stimulated
Arrhenius to give a sharper expression of his views, and the theory now appears
as the central feature of an admirable discussion of cosmical physics."

I hoped therefore by aid of the present method to eventually add a contribu-
tion of my own.

2. Apparatus.—This consists of a réctangular box, AA, 50 cm. long, 15
em. in thickness, and 10 em., or preferably less, in height, made of some ma-
terial impervious to water. Wood covered while warm with a thick coating of
wax and burgundy pitch answers the purpose very well, and is much lighter
than rigid metallic vessels. The front and rear faces of the box are of thick
plate glass. This must be kept clean on the inside, and suitable scrapers with
a vertical straight edge of soft rubber movable to and fro along the glass by

aid of a long horziontal rod should be provided within the box. The rods pass"

» Lehrbuch der Kosmischen Physik, Vols. 1. and ut. Leipzig, Hirzel, 1903.

128

es

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 129

out through perforated corks in the tubes e, one on each side, with additional
protection to secure an air-tight joint when not in use.

FIGURE 1.—CONDENSATION CHAMBER. SECTIONAL ELEVATION.

FIGURE 1¢.—THE SAME IN FULL

The air within the box communicates with the outside by three or more
stopcocks, of which B is very wide (more than } inch in bore) in order that
sudden exhaustion may be made through it. The stopcock C communicates
with the atmosphere at a place free from local nucleation, through a length of
#inch lead pipe; C furthermore communicates with the interior of the box
through a flat coil of the same lead pipe , p, lying in the bottom of the trough below
the water-level, ww. A coil of lead pipe in a water bath may also be inserted

130 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

on the outside of the box, the object being to heat the air to room temperature,
especially in winter.! Two thermometers, T, T’, with their bulbs respectively
in the air and the water within the trough, register the temperature. The end
of the influx pipe rises to a height qg, near the axis of the trough and opposite to
the outlet, C. Finally, the whole inside of the trough is lined with a double
layer of cotton cloth, 1, supported ona framework of stout copper wire. The
trough should be mounted with its longitudinal axis on trunnions in order that
the whole interior may be moistened in a single rotation, as shown in figure
1 a, in detail. The stopcock F, provided with a cotton filter, is often useful in
testing.

The horizontal diameter of the coronas is observed, the point source of
light being about 2 meters off on one side and a suitable goniometer about one
meter on the other side of the trough. The distances used were 85 cm. and
250cm. With the eye at about 1 meter from the fog chamber the apertures
of coronas are relatively independent of this distance and at the same time large
enough for satisfactory measurement.

An ordinary jet pump suffices for aspiration (with the cocks C and B open);
and with an added vacuum chamber provided with a vacuum or mercury gauge,
for sudden exhaustion (C and 6 having been closed), care being taken that the
connecting tubing beyond 6 is wide. These details are shown above.

3. Diffusion from two opposed surfaces.—The high values of nucleation
observed during the winter months will not be received without misgiving, since
the air during the very cold weather is nearly dry, and after being heated to
20° very far from saturation. Deficient saturation, however, would decrease
the size of the fog particles and, cet. par., increase the size of the coronas, in
this way showing the same result as excessive nucleation. Hence it is necessary
to estimate the time which is needed to saturate dry air in the apparatus de-
scribed in § 2.

Given a rectangular trough at the bottom of which is a surface of water
and at the top of which a surface of saturated cloth, the diffusion problem is
equivalent to the case of an indefinite air plate into which the vapor enters on the
two exposed sides. Thus if p be the vapor pressure relative to saturation at a
distance x above the surface of the water or a—x below the wet cloth, ‘at the

time f,
ax —(x/a)2kt 37% .—(3%/a)*kt

= Ava Ties
p= i—— (sine He uit +...)

where k is the coefficient of diffusion.
If diffusion takes place into a partially saturated atmosphere at an initial
pressure p,, the factor 4/z in the last equation is to be replaced by (4/7)

(1 aa Po) q
‘In case of a moderately long (20-30 feet), thin (} inch) influx pipe, experience showed

that both the internal coil, p, and the external coil are superfluous. They were there-
fore discarded.

BIR.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. I31

The question is most conveniently stated for the middle plane, x = a/2,
inasmuch as the saturation here is least. Since at 20°, k = .23 cm’/sec., about,
the results, if p, = 0 initially, are shown in table 1. If the initial saturation is
po = 1/3 or 2/3, the data, as the table shows, imply successively greater satura-
tion throughout than in the preceding case. As the relative saturation left
after any exhaustion would not probably be less than 2/3 even in the absence
of convection, the inferior limit of saturation shown by these data is excessive.

v= 5, R,=2/

- 10, Jos98 |
| | -40, pz'/3 =
ice

. | |
|
2 i
pP
| ZA |
0 i L
0 2 3 4 5 6 77, 8 9

FIGURES 2, 3, 4.—GRAPHS SHOWING THE PROGRESS OF DIFFUSION IN THE LAPSE OF TIME.

The upper graphs, figure 2, contain the relative saturation pressures as
ordinates at the times given by the abscissas in minutes, according as the initial

pressure is p, = 0, p, = 1/3, Or p, = 2/3.

TABLE 1.—DIFFUSION OF WATER VAPOR BETWEEN OPPOSED SURFACES;

@=11cm., * = 5.5 cm., k = .23.
Po = 0 Po = 1/3 Po = 2/3
t P t P
30 Sec. ST 3° sec. GO 30 SEC. -762
60 587 .722 60 .864
120 .866 120 -Q1O 120 -956
180 957 180 -O71 | 180 -986

Thus in 3 or 4 minutes the air plate in the trough may be considered satu-
rated under the most unfavorable conditions. In the aspiration of fresh air
through the trough the maximum rate was about 4.5 lit. /min., the usual rate

132 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

1.5 lit/min. The capacity of the trough, including parts above the cloth lining,
was about 11.3 liters. Hence 3 to 5 minutes suffice to renew the air, and each
particle remains in the trough from 1.5 to 4 minutes. The time needed for
adjustments after the influx pipe was shut off and prior to sudden exhaustion
was about 1.25 minutes. Therefore the total time for saturation was 2.7 to 5
minutes, which should fully suffice even in case of diffusion alone.

The conditions, however, are much more favorable as the influx and efflux
currents evoke considerable convection. The lightness of water vapor is itself
favorable to the same end. It is observed, for instance, on exhausting imme-
diately after the introduction of phosphorus nuclei, that filamentary condensa-
tion is in evidence, denoting currents upward axially and downward near the
walls of the receiver.1 These fog strands are the inevitable convection currents
due to the relatively low density of the vapor.

4. Miscellaneous tests —If there had been under-saturation the coronas
found on condensation should have been larger in non-saturated and smaller in
more saturated parts, which was not observed. Sudden exhaustion immediately
after shutting off the influx showed a somewhat enlarged uniform corona, but
even enlargement was not invariable.

Whether the fast or the slow influx specified was adopted proved to be
without marked effect for reasonable differences of time, 7. e., such as would not
imply time losses of nuclei.

The effect of a long influx pipe (10 meters of 41-inch lead pipe) and of a
short pipe 1 meter long could not be sharply differentiated, owing in a measure
to the cotemporaneous variation of atmospheric nucleation. So the presence
compared with absence of the coil in the water bath (24 turns each about 3.5
cm. in diameter) showed a negligible difference.

Experiments were made with regard to the usefulness of this coil for keeping
the influx air at room temperature both by filling the bath with abnormally hot
water (40° C.) and with broken ice. The effect on the apertures of the coronas
was in both cases of little importance. Hence, except on very cold days, the
water bath and coil may be withdrawn altogether. The atmospheric air after
traversing the 1o meters of influx pipe is already sufficiently heated to be intro-
duced into the condensation chamber directly.

Tests were also made with a U-tube loosely filled with wet sponges and
with a drying tube one meter long containing phosphorus pentoxide. In
neither case was a definite effect on the coronas ascertained.

Freedom from leakage was finally tested by filtering the air. The coronas
on sudden exhaustion showed a gradual decrease to complete evanescence.

5. Diffusion from a single surface-—The case is naturally less favorable if
the upper wet surface (double cotton cloth) is omitted. The computation may
be made from an expansion of Kramp’s integral, so that

2 L aF x
Dal (Gen “3X 1(4kt)32 a 5 X 2! (4kf)/? °° -)

! Smithsonian Contributions, No. 1373, 1902.

a

il ial

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 133

where is the vapor pressure relative to saturation at a distance x above the
surface of the liquid at the time ¢ for the diffusion coefficient k. = 23. The data
computed suffice to locate the curves which are sufficiently given in figure 3 for
x% =5cm. the middle plane, and x = ro cm. the top plane of the trough. When
the initial saturation is 1/3 or 2/3, the coefficient 2/7 is to be modified as stated
above. Figure 3 shows the corresponding results.

If it were not for convection, therefore, such an apparatus would be un-
suitable, for even after waiting 5 minutes, the air at the top, x = 10 cm., for an
initial saturation of 2/3 is but .8 saturated from diffusion alone.

One might therefore expect to obtain distorted coronas campanulate in
outline, small below and large above, whenever condensation is produced
within a few minutes after closing the inlet. Yet such is never the case if less
than a minute is allowed after influx ceases. Granting that 2 to 4 minutes are
needed on the average for a particle to pass through the trough, as stated above,
if exhaustion is made immediately after closing, the top layer is but 2/3 saturated.
Under these conditions there is in fact an unusually large green centered faint
corona with a horizontal band of crimson color running through it. Half a
minute later, however, the figure is quite regular again, showing that the con-
vection of light vapor must be very active. After 2 minutes subsequent to the
closing of the influx pipe, the air may be regarded saturated except in the
coldest weather.

6. Absorption and decay of nuclei—The losses in the influx ptpe are
difficult to determine because of the variation of atmospheric nucleation. The
observer is left in doubt whether a given difference is due to absorption in the
pipe or to causes without. The experiments incidentally made throughout the
long experience of Chapter [X showed no serious discrepancy.

The possibility of loss of nuclei on contact of dry air with the saturated gas
in the condensation chamber is an independent question. It is also to be borne
in mind that nuclei may possibly be produced by the sudden contact within the
chamber. No evidence is forthcoming.

If the nuclei after being introduced into the receiver are solutions, some
estimate of their persistence may be formed from my experiments on solutional
nuclei, by treating the loss as if it occurred at the boundary of the vessel only.
If the nucleation falls off from , to m in the time t, and k is the absorption
coefficient,

: — o-2(1/r-+1/)kt
n/n. =& PHD

where r and | denote the radius and length of the cylinder in which absorption

takes place. In case of comminuted pure water, k = 5 to 10 cm./min., and the
nuclei should quite vanish in a few minutes.

{= 1 min. n/n, = .154
2 .023
3 .003

134 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

If the nuclei are derived from very dilute solutions, like river water, an

average value k = .1 may be taken, whence if
? = Teamin 1 /— 2.90
3 .89
5 83
To 68
5° nS
I0o -02

The reduction within 3 minutes will not exceed 1o per cent., which would
usually lie within a given type of corona. The datum k = .1, moreover, corre-
sponds closely to the values found for phosphorus and other nuclei. Hence, if
the type of corona changes after 1 or 2 minutes’ waiting, it may be considered
certain evidence that the air is not saturated and the diffusion error predomi-
nates. Owing to the difficulty of avoiding either insufficient saturation or
excessive time losses, some of my observations in Chapter IX contain data for
two different aspirating currents, the faster corresponding to about 3 minutes’
sojourn of the nuclei in the receiver, the other to a time longer than 5 minutes.
In this way the effects of under-saturation which are most to be feared are
guarded against, while the faster current gives data falling short of the absolute
nucleation by not more than ro per cent. In the course of time this also was
abandoned as superfluous.

7. Effect of pressure difference——It is next to be considered whether the
pressure difference, 6p, used in the exhaustions is pronounced enough to catch
all the nuclei. This is of particular interest as a safeguard against low numbers
in the nucleations obtained.

The usual value, 6p = 17 cm., corresponds to the following pressure ratios
and adiabatic temperature reductions in air, ((76—p’)/76-p’—6p))="273 = 9’ if
p’ is the vapor pressure of water and 9’ the reduced absolute temperature.

10° Pressure ratio, 1.292 D = oGAeT
20° 1.207 263.4
30° 1.341 268.8

For comparison data were gathered with a larger pressure difference, 6p =
22 cm., for which the values are

10 I.414 245.5
20 1.416 254.1
30 1.432 261.5

Clearly, the coronas for the larger temperatures and temperature differences
must be smaller, cet. par., in view of the greater quantity of moisture precipitated.
The data for m, the quantity of moisture precipitated per cubic centimeter of
saturated air, have been computed by the method of C. T. R. Wilson and J. J.
Thomson and are given in the following table 2. Here #, is the initial tempera-
ture , t, the temperature before and ¢ after condensation.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 135

TABLE 2.—PRECIPITATIONS AT DIFFERENT TEMPERATURES AND PRESSURE
DIFFERENCES. vy = (p-p’)/(p-p’-Op); p = 76.

6p= 8.5 cm. 6p = 17 cm. Op = 22cm.
1 : ——
is is t mx ca y te t mx pe y t, t m X 108% y
ZC °C °C grams °C SG grams | °C °C grams
TO || —3-4 || +3.4| 2.1 1.12 |—18.3/— 4.5] 3.7 tea 127 SiO Ta || 42 39
20° | +5.8| 14.7| 2.6 I.Ir |— 9.6|/+ 8.8] 4.6 z.26 |—18.9/— 4.6] 5.5 1.38
Exes moe O 25 -7)\| 2.8 1.10 |— 4.2] 19.6| 5.7 E240 |S) ||) 70 |! 107 35
|

Since m = nzd3/6, if there are n fog particles per cubic centim. each of the
diameter d, and since sd = D where s is the aperture of the coronas with an
arbitrary goniometer and D the corresponding constant,

s¥m =.s24DWn,

which is constant for a given nucleation. Thus the relation between s’ and s
at 6p = 22 cm. and 17 cm., respectively, may be written

S/S =n) mm) 72 — 20 52-

In figure 4 the line s’/s has been constructed and the observations grouped
with reference to it, showing that the curve reproduces the experiments fairly
well. The exceptional cases are all too low; or, in other words, at the higher
pressure difference, 6p = 22, which requires a longer period of waiting after
influx ceases, relatively fewer nuclei are entrapped. From this one concludes
not only that from the medium, if saturated, all the nuclei are precipitated at
6p = 17, but that at the higher pressure difference the time needed for adjust-
ment is excessive and that the time loss of nuclei in the receiver frequently
becomes appreciable.

On the other hand, if 6p exceeds 22 cm. for the given apparatus, the con-
ditions of spontaneous condensation of dust-free moist air are initiated and
continue thereafter with increasing intensity for higher pressure differences.

8. Precipitation per cubic centimeter.—To determine m, I have heretofore
proceeded as follows: In a mixture of « grams of vapor, y grams of air, and
1—(x + y) grams of water, the absorption of heat due to a rise of temperature
ds at constant volume was taken as C(r1—(«+y) )+hx+rdx/ds + cy, per
degree, where C, c, and h are the specific heats of water, air at constant volume,
and saturated vapor, respectively, and r the latent heat. Since h—C= dr/ds
—r/9, h may be eliminated. Again the absorption of heat due to a volume
increase dv, at constant temperature is if y is the heat ratio r (dx/dv)du + y%c
(y—1)dv/v. If the expansion is adiabatic the total heat absorption is nil and
the equation thus obtained may be reduced eventually to

4 (rx/3 + (C(x—y) + cy)lgs)ds + ye(y—1)dlgs

130 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

As this is not a perfect differential I assumed the relation of v and 9
to be approximately that of air, v’*S = const., supposing that I could subse-
quently correct for the precipitated water by successive approximation. In
this way one obtains at the beginning and the end of the exhaustion for any
two temperatures $ and $8’ (using accents throughout for the latter case), after
integrating, x’/x = (9’/r’)(r/9—1g(9’/9)), where («—x’)/x is the mass ratio of
precipitated liquid to the original vapor. In my work thus far! the results
computed in this way and for 6p = 17 cm. were at 10°, 20°, 30°, m X 10° =
.59, 1-13, 1.85 grams, respectively, where the corrections for precipitated moist-
ure have been applied and 6p is an isothermal value. If op, the observed
pressure reduction, were treated adiabatically the corresponding values of
m X 10% would be .42, .76, 1.28.

The results for m so obtained will have to be rejected as they are much
too small (probably because the pressure coefficients were overlooked) when
compared with the more direct approximation of Wilson and Thomson. The
value of m is here found as an intersection by making the m values compatible
with the vapor density curve for water. These data have already been com-
puted for the pressure difference 6p=17 above, table 2, and will be used in
Chapter IX.

9. Relation of nucleation to aperture of corona.—A summary of the method
of deducing the nucleation (number of nuclei per cubic centimeter) from the
observed coronal aperture for an observed pressure difference (6p=17 cm.),
in the given apparatus, has already been fully explained in Chapter VII, § 26,
and needs but little further reference here.

Measurements of the aperture s would naturally be made as far as the
inside of the red ring or the circumference of the eventually white disc; but in
such a case they bring out very strong periodicity in the first place, and are
soon subject to large errors due to the increasingly vague and washed outline
of the disc. Hence measurement is more appropriately made to the dark blue
ring which limits the green coronas or to the dark interior of the green ring in
the crimson coronas. These lines are not only sharper but they reduce the
periodicity. It is understood that with air nuclei and 6p =17, the green corona
is seldom exceeded. Otherwise it would be necessary to increase the uniform
pressure difference, against which there is no objection other than the increased
practical inconvenience, provided the pressure difference for which saturated
air condenses spontaneously (above 6p = 22 cm. in the above apparatus) is not
exceeded.

to. Absence of electrification in cases of sudden condensation and of sudden
evaporation.—It has long been known as the result of most painstaking observa-
tions, that neither in cases of ordinary evaporation nor of condensation is there
an accompaniment of electrification. On the other hand, when a mass of
water is suddenly shattered, as in jets, the electrification is marked, while it is

1 Smithsonian Contributions to Knowledge, No. 1373.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 137

not obvious that the underlying cause is friction. The electrification soon
vanishes, however, whereas the nucleaticn persists. The question thus arises,
therefore, whether in ordinary slow evaporation the absence of an electrical
effect may not be due to the possibility of charges vanishing too quickly to be
noticeable. It seemed worth while, therefore, to examine the case for the
sudden condensation and the rapid evaporation of fog particles.

To test this question a graduated electroscope was introduced into the
condensation chamber. Toinsulate itinthe satu-
rated atmosphere, the stem was enclosed ina hard
rubber tube, as in figure 5, open below and only in
contact with the stem at the top by a sealing-wax
joint. The tube was thoroughly dried before in-
sertion into the condensation chamber, and the
instrument showed satisfactory insulation for a
half-hour or more, after which it was removed for
desiccation prior to new experiments.

Condensation was produced by exhaustion as
usual, and the nuclei used were obtained from air
as well as from phosphorus. The method of pro-
cedure consisted in observing the normal leakage
of the charged electroscope by observations made
every half-minute. Nuclei were then introduced
and the experiment repeated with the alternate
sudden production and sudden dissipation of fogs — Ficure 5 —Appararus to Detect
between each observation. IONIZATION PRODUCED By SUDDEN

CONDENSATION.

The two curves of leakage were not distinguish-
able, even when the fogs were of the densely opaque character due to phosphorus,
and the error of the method was about 5 per cent., with somewhat larger un-
certainty for the case of fog evaporation.

Pressure differences, however, were kept below the limit (much above
Op = 22 cm. in the given apparatus) at which saturated air spontaneously con-
denses without nuclei. This phase of the question, which in fact is rather the
more interesting one, is thus left at issue. If, for instance, dust-free air is always
slightly ionized and thus contains unstable systems, the increase of ionization by
sudden exhaustion in virtue of the unstable molecules referred to is by no means
excluded. The question has been touched above.

11. Conclusion.—In the above paragraphs I have endeavored to present
the complications to which the method of coronal registry of atmospheric
nucleation is incident, complications which were not anticipated and for which
I was altogether unprepared. In the course of my work I made several un-
fortunate blunders in endeavoring to reduce the data to absolute values, but
apart from these the greater number of discrepancies (as, for instance, the
periodic distribution of nucleation in terms of aperture) could not have been
foreseen at the outset.

138 : A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

With regard to atmospheric nucleation, it seems to me, in addition to the
remarks made in § 1, that the variety and importance of the phenomena which
are now attributed to the invasion of solar and cosmical dust into the atmos-
phere, such as certain variations of atmospheric pressure, of atmospheric elec-
tricity, of terrestrial magnetism, of auroral display, etc., induce one to wonder
why continuous and systematic records of atmospheric nucleation, other than
the series obtained at Ben Nevis during the period when Aitken’s observations
gave to the subject widespread interest, have not long since been included
among the records of observatories. Surely in discharging its remarkable and
varied cosmical functions, this dust from afar, if at all persistent, must some
day be detected undisguised.

CHAPTER IX.
THE NUCLEATION OF THE ATMOSPHERE OF THE CITY OF PROVIDENCE.

INTRODUCTION.

1. Preliminary.—In May, 1902, Mr. Harvey Davis, at my request, put
up an apparatus in the laboratory of Brown University for counting the number
of nuclei in the atmosphere, by measuring the coronas producible with such air
under appropriate conditions. The apparatus gave promise at once; but Mr.
Davis was unexpectedly called away before the observations became fruitful,
and the project was temporarily abandoned. Believing that an instantaneous
method of at least estimating the degree of atmospheric nucleation is a de-
sideratum, and must throw light eventually on the origin and character of the
nuclei in the atmosphere, I have undertaken the furtherance of the work my-
self, and the results obtained since October, 1902, after the indications of the
apparatus had become warrantable, are given in the present chapter. I may
add that Mr. Davis and Mr. R. Pierce, Jr., had been at work for some time on
the measurement of the daily variation of the solar constant (a project then
set on foot by the U. S. Weather Bureau), and that I had hoped from the co-
ordination of the two classes of data to reach conclusions of interest.

The chapter, therefore, contains nearly two years of continuous record of
the nucleation of Providence, R. I. The observations were made in the park
of Brown University, which is surrounded, however, on all sides by the city.
The density of population lies to the west and southwest. <A station entirely
away from the habitations of man would naturally have been preferable, but
this desideratum was at the outset out of the question.

2. Apparatus.—The apparatus and method by which the present results
were obtained are fully given in the chapters above, particularly in Chapter VIII,
and need not therefore be further considered here. In figure 1, however, I have
shown a simplified train of apparatus used in some of my early work (October 2,
tgo2, to March 15, 1903). Here d is the influx tube, 6 the water-bath at room
temperature, ¢t the thermometer, A the condensation chamber (the water w is
made to wet the sides by rotating A before expansion), k the vacuum chamber,
G the gauge, S the pipe to the suction pump. Stopcocks b and ¢ control the
exhaustion, together with a.

3. Classification of results —The results themselves may be embraced in
two groups. The first group of observations, from October 2, 1902, to March

139

140 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

15, 1903, were originally made with reference to an arbitrary scale of nucleation.
There are many reasons why it would be difficult to reduce them accurately to
the uniform scale of the present chapter. Among these, frequent changes of
apparatus, measurement of aperture to the red ring, and the absence of a variety
of details of record discouraged such an attempt. The reduction, therefore, is
approximate; but the values of nucleation are roughly absolute and quite

FIGURE 1.—DIAGRAM SHOWING GENERAL DISPOSITION OF APPARATUS.

satisfactory as relations, and they thus meet most of the demands made upon
them. In view of their preliminary character they will be given in charts
only. In these cases the condensation chamber was a large aspirator flask
(fig. 1), placed horizontally so that the measurement of aperture was made
parallel to the axis of the cylinder.

The second group of observations (by far the larger number) from March
15, 1903, to October 31, 1904, were obtained with apparatus perfected as ex-
plained above. These will therefore be given both graphically and in tables.
The latter will contain a variety of relevant data.

FIRST GROUP OF OBSERVATIONS.

4. arly observations —The charts Nos. 1-12, comprehending the first
group of results, follow. The nucleations, 1, are laid off vertically, and in
charts 1-8 they are given in full (” nuclei per cubic centimeter). In the re-
maining charts (Nos. 9, et seq.) the number of nuclei are given in thousands
(1 X 10%) for brevity. The temperature and winds (for 8 a.m.), taken from the
weather maps of Block Island and Nantucket, will often be entered, but as
the work progressed superfluous data were gradually dropped. In some of the
charts the changes of temperature, ét, and of barometric pressure, 6p, are also
given, with the occurrence of clear weather, ©, cloudy, @ and partly cloudy
® weather, sunshine S, rain R, snow Sn, cold wave c.w. or —S, etc. Refer-
ences to Fahrenheit and to centigrade are indicated.

The nucleation (chart 1) begins low on the 2d of October, 1902, with the
rain, but thereafter increases nearly fourfold with the bright weather of the
succeeding days. Note the prevailing winds from the north on the 5th and 6th,
clouds and rain usher in a second minimum. On the fair days succeeding the

pe

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. I41

nucleation is not as high as before until the roth, when a sudden enormous in-
crease occurs, cotemporaneously with fall of temperature. The high maxi-
mum is succeeded by an equally low minimum, brought on by the rainy weather
of October 11 and 12. The nucleation then rises during the succeeding fair
days, to fall to a fourth minimum during the rain of October 14. No doubt

-101° +67° 00° -7g° =8t'4.
£ Re |

a
ae Co op On.

Cuarts 1-48.—DaILy OBSERVATIONS OF THE ATMOSPHERIC NUCLEATION OF PROVIDENCE. NUMBER
or Nuc er (7) PER Cupic CENTIMETER, LAID OFF VERTICALLY. IN CHARTS I-8, 7 IS GIVEN IN FULL
In CHARTS 9-48, 2 IS GIVEN IN THOUSANDS PER CUBIC CENTIMETER. CHARTS I-12, ARE REDUCED FROM
AN OLDER ARBITRARY SCALE. Loca Winps, Rain (R), TEMPERATURE, ETC., ARE GIVEN IN THE CHARTS,
Arrows SHOWING THE DIRECTION IN WHICH THE WinpDs Blow. CLEAR©, PARTLY CLouDY @, AND CLouDY @
WEATHER, ETC., ARE INDICATED WITH THE USUAL SIGNS.

some variation is due to the variation of the temperature of the apparatus, for
which no correction! was made, but the rain and fair weather effects as a
whole are unmistakable. The maxima on the 4th and roth correspond to anti-
cyclonal conditions.

The apparatus was now modified in a way which need not here be instanced.
What is noteworthy in the next data (chart 2) is the occurrence of sharp minima
on October 16, 17, 21, 23, cotemporaneous with the passage of dense cloud
masses over the sky. In three of these cases the curve rises as soon as the
sky clears; on October 17 it does not do so, but the curve runs into the over-
cast condition of October 18. The pronounced minimum on October 19 during
clear weather shows that sunshine can not be a reason for an abundance of
nuclei. Similarly there is high nucleation on October 24 and 27, simultaneously
with an overcast sky. A number of night observations on October 18 give no
evidence of unexpected behavior.

On October 27 the apparatus was again modified, by substituting a long
cylinder for the aspirator flask thus far used as a condensation chamber. The
data (chart 3) begin with high nucleation for an overcast sky and fall off to
the rain storm on October 28. The high nucleations are very fluctuating,

1 Correction for temperature was applied at a later date, but thereafter again abandoned
as untrustworthy.

I42 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

which would be even more apparent if night observations had been made. The
cause is obviously convection, and the diagram necessarily presents marked

similarity to a wind curve.
The feature of the data from November 1-10 (charts 3 and 4) is the ten-

1 yy ke Fee eee

+h

CHARTS 3 AND 4.

dency to notched minima at mid-day or in the early afternoon. The maximum
on the 1rth coincides with a slight drop in temperature, but beyond this the
temperature is so nearly constant until November 24 that reasons for the vari-
ations of the figure are not forthcoming. The remarkable rise on the 21st may

CHART 5.

be noticed. On November 14 and 23 there are mid-day minima, but others are
not apparent. During the remainder of November (chart 5), rain minima
occur on November 25, 26, and at the end of November 29. The maximum on

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 143

November 29 is clearly associated with the drop in temperature, and there is
here one mid-day minimum but no others.

Throughout December cold waves associated with northwesterly winds
are productive of maxima. After the rain of December 1 the nucleation rapidly
rises, accompanying a second drop in temperature. The maximum is moder-
ately sustained until the 3d, where there is fall into the rain minimum, partial
recuperation thereafter, and then another drop into the rain of the 5th.

This rain soon changes into snow and from here the nucleation (chart 6) of
the cold wave begins, lasting from the 6th to the thaw on December 11. One

3e° 30°

oR ws

eS
‘i 1 | is

f° te

CHART 6.

may note the reduction due to very cold snow on December 7, and the curious
minimum accompanying the gale on the evening of December 8. The maxima
on the 9th and roth accompany northerly winds and a drop in temperature.
The very remarkable maximum from December 13 to December 16 follows the
fall of temperature after the thaw on December 15, introducing a rain minimum.
This interesting region is no doubt referable to the snow-storm on the evening
of December 13, the winds remaining northerly. The ground was frozen on
December 13 and 14. The region is not due to temperature. The maxima on
December 17 and 20 are less sustained, but they come with temperatures above
freezing and westerly winds. The absence of night observations is a dilemma
in these cases. The low nucleation indicated on December 20 continued for
several days (December 20 to 23) beyond the limits of chart 6, but is shown on
chart 7.

One may regard it as a general rule established by these charts, that the
nucleation increases when temperature suddenly decreases, but that tempera-
ture is not the sole or ultimate factor involved. To see whether this view
would work out in detail, I compared the observatory thermograph data for
Providence during the three months (kindly placed at my disposal by my

144 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

colleague, Prof. Winslow Upton) with the data for nucleation given in the
above charts. <A limited degree of similarity was in this way made out for
cold weather, both in the general march of temperature throughout longer
intervals and in its detailed variations. Thus the mid-day minima are obvi-
ously counterparts of corresponding temperature maxima. There were, how-
ever, many outstanding discrepancies, and it seemed probable that much of
this would be referable to local differences, as the observatory is differently
situated and at a distance from the laboratory. Beginning with December 28,
I have, therefore, been making daily temperature observations simultaneously
with the coronal observations, using a long mercury thermometer placed be-

CHART 7.

side the influx pipe. The temperatures are given on the tops of charts 7 and
8, and positive temperatures in degrees Fahrenheit are laid off downwards to
insure an easier recognition of coincidences, remembering that nucleation
maxima and temperature minima correspond.

In the fragmentary temperature observations before December 28 (chart
7) some relation is seen; thus the maximum between December 23 and 25 has
a temperature equivalent, the broad minimum on December 22 and the mid-
day minimum on December 24 correspond, etc. Turning to the cases under
detailed observation, the mid-day minimum on December 28, the fall on Decem-
ber 29, and, to a smaller extent, the data on December 30 and 31 (in the latter
case other factors are active) agree. The marked minimum on both cases of
January 1 and, to some extent, January 2 are especially striking.

In chart 8 the general maxima marked ‘“‘c.w.” may be noted, and the

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 145

One may point out the marked fall on January 11 and the

details are similar.
On this curve there is, however, one good

day minimum on January 14.

ig je =

ee

Sa

BB
SK Sore ais
Ss Ss
=
\
Seer S
=
i :
\ iy
San Se ie:

100,000 as i : |

75,000 \

SD. \S
PH vhs

25,000 é eel

B

in

CuHaRT 8.

example of the opposition so frequently to be observed in the summer time,

for on January 13 a pronounced minimum of the temperature curve occurs
Again, the marked fall of

simultaneously with a maximum of nucleation.

1903

CHART 9.

temperature on January 17-19, both as a whole and in its details, is unaccom-
panied by a corresponding nuclear effect, and this is true throughout the

146 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

remainder of January. Leaving the high rain minimum on January 21 (chart
9), the cold wave from January 23-26 produces but a small effect, while the

CHART 10,

high nucleation on January 27 is not foreshadowed by the corresponding march
of temperature.
5. Plate glass apparatus—Apparatus with plate glass windows was first

1903

CHARTS II AND 12.

installed on January 29 and thereafter used permanently. The beginning of
February (charts 9 and 10), with its frequent dark fogs and relatively warm

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 147

weather, shows lowered winter nucleation. The data are numerous because
many subsidiary experiments were made, but little of new import suggested.

The method of measuring coronas to the outside of the first red ring was
adopted on February 16.

The cold wave beginning on February 18 (chart 11) shows moderately
high nucleation until it breaks with the thaw on February 21. Thereafter
gradual but fluctuating fall of the temperature and nucleation curves appears
from February 22 to the intense rain on February 28. During much of the
time the ground was covered with snow. Mid-day minima are of common
occurrence.

March 1-15 (chart 12), with its relatively high temperatures, has nearly
wiped out the characteristic winter nucleation. The daily nucleation is not
very variable and the rain minima are curiously high. The temperature effect
is quite vague, there being as much opposition as correspondence. This sudden
drop of the winter nucleation in March, 1903, is particularly noteworthy and it
will not be repeated in 1904. The Sunday nucleations in March are not ex-
ceptionally low until the 15th, when an abnormally small value for a clear day
appears. As there is a general cessation of work in factories on this day, the
corresponding nucleations are to be carefully scrutinized throughout the period
of warm weather following. During cold weather there is no reduced nuclea-
tion on Sunday.

SECOND GROUP OF OBSERVATIONS.

6. Arrangement of tables and graphs.—On the 12th of March the same
apparatus was installed in a new position, and on March 15 the results referred
to above, § 3, as the ‘‘second gruop,”’ begin. The scale is uniform throughout,
though somewhat enlarged, and the plan is the same as above. Local winds and
temperatures only are entered.

The data are also given in tables, as it is believed that the method is now
sufficiently definite to make these available for other purposes than the im-
mediate ones of the present chapter. The tables contain the date and hour,
the state of the weather, the temperature of the atmosphere where the nuclea-
tion is observed in degrees Fahrenheit, and of the air within the apparatus in
degrees centigrade; furthermore, the aperture, s, of the coronas when the
goniometer radius is 19.5 cm. and the distance of the goniometer and source of
light 85 cm. and 250 cm., respectively, from the condensation chamber. Thus
the angular diameter is p=2 sin ‘(s/39). The remaining columns contain the
coronas with the colors recorded from within outward, abbreviated as stated,
and the absolute nucleation, , or number of nuclei per cubic centimeter of air.
Certain obvious remarks will be added.

The correction for the temperature of the condensation chamber (pur-
posely kept as nearly constant as possible) was not applied. This would have
enormously increased the labor of reduction, without materially changing the

148 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

distribution of the nucleation observed. It is to be remembered, moreover,
that the true law of reduction is periodic, and unless the periodicity is fully
worked out for the small coronas, the temperature correction has little meaning.
It is because of this complexity that the tables are very specific as to the color
type of the coronas, their size, the temperature of the fog particles, etc. In
any special case, therefore, the rigorous reduction (if of interest) may be at-
tempted, but at present there is no call for it.

As to the accuracy of the results so far as the readings are concerned, an
error of 2000 to 3000 nuclei per cubic centimeter is unnecessary in the ex-
treme case of large coronas; whereas the whole range of variation is about
from 2000 to above 100,000. For the small coronas the reading is correspond-
ingly sharp, and the same is true if a larger goniometer is used.

TABLE 1.—Continuous record of atmospheric nucleation from March 15,
1903, to November 1, 1904. The time (#) is given in hours and tenths of an
hour, the temperature of the fog chamber in degrees centigrade, the local at-
mospheric temperature (approximate) in degrees Fahrenheit. The aperture
of the coronas is s, where the angular aperture is p=2 sin s/39, as the arms of
the goniometer are 19.5 cm. long. The last column but one shows the number
of nuclei, , per cubic centimeter of air, using the table in Chapter VII, § 26,
for the reduction from s to m. Temperature corrections were not applied.
The pressure difference was uniformly 6p=17 cm. of mercury. The distances
of the eye and the source of light were 85 cm. and 250 cm. (1:3), on opposite
sides of the condensation chamber. The weather abbreviations are f fair, f’
partly fair, c’ partly cloudy, c cloudy, R’ slight rain, R rain, Sn snow, 5 sun,
etc. The coronal color abbreviations are c (crimson) deep red, r orange red,
br brown, o orange, y yellow, g green, b blue, v violet, p purple, etc. A ver-
tical line denotes an indeterminate color, as wr|g; an accent an approach to
the color, as r’ reddish; a capital, a deep dark color, as B deep blue, etc. Mixed
colors are written together, as bg, blue-green, or, orange-red. Subsidence
measurements when entered (November, 1903) are usually abbreviated a/b,
meaning that a seconds are required for a fall of the fog line of b centimeters.
Sometimes the abbreviation is b/a, as will be stated (April, 1904).

o oa |
alae :
BY | Be i s
K Sp | 84 5 k
f a : ge ga 3 Bi & Remarks.
. ; J q
gi) So Be) Bs |] gee |e 85 gE
o x oo Ps is OA! OF & 3'0 =
Qa a e S aq | ad <q OO AZ
1903 |Mar. 15/ 10.3] £ |N.E.| 20.5] 49.2 1.8| w br B 7000
10.6} f 20 — 1.8 Do 8000
12.8] f 2I | 50.4 1.8 Do. 8000
1.0| f 21 = r.9 Do. 8500
r.2)| f 21 — 2.0 Do. T0000
1 Ail) ch 21 — 2.0| wbrBg T0000
Gaal iat 20 | 44.8 1.9|wbrBegr 8500

Year.

1903

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 149
TABLE 1—Continued.
ey | ee
j Be] Ba o =
; 3 ol as 5 ce 3 ay 3 Remarks.
o oO 2 so} Ag Qo re e K d
8 & o a Ba EE a BS 5
A 6 5 Elad|aa| < 6d Z
Mar. 15] 5.4| f 20 1.9 Do 8500
6.4] ¢ 20 | 42.0 1.8 Do 8000
eRe — | — 1.8 Do. 8000
Mar. 16] 9.3] c N.E.| 18 | 41.4 2.4; wbrBeg 18000
9-5] Cc 18 _- 2.6 Do. 22000
Dist C 19 | 42.6 2.4 Do. 18000
3) |PC 19 — 2.5 Do. 21000
AES) TQ || 42.5 2.8; wbrBp 28000
AaTinc 19 -— 2.9| wbrBr 31000
5-9] c 20 — 2.5|wbrBg 20000
Garlic 20 | 40.3 2h5 Do. 20000
Mar. 17] 9.4/c N.E.| 20 | 45.8 3.6] weg 57000
9.6) c¢ 20 — 3-5| Woe 53000
TOL |"C 19 — 3-1| w br Bip 40000
ONS 20 | 53.4 2.9| w br Bp 31000
TEAC aL — 4.0] Wrg 78000
12.8 c 21 —- 3.0| wbrBp 37000
2.7| ¢c S.W.} 21 | 58.8 3.1| w br Bip 38000
Gar)|\Kc 23) 56.4 Bak Do. 38000
6.3] ¢ 23 -— Za Do. 38000
Mar. 18] 9.5|c¢ N.W, 20 | 49.9 2.1] wbrBegr 12500
9-7] c 20 — 2.3| w br B 16500
1z.8| f (S) 21 — oes Do. 15500
2 vie et 21S 4L0 2.3 Do. 16500
Aes lief Baw ogas 2.3 Do. 15500
4.7| £ 22 -- 2.3 Do. 16500
5-8] f 22 | 50.0 2.2 Do. 13500
Sa] ek —} — 2.3 Do. 15500
Mar. 19] 8.7] c S.W.| 19 | 44.8 3.0| gbrBp 35000|
8.81 ¢ —}|— 3.0 Do. 35000
2.6| f (S) 2I | 66.8 2.9| y br|b 31000
3.0| f 21 _— 3.4| w br 49000
Ae Tee 22 — 3.0| w br 35000
4.2| f 22 | 64.6 3.0| w br|b 3.5000
Sap) tt Dia aes 70 3-5| Weg 55000
5-8] f 22 = 355 Do. 53000
Mar. 20] 9.2] f’ (S) |W 20 | 62.4 3.0| wbrB p 35000
9.4| f’ 20 _- 3.8| wre 66000
10.0 | f’ 20 — 3-5| weg 53000
12.8} f’ 2t | 68.5 2.8| y’brbp 30000
TsO)! 21 — 3.5| wre 53000
Astle 21 — Ba5 Do. 53000
Bialik 22a ey Lee 2e7\\.e" br Bip 27000
2.8] f 22 — 2.7| w br|b 27000
Goel at 23) (54769 2.8| wbrBp 28000
Ge ee 23 —_ 2.8 Do. 28000
6.0] f 23 — 2.8) welg 28000
Mar. 21] 9.6]cR E 20 | 46.2 2.4| cor TgO0o
9.7|R 20 — 2.2 Do. T4500
IO:7|'C 20 | 48.4 3.0) wbr Bp 35000
12.2| Cc 2t | 49.8 2.4| w br b’ 17500
12.9| c¢ 2t | 52.2 2.3| w br Bir 15500

150 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

# Bo
B4!] 3s 6 =
& 3 o
a é : S 5 5 ay & Remarks.
: é 3 og | ak] ad 2 She
¢| £ |€| F | 2 |eel ge) 2] ee 5
- A & E B |ad|/aa| < SS Z
1903 |Mar. 21] 2.9] ¢ 21 54-9 2.9| wbrBr 31000
3. AC 21 ae 2.6| wbrB 22000
SevAlKC 23 | 56.0 2.9| wrg 31000
5-9| ¢c 23 a 2.8 Do. 30000
Mar. 22] 9.5|cR’ |N.E.|22 | 51.2 2.0; wbrBg 11000
10.0] C 22 | 52.7 2.5|wbrBg 21000
Ts Si Oke 22 Ns2e4 2.4 Do. 17500
TAD ECHR: N.W.| 22 —_ 225 Do. 20000
GIGI 2 Taal ile 2.4 Do. 17500
6.6| ¢ 21 -= 2.4 Do. 17500
Mar. 23} 9.0] R N.E.|20 | 44.7 1.8] cor 7000
9-4| R 20 — 2.0 Do. 11000
12.8] c 2r | 47.6 2.1 Do. 11500
L250) |"C 2I a ee Do. 12500
2.8] R 21 | 49.0 2.9| r br|bg 33000
3.0| R 21 = 2.9| wbrBp 33000
3.4|R 3: 21 — 2.9 Do. 33000
Gea |e Re 22570 2.5| wrg 21000
6.3) cR 22 aa 2.6) wbr Bg 22000
Mar. 24] 9.2|c S.W.|20 | 57.2 2.8| w|Bp 28000
9.4| ¢ 20 = DH) Do. 25000
12.8] ¢c ar | 63.2 2.2| wbrBg 13500
12.9] Cc 21 = 2.2| wbrBg 13500
4.0] Cc 23 | 60.5 2.7| y'tg 25000
eT NIc 23 —; 28|¢’Bp 28000
6.0] c 23 || 55.0 2.7; wbrBg 27000
6.1] c 23 — 2.8| wrg 28000
Mar. 25] 9.5| f N.W.| 20 | 51.2 2.6| w br 22000
TOSS |b 20 | 53.0 2.8] w br 28000
11.6] f ZON| sey 2.6] wbr 24000
12.5| f 20 | 51.3 ls Ly re 28000] |
P2zgiiic 2r |) 5 ro 2.2| wbrB 13500
3-211 ¢ 2m — 2.2| y’ br B 13500 |
5.8| c 2am ao 2.1] Cor 11600
6.0| ¢ 2I — 2.3 Do. 15500
Mar. 26] 9.5) { Wen enon ages 2.8| w br b 30000
9.7| f —}|— 2.8| w br B|p 28000
12.9 £ 18 | 52.8 2.8 Do. 30000
12.8| f 19 — 2.8 Do. 30000
2.5|f 20 | 54.6 2.7|y' rBg 27000
2.6| f 20 — 2.9] y’r 31000
5.8] f 2Oy eS 2e7 2.7| Wrg 27000 |
eA 3.6| weg 57000 |
ae aa ie Do. 46000 |
Mar. 27] 9.4| f W. /|18 | 55.7 3-5| we Bg 53000
9.8} f 18 _ 3-6| wrg 59000
ETO} Me 18 | 62.8 2.8) wbrBp 30000
TES £ 19 —_ 2.9 Do. 31000
12.7| f TON W Oye2 2.9| w br bg 31000
2.8] c’ 20 | 68.5 2.8) y’ br bg 28000
5.8] c 21 — | :27|wbrBp_ || 25000 |

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 151

TABLE 1—Continued.

& ae)
Bo mR 4 SS
: €2| 82) 2 5 eae
a of Va 3 Sy emarks.
eo] 2 la) es |2\Felge) 2) 82 E
A iS = B jad|/ed| < SS Z
1903 |Mar. 27] 5-9] c 20 02:8 3-2) wre 42000
Mar. 28] 10.2] f Wi iat. {52.8 2.0] cor 11000
10.4| f 21 — 2.0| wbrBg 11000
12.8] f 22. )\)'58e3 2.3| wbrBg 15500
1.0] f 22 oa 2.8| ybrBb -30000
2.6| f BB S043 2.6] wrlg 22000
253k N.W.| 23 = 2.6] w’rg 24000
evict 23 | 47.0 2.2| cor 14500
5.9) £ 23 — 2.2| wbrB 13500
Mar. 29| 10.1] f N.W.| 20 | 36.5 3-1| wrbg 38000
LOW if 19 — 3-3| weg 45000
Tisai 20 | 42.0 2.7|wrg 27000
ig] E 20 aaa 2.8| wbrBg 28000
4.3] £ 20 | 44.4 2.6| wbrB 22000
4.4] f 20 — 2.7| y’ br bg 27000
to) [Pak N. 20) 4355 2.9| y br|bg 31000
6.6| f 20 | 51.0 2.8) wrg 28000
Onl t 20 — 2.8) wbrBp 28000
Mar. 30] 9.9/c Sobyen tye 43%3 2.7| wbr Bp 27000
10.9| c 17 — 3.8| y’ rg 66000
LEW sI6)|| KC 17 | 44.5 3-7| yt bg 62000
t= 5|C 17 = 3-7| wrg 62000
12.7/ Cc 18 | 46.8 2.8| w br Bp 28000
12.9] Cc 18 == 3-5| wrg 53000
ID] ¢ 18 a 3-5| wre 53000
Dea 19 | 48.3 3.2| wbrB 42000
Ba 7i|nC 19 = 3.4| wbreg 49000
Cac 20 | 46.3 2.3) wbrBg 16500
5-8} c 20 — 2.3| wbrBg 15500
Mar. 31] 9.4|cR’ JN. 21 — 2.9| wbrBr 31000
9-5]; Cc 2 eA 2-7 2.8) weg 30000
12.9| ¢c 21 | 46.6 2.8| y’ brig 28000
ef0)|| 2 aa Defi Do. 27000
Bese W 2 TSS 38 2.7| w br 25000
4.0| Cc 21 -- 2.7| w br|bg 25000
a5 |e) 22 -- 2.3| w br Bip 15500
5-6] f Zoe lS Ger 2.8) wbrg 28000
UNI Dg) .Q=4))\ if Wise | to) 5227 2.0} cor T1000
__ | ( 2.8] ) w br B bg 28000
O52 Ee ( 2.4] { wbrB 18000
12.5) | Cc 20 | 56.6 2.7| w br B 25000
12.6] ¢ 20 | — 2.2| w brig 14500
Bele 22145022 2.8| wbrBp 28000
(fs: 53000
a8) Aone. 138 kwrg { 28000
Wei a 22 49.0 2.8) wre 28000
7.8) £ 22 — 2.7| wrg 27000
pin 2) O20) let S.E. |19 | 46.5 2.8| wrg 28000
0-3)|£ 19 —_— 2.9| wbrBp 31000
rigging 20 51.6 2.4| cor 17500
2-2) |e 20 ~- 2.7| wbrg 25000
12.8| f 20 = 2.5|wbrBg 20000

152 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

e Eg
su] 3% a :
8 ; gs Es 3 « E Remarks.
g| 2 | @] 8 |B IERI EE] 3 bs g
oO x aoe oo or a °° =
> A a = EF led|}eq] < SO Z
1903 |Apr. 2| 5.3] f 5 22) A760 2.7| Wwrg 27000
5.4| f 22 — 2.7|wrg 27000
Se Salt aa 2.8} wbrBp 28000
Apr. 3] 9.4] cR’ |S.W.|21 | 57.7 2.5|}wrBg 21000
10.2] Cc 21 at 2.7; wbrBg 26000
10.4] Cc 21 = 2.7; wbrBp 26000
10.6] c 21 = 2.8| w br 28000
12.5 inc 22 | 62.3 2.6|wrBg 22000
122 0)|"C 22 aa 2.8) wbrBg 28000
BESiEC 23 | 60.5 2e7|| y¥ apie = 24000
4.2] Cc 23 = 2a Do. 24000
5 TAlee 23 | 60.0 2.8}; worG 30000
6.0 24 — 2.7 Do. 27000
Apr. 4] 10.2} c R’ |S.W.|22 | 60.4 2.8) w|B p 28000
10.3| Cc 22 — 2.8 0. 28000
12.0; cR N.W.| 2 60.1 2.9| w br diff. 31000
12.0] c 23 — 2.8| wbrBp 28000
ZEOunc 24 | 43-3 2.8/yrg 28000
4.0] c 23 aad 2.8| y’rg 28000
Sez ne 24, 3 2.5| w br 20000
5-9 |IC 24 ae 2.5|wbrBg 20000
Apr. 5 9.6} f N.W.| 21 | 33-4 3.0/ wbrBgyr]| 37000
9.7| f 21 — 3.2| welg 44000
rr] f 2I | 40.0 3.2| Weg 44000
1.4] f 21 — 3.1| w br Bip 38000
4.4| f 21 40.0 2.8} y’r bg 30000
4.6} f N 21 = 3.8] y’ brb 68000
Aha 2I | 39.0 4.1 Do. 83000
Bela ae 21 —_ 2.7|ybrBg 27000
5.6] f 21 3.2|wrBg 42000
6.3} f POX || eiifont 3471/29 Bap 27000
(wrbrg
6.4| f —|— Ber lwbrB A 40000
ADTs (O) | oesiet S 21 | 37.0 3-4, wrBg 49000
9-5| f 20 — 2.8| wbrBp 30000
mire Ay) ee 20 | 42.0 2.9| w br bg 31000
Tec 20 — 2.9| w br B|p 33000
3.0| f 22 | 42.0 2.4|wrBg 19000
ey) a 21 — 2.71 y' TE 25000
Bem ee 21 — 2.6] wrbg 24000
5-6] c 23 | 40.5 3.0| wy Bg 35000
Apia] | er0:8) IC Ron als 22 | — 2.9| w|Bp 31000
D2 Aul nuke 22 S25 De) Do. - 25000
12.5| R 22 — 2.7 Do. 27000
2.8/ R 23. || sae 2.6|wrBg 24000
2.9|R 23 _- 2.7| y’ rig 25000
5-5| R 24 | 53.6 2.6| wrg 24000
Baz mes 24 — 2.6| wbr Bp 22000
Apr. 8} 9.3) R 2 Sr 2.4| wbrBeg’r 17500
9.5|R 22 _- 2.3 Do. 16500
11.8| R E 2g) )| hana 232) COE 13500
12.1] R 23 — 2.2 Do. 14500

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

153

: Re l|ee| 2 : cette
_ S 5 of Oo 3 3 2 emarks.
og ae a ee ee
cele le | & | BE led)aa |. = SS Z
1903 |Apr. 8 | 4.2|R S. 24 |54-5 3-2|/wbrBg 44000
4.3 |R 24 — 2.8|wbrBp 28000
4.4|R 24 = 2.7 Do. 27000
pr. 0 |" gat it Ware 2g) 11/5'5:0 2.2)wbrBgplr| 14500
9.4 |f 22 = 2.6 | y rig 22000
9.6 | f 22 a 2.6|/vrg 23000
12221) 0 peor 2.7|wbrBp 25000
12.3 | f 23 — 2.6 | wrlg 24000
3-9 |f 24 | 63.4 2.7|gbrBp 25000
esta ft 24 — 2.7) Do. 27000
5.4 |f 24 | 61.1 227 Do. 27000
es | (6323) DE

sole 4 eines) = pees
APIO oq i |N, 22 |'54.0 2.2 | w br Bir 14500
CO)(6) || 21 = 2.6/wrBg 22000
9.9 | f’ 21 = 2.6 Do. 22000
12.4 | f’ (S) 22m | Soar 2.9 | w or bg 31000
2G tes) 22 aa 3-3 |wrbg 45000
Te 7 | ete 22 ao 28!gBp 28000
Bent) |INaWel23, |/6n.7 2.5 | y’ br Bir 20000
Beau eas) 2 a 2.6|wbrB 22000
5.8 |f 23 S77 2.2 | cor 14500
5-9 /f 23 = 235 |||COD 20000
FD E eat ORAS (it N. 21 2.4 2.8 | w br B|p 28000
g.6 | f 21 a 2.8 | w|B|p 28000
Tyo sete | 22 57-2 2.8 | w br 28000
Tah 22 = 3.3 | w br bg 45000
Aes [aE 22 a 2.8|wbrBp 28000
Qe Sal ah 22 | 59.4 2.2|wbrBgr 14500
3.01) £ W. | 22 — 2.5 | y olg 21000
eA 23 156.7 2.2 | cor 14500
COM te 22 = 2.6|wrg 24000
Apr. 12] 9.8/f£ N. 20. | 51.5 2.3|wbrBegr 16500
10.3 | f 20 aad 2.8|¢’ br Bp 28000
10.8 | f 20 a 3.3|wrBegr 47000
Legit 20 | 55.0 2.8 | w|B p 30000
abel 20 aaa 2.8 Do. 28000
“8 Ox 24|/wrBg 17500
ie i lee wbrBg 28000
4.0|f 21 57-0 2.3 |y’ r bg 16500
4.1 | f 21 a 2.5 Do. 20000
6.0 |f 2a aSARg 2.4 | COr 19000
Apr. 13] 8.9 |f N.E.|20 | 49.8 2.7|woB bg 25000
g.1|f 20 — 3.0|wbrBp 37000
9.3 |f 20 — 2.8 Do. 28000
12.0 |f 20 154.6 2.6 | wrlg 24000
reo eva i 20 — 2.6 | w olbg 24000
BeSnice 21 — pat ||XXore 12500
3.6 | f 2m | 53-0 223 Do. 16500
Get |] at 22 | 46.5 2.0 Do. T0000
SzO) Ef 22 — 2.0 Do. 10000
Apr. 14] 9.2|/¢ E. 20 | 44.7 2.1|wbrBegp T1500

154 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.
TABLE 1—Continued.
y Dp 4 1.0, A =
; ; 3 . BE i: | 3 a 3 Remarks.
si 2 [2] 2 | 2 28) gs| es) eee
> a eG = EB lediad| < OO Z
1903 |Apr. 14| 9.4|c¢ 19 - 2.5|wrBg 20000
OFC 19 — 2.2 | cor 14500
9.9 |c 19 45.8 223 (COT =
10.1 |c 19 | — 2.1 | Cor — dp=22
10.2 | Cc 19 — 2.1 | cor —
12-3) \C 20 46.4 | 2.1 |\cor 11500
12.4|c¢ 20 |— | 22\wbrBgp 14500
12.5 KC 20 | — 2.9 | wbrB —- 8p=9
12.6 — |— 1.8 | cor — 8p =22
255) {iC 21 46.2 2.35| (COL 15500
3.6\c 21 = 2.2 | COr 13500
6.0} c 21 43-9 2.0 | cor T0000
6.1 1c 21 == 2.0 | COr 10000
Apr DSi 0:30 (Cun N.E. | 21 — U7 COL dp=22-
9.4|R 19 | 43.7 2.1 | cor I1500
9-5|R 19 — 1.8 | cor 7000
9.8|R 19 — aya COT — 8p=22
12.1) R 20 «| 44.5 2.0 COG T1600
Loan 20 = 2.0 | COr 11000
27 WR 20 | 44.7 1.8 | cor 8000
4.2|R 21 — 1.9 | cor gooo
52) Re 2 ASe5 1.8 | cor 8000
: 6.0 | R’ 21 — BD COD 12500
Apr. 16] 9.5 /c N.E.|19 | 42.0 2.1 | cor II500
On7alee 18 — 2.1 | w br B|p 12500
1257 |.¢ 20 | 43.4 2.2 | cor 13500
12.9|¢c 19 — 2.3 | w br B|p 16500
Op3aikc 21 40.5 2.2 | cor 13500
6.4/¢ 21 — 2.2 | cor 13500
Apia yl Oselc N. 20 | 42.1 2.5 | cor 20000
9-61¢ 20 — 2.4 | w br B|p 19000
DEQ C 2m | 48.1 3.1 | w br 38000
12.0 | C 21 — 2.9 | g’|B p 33000
TEAC 21 — 2.7|wrB — dp=22
12:8 j\'c ai _- 2.8} g’|Bp 20000
Beqaike 21 45-8 2 SalCOG 15500
Be7alic 21 — 2.6 | y rlg 24000
“eae |e 21 = 2.6 | y rlg 24000
BOC 22) | 45-7 2.7 | y’ tlg 25000
6.0 /c — | — 2.6 Do. 24000
Apr. 18] 9.0 /f Wie) mo) hisexs 2.9 | w br B bg 33000
9:2) | t 19 _ 3.2|wbrBegr 42000
9.4 |f 19 — 2.9 | g’ br Bip 33000
9-7 |f 19 = 3-4| weg 49000
9.8 | f 19 — 3.0 |.w’ br B/p — bp=22
9.9 | f 19 — 3.8 | welg 66000
10.0 | f TOMES 3.3 | welg 47000
21h | Nie 20 |57.8 2 Sa COL 21000
12.2 | { 20 — 2.7 | w br Blp 25000
4.0 | f 20: | 50:0 2.9 | w br B|p 31000
eral eh 20 | — 3.3 |W a 45000
4.4 ]f 20 = 2.8|wrig 28000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 155

TABLE 1—Continued.

B .| oo
24 23 6 =
: el 7) | eee
E <4 . Sy D 3 3 emarks.
ee | s | & | eflee| & g§ g
ee jal = | E léa}ece] 2 88 Z
1903 |Apr. 18} 5.9] f 20 = 2.9|wbrBp 31000
6.0] £ 2I | 45.9 2.8 | w rlg 28000
8.8) f 22 | 52.0 2.3|wbrBg 15000| Night.
Oeste 21 aa 2.4 Do. 17500
10.4| f£ 22 | 40.5 2-4 Do. 17500
10.5 | £ 22 = 2.7 | wrlg 25000
Apr. 19] 10.4| f N.W 22 | 49.8 2.2 | cor 13500
10.6) f 2I — 2.5 | cor 20000
2b |e rea Ser 24a |\COr 17500
12.5| f 21 = 2.6 | wrlg 24000
Bele 2r || 56.7 2.6 | y’ rg 22000
3.8| f 21 — 2.0 Do. 22000
6.4| f 22 || 54.6 2.5 |\cor 20000
6.5| f 22 — 2.6|wbrBg 23000
Apr. 20] 8.9] f W. HON P5355 2.6|\y’rBg 22000
9.0] f 19 — 2.8 | w rlbg 28000
inva || at 20. | 60.2 2.7 |y’ tg 27000
reate(0))| 9 20 = 2.7|g’ br Bp 27000
a5 lee 22 | 64.5 2.5 | Cor 21000
Be oi |e 21 a 28/yrBg 28000
eta |e 22 | 63.2 2.8|/wbrBg 30000
Beek 22 = 2.8|wbrBp 28000
Apr. 24| 10.6] cf Wie itz |'soz0 2.4|/wrg 18000
11.4|¢ 20 aaa 3.2 | wr|g 42000
ste 2ON | SO=7) 2.8|g’ br Bp 28000
12.4| Cc S 20 | 55.8 3-3 |y tle 47000
TE IMG 21 ae 2.6|g’ br Bp 24000
Anais 22 | 55.0 2.5 |y’ tle 21000
4.6} c’ 21 + 2.6 | yrlg 24000
5-0) 21 — 2.7 | wrlg 25000
Apr. 25| 9,0] ¢ W ZO) ||| S07 2.7 | COr 25000
Opn |c 20 = 2.6|/wrBg 22000
12.4| Cc 21 | 60.5 2.8 | w brig 28000
12.5 | Cc 21 — 2.7|wbrbg 27000
Was 21 — 2.8/g’ brBp 28000
4.2|c¢(S) |S. 22 | 59.7 2.5 | cor 20000
— |f 22 — 2.5 | cor 20000
5.6| f 22 2.5 | Cor 20000
5.8| f 22 | 58.6 2.8 | cor 28000
Apr. 26] 9.9| c’ N.E.| 20 | 57.8 2.7 | COr 18000
10.4| c’ 20 == 2.4 | Cor Tgooo
TAN) es 20 | 62.0 1.9 | cor 8500
12.8| c’ (S) 20 2.0 | cor 10000
Lets c 20 = 2.2 | COr 13000
6.2] ¢ 2 54.8 2.1 | cor 12000
6.5] Cc 2 = 2.1, ||COR 12000
Apr. 27| 8.9| f N. 19 | 50.4 2.4 | cor 19000
g.1| f 19 = 2.4 | COr 19000
12.3] £ 20 | 61.4 2.1 || COT 12500
12.4| f 20 — 2.4 | COr 19000
12.6] f 20 — 2.4 | cor 19000
BrO)les N.E.| 21- | 61.6 1.8 | cor 8000

156 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

£ Es
Su x 4
: ea ee |e .
7 F ‘ 2 = ge | oa 2 a 2 Remarks.
g.) 8 | 8 | 4800) Besa me BS 3
i a a = E ledj/ed}| < OO Z
1903 |Apr. 27| 3.8/f 21 = 1.8 | cor 8000
5.8 |f = 1.8 | cor 8000
6.0 | f 22 58.5 2.1 | cor 12500
Apr. 28} 10.4 |f N 20 | 65.8 2.1 | cor 12500
10.5 |f 20 — 2.4 | Cor 17500
10.8 |f 20 = 27|/wrbg 25000
Tete aie 20 | 68.4 2.8\¢’ br Bp 28000
12.5 It 20 | 70.0 2.3 | cor 15500
12.6 | f 20 — 2.6 | cor 22000
Sema ie S 22 | 63.5 2.5 | COr 20000
Secu lae aa 2.8|/wbrBg 28000
ash lak 22 | 62.5 2.6 | wrlg 24000
Cali 22 aad 2.9|wob 31000
Apr. 29] 9.5 |£ N.W.} 20 = 2.7|wbrBg 25000
9.8 |f 20) || 75-2 2.7|)weBg 25000
Toate W. 21 80.0 2-2 | COr I4000
12.3 |f 21 — 2.1 | cor 13000
5.0) | 2 ORs 26|/weBg 22000
Se Sule 23 — 2.6 Do. 23000
Apr. 30] 9.4|f Ss 22 | 68.5 2.2 | cor 14000
O-5a it 22 — 26/wcBg 22000
9.8 |f 22 a= 2.4|welg 18000
12.4 |£ 23 | 72-7 2.2 | Cor 14000
r2.5|f 23 — 2.2 | cor 14000
Beale 2 74.5 2.6 | cor|g 24000
BiOule 23 — 2.6 | w rig 24000
BASe (st 2d) 65.5 2.4 | w tg 18000
5.6 |f 24 — 2.8 | w rig 28000
5.8 — = = Do. 28000
May 1 | 9.2|f W 21 | 49:5 2.2 | COr 14000
9.4 1c 21 — 25!wcBg 20000!
12.2] 4’ (S) 2m | 56.6 25|wrbg 20000
Eo eGe |e 21 e 2.5|wbrBg 21000
4.6 |f N.W..|'22 | 55:2 228|(COL 14000
Ae ee 22 = 2.5 |welg 21000
6.0 | £ 22 | 52.8 2.1 | cor 13000
Gant 22 — 2e3) | COL 17000
May 2] 9.3 /f E 20 | 46.8 1.9 | cor 8500
9.4 |f 20 = 2.3 | cor 16000
12.0 |f 200) 510 2.1 | cor 12000
Tae ae 20 — 2.3 | cor 16000
Ben k |20 | §3.2 2.5 | cor 20000|( Clear air after
B03) | 20 — 2.3 | cor 16000 western bliz-
Bese lpe 20 | 49.8 1.8 | cor + 7000 zard.
Reon ae 20 — 1.8 | cor 7000
May 3 | 9.9|¢ S.W.}19 | 52.9 2.5 | cor 20000
LOuL IC 19 — 2.47 | COr 25000
12.4] ¢ LO) eSaeT 2.7 | cor wig 25000
12.8 }¢c 19 — 2.7\gbrBp 27000
6.3 |¢ S. |zo | 84.0 2.1 | cor 13000
6.4] ¢c 19 — 203)| (COLE 16000
May 4 | 9.0/cR’ |N. 18 | 51.8 2.2 | cor 14000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 157

TABLE 1—Continued.

g a
Bal ee B S
: se) #e| g :
3 ; ’ 3 a a Qa s s a & Remarks.
eels | & | & |8e| Ge) 8 53 5
- A & EB jadjad| < 88 Z
4 —— =
1903 |May 4 | 9-1} ¢ 18 ao 2.8 | cor 28000
g.6]c 18 2.6 | cor 24000
Tega BOM NES Sei 26/wcBg 22000
12.4 |c 19 = 2.7\/wrbg 25000
Gum uRe 2 53.3 2.5 | Cor 20000
6.2 |c 21 aa 2.7, || COr 27000
May 5 | 9.0]/¢ N TO) | 5255 2.0 | cor 10000
9-4] c 19 — 2.5 | cor 20000
12.4 |Cc 20 | 59.0 2.4 |welg T9000
12).6\|\¢ aad 2.7|wbrBg 23000
BOC 22 52.9 2.2) | COG 14000
6.0/c 22 = 2S al 25000
May 6 | 9.3 ]|¢ N. |20 | 52.0 2.2 |\COr 14000
9-4 |Cc 20 — Qt ai COL 12000
m2 KC 2h OSG. 2 2-1 (COL 12000
TSH (AC 21 — 2.6 | cor 23000
eden ie 22 59-5 2.0 | Cor 10000
6.0 | f 22 — 2.2 | COr 14000
May 7 | 9.4|f E 20 | 58.8 2.7 | w br B 27000
9-5 \f S ne || == 2.8 | w brig 30000
12.4 |f 20 | 65.5 2.7|wbrB 25000
12.4 |f 20 — 2.8|wbrBeg 28000
4-5 |c Zou 5722 2.8 | w brig 28000
4.6]c 20 aaa 2.8 Do. 28000
May 8 | 9.6|f N.W.\ 19 | 65.0 2.4 | w br Bir 18000
9-7 |f 20 = 2.2 | cor 14000
12.4 |f 2m 722 2.4 | Cor 19000
12.8 | f 20 — 2.2 | cor 14000
5-6 | f Soe zen on.3 2.2 | cor T4000
5-8 |f 2 = 2.5 | COL 21000
May 9 | 9.0/f NEES 20: «|| 70:2 2.2 | COr 14000
Oeguibe 20 = 2.2 | COr 14000
12.4 |f 21 Alte Tey) COE 6000
12.7 |f 21 = 1.7 | cor 6000
Bott ak 22 | 70.3 1.8 | cor 7000
BET hE 22 — Teal COL 6500
Beale 22 | 67.2 1.6 | cor 5000
Reon 22 = 1.8 | cor 8000
May 10] 1o.1 | f S.W. | 21 2.6 | cor 22000
TOs3) |) Li am 0255 2.8 | w brig 28000
12.2 }\\t 2r | 66 2.7 | wbr 25000
Hoe Aa eL 21 — 2.4 | cor 18000
6.2 | f 21 | 59-3 2.0 | cOr T1000
6.3 | f 21 = 2.1 | cor 13000
May 11] 9.1|f S 19 | 64.4 2.9 | g’ br Bp 31000
9-5 |f 19 = 3.3 |weB 45000
11.9 |f 20 | 70.0 2.9|w brBp 33000
TAG pak 20 — 2.9|w brBp 33000
eye per Oues5 2.7|wrBg 25000
5.8 /f 21 _ 2.8 | w olg 28000
May 12] 9.1 |f S TO) O34. 2.8|¢’ br Bp 30000
! g.2\f 19 — 3.3 |wrlg 45000

Remarks.

bp=22
bdp=22
dp= 22

158 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.
TABLE 1—Continued.
Beast ee ict .
B onl pace ° =
: : : a sce eal ee £ ag a
e | 8 |e) 48 ee ecgeel ee ee 5
a = S = EF jed/a<d] < 00 A
1903 |May 12| 12.2 |f 200723 3-1|wbrBp 38000
12.4 |f 20 oars 3.0|weBg 35000
12.9 | £ 20 = 2.8|g’|Bp 28000
5-Ont 21 64.4 2.4 | cor 18000
5-4 |f 21 — 2.6 | wrlig 22000
May 13] 9.4 |f 1g | 62.9 28/wrBg 30000
9.5 |£ 19 — 2.8 Do. 28000
12.1 | f 20. || 67.1 2.6/weBg 22000
12.5 |f 20 ae 3.0|wbrBg 35000
5.8 |f 22 | 63.6 2.8 | wrlg 28000
6.0 | f 22 = 2.5, | COLT 21000
May 14] 9.2 /f S.W.|20 | 70.1 2.9|wbrBg 31000
O- 5h 20 = 2.7|wbrBp 25000
Beaalht 22 78.5 2.3) COL 15000
Beal 22 oa 2.3 | w br B|p 16000
Cea 23 (73:4 2.6 | w rig 23000
5.4 |f 23 = 2.6 Do. 22000
May 15] 10.9 |f N.W.|20 | 72.8 2.1 | cor 14000|
Tae alae 20 aaa 2.2 | cor ‘14000
Pager 20 75.2 I.9 | cor 8500
12.9 |f 20 a 2.1 | cor 12000
4.5 |f 22 oa 2.2 | Cor I4000
4.9 |f 22 7Ou7 223 |\COr 15000
5.0 |f 22 aa 2)1)| (COT: —
5.8 /f eee 2.3 | COr 16000
May 16] 10.8 | f N.E.|20 | 66.0 7 COD 6500
10.9 | f 20 aaa 2.0 | cor I0000
LD. Sie 20 | 68.5 1.8 | cor 7000
Tote (ae 20 == 2.0 | cor T0000
BeOu ie 22 | 69.4 2.2 | cor 15000
4.0 22 = 2.5 | cor 20000
5.8 \f 22 | 63.8 1.8 | cor 7000
5-9 |f 22 aa 1.8 | cor 7000
May 17] 3.3/4 S.W.|20 | 82.0 1.8 | cor 7000
Bette 20 — 2.0 | cor I 1000
Seoul 20 a= 1.8 | cor 7000
BxO ae 20 = Des GOL —_
Agee 20 — T.78| (COG =
May 18] 9.0/f W. |21 | 78.0 26/wrBegr 24000
g.2|f 20 — 2.7 |wrlg 26000
95h 20 — 2.3 | cor 17000
12.2) |) t N. 22 | 86.0 2.0 | cor 10000
12.3 |i 22 — 2.0 | cor 10000
12.4 ||t 22 = 1.9 | cor 8500
4.6 | f N.W.| 24 "| 87.0 1.8 | cor 8000
4.8 | f 24 = 1.6 | cor 5000
say pe 25 82 2.4 | cor 18000
5.8 |\f 25 aa 2h Mi COL 20000
May 19 |11.9 |f N.W.| 25 89 2.3 | cor 16000
12.4 |f 25 = 2.1 | cor 12000
12.6 | f 25 go 2.0 | cor 10000
D2eO) ee 25 — 1.6 | cor 5500

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 159

TABLE 1—Continued.

8 BS
B4| Bs 4 S
2 ee be | § a 3 Remarks
“ 3 a 3 os aH a3 £ 5 yg 4 emarks.
meee | © | 2 jee) BE) & ae 3
a qa & 5 =e lad] ac < 386 5
1903 |May 19] 2.0}c 25 aa 2.2 | COr T4000]
2.8/c 25 — 2.0 | cor I1000| Storm coming.
Bron Cc 25 — 1.8 | cor 7000| Thunder,
Bakalc 25 84 1.8 | cor 7000
3:8) |\C 25 == 2.4 | cor 18000
4.0/cR a5 | 78 1.8 | cor 7000) Hail.
Ase at S)) 25 76 2.0 | cor IIO0O
Se) |] ai? 25 70 2.3 | w br Bip 15500
Geli 25 = ZeTaGOr 11500
May 20] 9.4/f N.W.| 23 78 — |cor —
Orsi iit 23 a 2.7 | w br 25000
9-9 |f Ss. 23 — 2.5/wr 20000
10.0 | f’ 23 — 2.8 | wrlg 28000
Tse |i 24 80 2.2 COG 14500
ORAM (ly 24 ae 242) |||COE 14500
Sera 26 82 2.3 | cor 15500
3.4 |£ 26 = 255 KCOL 20000| Thunder.
Bank 27 70 2.9 | wrlg 31000
T5e(0) || 18S meet = 2.6 | wryg 22000
May 21| 9.7 |f N.W.| 24 79 ae a =
g.0 | f 2 aa 2.1 | cor II500
12.5 | £ 25 82 2.0 | cor I 1000
12.00) if 25 = 2.4 | COr 17500
ono. et 25 a 2.2|g¢ br Bip 14500
3.0 | f 26 80 2.5 | w br 20000
een |e 26 = 2.4 | w br 17500
4.8 |f 26 76 2.4 | cor 17500
4.9 |f 26 — 2.6 |wrlg 22000
May 22) 9.3 /\f W 24 73 2.6 | wrig 23000
9.9 |f 24 = 2-3 | Cor 16500
T2RON| i B 83 2.2 | COr 13500
mene | it 25 = 2.5 | cor 20000
Ziegh | \ak 28 82 2.0 | cor 10000
4.9 |£ 27 = Dea COG 11500
5-9 |f 28 80 2 TahCOT I1500
6.0 | f 28 aa 2a COG 11500
May 23] 9.2 |f N.W.,) 22 64 lost — —
9-6 | f 20 — 2.9|/2e’Bp 31000
9.9 |f 21 aa 2:9) | SAE Bp 25000
. Teneo} eh 2 69 240) ¥/ (br Damen 31000
weeAn if 22 69 2.6 | wrlg 22000
Te.) | 22 a 2.4 | w br 19000
2.8 | f 23 69 2.4 —_— 19000
Bese 23 — 2.8 | wrlg 28000
5-8 | f 23 65 2.2 | COr 13500
Seon line 23 — 2e2) COL 13500
May 24] 9.7/c N. E.| 22 57 Taya COT 6000
MOe teal ity 22 —- 2.0 | cor 10000
meee 22 62 2.0 | cor 10000
TPS lek 32 _— 2.1|/wrg 11500
6.0 | f 21 56 2.0 | COr 10000
6.1 | f 21 — 2.3 | cor 15500

=o.

160 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

: Be) ea] 8 :
- c 5 a & ge| 82 é a 3 Remarks.
3 is g $ Bo) ee Bey ise ae E
o s S SS is GO|) oF oo ome) S
a a a = Fe |aaq | aa <q OO a
1903 |May 25] 9.3 /\f N.W.| 20 64 2.7 | wr|g 27000
Onsale |N.E. | 19 = 2.7|wob 26000
mason Ni 21 66 2.1 | cor II500
12.6 |f 20 aa 2.1 | cor 11500
GeO) | 22 59 24|/wrBp 17500
Seale 22 = 2.4 Do. 17500
May 26| 9.1/f S 19 59 2.1 | cor 11600
9.4 | f 19 =— 2.6 | wrlg 24000
TOszuiih 19 = 2.6 Do. 24000
L220 | £ 20 61 2.2 | COr 13500
12/2; |b 20 = 2.3(|wbrBp 15500
12.4 |f 20 = 2.2 | Cor 13500
3 4h 20 61 2.2 | COr 13500
Sera 20 2.3 | COr 16500
Gest 2I = 2.5 | cor 21000
5-6 |f 20 58 2.5 | COT 20000
May 27] 9.2/c¢ S. |20 | 64 2.7 | wrlg 25000
9-4 | Cc 25 = 2.9|wrB 31000
10.0 | c 20 el 2.6 | cor 22000
L222 iT 72 2.6 | cor 22000
12.3 |/c 21 = 3.0|wbrBp 35000
G2) Sul | 21 — 2.7 |wrlg 27000
200 lax 70 2.3|wbrBp 15500
Be TAIG 2I — | — = a
5.8 | i 22 | 68 2.3 | w br B]p 15500
6.0 | f 22 = 2.3 Do. 16500
|May 28] 9.6] c’ S. W.| 21 | 72 ?2.6|¢’ br Bp 222000
9:9: 2t | — ?2.6 Do. ?22000
Tata Ould 22 | 68 2.8 | y’ rig 28000
12.0|c 22 — 3.0/9’ br Bp 35000
Beaae 23 69 2.7 |wryg 25000
3-5 | ¢ 23 = 2.7 | We 26000
May 29] 9.5 |f S ar | 71 2.7 |wrlg 25000
10.0 | f | 20 = 2.8 | y’ or bg 30000
Tao) | 22 | 76 2.4 | Cor 17500
12.4 |f 22 = 2.6 | wrlg 22000
A. Fe |24 | 71 2.3 | w br Bip 15500
Be onl 230 iar 2.5|wbrB 20000
6.0 | f 23 — 2.7 | w rg 25000
May 30] 9.8/cR’ |S 20 | 56 2.0 | cor 11000] Rain at night.
TOs vC 21 — 2.0 | cor 10000
10.8 | c 21 — 2.0 | cor 10000
T2-Ag|KC 25 60 1.8 | cor 7000
2). Gite 20 — 1.8 | cor 7000
5.8/f 23 On: 2.2 | COr 13500
6.0 | f 23 | — 2.2 | COr 13500
May 31] 12.8/f N 20 | 66 1.7 | cor 6000
Lea et 20 = 2.0 | cor IOO00
6.4 |f |20 | 60 Ta7akCOr 6000
6.4 |f 20 | 60 Te 7| (COD 6000
Oxsulit 20 —_ 1.9 | cor 8500
Jamey eras sack N.E.|19 | 68 2.0 | cor 11000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 161

TABLE 1—Continued.

B eo
BY leg 4 S
s Bere aly ce .
; ‘ Ss = ge | 82 3 3 yj & Remarks.
és | 3 E 8 | &.|/S&/ee] § es
ep} @ |e| & |E jétiée] 2 88 Z
es aa e
1903 |June x | 12.7] f 5 Oe 2.3 | cor 16500
4.6] f 2 |G 2:2 |S -br Bp 13500/Maine forest fires.
BeAl oe 2 | 2.6 | wrlg 22000
5.6| f 21 |64 2.6|wbrBgr 22000
June 2] 9.4|f N.E./19 |— lost | small cor —
Oe7 \iek 19 | 67 1.7 | cor 6000
12.2) f 20 | 69 2.5 | w br 20000
12.4| f 20) | || 2.8 | wrlg 28000
oyfa| 28 S 22 |(65 2.7|/wrlg 25000
6.0) f 22 2.7 |wrig 26000
7.6| f 22x 223 ||| COT. 15500
Wao) |e DE 2.3 | COr 15500/ Night.
10.2] f 2r |'56 2.4/wreg 17500| New revoluble
June 3 | 9.3] f W 20 |72 2.0 | Cor r1000| mounting.
9.4| f 200 = 2.0 (COL T1500
r2).2|| £ 24 |81 1.9 | cor goo0o
12.6| f 24 |— 2.0 | cor 10000
eal el 25 |83 2.0 | cor 10000
Bea) ee — |— 2.0 | cor 10000
Ee7 ale 25 |80 2eG COD 20000
REO 25 |— 2.9 | wrig 31000] Red moon.
June 4 | 9.0] fog N. 23.1 | 62 3.0 | wrlg 35000| Red sun.
OsSil es 23.1 |— 2.9 | w brig 31000
g.8 |red sun agen |'02 2.8 Do. 30000
12.0 24.1 | 66 2.5 | w br 21000
12.2 |fog, yel, 21.1 |— 2.4|wbrBp 17500
2.4|red sun 22.1 | 67 2.4|wbr Bp 19000} Forest fires on
2.5 fog 2250 || — 2.5 Do. 20000| Cape Cod.
red sun
2.6 | fog Dye) | 2.4 Do. 17500
red sun
g| fog 23.1 | 63 2.3 | w br B| 16500
6.0| fog Bartel — 23 Do. 16500
June 5 | 9.5] haze, 19.1 | 60 3.0 | wl|Bp 35000
sun yel.
OejalmDosy Ne Be cS. | — 2.6 | cor 22000| Forest fires on
9-7| Do. ie {| 258 COT, 20000| Cape Cod.
neal] — IDYo 20.1 | 68 2.0 | cor 10000
12.3| Do 20.1 | —— 1.8 | cor 7000
Bey Do 22.1 | 68 2.2 | cor 13500
3-8| Do. 2) Te | 222 (COLT 13500
5o53||. IDYoe 21.1 | 65 2.4 | COr 17500
5-7| haze |
sun 2201) —= 2.5 | cor 20000
ruby
June 6 | 9.4] haze |S.W.| 19.1 | 64 2.8}¢ br Bp 28000
9.5| haze LOer | — 2.8 | wrlg 28000
12.6] c 22.1 | 68 2.2 | cor 13500
E29) € OPE ee 2.5 | cor 20000
Bic 22.1 | 71 2.4 | cor 17500
azine 2201 ||—— 2.6 | cor 22000
3.9] c 22.1 | — 2.4 | cor 17500]

162 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

2 Eo
paleo Z s
5 fe )fea1 2 5 oe
; 3 s = as Q 2 S s ai Oo emarks,
| 2/2] 3 |EIER/BE| 2 | Be | 8
os A = = B ledq|az| < 88 Z
1903 |June 6 | 5.8]c 22.1 | 69 2.3|wbrBp 16500
5:9) (C 22.1 | — 238 Do. 15100
June 7 | 10.3 | R’ S. 20.1 | 65 2.7|wrg 25000
10.8 | R’ 20.1 | — 2.0 10000
i2.0|(¢ R’ 21.1 | 66 2.2 | cor 13500
D2 akc Qe hy | 2.0 | cor 11000
4.6 | f’ 21.1 | 68 2.0 | cor 10000
4.8 |f 20.1 \\—— 2.1 | cor T1500
6-13) £ 21.1 | 66 2.1 | cor 11500} New axis and
6.3 |f 21.1 |— 1.9 | cor 8500) support.
June 8 | 10.4} c¢ S 20.1 | 72 2.1 | cor T1500
DL £3} C 21.1 | — 2.5|y’ br Bg 21000
LE 42 h¢ 21.0 |73 2.3 Do. 15500
T2e51e 22.1 |7s 2.0 | cor 11000
12.8/¢ 21.1 | — 24|woBgr 17500
2 Suc 22.1 | 70 2.1 | cor T1500
2 7Ke 22.1 | — 1.9 | Cor 8500
BTC 22.1 | — 2.2 | cor 13500
6.8, R 23.1 |67 2.1 | cor 11500
6.9 R 23.1 | — 2.2 | cor 13500
June g | 9.0|c wet |S 21.1 | 68 2.3 | cor 15500
O.2))/¢ 21.1 | — 2.1 | cor 11500
12.4/¢ 22.1174 27\y'rBg 26000
L255; |\C 22.1 | — 2.9 Do. 26000
2. e 23.1 |72 2.1\/y’rBg 11500
B-Silte 23.1 |— 2.3 | w br Bir 15500
Se Sec 23.1 |67 2.35 |(COG 15500
5-6 ]e Sail 2-2) || COF 13500
June ro} 9.3 /c S. 21.1 | 70 1.4 | cor. small. 3000
9:5: /€ 21.1 | — Ta5 Do 4500
12.0|c 22ea ns coy Do 6500
12-2; |\¢ 22.1 |— Teo) Do 7000
3-4 ]¢ 2B eC |7a 2.0 | cor 11000
3. OIC 23.1 | — 2.3 |wo Bir 15500
5-8]c 24.1 |67 1.8) | Cor 8000
6.0 |¢c 24.1 |— 2.0 | cor 11000
O-5 ie R 22.1 | 70 1.8 | cor 8000
June rr | 9:6) R E. 22.1 |— 1-7) | cor 6000
L2- 251C 22.1 | 70 1.9 | cor 8500
T2-3ahe 22¢1, | — 1.9 | cor 8500
4cOWe 23.1 | v5 1.6 | cor. small 5000
4.1 /c¢c 23.1 | — 1.6 | cor. small 5000
6.1 /c 24.1 | 67 Ley (COL 6000
Or2e ie 24.1 |— 1.7 | cor 6000
|Juner2] 9.2/R E. 22.1 | 69 1.3 | cor 2500
9-4|R 22.1 |— 1.3 | cor 2500
12.7) | 23.1 | 70 1.4 | cor 3500} Violent wind
12.8|/R 23.1 | — 1.6 | cor 5000| storm.
Ase ne! 23.1 | 66 1.8 | cor 7000
4.8] c S 23.1 |— 1.7 | cor 7000
6.0 }¢ 24.1 | 65 1.7 | cor 6500
6.1 ]¢ 24.1 |— 1.7 | cor 6000

| Year,

1903

Date.

June 13

June 14

June 15

June 16

June 17

June 18

June 19

June 20

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 163

TABLE 1—Continued.

Be . | a9 ‘
See. al) wee S
5 fa/fa] § 5

a cra d ad | ae 3s vi 2 Remarks.
Pee 2 eelee| 2 | es i

& = Ee laedl/az| < 88 Z
9.6| R 8. 21.1 | 68 222n\nye tS) Sp 14500
9-9| ¢ ee |e 2.6|wbrBegr 22000
10.4| C 208 | 2.9|wbrBp 31000
TEeO) |EC 22D | 97/2 2.2 | w br 13500
12.2| f’ (S) 22 2.2 | w br 13500
4.6| £ (S) 23.1 | 70 2.2 | y olg 13500
4.8| f AR It | —— 2.4 | w br b|p 17500
Bevan 23.1 | 68 2.4|wbrBe’b 17500
6.0} f gio | —— 2.1|}/wrbg T1500
10.0] f S. 21.1 | 68 1.8 | cor 8000
gia) 38 22.1 | 70 1.9 | cor 8500
m2) |e ei | 1.8 | cor 8000
AeaN 21.1 | 69 1.3 | cor 2500
4.8] f itgai || La3ui\COr 2500
Ora} |eE 22.1 | 66 1.7 | cor 6000
6.4] f£ 2 ere L7) | Cor 6000
TOs | Cuv N. 19.1 | 58 2.1 | cor 12500
10.4| R 19.1 |— 2eTa||COE 12500
TTS ale en 18.1 | 58 2.1 | cor 12500
Tete 7 eC Me || 2.1 | cor 12500
Aer 22.1 | 56 1.8 | cor 7000
4.8] c 225 1)|\—— 1.7 | cor 6000
6.3| R’ Boor IGG 1.6 | cor 5500
6.5 R’ 22.0) | —— 1.6 | cor 5500
9.5| R’ N. 19-1 |'52 1.7 | cor 6000
On| RY TOL | 1.8 | cor 8000
12E 3G TST} 515 1.9 | cor 8500
T2)4"|\(C 18.1 | — 1.9 | cor 8500
6.5} ¢ LET TESS 1.8 | cor 7000
Ouny| et E. 17.1 | 64 3.3 | w ple 45000
9-3} f iy fole | 3-4 | w pig g’ br 49000
4.4| f’ (S) 18.1 | — 2.2 | COr 13500
4.4| f 18.1 | 65 20 ||\COr T1500
9-7| c E. 17-1 | 60 2.3|wbr Bglr 15500
10.1/ Cc OT | 2.4 | COr 19000
TT O| eC 17.1 | 65 2.3 | cor 15500
1220) C Te Fao | 2s COw 15500
4.4| ¢c S. 18.1 | 62 2.5 | wt be r 20000
4.8] c 18.1 | — 2.5|wrBeg’be 21000
6.0} ¢ 19-1 | 59 23) COT 16500
6.2|c¢c 1O}s3 || —— 2.4 | Cor 17500
9.6] ¢c S. Teer |, O2 2.4 | COr 17500
9.8] c Te 2.4 | COT 17500
12.8] c 18.1 | 63 2.5 | w br B g’r 21000
12.9/¢ — |— 2.7 |welg 25000| Fire on campus.
5-8|c 19.1 | 61 2.0 | Cor 10000
6.0| c LOSE || — 22 Ty (COL 11500)
g.8/ c S2 7k) | OS 2.2 | cor 13500|
9-9] ¢ 70 | 2.2|woBlp 14500
12).8)|| 1c 19.1 | 69 2.3 | w br Bip 16500
12.9] ¢c Tost | —— 2.5/wbrBeg’r 21000
anc 19.t | 64 1.9 | cor 8500

164 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.
TABLE 1—Continued.
2. Wes :
S9/38 2 s
' 3 : 5 E oe 3 S 8 Remarks.
¢| £18) € | 3 Beles) a ee :
a a a = B ladile<| < oO Z
ere
1903 |June 20] 5.8| c¢ 19.1 | 63 2.5| cor 20000
6.0] ¢ 19.1 |— 2.4| w br B|p _ 19000
June 21] 10.5] R E 18.1 | 62 1.8] cor 7000
10.6] R 18.1 | — Teal COT 6000
T222)|bG ue 18.1 | 62 125 | COE 4000
T223) | sCuRe 18.1 | — 1.5] cor 4000
An 3i| nek 18.1 | 60 1.8) cor 7000
4.8| R Oe 1.8] cor 7000
6.4) R 19.1 | 58 1.8| cor 7000
6.6} R 19.1 |— 1.8| cor 7000
June 22] 9.5|cR’ |N.E.) 17.1 | 57 e7)|\ COG 6000
9-7| c 7 1.7| cor 6500
£2:0)|'C 18.1 | 59 1.8| cor 7000
2201|C 19.1 | 60 1.9 | cor gooo
5.6] c 19-1 | 59 2.8| w|Bp 30000| Fire on campus.
SOC 1g. | — 3.6} wrg OOO) sae aie us
June 23| 9:6) ¢ E. L7ak |'159 1.6] cor 5500
9-8] c 17-1 |— 4.1| gBp 100000| Fire on campus.
L255) [1c 18.1 | 58 1.8| cor 7000
1.0] c 18.1 | — 1.8| cor 7000
5-8] R 19.1 | 54 2.0} Cor 10000
6.0] R 19-1, |— 1.9| cor 8500
June 24] 9.8] c Ne Eee 55 1.6| cor 5500
9-9] ¢ 17.1 |— 1.8| cor 7000
TER fae 18.1 | 57 1.6| cor 5500
5-8] c 19.1 | 54 1.8) cor 7000
6.0] c LO. |j—— 1.9| cor 8500
June 25] 9.6} c N. LOst | 5H 2.1 COL 12500
LO: 7) (C 16.1 | — 2.0] cor 11000
12.8] c 17.1 | 59 2.3] wo Bg’r 15500
12.9] C 17.1 |— 2.5; wbrBegr 20000
2.7 Cc 18.1 | 60 2-1) COE 11500
3.8] c 18.1 | — 2.1]; wobg 11500
CAI Che 18.1 | 59 2.3| wbrBg 16500
6.0} c Tet ||—— 2:3 Do. 15500
June 26} 9.5] f (S) 16.1 | 70 2.4| y’brBg’r 17500
9.8| f W. {17.1 |— 2.3| wbrBegr 15500
2.2) 18.1 | 74 2.8| w orlg 28000
12.4| £ (S) 18.1 | — 2.8| worbgr 28000] Fire on campus.
4-34 20.1 | 76 2.5| wbrBp 21000
6.0] f 20.1 | 73 2.3| wbrBgr 15500
June 27} 9.7) f W. | 18.1) 75 1.8| cor 7000
9.8 | f 18.1 | — 2.4| w br Bg’r 17500
12.0| f 19.1 | 81 2.5 Do. 21000| Fire on campus.
4.4| f 20.1 | 80 2.7; wbrbeg’r 25000
6.4| f 2z.t |\75 2.3| w brig 15500
June 28] 10.3} £’(S) |N. 1g.1 | 72 3.0| y brig 35000
10.8| f 19.1 | — 2.8 Do. 30000
Tee ae 20.1 | 75 2.3| wbr Bg’p 16500
TeAul ib 20.1175 2.3| wbr Bg’p 16500
5.0| £ 5. 20.1 | 69 2.2| w brig 14500
6.8| f 20.1 | 65 1.9| cor 8500

A CONTINUOUS RECORD OF ATMOSPHERIC

TABLE 1r—Continued.

NUCLEATION.

165

Sl) eno /
Pu = 5 4 mS
eG eaeeal eee 2 a
: : a ; ge 3 3 Sy & Remarks.
eee ie | 8 | 2 ee ea! & e8 E
ea |e| = |e lét 6a) 2 388 Z
1903 |June 29| 9.6) f% (S) 18.1| 70 1.8 | w br Big 8000
HEA)! £4 E. 20.1| 72 1.7 | cor 7000
3-4] Cc =] Ws) 1.7 | COr 7000
Feoll CPi 21.1] 61 (2.2)wbrBegr) 13500
(2.2 | wolg j
June 30] 9.6] f’ N. 19.1} 68 2.9 | g’|b’p 31000
9.8 | f’ 19.1 | — 2.9|wbrbgr 21000| Runs to w r g for
large d.
10.0} f” 19.1 | — 2.9/2’ bp 31000] Fire on campus.
Toei ete 20.1| 72 2.9 | wr|g 31000
Bai liak 2s es 2.6|wbrB 22000
Ga || ak Ss. 22.1) 71 2.7|wrlg 25000
Sebi) aL 22.0 || — 2.8 Do. 28000
July 2 9.3| R S. 2051 || 72 sen eabep 38000
TOES) |iee ert N77 3-1 | g’ br Bip 38000
12.7 | £/ (S) 22.1] 85 2.4 | w br Br 17500
253) | 22.1| 87 2.4 Do. 17500
Aral 23-1| — 2.9 | w olg 31400
y (2.9 wBp)
Bes LS) 24.1| 87 lao lwre § 31400
July 2 | 9.5/f We. | 22.0) 87 3.0|g’ Bp 3.5000
9.7| £ (S) 228th 3eO Do. 35000
Ts 23-1] 90 2.4|wbrB gir 19000
TS) tek 24.1] 92 2.5 | w br Br 20000
Angi let 25.1] go 2.7|/wrg 26000
Bestia 25.1| 89 2.4 | COr 17500
July 3 Oe7alee We 23-082 2.7|wrg 26000
TOs) | L 23 tl O2 2.3 |wbrbg 15500
1.0| f 24.1} 84 2.3 | cor 15500
2.8/ f£ 25.1] 85 2.3|wbrB g’r 16500
4.8) f 26.1| 84 2.3 | w brig 15500
2.7|/wrg
EEG) |e 26.1| 82 173 reel 26000
July 4 g.8| f N. 23.1] 74 2.5|wbrB g’r 20000
repro |e 24.1] 78 2.2 | COr 13500
12.6| f 25.1| 79 2.0 | Cor 10000
3.0| £ 25.1| 81 2.2 | w brig 13500
5.4] f —| 71 2.0 | cor 10000
July 5 OFAN ia (S)) | Sawin |!23i0 | 74 2.2|wbrbgr 13500
11.8] f 24.1| 78 2.2|wylg 14500
Goals te DAT ae 77 Len COL 6000
6.4| f Qari) | 73 1.7 | cor 6500
July 6 9-3| R S. W.| 22.1] 66 2.6|wbrBgr 22000
10.3] R 23.1| 66 2.9|wBp “| 31000
120)" RY 23.1| 68 2.5 | w br Bir 20000
6.0} f 25.1| 71 2.3 | w olg T5500
July 7 Ouse N.W.| 23-1] 79 2.7|wrg 25000
a2) 24.1| 83 2.5|wbrBgr 20000
neers 24.1| 85 2.1 | cor 14500
BeSi lates 25.1] 85 2.3 | w br B g’r T5500
eee 25.1| 84 2.1 | w brig 14500
July 8 9.8| f W.. "23.1 | 83 2.2 | cor 13500

166 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

vo oO.
B4|25] 4 s
H Sp a 2 re
& d a8 3 o 3 ey & Remarks.
go] SE) ea ee eee ie 88 g
o 5 = om | 8 a eo 3
m A S s EB |ed/e<|] < oo Zi
ie
1903 |July 8 |11.9| f — | 87 2.2|wbrbgr 13500
Beale 26.1 | 89 2.4 | w br Bir 17500
f G@allise 6|/wrbg
57 20.1 hese w rig 22000
July o)| 19-71\ N.W.| 24.1 | 85 1.9 | cor 8500
Lowonhot,clear
12.1] f 26.1 | gt 1.9 | cor gooo { oe
3:0) | £ 27.1 193 Ta5 |\cor 4500
5.8| f 27.1 | gt 2.0 | COT 12500
6.0] f 28-108 2.1 | w brig 12500
July ro} 9.2} £(S) |S. — |88 2.9 |g’ Bp 31000] Old apparatus.
9-4] f 26.1 |— 2.9 | g’(br) Bp 31000|| “
9.9| f 27.1 | go 3.0 Do. 35000| New apparatus.
TOS) el 27.1 |— 2.8] wrlg 28000
Tie 3h | ee 27.1 | 93 2.8|wbrBgp 28000
12.4| f' 27.1 | 94 2.8 | wrlg 28000
12.8| f’ 27.1 |— 2.8 | g’ (br) B|p 28000
20h 28.1 | 92 2.4|w br Blgr 17500
BaD \c 28.1 | 87 2.8] wrilg 28000| Coming storm.
5.3| c R’ 28.1 | — 2.8|g|Bp 30000| Blown over.
July 11} 9.5) f 26.1 | 84 3.6 | wrlg 57000| Empty of water.
Onze W. | 25.2 |— 3.0 | w br Bip 35000] Water put in.
9.9| f 25.1 | — 3.7 | wrlg 61500
10.4| f 26.1 | — 2.8 | g’|B|p 30000
10.9| f 26.1 | 88 2.4|wbrBgp 17500| Old apparatus.
July 13) 9/6).c RR! Wie 26m i74. 2.3|wbrBg’p 16500
9.8| R’ 26.1 | — 2.3 Do. 16500
12.9|cR 27.1 | 74 2.3 | w br B|p 16500
3.81 ¢ Ziel 73 2.0 | cor 11000
Sa7ale 27s | 73 3.0] w br B 35000
5.8| Cc 27.1 |— 2.8! wrlg 28000
July 14] 9.4] f (S) 25.0 1176 2.4 | cor 17500
9.8| £/ W. | 24.1 |— 2.4|wbrBg’r 17500
11.2 |) £ 25.1 (77 2.7 |wrlg 25000
P2.5)|\£ 26.1 | 78 2.7 | w br B|p 25000
2.9| £ 26.1 | 79 2.4 | cor 17500
4.3] £ 205) 77 2.3|wbrBegr 15500
Sead 26.1 | 74 { e. ee 28000
July 15] 9.8] f W. | 23.1] 72 2.2 | w br Bir 14500|Rain during night.
10.6) f 23.1 | 74 2.4|wbrBgr 17500
PTO) 24.1 | 74 2.8|/wrilg 28000
Are 3i ips Re 25.1 | 69 2.2|wobgr 13500
Semit 25.1 | 68 2.3 | w|B\p 16500
July 16| 9.3] f W. |22.1]70 2.5 | w br Bir 20000
1223) \\h 23.1174 2.4 Do. 19000
5.6] f 23.1 |\72 2.2 | w br Bir 14500
July 17 | 11.4] f W. 79 2.3 | cor 15500
b2.2))) f 22.1 | 78 2.2 | COT 14500
5.6| f 22st 79) 2.5 | cor 20000
July 18 | 10.9| f” S. W.| 22.1 | 76 2.2 |g’ br Bir 13500
Te eae 23.1 | — 2.2 Do. 14500

| Year.

1903

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 167

TABLE 1—Continued.

July 20

July 21

July 22

July 23

July 24

July 25

July 26

July 27

Seal ee :
Ia nee bad S
: eo lee |) 2 :
. a cs Sears 5 Ss 8 Remarks.
E oP seen Be |) Ss, a
& = Ee ledi|adq| < OO Za
T2eA ie 23.1 | 76 2.3 | w br Bir 15500
403) 6 RR! 23.1 |70 1.8 | cor 8000
Sei5t liek 23.1 | 67 2.2 | cor 14500
10.4| R’ N. 22.1 | 66 1.5 | cor 4000| Rain all night.
eet Xe 23.1 | 66 1.4 | cor 3500
Benne 23.1 | 68 I. | cor 1800
SeOi|ac 68 1.5 | cor 4000
Tie 2 |e N.W.| 21.1 | 74 2.6 |wrlg 24000
025 |e 22st —— 2.5 | w br Bir 20000
12.9| f 22.1 |74 2.8 | w olg’ 30000
TO) 2200 |= 2.8 | g’ br Bir 28000
Ae Taek agin p72 2.6|wrbrBgr | 22000
Basia 23.1 | 70 2.5 | w br Bir 20000
9-7| f N.. B.| 22.1 | 72 1.4 | cor small 3500| After cloud-burst
10.3] c 22.1 |— Te5) | (COr 4500| and thunder-
TON 230 170 1.6 | cor 5000] storm.
TAO} EC 23a8 | 73 E.7) |(COr 6000
Ba 7h CLS Dyer NG os 1.4 | COr 3500
BIOENG 24.1 | 70 aly | (he 4000
9.9} ¢ iN. 22.1 |70 1.5 | cor 4500
11.8] f S. — | 76 2.0 | cor T1000
3.0] f 24.1 | 76 2.4 | w br b|p 17500
3-4| £f 24.1 | 76 2.4 17500
GaSilEC 24.1 | 74 1.8 | cor 8000| Storm coming--
10.4] f Se Blea n 1) 77 2.3 | w br Br 16500] violent rain
Tere fale 22.1 |— 2.4|wbrbgr 17500| storm.
Te2RONCY 23.1 | 74 2.4| Do.B 17500
BEo}|EC 24.1 |77 2.1 | wrlg 11500| Storm clouds.
5-8] f’ (S) 24.1 | 77 2.2 | w br b’ g|p 13500
7 2.7 |wrlg
9.0] f 24.1 | 72 as ee 26000
10.3 | f N.W.| 22.1 | 79 2.3. | COT 15500
11.0| f D2 ee 2.2|wbrBe’r 14500
1.0| f 24.1 | 82 1.7 | cor small 13500
Batali 24.1 | 83 2.2 | cor 13500
5-9| f 25.1 | 80 2.0 | cor 10000
TAK) |e S. W.| 22.1 | 83 2.3 | w br Bir 16500
1 2e47\ 23.1 | 84 2.1 | cor 11500
a8) nik 24.1 | 86 25|/wrBg 20000
Anish 24.1 | 86 2.5|wbrp Bir 20000
Sasi let 25.1 | 84 2.2|wrbrBer 13500
a7 le W. | 23.1 | 80 2.2|wrp Beglr 13500
10.8] ¢ 24.1 | 84 2.0/wrpBg TOOOO
1.0] c 24.1 | 86 1.8 | cor 7000
4.4|f 25.1 | 86 1.4 | cor 3000
aula 25.0) ||— 1.4 | cor 3000
eraled 25.1 | 84 T.7-|| COT 6000
1.5 | cor 4000
ae Ziscty | 47 1.5 | cor 4500
10.4| f N.W.| 21.1 | 68 2.4 | w br Bir 17500
Toe) lt 22.0) | 7/2 2.4 | w br Bir 17500
35 | a Zee 73 2.4 Do. 17500

168 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

& Eo
Ee lee a S
5 Bh eau! was 5
; ’ 3 — ge ga zs 3 P 3 Remarks.
g| € || 8 | 8 |e) eels Be 5
> A a) £ B jadjae<| < OO Z
1903 |July 27] 5.9|f 23.1 | 70 2.3 | w brlg|r 14500
July 28} 9.8} f N.W.| 21.1 | 70 2.9|\g’brBegr 31000
11.0| f 21.1 | 72 2.6 | wrlg 22000
reba eh 22.1 | 74 2.3 | w br B glr 15500
3.0| £’ (S) 23.1 | 76 2.3 Do 16500
pea en 23.1 |74 2.6 | wrlg 22000
6.0| f 2360 2.6| Do. 22000
July 29] 9-5|¢ S. W..| 22.1 | 75 2.5|wbrB 21000
12.3] Cc 24.1177 2.2 | wtlg 13500
22r1\'e 24.1 | 78 2.3 | w br B\p 16500
Gail |x 25.1 |75 2.3|wbrBg 16500] Storm coming.
July 30] 9.7) f S. W.| 23-1 | 84 2.8 | wrlg 28000] Afternight’srain.
10.3] f£ 24.1 |— 2.8|wbrBp 30000
12.9| f 25.1 | 89 2.4|w brB 17500
Broun 26.1 | 87 2.3|wbrBg 15500
4.9| f£ 26.1 | 85 2.2|wbr Beglr 13500
Gacy i 27.1 |— 2.1 | wrlg 11500
July 31] 9-5) f W. | 24.1 | 75 2.1 | w brb 14500
reaee(o)|| 38 24.1 |78 2.3 | w br Bg|r 15500
3.0)| 17 25.1 | 76 2.3 Do. 15500
4-4] Cc 25.1] 72 2.2 | COT 14500
5.6] ¢ 25.1 | 71 2.1/wrbgr 11500
Aug. 2) 974d N. 22.1 | 70 1.7 | cor 6000
Quyalne a 23) ||COL 15500
12.4] f 23.1 | 74 2.3 | w br Bir 15500
Selah 24.1 | 76 2.1|wbrB gr 11500
BAG ee 25.1175 2.1 Do. 11500
9-3] f 24.1 | 66 2.4 | w br B|p 17500
Aug. 2 | 9.7)\\f N. 23.0 73 1.7 | cor small 6500
11.4| f 23.1 | 70 ‘1.9 | cor 8500
1.4| f 24.0 178 1-42) COL 6000
4.8| f£ 23.1 | 78 1.5 | cor 4500
6.0| f Beer is 1.3 | cor 2500
O51) : 23-10) 70 1.8 | cor 7000
9.8| f 22 eta 1.9 | Cor 8500
Aug. 3 | 10.8] c’ N. 22.1 143 3-2)/wrB 44000
£2-14)| .¢ 23.1 | 76 2.3 | w brig r 15500
27 Ge 24.1 | 76 2.2 | cor 13500
4.8| c! 24.1 | 72 2.0 | w br Bir 10000
Brora 24.1 | 70 1.6 | cor 5000
Aug. 4 | 9.6] c¢ N. E.| 22.1 | 66 2.2 |w br bg|r 13500
12.3 |¢ 22.1 | 68 2.2|wbrbgyr 13500
3-0 Cre, 23.1 |65 2.1 Do. | 12500
5.8| R 23.1 | 62 2.0 Do. 10000
Aug. 5) | 12.3) E 21.1 | 60 1.2 | cor small 2000
1.0| R 21.1 | 60 1.5 | cor 4000
A ICURY 21.1 |— 2.0 | cor T1000
6.0| c¢ 21.1 |60 1.5 | cor 4000
Aug. 6 | 12.3] ¢ N. E.| 20.1 | 69 1.4 | cor 3000
12 0C 20.1 | 68 1.1 | cor small 1800
4-4} Cc 21.1 | 67 2.1 | cor 11500] Thunder.
See Bite Tes O15 2.0 | Cor 10000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

169

e 2 |eae ;
: Ba) | 59 2 =
b 85 3a oO 7
a Baer Seal) oun’ 5 3 g Remarks.
KB 3 3) Ss se) Aad | ao a ge I
oes | fe | Se ee | bE | 8 ES 3
oe Q - = FE ladl/aq| < OO Z
1903 |Aug. 6 | 6.0/ cR 22.1 | 65 2.3 | cor 15500
Aten 7 oF c 9. W.| 20.1 | 72 2.0 | cor T1000
12.5| f’ (S) ame || 717 1.8 | cor 8000
BEBE IGS) irae O77 2.2 | wrlg T1500
4.9| £ 21 | 74) 2:0 | COT 10000
6.0| f ain | 9 2.1 | cor 11500] Clear, windy.
Aug. 8 | 9.5] f N.W.| 19.1 | 65 2.8 | wrig 28000
9.9| f 19.1 | 67 2.8| Do. 28000
11.0| f 19.1 | 70 2.9|wbrB 31000
rerigky || ae LOAD 7,0 2.6 | wrlg 24000
BeBe: 21.1 | 68 2.3 | w br Bir 15500
523 || £- — | 66 2.4 | w tg 19000
NUS =EO he |) 10-0)|) COR! 20.1 | 62 2.0/corwbrbgr| 10000
sp AS} | NC 20.1 | 64 1.7 | cor 6500
Te21|eC 21.1 | 65 1.5 | cor 4500
4.4] Cc 20.1 | 66 Tey 6500
6.0] c 21.1 | 65 1.6 | cor 5500
9.7} c fog 21.1 163 1.9 | cor 8500
Aug. 10 | 10.5] f W. | 20.2 | 75 2.8 | wrlg 28000
Tze lhe 20.1 | 80 2.1 | cor T1500
2.7)\\ £ 22.1) 77 2.2 | cor 13500
5.6] f 22.1 |77 20/wrBgr T0000
Nee Teron 7) |(C° S. 20.1 | 73 2.8 | wrlg 28000
10.1] ¢ 20.18 ||— 2.3|wbrbgr 15500
11.8] c 21.1 | 74 2.4|wbrbgr 17500
12.1] ¢ 21.1 |— 2.4 Do. 17500| Fresh water, etc.
BET Cee 22.1 | 70 2.8 | wig 28000
6.1] c 23.1 | 69 2.2 | cor 13500
Aug.12]| 9.4] f W. | 20.1] 74 2.4|wbrBegr 17500
9.9| f — |— 2.4 Do. 17500 \
Dace eb 21.1 | 76 2.5 | w br Bir 21000] New inlet pipe.
Tee 22.1 |— 2.3 | cor 16500
Berle 23.1 | 78 2.1 | cor 12500
5.6] f 23.1 2.3 | w br Br 16500
Aug. 13] 10.4] f W. | 20.1 | 69 2.2|wbrbgr 13500
10.8] f£ 20.1 | — 2.3 | w|[Blp T5500
Zi 20.1 | 72 2.7 | wre 26000
pO) (et 20.1 |— 2.6 | w br 22000
Aug.14] 9.6] f W. | 20.1} 71 1.8 | cor 7000
10.9| f 20.1 | — 2.2 |wrg 13500
real 20.1 2.4 | w br Bir 17500
Aug. 17| 4.0] £ N.W.| 21.1 | 75 2.4 | w br Bir 17500
5.8| f 22.1174 2.2;wrbgr 14500
Aug. 18} 10.5] f N.W.| — |75 2.9|g’ br B 31000
12.0] £ 20.0 | 78 2.8] wrlg 28000
3.2| £ S. 22.1 | 76 2.5|wbrbgr 20000
4.8| f DRRma gS 1.9 | cor 8500
Aug. 19} 10.5 | f 22.1 | 77 3.1 | w b/blp 38000] New apparatus.
10.8 | f S. 22.1 |— 3.2|Do.brw Bp} 42000
12.5| f 24.1 | 79 ee eet 35000
3.2| 25.1 | 80 2.5 | w br blr 20000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

170
eg) 2 1214
Pu eae ee
1903 |Aug.19| 3.8/f
4.3 \f
On| ee
Aug. 20| 9.1 }c
r0.2 | €
Te sO
127 4\(C
a 2uilwRe
4.0 | R’
Seal eed
Aug. 21] 9.2 |/f
De salt
D2. Zale
Arriah
6.0 | f
Aug. 22) 922) |)t
LO.E || f
12.0 |f
Ta. |b
Aetalve
Se Sait
6.0 | f
Aug. 23] 10.1 | f
TO-5 || |t
oe yale
feral
5-9 |f
Aug. 24} 9.6|c’ R’
9.8/5
12.2|f'S
12.4 | f/
Sea |a
Geer {at
Aug. 25]10.3/cR
125) |r
3.0) ||(C"
Ges elkGaks:
Aug. 26] 9.8} c
10.2] Cc
T2320 iC:
12.8} c
Sesalic
Bevee
Aone
5-5, [ee tore
Sa
Aug. 27] 9.8|f
10.5 |f
Tsu ee
meCeON| Ce
12.9|c

| Wind.

Jaa |T

W.

N.W.

N.W.

N.W.

N.

TABLE 1—Continued.

emperature
Apparatus.

tb tb YH
=~. bw

HH HR RR HOW

bo NNN
WwW AUN

NN
Oo W
HoH

2 Gal
2i5e0

2 aT
24.1
24.1
26.1
26.1
20.1

24.1
24.1

25.1
25.1
23.1
23.1
23.1
24.1
25ar
pyeat
22.1

24.1
25.1
22.1
220
20.0

23.1

Zor
Dyin

2 Kio
2Tor

22k
22.1

Temperature
Atmosphere.

74

Aperture s.

HH HH NW
sh nwo aod

to
Lan!

bob
ao

Ww

HOOnMaARR

wBHYHNKNKD HH HH HHH HDD DHNDHNDNHWWWWH HH DH HKHNHNN W WH NHN DN HD

BR AWS WYAKE DUAN HO DWHAHAUNNHANHOS DAHDHHE OAL O

°

Corona
Colors

cor
w br bip
w rlg

w rg

w rig
cor

w tg

w br blr
cor

cor

cor
cor

cor

w br blr
gbrBr
g blp

w br Bir

Do

cor
cor

w br blr
cor
cor
cor
cor
cor

w rg
wpbg’r
w r\g
w rg
cor
cor

g’ bp
cor
cor
wrg
cor
cor
cor
cor
cor
cor
cor
cor
cor

g’ brb
g’ brb
w tg
w br blr
cor

{ above
" Uwalg

Number 2.

21000
31000
35000
38000

35000

17500
22000
17500
II500
11500

8000

§50°

9500
53000

47000
55000
53000
13500
T1500
28000
10000
13500
31000
4000
4000
6000
5500
3500
3000
4500
6500
8000
31000
28000
23000
17500
17500

Remarks.

Open trench at
Wilson Hall.

Window open.
“ae ae

Violent rain,

New pipe.
New adjustment.

Glass clean, new
water.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. I71I

TABLE 1—Continued.

“i wo ae | io .
x Bence ea lope: 5 a g Remarks.
v 2 og Aad | Ao £ ag m
2 g 3 Bee ea Hee hs £8 g
A & = B ljaclaa| < 68 Z
1903 |Aug. 27] 3-4/¢ 23) tal 74 2.2 | COT 13500
3.8} ¢c 2B To 2.2 | COr 13500
5-8] Cc 23.1 | 69 2.0 | cor 10000
6.0 ]}c¢c eee 2.0 | cor 10000
Aug. 28] 9.6/c N. E.| 21.1 | 69 2.2 | Cor 14500
1253 |/C Dost 7a 3-3 | cor above g’ | 45000
T2e4a\iC eee lhe 3.0|g’ brbp 35000
Bean Cu RY 22.1 | 66 2.7|wrlg 25000
3.8] c DD ait |\—— 2.6 | cor 22000
eye Ose —= 63 I.g | cor 8500
EEO uC Baer 1.9 | cor 8500
Aug.29| 9.9/R E. 20.1 | 63 2.7 | wrlg 25000| New pipe.
10.1|R 20.1 | — 2.2'\| COr 13500
12.8 | R’ 2A || Oe Tas) || COL 4000
Bean RS 22063 1.5 | cor 4000
Gea EA 22.1 | 61 2.1 | cor 11500
Aug. 30]10.8/R E. 20.1 | 64 1.7 | cor 6000
Desi |e 21.1 | 63 1.6 | cor 5000
pote eR: 20.1 | 61 Gag) ||| COL 2500
Aug.31] 9.7 |R’ iN. 19.1 | 61 1.8 | cor 8000] Stagnant air
920) | BO ea 1.9 | cor 8500
2c 20.1 | 62 1.9 | cor 8500
4.0} C 21.1 | 63 1.9 | cor 8500
Sasi 21.1 | 62 1.9 | cor 8500
Sept.1 | 9.6/f W. |19.1|70 2.9 | w br b|p 3 1000
MOns) pe 19.2 |\— 2.9 | cor 31000
T2951 ||) ein P75 2.4|wrb 49000
ge
Teo ie 21.1 | — BES jee 45000
Beanie S — 174 2.6 | cor "| 22000
5-8 | f 22.1 | 70 2.6 | wrlg 23000
Sept. 2 | 9.9 | c’ W. | 20.1 | 70 2.3 | cor 15500
LOO) | (Co 2050 | 2.3 | cor 15500
12.9 | c’ 20.0 |— 2.8/wrie 28000
4.4/C POTENT Ie) 2.7 |wrg 26000
Seo uEC 22.1 | 68 1.9 | cor 8500
Sept. 3 | 10.0 | f 20.1 | 67 1.5 | cor 5500
11.4 |f 20.1 | 72 2.8 | w br B|p 28000
Der | 21.0 | 73 3.0 Do. 36000
2.8 \f S — 175 2.8/wrg 28000
4.7 |£ 22.1 | 73 278) aDOr 28000
Segui 22.1 | 68 2.8 | g|b|p 28000
Sept. 4 | 10.3 | f S. 21.1 | 72 2.5 | cor 20000
12.4 |f 22.0175 3.0 | wlb|p 35000
BEAg Gh 23.1 | 76 2.5 | cor 21000] Fire on campus.
Boy |lae 24.1 | 74 2.5 | cor 21000
Sept. 5 || 9:7 |/c S. W.} 21.1 | 72 2.6 | cor 23000
1072524. || © 22) Te 77 2.6 | cor 22000
BEOn|EC 2Beia 7/0 2.6 | cor 22000
4.0 Babies 75 C7 ||(COG 6500| Thunder.
EeOm|ked 23.1 | 67 2.3 | cor 15500| Rain over.
Sept.6 | 9.9/f Ne liter | 65 2.8 | cor rlg 28000

172 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

Eo :
ti Seale 2 =
; a 3 a ao = 8 i eS Remarks.
¢| 2 |8 | & | 8 |be)92 | 8 ae g
m Q = = = laa laa < OO Z
1903 |Sept. 6 | 10.8 | f 21.1 | 67 2.8 | w rg 28000
12.8 | f 22.1 | 69 2.4 | cor 19000
Ses he 21.1 | 68 2.1 | cor 11500
Sept. 7 | 10.7)|£ N.W.| 20.1 | 66 2.3 | cor 16500
i 68 { 3.2 | w br
1.0 ]c 20d 12.8 | Do. \ 42000
Zan C N. E.} 22.1 | 69 2.8 | cor 28000
4.2 ]¢ 21.1 |— 2.4 | Cor 19000
S.Sc 22.1 | 66 2.4 | cor 19000
Sept. 8 | 10.4 | f N. 19.1 | 63 2.9|wbrb 31000
12.4 |f 19.1 | 65 3.0 | w br blr 35000] Repeated s=3.0.
2.8/\f 20.1 | 66 26000 i S=2.7.
3.0 |f 20.1 | — 2.7|wrg 26000
5.8) £ 21.1 | 61 2.7 | cor 25000
Sept. 9 | 10.2 | f Ss. 18.1 | 64 3-3 | wp 47000
10.4 |f 18.1 | — 3.4|wp 49000
LD-0) it 19.1 | 67 3.1 | w br bir 38000
1.0] f 19.1 |— 3.1 |g’ br blip 38000
2.8 | f 20.1 | 67 2.9 | wrig 33000] Repeated s=2.7.
5-4/f 21.1 | 64 2.8|wrig 28000
9.6\c 20.1 | 61 3.0 | wrlg 35000] Night.
Sept. 10] 9.9} c 19.1 | 72 Glass broke.
10.71},c 19.1 | 73 3.0 | w br bp 35000] Old apparatus.
Deez |0C 21.1} 74 2.9 Do. 31000
12.4|C Beet VA. 2.6 | cor 22000
Wes ie 21.1] 70 2.6 | cor 22000| Repeated s = 2.3.
Sept. 11] 9.5 |f W. | 27.2} 78 2.9 | wr|g 31000
Tee yl 22.1 | 82 2.7 | COr 25000
264) | 23.1 | 82 2.9 | cor 31000
2.9 |£ 24.1 | 83 3.0 | w(br) bir 35000
50h 24.1177 2.8 | wrlg 28000
Sept. 12| 10.5 | f W274 4.0|g’0b 75000
De galict 27D ska 975 4.0| Do. 75000
B25 pt 22.1 | 76 3.3 | w br 45000
Bega it S 23505 a7 2.8 | wrlg 28000
Broa 24-5 | 42 2.8|wrig 28000
Sept. 13} 11.0 | f S. W.| 21.1 | 79 2.8 | w br bir 30000
1.7 23.1 | 83 2.0 | cor 10000
6.8 | f 23-1 | 76 1.6 | cor 5000
Ont 23.1 | 74 2.0 | cor 10000
Sept. 14] 9.7 | f W. | 22.1 | 83 Bese ao Deennes 45000
12.4 |f 25.1 | 88 3.0|wbrbgr 35000
2-5 alii 25.1 | 88 2.7 | cor ; 25000
Savalas 27.1 | 84 2.8 | wrlg 30000
Sept. 15] 9.7. |f W 24.1 | 81 3.0|/wrbg 35000] Repeated s= 3.0.
G2 shi Re 26.1 | 86 3.6 | wrlg 57000 i S=3.2.
2.5) £ 26.1 | 86 3.0 | g’|blr 35000
Bae 26.1 | 77 2.8 | wrlg 28000
Sept. 16] 2.3./¢ S. 24.1 | 78 2.0 | cor 10000
BeTRIC 25.1 | — 2.0 | cor 10000
GsHiIKC 26.1 | 76 2.0 | cor 10000
Sept. 17! 9.5 | ¢ S. W.| 24.1 | 76 2.2 | cor 13500

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

173

ami ake b
o O)
: 24/238 2 s
K 85 oS vo :
g BASS eo G1 8 3 ,
‘ é J = : Pa a 5 Ee 2 Remarks,
mel a |e | Fi| Ee jedlce| 2 88 Z
1903 |Sept. 17} 12.3 | ¢ 26.1 | 80 245) | COr 20000| Wind storm.
BaSuIC 26.1 | 78 2.8 | w rig 28000| Gale,
Ges hlicek: 26.1 | 74 2).2)\\(COr 14500
Sept. 18] 9.4 | c’ N.W.| 212.1 | 65 2.6 | cor 24000
12.4 | c’ 22.1 | 70 2.7 | wrlg 25000
6.0 | c’ 22.1 | 64 2.7 | wrlg 25000
Sept. 19] 9.7 |f N.W.| 20.1 | 62 4..;}g¢bp 80000] Repeated same.
12.8 | f 21.t | 67 3.0 | g|blr 35000
Bo) | Key 22.1 | 68 2.6 |wrig 24000
Gas m Ce 22.1 | 63 2.8 | wrig 28000
Sept. 20] 10.4 | c’ N. E.| 20.1 | 63 1.5 | cor 4000
TLE QUIIC! 20.1 | 63 1.8 | cor 7000
Te mae 65 1.8 | cor 7000
5-4] Cc 21.1 | 60 1.9 | cor gooo
6.4 | Cc 21.1 | 58 2.1 | cor 11500
Sept, 21) (9-3) |\c N. 18.1 | 59 2.4 | Cor 17500
LEONI 20.1 | 66 3.0 | w br Bir 36000
Ben \ak 20.1 | 67 3.1 Do. 38000
Bae) |e 20.1 | — 3-0 | g’ br blr 35000
5.9 |f 21.1 | 63 3.0 | g’|b|p 35000
g.o|£ 20.1 | 56 323) COE 47000] Night.
Sept. 22] 9.5 /f 19.1 2.6 | cor 22000
Stolk Wr. | 23.2178 2.8 | wrlg 28000
5.2 \f 23.1 | 75 2.9 | cor 31000
6.0 | f 23.1 | — 3-1 | w|b|p 38000
Sept. 23} 9.4 |f 8. E. | 19.1 | 67 2.8 | w rig 30000
12.4 |f 21.1} 74 3.0 | w olg 35000
6.0 | f 23.1 168 3.3) cor. 45000
Or2iiih 2B ete 2.9|wbp 33000} Repeated.
pepey egg oR: |W. | 20.0 | 65 2.8 | wrl|g 28000
eae |C 21.1 | 69 2.8 | Do. 28000
2.9|c 20.1 | 67 2.8 | g’|b|p 30000
6.0 | f’ 271 Ox 2.7|/wrg 25000
r {3-7 | wrlg )
Sept. 25] 9.2 | f N.W.| 17.1 | 56 es Bi len see 34000
O25 lf 17.1 |— 22S) Dor 66000
ret 63 3.4 | w br blg 49000
Beaute 20.1 | 65 3-2 | wlb|p 42000
6.0 | f 20.1 | 60 3-4 | w rg 49000
9.4/f 20.1 | 55 3.0 | w|b|p 36000
Sept. 26] 9.6] f Se 18.1 | 63 3-3. | wrlg 47000
9.8 | f 18.1 | — 3.4 | wrlg 51000
12.9 |f 22.1 | 69 2.7 | w rg 26000
4.0/f 21.1 | 70 2.9 | wib|p 31000
6.0 /f 21.1 |64 2.9 oO. 31000] g’/b/p
Sept. 27] 9.8|ftoc |S. W.| 19.1 | 70 2.1 | cor II500
TAL} || IRN 21.1 | 69 2.0 | cor 11000
Asai 20.1 | 71 2.7 |wrlg 25000
O=5 1c 21.1 2.3 | COr 15500
Sept. 28] 9.5 |f 18.1 | 61 2.8 | cor 28000] Rain and storm
12.0 /f 18.1 | 62 2.9 | cor 31000] at night.
5.8} f 20.0 157 2.9 | wo bg 31000

174 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

; 23,88 3 =
3 ; 5s Se 3 sy 3 Remarks.
: oO ~ ? Ag ao Ee Gu
3 2 g 3 S |/§a/8e] 3 a §
~ | A |a | B | FE ladles] < 88 Z
1903 |Sept. 29] 9.5 |f 3.0 | w br Bir 35000
12.5 333 Do. 45000
3.8 | f 3.0|wog 35000
6.3 |f 3-2 | w br 42000
g.o |f 3.1 | wb 38000] Night cold.
Sept. 30] 9.3 |f W. 3.8 | wrlg 68000
9.8 |f 4.0 | g’|b|p 75000] Cold.
12.4 |f 3-6 | wrlg 58000
Balk 3.1 | w br blr 38000
6.2 | f 20.1 | 60 2.8 |g blip 28000
Octan: Qo 7) (it! S-Wil tyr 165 3.1 | w br blr 38000
1.0 | f 19.1 | 74 Ber Do. 38000
ae Fait 21.1 | 73 26)/wog 22000
Geo) || 21.1 | 68 3.1 | w br blr 38000
Oct. 2 9-4 | Cc W. | 19.1} 67 2.3 | cor 15500
1.0 ]C 21.1} 73 2.2 | cor 14500
4.0 | c N. E.| 21.1 | 68 1.8 | cor 7000
5-0 |c 22.1 |65 1.8 | cor 8000
Oct. 3 9-3 |f N. E.| 18.1 | 59 1.8 | cor 8000
10.0 | f 18.1 | — 2.5 | cor 20000
12.0 | f 19.1 | 66 2.8 | g’ b Bir 28000
B30) |i 21.1 | 68 2.7 | cor 25000
5.8 |f 21.1 | 59 2.3 | cor 16500
Oct. 4 |10.4/f S 19.1 | 62 2.9 | cor 31000
MOs7 |e 19.1 | — 2.8 | g’|B|p 28000
12.4 |f 20.1 | 64 2.9 | g’|blp 33000
6.2 |f 19.1 | 62 2.9|wrg 31000
Octis 9.3 | fl W. | 19.1 | 69 Bho | within 42000
9-9 | f 19.1 | — 3-6 | w rig 58000
12.4 | Cc’ 21.1 | 76 3.0 | w|B|p 35000
Gara Cul 23.1 | 67 2.8 |g’ blp 28000
6.0|R 22.1 | 67 3.2|wreg 42000
Oct. 6 prt || 3 N.W.| 19.1 | 65 2.2 | w|B\p 13500
10.0 | f” 20.1 | — 2.2} Do. 13500
T2.0 \\t 20.1 | 70 2.2 | COr 14500
3.5 |f S. E. | 22.1 | 69 2.7 | cor 25000
6.6 | f 22.1 | 61 2.5 | cor 20000
10.9 | f 21.1 | 58 2.0 | cor 11000] Night.
Octs7 9.2] Cc E. 20.1 | 63 2.1 | cor 11500
12.6 }c 21.1 | 64 1.8 | cor 7000
2.9 |c R’ 21.1 | 64 1.8 | cor 7000
6.0 |c R’ 22.1 | 60 2.1 | cor 11500
Oct. 8 9.7 | Cc S. E. | 20.1 | 66 3.0/wrlg 35000| Repeated s=2.5.
D252 iC = ||o8 2.9 | wre 31000
2.8 | c’ 21.1 | 68 3.0 | w|B\p 35000
6.2 | ce” 22.1 | 62 2.2 | cor 13500
Oct. 9 g.1 | f E. 20.1 | 65 1.6 | cor 5000) Rained at night.
1.0 | Cc 22.1 | 66 1.6 | cor 5000
2 -Ou 22.1 | 66 TS eeor 4000
6.0} Cc 22.1 | 61 1.7 | cor 6500
Oct. 10 | 9.1 | R’ N. E.| 20.1 | 55 1.8 | cor 7000| Floods in New
12.8 | R’ 21.1 |57 1.7 | cor 6000! York.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 175

TABLE 1—Continued.

B a
Secale 4 <
u SE i] o =
: a . 5 G o 3 3 3 Remarks,
: iS 3 J Ag | ao z ag a
as | 3 E S| Ss Hees | -s 22 E
o @ 3 = oa] os a 06 5
m a a = E lad laa < oO Z
1903 |Oct. 10 | 6.0} R’ Bio EsTm|Si 1.8 | cor 7000
©ciser) 9-9) Re N. E.} 20.1 | 55 2.1 | cor 11500
12.4| R’ Ain (G5 1.8 | cor 8000
5.0| R’ 20.1 | 54 1.7 | cor 6000
6.2} R’ 21.1 | 54 1.9 | cor 8500
Oct. 12 | 9.4] R N Ig.I | 52 2.3 | cor 15500
12.8] R TQ sass 2.9 | y olg 31000] Repeated s=3.0.
MBN aie | 5/5 2.7 | wrlg 25000
Gsaiixc Bea G5 2.5 | cor 20000
Ocir3)|955)|¢ N. 19.1 | 57 2.5 | w|Blr 20000
12.0) (C 20.1 | 59 223) |9Do; 15500
3.6|c 21.1 | 60 --| — oo
6.4] c 20.1 | 56 2.8 |wrlg 28000
Oct. 14 | 9.4] f N. 17.1 | 59 3.7 |wrlg 64000] Fine weather.
g.8| f Tey | 4.1 | g’ ob’ 30000
rae 7a ec 19.1 | 68 3.6 | w br »| 57000] Sun spots.
220) | 20.1 | 68 3.1 | w|B\p 38000
Tei lack 20.1 | — 3.3 | wbr 47000
3.2| £ 20.1 | 68 Bete lla 38000
4.8| f 20.1 | 64 2.6 | cor 22000
SiS) [ee 20.1 | 63 2.4 | cor 17500
Octers | 9-4) £ 18.1 | 55 2.9 | wiblr 31000
. TAME 19.1 | 61 3.8 | wrlg 66000
2.8} f 1g. | 61 2.8 | wrlg 28000
6.0] f 20.1 | 55 3.0|2¢’ Bp 35000
Oletts 00 || Way) || ay 8. 18.1 | 58 Beat lw Die 45000
10.0| f 18.1 | —- 3.0 | w|Blr 35000
12.0) f 19.1 | 65 3.0 | g|Blp 35000
424i 21.1 | 64 2.7|wrig 25000
6.3] c 20st |103 2.7 |wrig 25000
Octen7) | 0:8) |7c S —— |i05 2.71/wrg 25000
silk: 21.1 | 66 29/gBp 31000
Bee 22.1 | 66 2.9|wlbgr 31000
6.0} R 22.1 | 61 2.5 | w br 20000
Oct. 18 | 10.0] c’ W. |19.1 1/59 2.3 | Cor 15500
TRS y CAR 20.1 | 60 2.5) || COL 20000
4.4| £ 20.1 | 54 2.5 |wrlg 21000
Osa) 20.1 | 49 2.6 | cor 22000
Octyroy io-3) | WwW 18.1 | 48 3.9 |yobg 70000
9.8| f 18.1 |— 3.5|wre 55000
nef C 19.1 | 53 3.1 | wib|p 38000
Belle 20.1 | 55 3.0 | Do. 35000
6.0| f 21-1 | 52 3.1 | w br 40000
Oct: 20) 0.4) £ S. W.| 19.1 | 62 2.8 | g’|Blp 30000
MO) ek 20.1 | 67 2.9] Do. 31000
2.0] f 20.1 | 68 2.9 | cor 36000
6.5] f 22.1 | 60 2.9 | g|Blp 31000
Oct. 21 | 9.4| f N.W.| 18.1 | 57 3.2 | wlb p 42000
9.8} f 18.1 |— 3.6|/wrg 57000
Te24| ot 20.1 | 58 326) SDo: 59000
2.7|f 20.1 | 59 3.1 | w\blp 38000
6.0| f 20.1 | 54 Bet DOF 38000

176 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

& ad
‘ g .
E So sees .
= Seine Sai peice ees By 2 Remarks.
Bu |g ei} 8 | €/g8lee| & 85 5
2 3 s s i S| 35 a 30 3
a a a = Ee |ad| ac < oO A
1903 |Oct. 22| 9.3 /f N.W.| 18.1] 50 2.8| wrlg 30000
1.4 |f S. 20.1| 57 2.7| cor 27000
4.0 |f 20.1| 56 2.8} wrlg 28000
Sagal 2 TE SS 3.2] wrig 44000
Oct: 23] 953) 'c” S. W.| 19.1] 65 3.0| cor 36000
12.01|'C 20.1| 69 3.0| wr|g 35000
As2niC Re! ae Si 2.4] COr 17500
5.8]¢ 21.050 2.4] cor 17500
Oct. 24] 9-4] ¢ N.W.| 19.1] 48 3-0] g|blp 35000
12.0: | Cc” 20.1| 51 3-4| W pcor 49000
Bu5n ce 20.1| 51 2.9| g|b|p 31000
O:0} (C7 20.1] 48 2.9|gBp 31000
Oct. 25 | 10.4 | f N. 18.1] 45 4.3] gy o bg goooo
10.8 | f 18.1} 45 4.0| wog 75000
12.3 Cc 18.1| 48 3.0] g|b|p 35000
5-0 ]c 17-1| 45 2.8| wrlig 30000
6.5 ]¢c 17.1] 45 2.81|| Do. 30000
Oct. 26} 9.4/f W. | 16.1] 46 3-1| wBg 38000
10.0 | c (S) 16.1 | — 2.7| cor 25000| Clouds.
ETT 16.1] 44 2.4| rop cor 70000
Be4n [Ls 19.1| 43 2.9| w\b|p 31000
6.0 |f 20.1} 38 2.8| w\b/p 28000
Oct. 27] 9.6|f W. | 19.1] 39 BEB ELOD sete 47000
12.8 | f 2I.1| 42 3.4| wrlg 49000
CLONE 22.1| 38 3-2| wlblg 42000
Oct. 28] 9.4] f E39 4.0] yo bg 75000/ Snow.
9.9 |\f N.W.| 19.1] — 4.1] Do. 80000
1.4 |f 20.1| 46 3.7) wrg 62000
gaze 21.1] 46 3.1| w Blp 40000
6.0 | f 22.1] 41 Z|) Do; 38000
Oct. 29] 9.3 |£ 19.1] 45 3.4| wp dir 49000
2.8/f£ 23.1| 62 Bos|| . Die 53000
6.0 | f 23.1] 56 3.5| wrlg 53000
Oct. 301 09-5, |/£ W. |e2r.r] 59 3-1| wlb|p 38000
27 |e 22.1| 68 2.5)| (COL 20000
Bula ab 23.1| 68 2.7| w|b|p 25000
5.8 \f 23.1| 62 3.6 wrg 57000
Oct31 |) o-2) it W. | 22.1} 63 2.6| cor
9.9 |f 20. 1)|'-— 3.2) wr 42000
7 12.2 |f 21.1| 69 2.9| g’|b’|p 33000
Fg |\5% 22.1| 68 3.0| wibl|p 35000
Bean = |) on 8.0) Do: 35000
Nov. 1 | 10.8 | f N.W.| 19.1| 62 3.0| w|blr 37000
12.0 |f 19.1 | 63 3.0| cor 35000] Apparatus
12.4 |f 20.1 | — 2.9] wrlg 33000| cleaned.
4.5 \f 20.1| 61 2.5| cor 20000
6.3 |f = Gy 3.0| g’ bip 37000
Nov. 2] 9.4/f W. —|58 3.7| wrg 61000
r.0|f 20.1] 65 3.2| w blp 42000
4.1 /f 21.1] 64 3.0| wolg 35000
5.8 /f 21.1] 60 3.5| wrig 53000
Nov. 3 | 9.2\f W. 17.1} 58 3.4| wrlg 49000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. : 177

TABLE 1.—Continued.

& ao
Bu] 35 4 x
s ge) ea| 2 :
P oS] ¢o 3 3 4 :
« Z 2 3 z Gd ad E | aé 4 Remarks
pameen le | & | & led|a<| =< 388 Z
1903)|Nov. 3 | 12.5] f 19.1 | 65 3.2| w|blp 4200¢
leesesiliet 20.1 | 66 2.7| wrlig 2500
ead 200 |) OF 3.6 | w rg 57000
Nov. 4 | 9.3] £ W. | 19.1] 62 2.9| wolg 3.1000
Tea} |e 20.1| 70 3.1 | cor | 38000
2.6| f 2era |e 7 1 2.9| wr|g 31000
: Heese salt 21.1] 66 3.5 | welg 53000
Nov. 5 | 9:3)cR’ |S. WJ 19.1| 60 2.9 | cor 33000
|zr.6| cR 21.1| 62 2.8| w rig | 28000
| Ae |e 22.1 | 64 3.2 | Wp cor | 42000
5-9] Cc 23.1| 62 2-5) |) COL 20000
Nov. 6 | 9.1/c¢ 19.1 | 40 3.0] w olg 35000
10.1] Cc 20.1| 40 2.0 |) =Do: 35000)
122A) |\1C 20.1} 38 3.0| g Bip 35000) Snow.
3.2| Snow 20.1 | 37 3.0 | cor 35000
Ge7/|| © 22.1 | 37 3-0] w|blp 35000)
Nov. 7 | 9.6] f 20.1| 36 3.4| Wp cor 49000
10.0| f W. | 20.1) — 3.1 | g’ blp 40000
Tega et 18.1| 40 aa oO. 38000
Basile 22.1] 41 2.6] wrlg 24000
5-9| f — | 38 3-5| wrig 53000
Nov. 8 | 9.9] f N.W,| 23.1] 44 3.4] wrig 49000
12.2| f 24.1| 50 2.9| g’ |b’|p 31000
Beni t 24.1| 48 2.9| w|b|p 31000
Nov. 9 | 9.2] f W. | 22.1} 48 3-9| yg o bg 70000
9.6| f 22ea | 4.2| g’ 0b’ 85000)
12.9| f 22.1| 60 3.6| w rig 57000)
4.0] f 23.1] 58 3.0] wlblr 35000, Repeated s=2.8.
6.0| f 2361/52 3-4) wre 49000
Nov. 10] 9.6| f S. W 20.1] 53 3.0| wlblr 35000
L250 |: 21.1} 58 3.0) || Do: 35000)
3.8| c 2u.0| 58 3.9| wrlg 70000
5.9| f — | 53 2.8| wlblp 28000
Nov. 11] 9.5] f N.W,| 18.1] 50 2.8| w\b|p 28000
Teepe 19.1] 57 2.2| cor 13500
12-9) jet IN ey eeu Za8 COL 15500
240)|| £ : 19.1| 56 2.2] cor 13500
Sale 20.1 | 49 27 COL 25000
Nov. 12] 9.5] f Wis |r Sar |52 2.2| cor 14500
TAGES) 18.1] 56 3.4] y’ ple 49000
BEOllieE TORT |Ne5 07 3.0] wlblr 35000
6.0| f 19.1| 50 je = ae 70000
Nov. 13] 9.5] f S. W) 17.1| 52 3-7 | y’ plg 61000
eye 18.1] 59 3-3 | wrlg 45000
4.8| f —/56 | 3-3) wrig 45000
cealet 19.1] 54 3.3 | w clg 47000
Nov. 14] 9.5|c¢ N.W,| 18.1 | 48 2.7 | w|b|p 25000
Tele 5 le 18.1 | 48 3.0 | cor 35000
Aveta ate W. | 18.1] 47 3.2] w p cor 42000
5-9| f 18.1} 44 3-.9| g’ o bg 70000| Repeated same.
Nov. 15] 10.2] f W. | 27-1] 42 3.5 | wclg 53000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

2 B¢
uo See a
Z & Ou on rs av
| 2 |2) 8 | EER BE) B| 88
> a | 4 2 B |adc|a<| < 68
1903 |Nov. 15] 12.5] f cea (ead 2.8| g\b|p
7. All| £ 16.1] 39 3.4| w rig
Nov. 16] 9.3] c N.W,] 16.1] 40 3.2| wp
Tne Sale kee 16.1} 42 2.8| w olg
3.4| R 18.1] 43 2.7| w br blr
Be nex N. E| 18.1] 44 27 (Oe
Nov. 17) 9.5| R N.W|| 17.1| 40 re g’|b|p
12.2) R 18.1 Seon
SS Vets mers
3-5|R 19.1 | 57 34| weg
\Nov. 18 9.7| ¢ N.W 18.1} 43 2.8| w olg
T27 EC 17.1| 44 2.8| Do.
al 20.1 | 39 2.8| g’|b p
Nov. 19] 9.4] f W. | 19.1| 34 3.3| wp be’ r
Toes it 19.1| 40 Beai|) DO:
Beater 20.1 | 36 Zio | | IByoy.
Nov. 20 9.5] f N.W)| 21.1| 32 3-1 | w|B\p
11.9| f N. 20.1| 36 3.4|wpbeg' tr
Ani let 20.1| 36 Sesh) 1DYee
6.1| f 21.1| 33 3.8| wrlg
Nov. 21| 9.4| f N. 21.1] 29 3.2| w br bir
Tasik 21.1| 38 B02) Dor
4.4| f 21.1] 36 3.2|wpbegr
Gr) |t arn 35 2.8|gbp !
Nov. 22) 9.8|c N. 21.1| 36 { 3.1 ea
os 3 (3-0 w|blp
r.0| c aire || a0 2.7 | wtlg
5-01] ¢ 22.1| 39 3.2|wpbegr
6.6\¢ 19.1| 38 3.2|wpbegr
Nov. 23] 9.3] ¢’ N. 21.1| 37 3-8| y’ rlg
10.0] Cc PM || —— 4.0|gbp
12.8| Cc 2 al Aes 2.7| wrig
326) |/8C S. 22.1] 43 2.7| wrlg
GeAinG a 22.0 | a3 3.1) wpbegr
Nov. 24] 9.3] f Wear 22m a2 3.3)wpbeg r
12.8 )°f7 20.1| 44 2.8! cor
Sey lac 20.1| 40 2.7| w tg
6.0| c 2iTeaT|| 3)5) 2.5 | cor
Nov. 25] 9.3| f W. | 22.1) 26 3.0| g’ blp
Onyalit 22.1 | — 3-7| wrlg
11.4] f 23.1| 31 3-9| y’ orlg
3.2| £ alee 2.9| g’|blp
Basil L 21.1| 26 2.9| Do.
6.0| f — | — 230) s Dor
Nov. 26] 10.0] f N.W |] 22.1] 25 3.6 | w rig
1.0| f 22.1} 29 ae eo Ps
| 4.0| yo bg
eo tae 22.1| 27 3-5| wrlg
Nov. 27] 9.4] f N. 21.1 | 20 3.8| w olg
12.6] c 2 eT ear 3-9\|gbp
3-0] c | 21.1] 27 3.5| wrig

Number n

28000
49000
42000
28000
25000
25000
28000

53000

49000
28000
28000
28000
47000
45000
45000
40000)
49000
45000
66000)
42000
42000
42000
28000

38000

25000
42000
42000)
66000
75000)
26000
25000
40000
45000
30000
25000
21000
35000
61000
70000
31000
31000
31000
57000

78000

53000
66000
70000

55000

Remarks.

Cold wave.

Repeated s=3.6.
130/10 *
Repeated same.
59/5

* Subsidence data, meaning that in 130 sec. the fog line has fallen 10 cm.

| Year.

1903

Date.

Dec.

Dec.

Dec.

Dec.

Dec.

Dec.

Dec.

Dec.

Nov.

Nov.

Nov.

Nov.

27

28

20

30

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

179

o Oo.
ae 38 4 s
eee) 2 |) a log
: Ss : aw a 5 a Remarks,
a) g | =| Fel ge/| & 88 @
& == BE lad|ad| < 5d Z
Bene Qe ST | 4.2} gsyo- 85000] 69/5
esa lne 2A || 4.0] w orl|g 75000| Repeated same,
58/5.
9.3| f N. 20.1| 26 4.0] w olg 75000| 56/5
Te2)|) ib 2120} 30 2.9/ ge’ bp 33000
Beal |e 21.1| 31 3.2| w br— 42000
(oie |] 21.1] 28 3.8) wre 66000] 59/5
10.6| ¢ N.W|| 21.1] 24 2.9| g’|blr 31000] 28/5
coer CS 21.1| 27 2.9 Do. 31000] 43/5
1.4| c Sun 21-0) :20 2EOli|ee DOr 31000] 39/5
5-2| ¢ Sun Bree 2/7 3.4| wrlg 49000
5-9| c Sun Pyne 3.0] w|b|p 35000] 41/5
6.5] ¢ Aisi || — 3.0 O. 35000
Oeste NVienle2oens (25 3.0| wrg 35000] 47/5
E2eO)|\ 19.1 | 32 3.7|/ wre 64000| 49/5
pais 19.1| 34 3.0} g|b|p 35000] 39/5
cele 20.1] 30 4.0| wy r bg 75000] 65/5
9.3 | f N.W|| 20.1] 26 4.0| wrg 75000] 63/5
me) OST NGS 3.0| g|b|p 35000] 32/5
3.4| f 19.1 | 34 2.9| w rig 31000] 26/5
6.0] f Ig.1| 30 2.9|wrg 31000] 41/5
9.4| ¢ Sun 20.1 | 30 3.6) wre 59000] 53/5
12.8| c¢ N. 20.1 | 36 3.6|wrg 57000] 48/5
Bee CuRe — | 34 3.8| y’og 66000] 47/5
6.0} c 20.1 | 34 3.7| wre | 61000! 40/5
9.5 | c¢ N. Diiait || hae 3.5| weg’ | 53000] 58/5
11.8] c 20.1 | 33 4.0| y o bg | 75000] 66/5
Boll iC RY — | 34 3.1] glb|p 38000] 36/5
Re) © 22.1 | 33 3.9|gbp 70000| 70/5. Repeated
6.0] c 22eTs |e 4.1| g’ o bg 80000 S=3.2
9.5| Cc N. 21.1| 34 3.6| wr g” 57000
12.0| Cc 22.1| 137 3.6| w rig 57000
Bei lita 22.1} 39 3.2 | wlbl|p 42000
6.0| f 21.1) 35 3.8| w rig 66000
Geanlecs W. — | 39 4.2| g’|B p 85000
9.7| c oie | 4.2| Do. 85000
TONS) |G — | 44 3-5] up 53000
4.4| Cc 22.1| 40 3.0| wrlg 53000
Groj| 12 23.1 | 40 3.5 | wrlg 53000
t0.0| f N.W| 23.1] 33 3.2 | g|blp 42000
12)3)|| £ 23.1| 38 3.1 | ¢ br blp 40000
ALGAE 23 ata 5 3.0 ak 35000
= f 3-9 | w olg 0000
9.3| f We iiezerr taille 7
1 2e92\\ £ 22.1| 39 3.5|wplgr 53000
Aeeyl| at 22.1 | 37 4.1 | gy’ o bg 80000
6.1] f gaat | 315 4.1} y’ r bg 80000
9.6| f Wien |eizen 3/5 3.5 | wp cor 53000
Teen | 22.1] 43 2.6 | cor 22000
12.8] f PEE || 2.5 | cor 21000
4.6] f 2iat | Abe 3.6 wre 57000
6.0] f 22T |) —— 3.6|/wrg 59000

180 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

& £9
2a Sh -
5 ae a 3 5 Remarks /
; s ; of | oo 3 a BS emar
¢| ¢ |8| 8 |B] eel ge| 2 | ee E
> A & 2 EF jed|ad| < 6d Z
1903 |Dec. 9 9-5| c N. 20.1 | 36 3.0| cor
12am EG 21.1| 40 2.9 | cor
4.7 | Re N. Ej 22.1] 42 2.6} cor
6.0] R’ 22 2.1 | cor
Dec. 10] 9.3] f |W 22.1 | 39 3-3 | Wp cor
12.8| f 21.1] 43 3.5| wre
5.8| f 2200 a7 2.01 COT,
\Dec. 11.) 9.4] f W. | 21.1] 30 3.2| cor
12.4| f 20.1 | 36 3-5 | wrg
3 0r\\f | Are shy ae w Plg
| 4.1| g obg
6.0] f 20. til 3A leelteibe \
Dec. 12] 9.9|f W 20.1 | 30 4.0] y'og
10.5) 2 Zona la Aa Tal VaOue:
Oi] C 20.1 | 37 3.8) y’og
4.0} e(S) | — | 37 3.8| wy bg
630 re 21.1} 34 3-5| Wr
Dec. 13| 10.4! R |S. WJ 19.1] 54 2).2)\\| COL
esas 20.1 | 41 2.2) cor
6.0} f Cold arm 32 2.6 | cor
7.0| £ 20.1} 30 2.6 | cor
Dec. 14] 9.8) f W. 16.1 | 24 4.1] g’ o bg
10.1| f ae | 4.2) wre
ete 16.1| 27 3.7| w tig
6.0] f yeaa 4.1] yo bg
Dec. 15| 9.0] ¢ W. | 18.1] 20 4.0] wog
12.6] c’ 18.1| 26 4.0] wog
Bene 18.1] 25 2.9 | w|b|p
6.4] f 18.1] 22 3.4] W p cor
Dec. 16] 9.8] f Wise ereernes 3.61 w clg Repeated same.
D2e5\ ee 18.1} 27 3.6) wpbeg
QO) 18.1| 26 3.7 | w 7
6.0| f 18.1| 2 4.1] gbp
41} gbp)
Dec. 17) 8:7) W 18.1] 15 3.8|wrg
{( 4.2] wo bg
E263) 18.1| 25 levtlore \
4.6| f 19.1] 28 4.2| wo bg
6.4| f TOC 27 4.2} Do.
Dec. 18] *9.5,| £ N.W) 19.1| 16- 4.4| g’ o bg’
9.8 | f 19.1 | — 4.0| wrg
1.0| f 19.1| 18 4.3| g’ br b
6.0| f 20.1| 15 4.2| wo g’
Dec. 19| 9.5| f W 18.1 | 21 41}/gbp
10.5 | f 18.1| 24 4.0; wog
1.0| f 19.1] 30 4.0] wWog
6.0} f 20.1 | 28 4.4| y’ 0 g’
Dec. 20] 10.3} R 20.1 | 45 3.7| weg
12:9)| Re 21.0) Ay 2.8) ¢’ Bp
6.0| R 2'2 sa Si 273) |NuCOk
Dec. 21] 9.8) f We | 225m \o 2.9|¢’ Bp
£2.53 0 22.1| 40 2.9|¢’ Bp

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 181

TABLE 1—Continued.

o oO .:
24/25
5 ae ee
3 3 | BS] ad
a | 3 e | & & | £2] £e
o a a = B |aa]/aa
1903 |Dec. 21] 6.3] f 2371/37
Wecw22)ossnc Re | | 20.1 | 39
r2.8| c W. 21.1| 40
6.0| f 21.1| 28
Dec. 23] 10.1] f S. W 21.1] 28
10.5 | f 21.1 | —
TAGS fae Seen Si
4.0 Pita ill 63
6.0} f 22th 5
Dec. 24] 10.1] f S. W| 22.1] 38
10.8| c 22.1 | 43
12.7| Cc 22.1! 47
Dec. 25] 10.9| c W. | 23.1) 47
1G 4|[- 23.1| 43
Dec. 26] 10.2] c N.W| 23.1] 35
1.3 | Snow ipa |i
1.6] Snow | 22.4) —
4.0] ¢c 23.1] 20
Sea ic | 23'ar |e
6.0} c ZiQieT ety
Dec. 27| 9.4/ ¢ S: W.l2r.r | 13
10.8| Cc 2 ata) 15
12.9} Sun 2. Ole el}
ERetlee 21.1| 22
7.0/¢c 2m T | 23
Dec. 28 | 0.6)| £ N.W, 20.1] 17
TO) eh 20.1| 19
acel| i 20.1| 17
Garin Pyitsne||| at
Dec. 29} 10.9] c ebro =a) eL3
12.8] Snow | 18.1] 16
Besiine sun? 18.1| 20
6.2} Sun 19.1] 20
Dec. 30] 9.8] f N.W,| 19.1] 19
HOw | et Oe |
12.8] f W. 20.1 | 29
a7 pe Ws |] Dosie || 23}
6.0] f 20.1| 25
Dec! 32 9:9) £ W. | t9.1| 22
1:2) f 21.1| 24
6.5} f 20.1 | 22
1904 |Jan. r | 10.4| f W. | 20.1] 2
esa lee 2iet|| 33
5a) liek 2ieL\|| 20
Jan. 2 9.8} ¢ N. 20.1 ||| 9
10.2] c 20.1] 8
Ta || Pits ||
3.6} c Sun 20.1 | II
6.0] ¢ 20.1 | 12

—_——

Ww - FPReOHHHROWHRWOWHEEwWWWS HEHEHE HHPwWWWWwH HWWwWHEWWNH KH KHWHH HHH HP BPW WH WW

oo

Apertures.

HHS AKKWOHOMWOHAMABRHOHHKHKNADAOHOOROS

Ob CORN NKHHOKHHOHH ON

H

Corona
Colors

aaa
Sou 4
“0g of

que. a oatoa St uogioa el
2 oT GQ oT OQ *
O oe 09

OQ
was Oo
TTstoeoeto0o0 0 OTO
Ceo Cire rir

gogo
Sag

wt

|

Remarks.

| Number n.
|

53000
57000
38000)
38000)
75000
80000)
85000
75000)
88000)
85000
85000
49000
3 1000
30000
17500| Repeated same.
66000
70000
75000
49000

35000

80000
73000
49000! Repeated same.
57000)
57000
55000)

88000

80000
80000
75000
38000
49000
53000)
80000)
75900
62000)
35000
83000
66000
64000
85000 Repeated same.
42000
59000
44000
80000
80000 Snow.
goooo| “

70000 Repeated same.
75000)

182 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

a. | feo :
ee) a2) 3
= ; as 8 & e 4 2 Remarks.
eo NS | 6 ee ea cere as gs E
5 3 | 8 2 = of] o5 a O° 3
> A & = Bjad|ed|] < 58 Z
1904 |Jan. 3 | 10.6| f’ 18.1] 8 4.4| v’|b|p 100000)
10.8 | f’ N. LO-E) || 4.2} gbp 100000
12.0] f EQnL | ey 4.3] v’ p cor 100000
Teal 1g.1| 9 4.0; gbp 100000
te 19.1] 10 1/{gsbp } 100000
5| f 9 4 12 ie
5.6| f 19.1| 4 4.0|wrg 75000
{ 4.3| wo bg
6.5| f TOSU | 43 Mee ia \ goooo
Jan. 4 9.6] f W. | 17.2) -5 4.3|/g¢bp 100000
10.0| f£ sii fei| | £6) 4-3)|) 2 Do: T00000)
12.4] f L7s0)|) ot 4.2|¢bp 100000
ee 7A 17 eer 4.1) g¢bp 100000
323 ic 7s me 4.0| wog 75000
| a.7|¢£ 17.1} -$ 4.0| WOg 78000
6.0| f 17.1 | —2 4.0; wog 75000)
Jan. 5 9.4| f W. || 16%2)| 0 4.1|gbp 100000
> 41\/g¢bp
9.8| f TOV) Be oe eet 100000
12.6] f 16.1] 13 4.1|gbp 100000
4-5| £ N.W|] 16:1] 14 41\gbp T0000
6.0| f L7en [LS 4.0| g bp goooo
Jan. 6 9.7| f S. 16.1] 9 4.0\gbp 100000
DrOWah 16.1] 18 4.0| Do. 100000) At first v b p.
LoAG| | E D7 ty) 25 4.1| wo bg 83000
4.0| f S. W| 18.1] 25 4.1|/gbp 100000
6.4| f 18.0) 24 4.3| wo bg 93000
Jan. 7 Oot W...) } ren! 25 4-3|¢bp 100000
12.4| f 19.1) 29 2.7} g’|blp 25000) Repeated same.
3-71 c Ss 19.1 | 29 3.9| wog 73000 +: 4
6.0| f Fog 20.1| 26 3-9| wo bg 73000
Jan. 8 Os5ec N. 20.1| 21 4.1] g’ onot reg 80000
10.0| ¢ 19.1 | — 3.9| wog 70000
12.3) € 19.1| 29 3.2)|(cor 42000
3.3 | Snow 20.1 | 29 3-9| wog 70000 Repeated s= 4.0.
5-7| c Snow 21.1) 30 3-9|wog 73000
Jan. 9 9.7| Snow |N.W| — | 32 3.1 | up cor 38000
T2ereCe 21.1| 36 3-1] glblp 38000
4.0| Sun’ |W 22.1| 36 2.8| wo g’ 30000
620) || 22.1] 34 2.6 | cor 23000
Jan. 10| 10.5| f Wie a 23h 33 3.3 | up cor 45000) Repeated s=3.2.
12.8| f 22.1] 34 3-0| g|b|p 36000
5:4) £ 23.1| 28 3.3 | up cor 45000
6.6] f 23.1| 27 3-1] wlb|p 38000
Jan. 11] 9.2|f N.W]] 22.1] 21 4.2} wog 88000
10.0] f 22.1 | — 3.6| wre 59000
1.0} 4 21.1| 29 44|wog 95000)
3.44 £ 22.1] 30 3-1| wiblp 40000
A=5i) £ — | 27 Ber oO. 38000
6:3) ab 22.1| 26 3.6| wre 57000
Jan. 12| 9.6) f N. 19.1] 28 4.3| wo bg goooo
11.6] ¢ 19.1] 34 4.2} gbp 85000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 183

TABLE 1—Continued.

& Ho
24/25] ¢ :
‘ 85 a o
4 ° BS 5S 3 Sy 8 Remarks.
meee Pe | e | 2 heel ee] B es Z
3 3 5 2 elas cul es | a: Bed 5
~ A Ss = Flaed|ad| < oo Z
Rood ane 12) 1.3'| £ sea |e 3 3.4| wrg 49000
Naneets | Ou4) iC N. E.) 22.1] 35 2.8| w brig 29500
a5 eGu E. Bioeth au7 2.8) wo cor 28000
Be mlmculs: 23.1] 38 2.6|/wrg 22000
Oral uk: | 22.1 | 40 2.6| wre 23000
Jan.14| 9.5) fS’ |S. Wij 21.1] 34 3.4, weg 49000
p26 tat 19.1 | 37 4.2| wo bg 85000!
egal TOE) 37 3.7| wre 62000
Beas 20.1} 36 3-.9| Wog 73000
5.6| f 2m) 25, 3-7| wre 62000
Jan. 15) 9.4) f AVV iso fare et feo ae |e" bribes: 80000
Tg} | IE 2 Tel 30 3.8) wee 68000
Besulet 21.1| 28 3.4| wre 49000
Grr lt 21.1) 24 ee wiblr } 36000
2.9 | cor
Jan. 16] 9.8| f’ S. W| 21.1] 23 3.4 | up cor 49000
rise fa Bie | 215 3.9|wog 70000
4.4| c Snow|S. 21.1 | — 3.1 | cor 38000
6.2! R 20.0 || 38 2.8 | w|blp 28000
Jan. 17| 10.6] f Wis |p2 2h 1028 2.9| g’|b|p 31000
eG 22.1| 29 3.3 | w br cor 47000
Caza 2251|| 22 3.0| w\b/p 35000
6.5 | f BET e2O 3.0| w\b\p 35000
Hanes ors lit N.W,| 20.1! 10 4.1} gbp ITO00000
O97] || 4% 20.1 | — 4.1] gbp 100000
Tesi |e 19.1| 14 4.1|gbp 100000
4.0| f 20.1| 14 4.1) gbp 100000
5.9.| f 2OnTy (ere 4.2| wo bg 85000
Jan. 19| 9.6/f N.W 18.1] 0 4.4|g¢bp 100000
10.0| f 18.1 4.3 | Do. up 100000
11.9| f EOS 4.1|/gbp 100000
3.8| f 18.1] 13 4.0|wog 75000
6.0| f MOL) |e 4.1|¢bp T00000
Jan. 20] 9.5| Snow |N. 18.1 | 16 4.6 | wp cor tooooo | To wbhp
9.8| Snow 18.1| 18 A24\ DO: 100000
11.8| Snow |S. W.| 17.1) 24 4.2|gbp T0000
2.0] Cc 18.1| 26 4.1/gbp 100000
Aare 18.1 | 29 4.2| y’ o bg goo00
6.0] ¢ 1g.1| 28 4.0| wrg 75000
Jan. 21] 9.4] c¢ Nee?) rors |\'32 4.0|wrg 75000
11.5 | Snow 19.1 | 33 2.2 | cor 13000
11.9] Snow TG e-1 | oe 2.8 | cor 28000
3.8| Snow 19.1] 32 2.8/9 bp 29500
6.0| Snow 20.1 | 32 3.6| weg 57000 | Repeated s=3.7.
Jan. 22| 9.8/R N. EJ) 21.1} 38 2.5 | cor 20000
12.9| R 21.1| 36 2.9|wog 31000
Bea Re 22.1} 30 2.9|gbp 31000
6.0} R’ 21.1 | 30 3.1] w br blr 38000
Jan. 23! 9:8\eFog |S. 22.1 | 44 4.0|/wrg 75000
to.1| Fog | — | — 3.5|wege 53000
1.1| Fog |S. W.| 22.1] 50 3.4] w br cor 49000
4.8; Fog Vimeo a At 3.2) Do. 42000

184 A CONTINUOUS RECORD OF ATMOSPHERIC’ NUCLEATION,

TABLE 1—Continued.

8 EG :
; 3 : BE 3 o Z | 8 oi 3 Remarks,
¢| 2 | 2) 8 | 2 eel ge) ee g
Ss A a = EB laed}aed| a | 58 Zz
1904 |Jan. 23| 6.3) Fog 23.1| 46 4.0| Wog 75000} Repeated same.
Jan. 24] 1o.1| f N.W|| 21.1] 41 2.8| wy cor 30000
12.4) £’ 20.1| 42 2.8| glb p 28000
4.7| £ W. | 19.1] 36 3.3 | Wp cor 45000
6.4| f 19.1] 34 3.4| we 49000
Jane 25)|) 9:0) Wears ae ny 4.1) gbp 100000
12.4| f TO. | een 4.1|g¢bp 100000
BVA \\ab 16.1| 22 4.2|wog 85000
5-8] f 7 otal 22 Atl) Doe 80000
Jan. 26| 9.3) f N. 18.1] 19 . 2.8| w|blp 28000
DE 7c |E. 18.1| 26 3-7| wre 62000
| 3.4| c Snow 18.1 | 28 2.7| wre 26000
5.7| ¢ Snow 19.1| 28 2.7) | COr 25000
Jan. 27! 9.6] f W. | 20.1] 22 3.8| wre 66000
Te 2) |e 19.1| 24 4.2| wo bg 86000
1.0] f 20.1 | 25 4.1) wog 80000
SQ) ink 20.1 | 20 3.4| weg 51000
Jan. 28} 9.5] f 19.1] 15 4.3|wog 93000
I.0| Cc 1g.1| 22 4.1)wog 83000
6.1| c’ Fog 20.1} 24 3.8| wre 68000
Jan. 29] 9.5| Snow |N. 1g.1| 22 4.2| gy|blp 86000
11.4| Snow’ 19.1] 25 4.3| y’o bg goooo
3.2| Snow 20.1 | 28 4.3) Wy g 93000
5.8| Snow — | 24 4.4| wo bg 95000
10.6| Snow 2I.1| 22 3-5 | Wp cor 55000
Jan. 30/ 10.0] f |N. 20.1 | 22 4.2} wog’ 85000
11.9 | f 20501) 27 3-7| wre 62000
4.3| f 20.1 | 30 2.8| w|b|p 28000
Gog et 21.1| 32 ce ee 100000
4.3| wor
Jan. 32) 10.7) £ S. WJ 21.1] 35 3.8|wrg 66000
1.0| f 21.1| 38 3.5 | wp cor 55000
aot 21.030 2.9 | cor 33000
6.3 | f 27. ths 0 2.9 | cor 31000
Feb. 1 9.7| c N.W| 21.1] 39 4.3) yo bg 90000
12.6| c W. | 20.1] 40 3.6/wre 59000
1.0} Snow’ 18.1] 40 3-0 | wlblp 37000
4.0| Cc TOL 33 3.0] wo cor 35000) Leak—broken
5 Ol|Kc 20.1] 30 3-0 | cor 35000] glass.
6.0] c 21.1| 29 3-0 | cor 35000
Feb. 2 | 9.71 f WwW 20.1 | Ir 4.2|g¢bp 100000
EE. 7 ak 20.1] 16 4.1 | “Do! 100000
Toi le S. 20.1] 20 4.1| g’ bp 100000
4.0| c S. W| 20.1] 21 4.2} wre 85000
6.5] c 20:5)|'—— —|— ==
Feb. 3 ges W. | 21.2] 22 4.1] wog 80000
12.7| f 24.1] 24 4.1] Do. 80000] After repairs.
3 Tale 2a) 23 35|/ weg 55000
6.2] f 21h | 22 3-4| wp cor 51000
Feb. 4 Qe7ulich W. | 20.r| 18 3-9| wre 70000
L220) ih 1g.1| 21 4.4| w br bg 95000| Repeated same.
6.0] f 21.1| 18 3-9| wrg 70000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION 18

TABLE 1—Continued.

Becelirealit a
3 gel gal e : os
a : OH YO 5 By = emarks.
Meee |e eelee) g | ak |
oO Ss =} = oA on a 0° 5
- A & = FE led}ad| < OO Zz
1904 |Feb. 5 9.6} f W. — |} a1 3.6| wre 59000
m2 aol TO aie F215, 3.9| wre 70000
4.0} f 19.1] 25 2.7)| Wat £ 64000
Gqa6 |) 38 [9.1] 24 4.2| wy g’ 86000
Feb. 6 9.6] c Snow 20.1 | 29 3.1 | w br cor 38000
Tene.) |e TO tales 2021s Do: 43000
ZEON Zio) ate eee Basile Wo: 45000
Sols || GARY 20.1 | 29 2.8| gbp 28000
6.0] c 20.1 | 30 2:6)|| Dor 28000
Feb. 7 | 10.0] ¢c 2I.1| 39 3.6|/ wre 57000
12.5 | ¢ Fog 21.1| 46 2:9) ¢ bp 31000
eae WwW 23.1| 48 2.8) gbp 28000| Apparatus re-
6.2] ¢ 23.11 50 2.8| Do. 28000| paired.
Feb. 8 9-4|f Wie e220 | 24 Ba Tal sweet et 64000
11.9| f 220 |) 20 4.3| wo bg go000
Heal (et art 26 4.3| y’ o bg goooo
3.4| f 22.1| 24 4.3| y’ o bg goo00
6.2| f 22a | sO) 3.6|wreg 57000
3 f3-9|wrg
A : .W| 20.1] 10
Feb. 9 Osa) t N.W] 20.1 aan et 70000
12.0] f 19.1] 14 4.4] wo bg 95000
ep e ZOD) |) 15 3.6|/wrg 57000
6.3 | f Ig.1| 14 2.6) Do 57000
Feb. 10] 9.3| f. W TStr| 30 Anti) Seep 100000
pie lied 18.1] 19 itgae|) WADYoy 100000| 4o *
eal te 19.1 | — 4.1| Do. 100000] 41 *
4.8| f 20.1] 25 4.4|/wog 95000| 42 *
Gre} it 20.1] 26 3.6| wre 57000
h Feb. 11} 9.8] f N.W,| 19.1] 15 4.2| wo bg 85000
TDL0)|| 3 18.1] 21 3.8) wrg 66000
Re | ae 19.1| 23 3.1 | wlblr 40000} 43
BOs Cy 19.1 | — 3.1 | w br cor 40000] 44
4.6] ¢ 1g.1| 26 3.0 | w|blr 35000] 45
oO) LC 2050 | 923 3.7| wre 62000] 46
Beb: 12))| 925)|| £ N. TOS || avy 4.0| wrg 75000
mae Ne Ig.I| 24 44|wog 95000] 47
oes TQ) |(—— Sa eWwete 64000] 48
Seo 2OsTi| P27 4.2|g¢bp 100000] 49
ARH, 20.1 | 28 4.219’ bp 100000] 50
6.0] f 2Te Ty 27 4.1) wrg 80000
Feb. 13] 9.6] f N. 20.1| 29 4.1|/g¢bp 100000
10.0 | f — |— Asta DOE 100000
lee2 a |G 20.1| 33 4.4|/wog 95000
Ae2| © 2OuLI 31 3.71 wre 62000| Repeated same.
6.0; Cc 20.1 | 32 4.1] WOog 80000
Reb! 14) 10.5) ¢ N 21.1| 28 3.1] w br blr 38000
I.0| Cc Niet es 2.9|WwWog 31000
5-4] Snow 21.1| 30 3.0 | cor 35000
6.5 | Snow 21.1 | 28 2.8| wr g 28000
Feb. 15] 9.6| Snow |N.W] 21.1 | 29 (ae gbp} 100000
4.1| wog }

* These numbers refer to photographic slides of the fog particles.

186 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

S Ro
Su | Si 4 :
u se| as 2
& : 58 2 5 & ai & Remarks.
ey gee eae ial BS é
>| A |a&| BE | B lad|ea| < 85 Z
1904|Feb. 15| 11.5} c 2T-0)| 3x 3.0| g’ bp 35000
Becilet 21.1] 30 3.0/ x bp 35000
6.0| f 2B P22 35) | Ww 228 53000
Feb. 16] 10.0] f W 19.1| 6 4.1|/gbp 100000
12-4 20.1| 9 4.2|ebp 100000
Teale 20.1] 9 4.2] Do. 100000
As2)| tf 19.1| 9 4.3 | w yo bg 93000
5.9) £ — 8 3.6) weg 57000\Cube apparatus.
—— ie 3.6 Wwe g 6“ “
622) re re 57000
lFeb. 17| 9.5] f Ww — |12 3.9| wre 7oooo| “ r
11.6| f —|| 25 4.2| wo bg 85000| “* ii
12.6| f — | 19 3.9| wre 73000) “ .
6.2| f — | 18 3-3 | w br cor 45000| “ -
Feb. 18| 9.8] f N.W| — | 18 4.2| gy br b’ 88000| “ *
{cara emf — || 25 4.1] wog 83000] “ -
4.1| ¢ — | 30 3-6) weg 57000| ~~ ti
Arr \ic 20.1 | 29 3.8|wrg 66000| Long apparatus.
6.2] ¢ 21.1| 28 3.8 | w rg 66000
Feb. 19| 9.8] cSnow’|S. WJ 20.1| 26 4.0| wog 75000| Repeated same.
3-9) wog
Tre3 Wc 20.1| 27 ee os = » 73000
nates) |e 2 On tal 4.0]; wog 78000
nA vIKe 20.1| 28 3.9|wreg 70000 | cock
WAL Ne Si 3-8) weg 66000
12.8 | Snow 20.1| 27 4.0|wyg 78000 \ Fl
1.0| Do. 2050 |— 3.9|wog 70000 é
4-05 | Don sien eect 3.1 | w br blr 38000 \ Fl ‘
5e2i) Dor 21.1, 26 3.7| wre 62000 f
5-5 | Do. 2 t 3.6) wr cor 58000 Ck
5-8| Do. 2M 3.6! w p cor 57000 :
Feb. 20] 9.7 |f N.W|| 20.1 | 23 4.1| gbp 100000] Repeated same.
D2bAN | ih 20.0) | 3x 4.3 | wo bg 93000
1.0 |f 20.1 | 33 4.3|wyg goooo
Beales = | 30 3-4| w br cor 49000
Anat 2ite Tales )5 3.4 | w br cor 49000 i
Gah iid 22s a 4.3 | wo bg goooo \ Fl.
Oty ||ae 2 ORT 4.2) wog 85000
Feb. 21] 10.4 |f N.W.| 20.1 | 29 132 Wes a 73000| Fl,
4.0| wog
10.8 2050 | 35 4.1] wog 80000) Ck.
11.8 amare 3.7| wre 62000} Ck.
12.0 ATT | 3.8| wre 66000] FI.
T2k2 2 ee 3.4 w br cor 51000] Fl.
1.2 21.1| 36 3.1| wl|blp 38000) FI.
Beane 22°1| 35 3.1| w br bir 40000} Fl.
Ona ak 23.1| 35 3.1| wb’ p 38000) Fl.
6.5 |f 2ST oe 3.1| wb’ p 38000} Ck.
Feb. 22| 10.0 | R’ S. W 22.1] 49 3.0] g’|b’|p 35000
TON eRe 23.1] 49 3.2 | wp cor 42000
BE 2ui'C 23.1 | 43 2.4| cor 17000
1 FL

refers to the instantaneous valve (Chap. VI, 2 33); Ck., to a half-inch plug cock.

A CONTINUOUS RECORD OF ATMOSPHERIC

TABLE 1—Continued.

NUCLEATION. 187

B a
sion ||, 3.3 G :
5 #2| 28] 2 5
. e iF 5 = Seer eese 2 av 2 Remarks.
Meee fe |e lgelee| &| 88 | §
> A & = BE led|ea| < 66 Z
1904 |Feb. 22| 6.0] f 23-1 | 38 2.7| Wr cor 25000
Feb. 23| 9.9] f 20.1 | 36 4.3| g’ br bg go0000
ee ee 20.1] 4 4.1] wog 83000
TMG a |e E 20.1| 40 4.4| y’ o bg 95000
4.9|f 21.1] 40 3.5 | Wp cor 53000
6.2| f 1.1] 40 3.1] wiblp 40000
Feb. 24] 9.5| cSnow|S. E, 22.1] 36 2.8|wrg 28000] ~
12.0| Snow |N. E| 38 2.9|wog 31000
1.2| Do — | 37 3.4| Wp cor 5 1000
BeAultC N. 20.1 | 37 3.0| w olg 36000
6.0] c a5 3.0| WO cor 35000
Feb. 25| 9.8| f Wise eon |fno Aeiall SOND e 80000
12.1 | £ 19.1| 18 An) | YOre! goo0o
12.8| £ Ts 518) ||— 4.1|gbp 100000
3.8| £ 19.1] 19 3.7| WIg 62000
6.3 | f 20.1| 16 3.4|Wwcg 49000| Repeated same.
Feb. 26| 11.7| f MORE |e 3\2 4.1| Wor g 80000
1.0| f 19.1] 2 4.1| Wrg 80000
Bale |) ak Qe | P25 3.2| g br bip 42000
5.0| f 19.1] 25 3.5 | Wp cor 53000
Feb. 28| 10.5] f 8. Zour (32 2.9| g|bp 33000
TO4|/C Tes yl 315 2.9| g’|bp 3.1000
5.0| R’ S. 20.1 || 39 3.2| wiblr 44000
6.3) R 22.1] 40 3.3 | Wp cor 45000
Feb. 29] 9.5| ¢ N. Ej 22.1] 38 2.8| w rg 28000
ETS a Ac 22.1| 41 2.9| Wo cor 33000
Dee De miaeat7 2.5. | w br cor 21000
6.4] Cc 2B ra 338 2.4 | cor 17000
Mar. 9.8 | ¢ Snow 212 || 9315) 2.5 | wbr cor 20000
Tee E. 2a eae7 2.8| wr cor 28000
oil © 22.1| 38 3.2| w br blr 42000
6.0] c 22.1 | 36 2.7|/wrg 26000
Mar. 2 | 9.6] f N. E} — | 3 3.2 | WO cor 15000
ng) ak S. 21.1] 40 3.7| welg 62000
{ 3.0| g’|b[p
aril a BiTety ss ee Se 36000
Mar. 3 OaG)|| USS Ss 22.1| 44 3.0| g’|bp 35000
12.0| R 22.1| 46 2.9| g’|bp 31000
2.7| R’ Fog) 23.1| 48 3-5 | w rg 53000
4.2| R’ Fog 23.1| 48 2.8| w olg 28000
pay == |= 3.0| w olg 36000
Mar. 4 | 9.8| f N.W| 19.1] 20 4.2|gbp 100000
12.0| f 20.1 | 25 4.3 | y’ o gb 90000
Teale 2 Tet| 20 4.4| Do. 95000
Bi) 38 20.1| 25 4.1|/yog 83000
Ge Ot hee 21.1 | 22 4.1 |-y’ og 83000
Mar. 5 Oezhiict N. 2Our 7 4.1| wrog 83000
TeecH et 22.1 | 27 3.2 | w\b|p 42000
3.6| f 20eT| 27 29|wog 31000
Eee alee S. 20.1] 26 ee gbp 35000 \
2.9| wor 31000
Mar. 6 | 9.8/c¢ Se ealezorra | 32 3.5 | wrlg 53000

188 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

g Eo
Beas lear 2 Ss
5 S38 [cee & 5 R ike
, : a ee gE] 22 z ui & emarks.

| 2 |e] 3 | Eee ee] | 88 | 3

» | A |& = B |ed|aa| < 88 Zz
1904 |Mar. 6 |12.1 /c Ss. 21.1] 34 3.0| wo 35000
4.0 | ¢’ 2a |G 2.6| wr cor 22000
6.2 | c! pies oe 3.2 | w br cor 44000
Mar. 7 9.8 \c S 22.1] 44 3.2 | w br cor 44000
12.1 | R’ 21.1] 46 3.7) weg 64000
4.6|R 22.1| 44 2.7 | wr cor 26000
6.0 | R! 22.1| 46 2.50 COG 20000
Mar. 8 | 10.0 | f’ W 201 | 5 x 3.1 | wbr cor 38000
12.6 | f’ —— e555 350 Dor 35000
OnE it 222% AS 2.55| (COE 21000
Mar. 9 | 10.7 |f W... |P22s0/"40 4.0| wrog 75000
11.9 |f 19.1 | 35 3.9|wrg 70000
Beale 19.1 | 33 3.3 | w br cor 47000
5-5 \f 20.1 | — 3.2) Do: 42000
Mar. 10] 10.0 |f W. | 18.1] 27 era eae lien 100000

TAS ih 18.1] 30 3.9|wrg 70000| Cold wave.
T2.5 | £ LO |ea2 3¢5)| Do: 68000
5-8 |\f TOsL |r 3.4| w br cor 49000
Mar. 11| 9.7 |c N.W] 20.1] 34 2.8 | cor 30000
TeneAaic 20.1] 34 2.5 | w br cor 20000
3.8 | Snow 20.1] 31 2.6| wr cor 23000
6.6 | c 20.1 | 30 2.9} w br cor 33000]
Mar. 12| 10.4 |f N.W,| 20.1 | 32 4.1) gbp 100000
aoe 21.1 | 38 3.2 | w br B|p 42000
6.2 | f 21.1 | 36 Bean Dos 42000
Mar. 13) 10.6 |f N. area 3.5] wclg 53000
ease 2OnL Ness Be5iln DoE 53000
4.4 |f eT 7) 3.0) hou ep 36000
6.5 |f 2D. 0 ea 3.0| wbp 35000
Mar. 14| 11.2 |f N.W,} 20.1 | 37 3-9|wrg 73000
E228) | if 21.1 | 40 3.8| wre 68000
4.1%) Do.?

3.6 |c 2a aep Vee au 70000
5.O) |e! 2215 41]/wog 80000
Mar. 15|11.0 | Snow |N. 200 35 29|/wrg 31000
12.0| Do. Ziel le 3p 2.9|gbp 31000
325) |, Do: — | 38 2.9|gbp 33000
Graalic 21.1 | 36 270)|) Do: 31000
Mar. 16) 9.5 |\f 220 oi 3.6|wreg 59000
pele |e 21.1| 38 3.8|wrg 68000
5.0 |f 21.1 | 35 3.0 | g’|blp 35000
Grant DPT EBS 3.0| Do. 35000
Mar. 17! 9.8 |f N.W1| 21.11 30 4.1! g bp mixed! rooooo
DEO ge 2Tt | 33 4.1 | wo b’ 83000
TO eh Diets |S y7 BerTil) WACK: 64000
3.4 \£f 21.1| 40 3-0 | g’ cor 35000
5-6 |f 27.1) 3'5 3-4 | wp cor 49000
Mar. 18|10.1 |Snow |S.E.| 22.1] 35 2.9| w olg 31000
12.7 | Snow’ 22.1 | 39 2.8] wrlg 30000
BIO) || 1 22.1) 41 3-4 | wp cor 49000
5-0|R 22.1 | 39 3.1 | w br Bir 40000
6.5 | R’ 230 |e 3y7 2.8| w’ Bp 30000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION, 189

TABLE 1—Continued.

: | Year.

1904

Date.

Mar.

Mar.

Mar.

Mar.

Mar.

Mar.

Mar.

Mar.

Mar.

Mar.

Mar.

Mar.

19

20

21

22

23

24

25

20

27

28

29

30

| Time.

10.5

H
°

Dw
wn

ol

H

aS oe
PANNA ODAVNWOE

SROANHOSAKO

6.1

Z
8
=

f

f

f

f

f

f

f

f

f

f

f

c

Gay

CORY

R

ehtont

f

f

fi

f/

~

Stee

Ph eh ehiph re ehh ooo a00800hO0A0400000 0 mre rh rh

u a
p1ee ae) 2 ad 3
=e |es|ee) & 8 § g
EF jed|aa| = 68 Zz
W. 22.1 | 44 tee g|b|p 33000
Y P3208 war: 57000
eae (3-2 | g|Blp 38000
22.1 | 48 3-5 |wrig 53000
23.1 | 44 3.0 | w O cor 35000
W. | 22.1 | 49 3.0 | Wo cor 35000
21.1 | 48 28\|wrg 30000
20.1 |38 2.4 | COr 17500
N. mG st. |)-eXo) 4.0 |g’ bp 100000
17.1 |— 4.1 |g¢bp 100000
18.r | 45 4.0] Do. 100000
18.1 | 45 3-0 | wiblr 35000
18.1 | 40 320) Do: 35000
5. = 2 2.9 | wo cor 33000
20.0 | 42 3-0 | g’|B|p 35000
20.1 | 41 2.8 |e’ bp 30000
21.1 |44 28/wrg 30000
N.W.| 21.1 | 53 28/wog 30000
22.1 |59 4.3 |wo bg goooo
22.1 |59 3.8 |weg 66000
2/2 aiealioie 3.0 | wo cor 35000
22.1 | 50 2.9 | g’|B|p 31000
N.W.| 19.1 | 49 4.2 |wog 85000
10) || —— A.3) | Do: goo00o
Doin |) 3-5 |weg 53000
20.1 159 3.6 | wre 59000
o 20.1 | 45 3.6|wreg 59000
iS. TSex |S 4.1/wrg 80000
Saal lama 3.5 | Wp cor 53000
20.1 | 54 253)" Do: 45000
2 eta icy 28|wre 28000
aire 57 28\|wreg 28000
S. W.| 21.1 | 61 3.0 | g|Blr 35000
21.1 |— 3.5 |wr 35000
22.1 |58 3-1 | wlblr 38000
Pater || Ge) 2.1 | cor 13000
21.1 |54 2.2 | cor 13000
E. 18.1 | 40 3-3 | wp cor 45000
18.1 | 41 2.9|g’ bp 31000
D7 138 2.3 | WY COr 16000
17.1 | 36 2.7 | w tig 25000
IN.W.| 16.1 | 38 3.6 | w ple 57000
8. 16.1 | 41 2.9 | wo cor 31000
17.1 | 39 2.9 | g’|BIp 31000
17.1 | 36 2.8|/wbp 28000
N.W.| 18.1 | 37 4.1 | gy o bg 80000
18.1 | 42 3.4 | Wp cor 49000
W. | 18.1 | 44 3.0 | w|BIr 35000
20.1 | 42 3.5|weg 55000
N.W.| 21.1 | 42 2.8/9’ Bp 28000
8. 20.1 | 48 3.8|wrg 68000
20.1 | 44 29\2e’ Bp 33000

Remarks.

—S—_”

Repeated same.

Repeated s= 2.9.

190 A CONTINUOUS RECORD OF ATMOSPHERIC: NUCLEATION.

TABLE 1—Continued.

| o o .
‘ aad o
3 F Be Do 5 Sy 3 Remarks.
4 3 3 6 Jd | ad | ao P ee
a 2 g 3 & | gal] 8 £2 &
o Ss =i om | or a 36 2
a A a = Ee jadi|aed| < OO A
1904 |Mar. 30 6.0] f 2om0 | Ar 3.2 | Wp cor 42000
Mar 30) os) t EL S 18.1 | 42 2.9 | wo g 31000
EL-AWG 18.1 | 49 3.5 | Wp cor 55000
3 TA RS eso oe 2.9|Wwog 31000
6.5) R 20.1 | 40 2.4 | w br cor 17000
Apr. 1 9.51 R 20.1 | 4I 2.4 | wbr cor 19000
12.7) R 20% |iefir 2.6 | Do. 22000
6.8} R’ 22.1 | 46 3.0|wBp 35000
Apr. 2 |) 923i |i 27m (Ay 2.9/|g’ bp 31000
sips) Ce iene |i 3.0 | wo cor 35000
heel (ey 22.1 | 54 2.9 |wog 33000
Grail! 21.1 | 48 2.8 | w|b|p 29000
Apr. 3) | rossi .| 18.1 | 43 B07 Wwe Ee 64000
Teshle 18.1 | 45 BU7| DO: 61000
ATC 18.1 | 41 2:0) (Sb xp 33000
O:siIRC TO-r 37 4.1|}/wog 80000
Ouse 18.1 | 37 3.7 |weg 61000
Apr. 4 9.5| f 19. | 37 4.2 |y’ o bg 85000
Mn ick 18.1 | 46 3.8 |wrg 66000
Seale 1g.1 | 51 3.2 |W p cor 44000
3.2 | wibir
5-.4| f 20.1 | 50 ee ul 42000
Apr. 5 9.0} f 18.1 | 48 41|/gbp 80000
T2).50l ee OE 3.6]/wrg 57000
6.0| f 5. TOR 153 3.8 |weg 66000
Apis. OM) Onriliet E. 18.1 | 56 4.2 | y br bg 83000] 6.5/82 *
11.4| f S. — |60 4.1|wog 80000] 5.5/74
Bueilk 18.1 | 58 3.5 | Wp cor 53000
3.91) TOC 3.1 | wbrcor 38000] 5/37
6.6] f 18.1 | 50 2.9 | w\|b|p 33000| 5/44
Apr. 7 9-5|R 18.1 | 51 3.0 | w y cor 35000| 5/40
1.0] R’ 18.1 | 53 3.2 | w|Blr 42000| 5/47
4.4| f 19.1 | 60 2.3 | wy cor 15000] 5/30
5-5| f 19.1 | 59 2.7 | wc g 25000) 5/35
Apr. 9 8.0| R 21.1 | 46 28\|wreg 28000] Night.
9.2| R 21.1 | 50 2.8 | Do. 28000
Apr. 10} 10.7] f£ 2am | 2.2 | wbr cor 14000] 5/25
5 ae 21.1 | 56 28|/wog 28000] 5/40
GaAs ee 20.1 | 60 2.7 |wrg 25000] 5/35
Oy) || a8 20.0 | 57 2.2: | COT 13500] 5/32
Apre Ur|) (9.2) |7eeRY +o) NeWeltosn igo 3.3 | Wp cor 45000] 5/53
12.4| R 19.1 |53 3.0 | w|blp 35000] 5/49
Ani |iee 2020) 153 2.3 | cor 16000] 5/22
BON! 20.1 | 52 2.9 | cor 31000] 5/46
Oraiilic? 20.1 | 51 2.8|wbp 28000] 5/45
Apr. 12| 9.3| R 19.1 | 46 2.8 |wo 28000] 5/37
Dike 7a ekee 19.1 | 52 3.0 | wlb|r 35000] 5/51
5-0| c 20.1 | 54 3.0/¢’ bp 37000) 5/49
6: ri |e? 20.1 | 54 3.8/wreg 66000] 5/62
Apia rails osailacs 18.1 | 45 3.0|¢’ bp 35000] 4/38
Teen lNee 18.1 | 48 29|/wrg 31000] 5/44

1 Subsidence data, showing a fall of fog level of 6.5 cm. in 82 sec., 5.5 cm. in 74 sec., 5 cm. in 37 Sec., etc.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

Igt

glee] . |
4 2 | 3.8 3 =
, ' a , gs 3 5 gy 2 Remarks.
cole ead | 6 | = |#alee| & eee 5
> A & e EF jeci|ac| < 3S Z
1904 |Apr. 13] 6.4] c’ 19.1 | 44 3.0 | w br cor 37000
8.3] £ 18.1 | 40 3.4 |W p cor 51000| 5/55 Night.
to.r| f 18.1 | 36 3.0|wog 33000] 5/41
Apr. 14] 9.0| ¢ W. |17.1 139 2.8|/wreg 28000) 5/36
1.0} R-Sn., yee ayy 2.8 |wog 30000] 5/46
Reealiet 17.1 | 41 29\gbp 33000| 5/42
6.5| f 18.1 | 39 3.0 | g’|b|p 35000
Apr. 15| 9.4) £ W. |16.1 | 42 3.1 | g|blp 38000] 5/48
Benes S. 16.1 | 49 3.0 | w O cor 35000
6.4] ¢ 18. | 45 3-5 | wp cor 53000
Apr. 16] 9.5| Sn.-R |N. E.| — |38 3.1 | wp cor 40000| 5/50
12.0! R’ Tete 30 Avr |tgaDyp 80000] 5/74
12.7 | R? 17.1 138 3-6 |weg 59000
B20) C N.W.| 18.1 | 42 3-6|/weg 59000
6.0} f 18.1 | 39 2.9 | w’|b|r 33000
Apr. 17| 9.6| £ W. | 20.1 | 40 2.9|gbp 33000] 5/50
rors) 20.1 | 44 3.0 | w br cor 36000] 5/50
3.2| £ 21.1 | 50 Byo)\|| 1Dios 36000
5.6| f 21,1 | 49 2.9 | w|b|r 31000] 5/48
Apr. 18] 9.4] f Sy 22.1 | 49 2.9 | wo cor 31000] 5/42
12.7,| £ piven (ves 2.8 | wog 30000| 5/36
ey ul) ae — | |5r 2.8|wreg 28000] 5/35
Apr. 19] 9.7| £ S. 19.1 | 54 3.0 | wp cor 35000] 5/45
TOES S|! 19.1 | 59 2.9 | Do. 33000
3.1| c’ R’ |S. W.} 19:1] 59 290|wrg 33000] 5/35
Apr. 20] 9.8] c’Snow 16.1] 35 3.7 | wre 61000
12.9| ¢c 16.1} | 39 3.1 | wiblr 38000
BeTAl Cc T7et; | 40 3.01] ¢ bir 37000| 5/46
6.3] ¢ 17.1] 40 3.0 | v’ br cor 37000| 5/39
PNprereie |e 1O.7) |) £ N.W.| 18.1) 50 4.0 }/gbp 100000] 5/83
11.8] f 18.1) | 54 4.2|gbp 100000] 5/78
Be6)| fi 18.1) | 54 4.0/gbp 100000| 5/72
NPG 22)| 9.5 | £ N.W.| 17.1,] 52 4.0|wrg 75000] 5/58
12.7 || £ 18.1, | 50 4.2|wyg 85000] 5/63
5.0| £ S. 1g. | 49 2.9 |W y cor 31000] 5/39
6.3| £ Ig. | 47 2.9 | w O cor 31000] 5/33
Apr. 23} 9.0/f 5. 16.1 | 51 4.0 |wrog 78000) 5/66
cae |p Teste |G 7 3.2 | Wp cor 42000] 5/51
4.4| £ 17.1 | 50 3.0 | w br cor 35000] 5/42
Grazie 18.1 | 46 2.9 | cor 31000] 5/42
Apr. 24| 10.0] f 5. 2Oe0 |S 1 3.3 | Wp cor 47000] 5/55
12.5 | £ 21.1 | 54 3.0 'wbp 35000; 5/52
cect ee 21.1 | 50 2.9|wre 31000] 5/38
PONS NOs RI 1S. 22.1154 2.7 |wrg 27000| 5/36
LQ.) £” 2r.1 | 62 _ 3.3 | Wp cor 45000] 5/51
BEOi |e 2z.1 | 62 2.9 | wrg 33000] 5/38
Graiete 21.1 | 58 3.0 | cor 35000] 5/43
Ia\jove, @H0)||" We)sier || 10 N.W.| 21.1 | 57 3.6 | wp cor 57000] 5/55
12.3))\ fi S. — |6r 4.0 | wrg 75000| 5/64
Esl 20.1 | 56 3.0/gbp 35000] 5/48
Aor, Dipl) O83) Se N. E.| 20.1 | 45 1.8 | wy cor 7500| 5/23
12.3 | R 20.1 | 46 2.0 | wr cor T1000] 5/25

192 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

é Bo ’
Ba 25 6 =
L Be /aa 2 -
‘ a e 38 ae 2 g a pe Remarks.
¢| 2 | 8) 8 |B leeleal 2 | oe .
> A & = EF jad|eq] < co Z
1904 |Apr. 27| 3-4| R 20.1 | 45 2.0 | wr cor 10000] 5/25
6.0| R 20.1 | 42 1.9 | cor 8500] 5/24
Apr. 28| 9.4| R N. E.| 22.1 | 46 2.4 | cor 17000] 5/32
paces le 21.1 | 48 2.1 | cor 12000] 5/24
4.0| R Pater Hits 2.0 | cor 12000] 5/27
4.7|R 22.1 |— 1.9 | cor 8500| 2/5 2/10 2/7 *
4.9|R = 1.9 | cor 8500| 2/15 2/6 2/6
CUCHRe 20st) 53 — |cor — | 2/13 2/11 2/6 9
Beale a 1.9 | wrcor 8500] 2/16 2/9 2/5
5-9| R 1.9) ||) — 8500] 2/16 2/12 —
Apr. 29| 8.8] R’ 22 aba Si 3-7 |wre 61000] 2/18 2/23 2/22
9-3| R’ — 3-0 | w|b|p 35000| 2/20 2/18 2/16
OFC = |= 3.0|wrog 35000| 2/20 2/15 2/12
TIS EC Peet Is 2.8|wrg 30000| 2/17 2/12 2/10
3-5| R’ 22.0 |\57 2.2 | wp cor 13000| 2/13 2/8 2/5
6a|ee 2ST I SV7 3.0 | wbr cor 35000
Apr. 30| 10.0| c’ N.W.| 21.1 | 63 3.0 | w|b|p 37000| 2/19 2/14 2/24
11.0] c’ (S) | 21.1 | 63 3.0|/wrg 37000| 2/20 2/14 2/12
2.3| R 1S. E. | 21.1 | 64 29|wrg 33000] 2/15 2/14 2/11
6.0} ce’ (S) 2Ter |(62 3.0|wbp 35000| 2/23 2/15 2/17
May 1 9.6} ftoc |N. E.| 21.1 | 59 1.9 | cor gooo| 2/13 2/8 2/6
12.0] c 21.1 | 62 2.7|wrg 25000
Serc. 21.1 | 64 2.5 | cor 21000
6.4| c! err tal 55 253 4\\COG 16000
May 2 9.0] f N.W./| 18.1 | 59 3-4 | wp cor 49000] 5/51
rie || N. E.} 19.1 | 63 3.0 | cor 36000
3.3)| £ E. 19.1 | 62 2.8 | w rg 28000
5.4| f 19.1 | 60 3.1 | w|B|p 38000
May 3 9.2| f N.W.| 18.1 | 57 4.1 ]/gbp 100000| Repeated same.
Tei ee S. 18.1 | 63 3-8 |wrg 68000
Seay 17-1 | 54 2.8|wrg 28000
May 4 Orsi |S. T7.0 | 6x 4.0|y 0 bg 78000
Lee wiee 170105 3.8|weg 66000
emul iek 18.1 | 74 3.1 | w br cor 40000
653) 18.1 | 67 3.1 | glblp 38000
May 5 9.2| f Wie nS ar72 4.0|wy bg 78000
12-3) |b 20.1 | 79 2.9 | w oO cor 31000
3.4] f 22.1 |82 2.4 | cor 17000
May 6 | 9.4| f N. E.| 18.1 | 67 2.8 | w brig 29000
12.2) £ 19.1 |70 2.9 | w br cor 31000
Sea dick 19.1 | 68 3.0 | w olg 35000
6.3| f S. 18.1 | 61 3.0 | g|b|p 35000
8.8| f C820 53 3.0|wbp 35000| Night.
9.7| f TS.0 [52 2.7 |W rg cor 25000
May 7 9.3| { Fog |S. W.| 17.1 | 59 2.6 | wbr cor 23000
11.9| f 5. 18.1 | 67 3.1 | wlblr 38000
4.0] f 19.1 | 74 2.8 | w o cor 29000
See) et LOeas 7a 2.6 | w br cor 22000
May 8 | 12.0] f S. W.| 19.1 | 74 3.0 | g’ br cor 35000
4.0| f 20.1 | 74 2.9 | w ylg 31000
6.0| f 20.1 | 67 2.1 | cor 12000

1 Subsidence over consecutive distances of 2 cm. each.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 193

TABLE 1—Continued.

e | 86 :
: Salyer | 2 =
3 3 Ee oe 5 e | 3 Remarks.
: é ~ ~ ag Qo | P sia 2
Peles: | 2 8 | 8 |ee\/ge| 3 22 E
> Aa | a4 = EB led|a<| < 68 Z
1904 |May 9 9.5|R E 19.1| 63 1.8| cor 7000
72.3) || R 19.1} 65 2.0] cor 10000
3.6] R’ 20.1| 66 2.1 | Wr cor 12000
May 10] 9.4| f W 19.1| 70 4.3| y’ br bg’ 90000
9.8| f LO.) 4.2| yo bg 88000 | 2/36 2/24
Teo el 20.1] 74 3.0] cor 35000
Beil S. 20.1] 72 2.9 | w|blr 3 1000
6.2] c 2 Opera 7a 2.9} Do. 33000
May 11] 9.3|c¢ N.W| 20.1} 64 2.2|wrg 13000 | Rain at night.
12.6] c S. 20.1| 68 3.4 | Wp cor 49000
Abs istic 20.1 | 70 3.0] wlblr 36000
6.3| ¢ 20.1| 69 2.9| 28’ bp 31000
May 12] 9.4] f N.W| 19.1} 63 2.4| cor 19000
ae2ileh N. E 19.1 67 2.6] cor 22000
5.6| f S. 19.1 | 64 3.0| wlblr 35000
May 13] 9.2] f N.W] 19.1] 69 3.5| wre 53000
Agra 19.1] 72 2.0| cor r1000 | Repeated s= 2.1.
6.4| f 19.1) 66 1.9 | cor 8000
May 14] 9.0| c’ Fog 19.1| 63 25k |\ COL 12000
L2G |C 19.1 | 66 1.9 | cor 8000
Ge2)|"¢4 19.1} 60 1.8| cor 7000
May 15] 9.5|¢ N. Ej 18 1} 58 1.8 | cor 7000
254 | 18.1| 62 2.0] Cor 10000
6.2| R 19.1} 55 2.0| cor 10000
May 16] 9.5] ¢’ (S) 18.1} 60 3.2] Wp cor 43000 | Rain at night.
oer 18.1| 62 3.0| wog 35000
Bes Coe 18.1} 59 2.9;/wog 33000
6.0| f 17.1| 60 3.0| w br blr 35000
May 17| 9.3] ¢ S. Wi 17.1] 58 3-3.) wlblr 45000
TSEC 17.1| 63 Buen Do: 42000
DeeleCd Tests | NOs 3.0] wog 35000
6.1| ¢ 17.1| 62 3.0] wiblr 35000
May 18) 9.2) c R’ |N. E} 17.1] 63 1.9 | cor 8000
WAG) || IR 20.1| 63 1.9 | cor 8500
Bese 18.1| 58 1.9 | cor 8500
May 19] 9.5|c N Teta 29|/wrg 31000
T= 9)|| (C 18.1] 59 1.9 | cor 8000
3.0| R 18.1| 56 Deerll (ore 15000
5-6} R 19.1] 55 2.8) wre 28000
May 20] 9.1] c S. W| 18.1} 63 2.8|wreg 28000
12.5| c’ (S) 18.1| 69 3.2 | Wp cor 42000
3.0] R’ Wile | nSer |Or 2.8) wreg 30000
6.1| f S. =O? 2.8|/wreg 28000
May 21] 9.5|f N.W| 18.1| 67 3.8] y’ og’ 66000
Tet lt 19.1] 73 3.0| wre? 35000
4.8| £ S. 20.1| 69 2.9|wog 31000
6.0] f = |) 3.2| Wp cor 42000
May 22| 9.8/f 8. 19.1| 68 2.4 | Cor 17500
10.8| f 19.1| 70 2.4| cor b 18500
1283) (et 19.1] 71 2.2} cor 13500
May 23] 9.7] £ S. 19.1] 67 3.0| g blp 36500
Tesh | 19.1| 68 3.0 oO. 35000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE 1—Continued.

194
Ba

a SH

vo Re

. 2 z s : as
g| 2 | 8] 8 |S le
~| A |a| & | & | és
1904 |May 23] 6.3] f 19.1
May 24] 9.3] c’ S. W| 19.1
12s 5, |e 20.1

a2) (at 20.1

May 25] 9.3] f N.W,} 20.1
Tae 7liae 20.1

ay|| ae 20.1

6.3] f Pe

May 26] 9.3] f S. aie
12.4) i 22.1

6.4] f 22.7

May 27] 9.8] f S. W] 22.1
12.0| f 22.1

3.2| R 220m

Gest 22e1

May 28] 9.5] f W. | 20.1
9.7|f 20.1

1.0] f 20.1

Seale 220

May 29| 9.6] f Ss. 2a
to.1 | f 21.1

10.8 | f£ ores

12.0| f Dat

4.5|f S. W,| 22.1

6.4] f 23.1

May 30| 9.5| ¢’ rig
LO}|\(C 22.1

DErilic 22a

6.4; Cc Z2aT

May 31| 9.5|c¢ N. E,} 20.1
12.3| c’ (S) 20.1

265) £ (S) | 20.1

3 | tie =

SrA ie Zieh

une: || “9%o)|Kc N. EJ 20.1
TLC 19.1

Zieh ols 19.1

5.8|R 19.1

June 2 | 10.4] ¢ N. Tepe
12.9| c sett

Aaa ic 18.1

Grailke 18.1

June 3} 9.8/c N. EJ 18.1
10.8| c 18.1

Lies an) 18.1

Teoh —

geen (S) 19.1

6.0/ f 19.1

June 4] 9.5] f N. E|| 19.1
wen Ls 19.1

Atmosphere.

Ass CONT ST ST ST ST st Ts st ST ON
[Sora anio eee onto oe ees

Aperture s.

HPWND HN ND NH HWWWWWWWRNHWBWNHDHNDNDNDNHNWNHNDDHDDND
TIOHROADH DWOnNHHAUM OHWHR OMOW WOWW ONAT00 0

NN
H~T

NNNNNNNNHHNHHHN DD
WO MADMAN H DOMWOHO OOK OU

cor
cor
cor
cor
wrg
w rg

wrg
wrg

Do.
w\b|p
cor
cor
cor
cor
wrg
cona
cor
cor
cor
wre
w|b|p
wr¢g
wrg
w|b|r

cor

Number u.

31000
31000
31000
25000
53000
30000
15000
16000
31000
28000
15000
31000
53000
36000
17000
80000
45000
38000
35000
53000
50000
49000
42000

7000
28000
17000
19000

8500
17000
35000
26000

26000
12000

12000
28000
19000

8500

7000

8500
12500

7000

7000
13000
13000
24000
28000
28000
28000
31000
15000

Remarks.

Rain at night.

Rain at night.

Repeated.

Repeated same.

Rain at night.

New apparatus
same.

New brass appa-

tratus. ico

peated s=2.8.
Night.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 195

TABLE 1—Continued.

: Ba/Ba| 2 ‘
‘ 2 ie Z| = ge 2a 3 a B Remarks.
g 2 g 3 & | go 88] 38 es g
ai a | & E Ej edle<| < 88 Zz
1904 |June 4 | 5.0/ f 20.1| 72 21/wreg 12000
6.5 | f S. W 20.1] 67 25|/wrg 20000
Hamels tess |) £ SV a ELOnI 7 27/wreg 25000
4.0| f 20.1 | 83 2.4 | Cor 19000
June 6 Oe7i\c N. E| 18.1] 60 Tes COL 4000
10.5 | Cc 18.1 | 63 TiS) | COL 4500
I2.0|C 18.1 | 66 1.9 | cor 8500
SO} 19.1 | 67 1.8 | cor 7500
June 7 | 9:5) ¢ N. 18.1| 62 2.2| cor 13500| Rain heavy at
pees! 8. WJ 19.1] 70 2.8| wrg 28000] night.
Ber eRe 19.1 | 69 2.9|wreg 31000
Be 2allek 19.1 | 68 2.3 | wrg 15000
June 8 | 9.9} c¢ See} 20.2 | 72 1.5 | Cor 4000 | Rain at night.
II.0| C 20.1| 74 fas! |NCOT 4000
H2ESH NC E. 20.1| 73 1.6 | cor 5500
rac lhe 20.1| 68 1277, |\| COr 6000
June 9 | 10.0/ R’f |N. EJ 19.1] 65 2.6|wreg 24000
12.0! f’ NOs |— 27l\ Dox 26000
June 11} 9.6] f N. 18.1 | 69 2.6 | cor 22000
pat || 18.1} 69 Zit) COG T1500
2x0) | N. EJ 18.1] 70 1.8 | cor 7000
6.0] f 18.1 | 67 1.8 | cor 7000
Jumiehr2|| o.5)| £ 17.1 | 66 2.2 | cor 13000
rarer ll ee E, 17.1 | 68 2.5 | Cor 20000
BEO) [et oper ea 1.9 | cor 8500
June 13) 9.4| ¢ > du) eG/ete || OL 255) |RCOE 20000
BEOECS 17.1| 66 Quel wit: 25000
June 14] 10.0) f foe || Oe 3.9| wre 70000
10.3 | f — | — 3.9| Wrog 70000
Be) lie 17.1] 62 3.0| wlb|p 35000
5.6] f 18.1] 54 2.6 | cor 22000
June 15] 5.5/f S: 17.1| 67 2.6| wiblr 22000
June 16] 10.3| c’ SHAVE leer e712 3.7 | wre 61000
10.5 | — — |— 3.9| 2’ bp 71000 | Repeated.
Teo eA |e retsheae iy 4.1| gbp 100000
3-6) RY 19.1 | 74 2) | 42000
6.0} c’ LO.0 || 72 2.8 | wl|blr 30000
June 17} 9.8) f N. EJ 18.1] 72 2.7| w tg 25000 | Repeated s=2.6.
ne2e ©} [ile 19.1| 76 2.7| wrg 25000
S201) £ S. 19.1| 76 2.7| wrg 25000
See |e 19.1| 72 2.8| wib|r 28000
June 18) 3-1] £ W. | 19.1] 74 2.8 | wi/bir 28000
rere 19.1 | 79 223) COr 15000
Bee lat 20.1 | 82 3.0| wlblr 35000 | Repeated s=2.4.
. Ora) at 2I.1| 79 2.6|wrg 22000
June 19] 9.5] f N.W.| 20.1 | 74 3.5| wre 53000
9.8 | f Zoe | 4.2| wo bg 85000 | Repeated.
TOn3\| hi 20 ee 3.6) wre 58000
June 20} r0.1| f N.W.| 20.1] 78 3-9|wog 70000
10.8| f — | 80 3.9|wog 70000
12.4| f 21.1| 82 3.6| wre 57000
226)|t S: 21.1| 78 2.3) cor 15000

196

A CONTINUOUS RECORD OF ATMOSPHERIC’ NUCLEATION.

TABLE 1—Continued.

& a :
+3 23 38 2
.| # ol eee) Ge) ae 2
¢| 2 | 8) 8 | 2 eel ge ed eee
a AQ & = EB |ad| ad] a OO
1
1904 |June 20 5.0] f 22.0 | 2715 2.4| cor
June 21) 9.8} f S. Wal 22sn a7 2.8 wien)
1 oe: 2.9 | w|blr
10.4] f 23ar ele ae
2.0] f 23.0 (185 2.9| w br cor
June 22) 9.1] f W. | 23.1] 81 2.8 | wr cor
12.4| £ 22 T NOS 2.9|/wog
Desi nCuines 23.1| 80 2.8|/wog
Beste 24.1] 78 2.6| cor
June 23] 10.2] f N. BA 2250) 7a 2.5 | cor
T 2. hoe 22.1] 74 2.7 | Wr cor
3.41 £ — |75 2.7 | Wr COr
eS ie 22.1 | 72 3.0] Ww y cor
June 24] 11.0| f W. | 21.1| 77 3.0| w obr cor
12.9] f£ 2:2. TeO) 3-4) wp cor
4.6| f Zeta 2.8) wr cor
June 25] 9.0] f S. W.| 22.1] 78 3.0] g’|B|p
2 a5eiat 24.1} 86 3.1] wog
er aHee 24.1] 90 ? 2.6| cor
50k 24.1 | 84 2.0| cor
June 26) 9.0} f W. | 24.1] 87 ? 2.7| Wr cor
12.7 24.1| 33.6C.| 2.7 | wbr cor
Br oi|| HR 24.1| 80 1-3) COL
4.2! R off 25.1| 80 28|wrg
5-6| f Or 2.6 | w br cor
June 27} 10.5|f N.W,| 23.1 | 81 3.3 | wp cor
11.0| f — |— 3.1 | w br cor
12.9| f — | 84 3-0| w|b|p
3.8| f — | 84 2.7 | w br cor
Geet 24.1| 81 2e2)|Cor
June 28 9.2| c’ EB. Pehle 2.3 | cor
12.0] Cc’ 23. 0)) om 2.6} cor
Ke sieC Dye vare | rf ‘2.8)wrg
June 29] 10.3] R’ S. E.| 22.1| 64 2.2] cor
Be2ke N. 23.1| 70 BeQueCOrL
6.0] ¢ 23.1| 69 2.2) w rig
June: gol ronmiic R42 S: — |7o 2.4| w br cor
ime {| {8 24.1| 73 2.9| wrlg
5-8| R Ss. 24.1) 72 3.0} w bcor
July 1 g.0| R 2AnT | aie 2.3) wrlg
10.8| c 24.1| 76 2.7 | Wp cor
3-3] Cc 24.1 | 75 3.0| wre
5-5) | (¢ S. W 24.1| 74 Ze CO
July 2 9.5|f S. W|| 23.1] 75 3.2 | wp cor
12231) £ 23 ar | eT 3-0] g’|blp
July 15] 11.0] £ 24.1 | 83 3-1 | w br bir
Genet 24.1| 77 2.5 | cor
July 16} 9.7]|¢ 8. W)] 25.1] 77 2.6 | w br cor
T2a 50 hice 25.1] 85 3.0; wrg
5-8|f 25.1| 79 2.0] cor
July: 17 || cox | N. — | 84 1.6 | cor
Oey ate 25.1| 86 1.6 | cor

&
3
§
3
z

17000
29000

31000

33000
30000
31000
28000
22000
20000
25000
26000
35000
36000
49000
28000
36000
40000
22000
10000
25000
25000

2500
28000
22000
45000
38000
36000
25000
13000
16000
22000
30000
14000
15000
I4000
17000
31000
37000
15000
25000
35000
25000
42000
35000
38000
20000
22000
35000
T1000

5500
5500

Remarks.

Old apparatus.

Violent thunder-
storm.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 197

TABLE 1—Continued.

g RS
BY| BS 4 s
5 Se oe 5 3 R k
Ee 3 B 3 = ene ad £ SY ai 3 emarks.
Bel 8 E Sec Wesei ee |) 1S ES 5
ES a a = Eled|ed| 88 | Zz
1904 |July 17} 2.5] f’ a 1.5 | cor 4500
5-0] f 26.1} 85 1.8] cor 7000
July 18) 9.5) f N.W.) 24.1 | 84 3-6) wre 59000
10.8| f 25.1| 85 3-4] wp cor 51000
1.6] f 5 25.1| 89 ? 2.8/ wre 28000
July 19| 11.3] f S 25.1| 81 2.3 | wocor 16000
12.9| f 26.1 | 86 2.3) Dos 16000
2.4| f — | 87 2h Doe 16000
Beye 20. T | 2:3) Do: 16000
July 20) 10.2] f N.W|| 26.1 | 87 3-0 | w rolg 35000
2.0] £ W. | 26.1] 88 2.7 | w br cor 25000
Rea ll tt 25.1] 84 2.2] cor 14000
July 21] 9.6/f W 25.1| 81 3.0] wrlg 35000
se2AO) |p 25.1| 84 3.0] wog 35000
Be2il|iet 25.1| 84 2.3] wrlg 15000
BeOllt 25.1 | 82 2-2))| COL 13000
July 22) 10.0] c’ E 24.1| 77 2.3 | w rbr cor 15000
TezeGh 24.1 | 81 1.6 | cor 5000
4.6| R’ 24.1 | 71 1.6 | cor 5500
6.0] c¢ 24.1 | 69 To) COL, 7500
July 23] 9.7| R’ N. E.| 24.1 | 66 1.4 | cor 3000 | Rain at night.
ng2s35 | 24.1 | 64 1.4 | Cor 3000
BeraG N. 23.1| 66 1.6 | cor 5000
Gato) || 2 23.1| 63 1.6] cor 5500
July 24) 12.2] c N SIGS 1.4 | cor 3000 | Repeated s=1.3.
6.0/ R 23.1| 65 1-2)|'sCOr 2000 fi S=I.1.
July 25] 9.5) c 23.1| 72 2.3 | cor 15000
TORCH cls 23.1] 79 2.4 | Cor 19000
6.0] t = 7O 2.6 | cor 22000
July 26] 10:7] c' R’ |S. 23.1] 79 3-1 | wiblp 40000
mona cs 24.1| 80 Batali DOs 38000
4.0| f 24.1| 80 Zia || 2DYoy 38000
6.0} f 24.1] 77 29|wrg 31000
vez ors |icoR. |W. | 24.1 78 3.1 | g’|blp 38000 | Rain at night.
peal 24.1 | 80 3.0|wyog 35000
BEAN Ce 24.1 | 82 230) |) Do: 35000
6.0] f 2am 77 2.9|Wwog 33000
July 28) 2.1] c’ See leea tS) 2.8/wreg 28000
O35) |"c 25.1 | 76 2.4 | W rcor 17500
July 29} 10.4] ¢ WwW 25.1| 78 2.7 | Wp cor 25000
12.7 | £ 24.1 | 83 3.0| w|b|p SICCo
3.3|f 24.1] 82 3.0|wreg 35900
5-6| f 25.1|/ 78 2.4 | cor 17500
July 30] 9.7] f N 23-1] 71 3-.8| wre 66000
10.0} f = |= 3.6|wreg 59000
Te el 23.6) 77 3.1|wog 40000
4.2] f 24.1 | 74 2.7 | Wp cor 25000
Gere t 24.1] 71 2.6 | w br cor 22000 | Repeated same.
July 31} 9.4| f S. W,] 24.1] 78 2.2 | Cor 14500
re ear at 24.1] 84 2.0 | cor 10000
. 4.0] f 24.1| 84 Toba COL 4000
6.0| f 25.1| 80 D5) cor 4000

198

| Year.

1904

Date.

Aug.

Aug.

Aug.

Aug.

Aug.

Aug.

Aug.

Aug.

Aug.

Aug.

Aug.

Aug.

Aug.

I

Io

26

2

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

H .
Time.

AWwnwo Nn Wnwo
HWN

4
onowot ws

©
oO
ROO Oma wWO

12.5

HH

HoH
Nown OMNW NO

0 DRUAAKAE

| Weather.

COG G

~

~

Q

QO
~

COnQie yale fiesMa Seal sues) Qo0

~

Aw

~

Phra rA RA RR RPh OQ 00 0 0 0 Mr eee

| Wind.

n
=

&

rr

DAR A AZAAPDON ASHADHHGH BHH
Sf a ee

fie

pnnnZ Aaa
ee

\jodex

| 24.1

TABLE 1—Continued.

Temperature
Apparatus.

25a
25.1
26.1
25.1
Asem
26.0

26.1
24.1
250
2Ger

al

24.
24.
24.
23%
24.
2%
24.
24.
24.
24.
24.1

Se + + SA SH SA SHS HS

24.1

24.1

24.1
24.1
24.1
23)

soa

23.1
23.1
23.1
22.1
2350
2a
24.1

2350

23.1

Gagan

22.1
22.1
22.1

22.1

Temperature
Atmosphere.

68
71
70
68

ean
to
‘oO

pe

YNNWww v Nw ow NON H Www ry www HNNHNNNHNHNDNNNNHNNNDHNHHHHAHNNDNNNNDND
MANOWB OO WOH DAHWAHH OB COIR AHDWHAHOSHKH OH DKHHNDAWROHDRRO OUA DD HATE

Aperture s.

Number n.

17500
17500
17500
31000
27000
13500
13500
28000
17500
25000
10000
31000

3000

3000

2500

8500
17500

7000
24000
I3500
13500
13500
30000
I1500
10000
13500

8500
15500
15500
15500
16500
23500
19000
64000
66000
66000
17500
36500
38000
38000

6500
11600
23500

40000

30000
33000
36500
45509

35500
20000

20000

Remarks.

Rain at night.

After rain.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 199

TABLE 1—Continued.

& a :
Dw pe a 4
5 Be/ Ea] 2 : Le
E 5 a , os VD a Sy v emarks.
a) ¢i/e| 8 | EES EE] & B8 E
eaicose | F | F leat} ca] < 88 Z
1904 |Aug. 28] 10.0/ f Weezer let 2EQNNCOr 13500
LDS) at W. | 22.1} 78 1.8 | cor 7000
Beselet S. Zee Tales 1.9 | cor 8500
Aug. 29] 10.5] f We reat as 3.2| glb|p 42000
Teter W. — | 81 2.8 | cor 28000
3.0] c N.W 23.1] 81 2.3 | cor 15500
Bron t N.W,| 23.1| 76 2.3 | cor 15500
Aug. 30, 9.2| f E. 22.1 | 66 2.8] cor 28000
22h E. 22.1| 72 2.4 | cor 17500
Ges et E. 22.1| 68 2.4 | cor 17500
Aug. 31] 10.3] f Whe || 22-1 |) 72 3.7|wrg 61500
t0.5| f — | — 4.0] W ro g’ 75000
Teena Viens eoizena 7/4" Se) Cor 38000
Bestel S. W.| 22.1] 73 3.1 | g|blp 38000
Seale — | 7o 2.8|wrg 28000
Sept. 1 | 10.6] f’ H 21.1| 70 3.1 | wlblp 40000
12.6] f 20. ia! 70 2.7 | Wp cor 25000
: BEO!| CA S. 21.1| 68 2.4 | cor 17500
Sept. 2] 9.8] c S. W|] 22.1] 74 2.8|/wreg 28000
G25 2)| 0c Wi. | 22.1) 78 3.1| wiblp 38000
Baty || 8. = ||4 2.7 | Wp cor 25000
6.0) f H Ss. BigeTa |e wA 2.6 | cor 22000
SEptes | LO:3) | £ S. 23.1] 79 3.2 | wl|b|p 42000
B20 ek S. 23.1] 79 3.0 | cor 35000
Beaulet Ss — 177 2.8| w|rlg 28000
Geaeliet Ss. Bian tye 2.3.) — 15500
Sept. 4|10.4|c R’ |N. 24.1} 74 2.3 | cor 16500
Dea W. 24.1 | 82 3.0] COr 35000
6.0} f N.W,) 24.1] 74 2.1 | cor 11500
Sept. 5 | 10.1] f N.W|| 23.1 | 72 3.1 | wiblp 38000
oie are ke N.W.| — | 74 3.1! Do. 38000
6.6| f Wien e237) ||4710 2.2 | Cor 13500
Sept. 6] 9.c| f NewBal 2 rer 62 3.1 | glb|p 38000
TREE Ne E20 67 3.01, Do: 30500
4.0| f S. W|| 22.1 | 68 2.8| wrg 28000
5.8| f S. 22.1 | 63 3.0| glb[p 30500
Sept. 7 | 10.2| f W. | 21.1] 60 Bont || Ode 38000 | Fog at night.
234 || Wiese |ie2it- 03070 3 ECOL 38000
Bere Ss. — | 68 2.4 | COr 17500
(Dein | 5% N.W|| 21.1| 64 28/wreg 28000
Sept. 8 | 10.6] f W. | 22.1] 75 3.1| wlblp 38000
Pah W. — | 78 2.8| w rig 28000
5.0/f S. W.| 22.1] 77 3.1 | glb|p 38000
Sept. 9 | 10.6] c N. E|| 21.1| 60 1.6 | cor 5000
Io] cR N. E.) — | 60 1.8 | cor 7000
5.01 c N. E.| — | 60 2.1 | cor 11600
Sept. 10) 11.8] c N. 21.1] 64 1.9 | cor 8500
21a |C N. 21.1] 65 2.3| wrlg 15500
6.0] f N. 22.1| 65 2.7 | Wp cor 26500
Sept. 11/ 10.0| f Fog |N. | 21.1] 69 2.3 | cor T5909
12.2] ¢ N, | amr} 75 3.5 | Wp cor 55000 | Repeated same.
BEONnCe Ss. 22m 1.8 | cor 7000

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

200
TABLE 1—Continued.
Oo oO .
Sip eSi ; :
Vu D
e| ¢ | a] $ ll 2 |eeiee)) ieee é
$ 3 & LZ = oa | 55 a, Sto g
m A G = FE le<led| << Oo Z
1904 |Sept. 12| 10.2] f We B20 1G 2.9 | wo 31400
oar S. W.| 22.1 | 81 2.5 | cor 21000
4.2) £ S. W.| 22.1 | 8x 2.2 | Cor 14500
Sept. 13] 12.0] f” E. 21.1 | 69 2.0 | cor 10000
Sept. 27/ 9.8] f N. 21.1 | 66 2.9 | wg cor 31400
Busine N. E.| 21.1 | 69 2.2 | cor 13500
Selec a O2 2.2 | cor 13500
Sept. 28) 9.4| c¢ N. E.| 20.1 | 62 2.0 | cor 10000
BTA N. E.| 20.1 | 67 2.9 | cor 33000
6.1| f 20.1 | 60 3.0 | g|b|p 35000
Sept. 29] 9.8| f S. W.] 19.1 | 66 2.9 | cor 33000
rine || S. W.| 19.1 | 65 3.0/gbp 35000
Sept. 30] 10.0] f IS. W.| 20.1 | 72 3:0 | w’|bl|p 35000
TT.9) |b W. |21.1 | 76 3.0 | g’|blp 36500
5.2| £ IR N.W.| 21.1 | 66 2.6 | cor 22000
Oct 2) |/Hio:on tt W. |19.1 | 60 3.2 | wp cor 42000
12.4) £ ——O2 28|/wog 28000
5.8] f WE) | Rol liss 2.8|/wog 30000
Oct: "2" || 1016)|"e W. {18.1 | 58 2.4 | cor 17500
12.3)|C W. |18.1 | 60 3.5 | wp cor 53000
Oucilnc? 19.1 | 57 2.9 | witlg 33000
Oct=3 9.2| f N.W.| 18.1 | 54 3.4 | wp cor 49000
5.0| f 18.1 | 56 2.5 | cor 20000
Oct24) || Tost N. 17.1 |58 3.5 | wp cor 53000
D223) | af N. 18.1 | 59 3:0) Do: 53000
Octas 9.7 | f Wie yee (Gm 3.5 | wp cor 53000
12.3'| f S. 17.1 | 63 3.0 | wirtlg 35000
6.0] f 1g.t | 58 3.0) Dos 35000
Oct. 6 9.6|R N.W.| 19.2 | 55 2.8\|/wreg 30000
Teale 19-1 | 59 3.0 | g|b|p 36000
Ans) di N. Ig.1 | — 3.0 | w O cor 35000
Octyy | 10!8i\ a N. E.| 17.1 | 47 3.9 |wog 70000
5.0| f N. E.| 18.1 | 49 3.0 | g|b|p 36000
Oct. 8 | 9.9] c’ IN@ Es ego 15/5 3.5 | wp cor 53000
12.6} c S. W.| 17.1 | 58 3.3 | wibir 45500
Seyi 18.1 | 56 3.3 | w\b|p 45500
Oct. 9 9.6| ¢c 19.1 | 63 2.4 | COr 17500
1:0)|'¢ 8. 20.1 | 65 2. aN ICOL 15500
6.6} c 20.1 | 56 2.2 | cor 13500
Oct. 10] 9.4] ¢ S. TO em |5 7 21.3) |iCOx 15500
12.3|c S. —— a0 223; |;COr 16500
5-0] f Ss. 20.1 | 64 3.1 | cor 38000
Oct. 11 |) rom | S. 21.1 | 68 3-4 | wp cor 51000
TG EC NSW nla 2 2:3) | (COE 15500
ALaulic N. E.| 21.1 | 61 1.9 | cor 8500
6.0} c OTe ta 2.1 | cor 11500
Octal ors) eRe N. E.| 20.1 | 50 2.3 | cor 15500
T2.4'|'¢ N. E.| 20.1 | 50 2.2 | Or 14500
6.n enk 20.1 | 44 2.2 | cOr 14500
Oct, 13) ‘or; N. 18.1 | 44 3.8 |wreg 66000
11.8] ¢ N. 18.1 | 46 3.2 | w|blp 42000
3.9) 1c N. 19.1 | 48 2.9|wog 33000

Remarks.

Rain at night.

Repeated s= 2.2.

ils

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 201

TABLE 1—Continued.

fe .| &3 4 =
a un - =
5 ge|/ ee] § 3
g 3 : 3 Z a8 ate z s g ; Remarks.
o 3 5 g Ra Ho) && a Be 3
a A & = ames eee = 56 A
1904 |Oct. 14| 9.5/1 N. — | 50 3.6| wre 57000
12.4| f£ 20.1 | 58 4.1| wo bg 81000
4.0} f 20.0) 517 2.6| cor 22000
6.0| f 20e0 (N52 3.1) wrg 40000
Oct. x5) |) 9:5) £ N. 19.1 | 49 3.7 | Wp cor 61500
TORS) cb N. Ig.1| 54 3.1] g’|blp 40000
6.5|£ 20.1] 50 3.0|gbp 35000
Oct. 16] 10.0] f N.W.| 18.1] 55 4.1| wo bg 83000
128m |) & N. 19.1 | 62 3.9|wreg 71000
GaGa lee 19.1] 56 3.4| Wp cor 51000
Ociaeny| 195i) £ S. W.| 18.1] 56 3.9| wre 71000
m2e2)| ob Ss. 18.1} 65 eT Waele: 64000
520)|| £ N.W.| 19.1 | 61 3.2 | Wp cor 42000
Octe18)|| 9-7)| £ S. ——= "F600 3.8| w elg 66000
TrvAo)|| a8 W. | 18.1) 74 2.9| — 31000
@eciamol|| 19.5;| Cc N. 19.1 | 57 2.0 | cor 10000
6.2] Cc I9.1| 57 2.0] cor I1000
Oct. 20) 9.6| c¢ N. E.| 19.1 | 60 2.5 | cor 20000
T)| 8. E. | 19.1] 63 3.0 | cor 35000
4.1] c Ss. 200 |) 03" 2.8] cor 28000
Ocia2n|| 79:3) |\1c S. 20.1| 67 2.2 cor 14500
4.0| R S. — | 64 2.9| wrlg 31000 | Gale.
6.3] £ 21.1| 60 3.1 | w br blp 38000
Octy22)| 9:8) £ S. 22.1| 59 2.5 | cor 20000
T2eGs |b Ss. 21.1] 62 2.3 | COr 16500
6.0} f — 158 3.0] wre 35000
Oct. 23] 9.8] f Wien P2015 3.5 | Wp cor 53000
ieeat || ae 5. W.| 20.1] 58 3.5 | wpcor 53000
Beg t Poyi8|| 533 2.9| wre 31400
Oct. 24] 9.3} f W. | 19.1] 47 3.3) W 47000
4.4| f N.W.| 18.1} 51 Bar| COL 38000
Octy2s | 10:0) f Wire 18ers 6 3.11 g’|blp 38000
3.5| wre
1.0| f S. W.| 18.1] 61 | . aes \ 55000
Caste 19.1} 56 3.4 | Wp cor 49000
wie 0) ona |e S. W.| 18.1] 57 3.4 | wpcor 49000
m2 e An RY S. W.| 18.1] 58 3.1]; wbp 38000
6.0] c 19.1 | 54 2.7 | cor 25000
Ociz727) |, o:0)|| £ N.W.| 17.1} 44 3.1 | wiblp 40000
Tea N.W.| 17.1} 48 3.2 | Wp cor 44000
GEO |leCe N. (53 3.1] glb\p 38000
Oct. 28] 9.6] f N. E.| 16.1] 39 3.1 | gibl|p 38000
12.4| f N.W./ 16.1] 45 3.6| weg 57000
eon |e 07 Tas 3.7| weg 61500
Oct. 29 | 10.2| £ S. sey fete || Ge 3.0| w|blp 36000
Te) ef W. | 18.1] 57 27) Wace 61500
6.4) f 18.1| 50 3.6|weg 57000
Oct. 30] 9.9] f N. 17.1} 44 3.9|wrg 70500
E20) £ N. 17.1 | 46 3.4 | Wp cor 49000
PAN 18.1| Ar 200) |i 35000
Oct 0-3) £ N. 14.1] 35 4.3|g¢bp 90500
9.9| f Seal ores 4.3| y’ o bg go500
GEO) N.W.| 16.1 | 43 4.3| wo bg go500

202 A CONTINUOUS RECORD OF ATMOSPHERIC “NUCLEATION.

+. Successive monthly data.—The temperature effect during the remainder
of March (charts 13, 14) is rather in opposition to the above inferences for
winter. Thus the maximum on March 19-21 corresponds to a hot wave while
the other cases are vague. In other words, there is quite apt to be a rise of
nucleation with rise of temperature. The reason for this is frequently to be
ascribed to the rain minimum which in winter corresponds to warmer and in
spring and summer to colder weather. The Sunday nucleation during clear
weather is low on March 15 and high on March 29. The maxima occur for

CHARTS 13, 14, I5.

cloudy (March 17), partially cloudy (March 19-21), and for clear (March 26-29)
weather. On March 29-31 the wind and weather change is a complete reversal,
and yet the maximum is sustained. Cloud effects appear on March 19, 25, 27,
but on the 18th clear weather is ineffective.

8. Throughout the whole of April (charts 14, 15, 16) the nucleation is
moderate. The maximum on April 5-7 clearly corresponds to cold weather.
The falling off to the minima on April 14-16 is due to the rain-storm, and the
subsequent maximum (April 17-19) has no temperature equivalent. The
minima on April 26-27 are cloud effects, the air having been purified elsewhere.
Remarkably high rain nucleations occur on April 3-5. Night observations on
April 18 show no exceptional values. The cusps on April 1, 5, 8, 9, 10, 11, 12,
13 18 are peculiar results.

9. May (charts 16, 17, r8) opens with clear weather but with relatively
low nucleation, due probably to the western blizzard. On May 7 and 8 change

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 203

of wind and weather is without effect on the nucleation. On the 2zoth there
is a cloud effect. No specific result marks the hail-storm of May 19. On May
17 and 31 there are low Sunday nucleations for clear weather, but the other
Sunday nucleations are high.

CHaRTs 16,17, 18, 19.

to. In the beginning of June (charts 18, 19, 20) hazy weather and rela-
tively high nucleation are due, no doubt, to the New England forest fires, in
Maine and elsewhere, which occurred during a period of drought. Intensely
yellow fogs, red moons, and greenish suns were observed. The nucleation,
however, is by no means remarkably high, not more than twice the normal
summer values and scarcely one-fifth of the large winter values—ain spite of the
excessive fogs. The copious rains which followed the period of droughts grad-
ually reduced the atmospheric nucleation to very small values (June 7-16).
The sharp maxima on June 17, 22, 23, are probably due to local fires.

11. In the end of June and beginning of July (charts 20, 21, 22) the
nucleation is remarkably high and sustained, changes of wind and weather (June
30, July 1), and even of high temperature (July 3), notwithstanding. Marked rises
of temperature and of nucleation concur. Sunday minima with clear weather
occur on July 5 only, July 26 being partly cloudy. The effect of gunpowder
smoke on July 3 and 4 seems to be quite absent. Rain on July 7, 22, 30 (the
latter is even followed by a maximum) scarcely reduces the nucleation, but low
values follow the rain-storms on July 18 and 21. The thunder-storms on July
10 and 15 without rain seem actually to be compatible with a rise of nucleation.
Incidental night observations on July 23, 26, and August 1 and 2 show no

204 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

determinable peculiarity. Throughout the month the temperature effect is
vague or in opposition to the winter effect.

CHARTS 20, 21, 22, 23.

12. August (charts 22, 23, 24) at first shows low nucleation due to fre-
quent rains. The rest of the month is somewhat uncertain in consequence of
building operations on the college green. Thus the maxima on the 24th and
25th are due to fires on the campus. Low Sunday nucleations occur on August
2 and 23, the latter being phenomenal. Rapid increase after rain is frequently
noticeable, as on August ro, 21, 27, and similar effects may be found in the
earlier months. |

13. September (charts 24, 25, 26), in spite of high temperature, clearly
ushers in the high winter nucleations, just as March had removed them. The
suddenness in both cases is surprising: one naturally asks whether there is here
a return to the locality of the nuclei-producing civil activity which had left it
in March. But the data of 1904 do not bear this out. The high maxima on
September 1, 9, 12, 15, 19, 22, 25, 30 are all striking. The Sunday minima on
the clear days of September 6 and 13 are not low. The gale on September 17
and the thunder-storms on September 5 and 27 are followed by high nucleation,
particularly in the latter case, but the assertion that nuclei are produced in this
way would be unwarranted.

14. October (charts 27, 28, 29) sustains the high nucleation of September.
Sunday minima (October 4, ef seq.) quite vanish, but the winter temperature effect
is not yet resumed. The rain effect on October 5, 12, 17 is obscure, and in the
latter case even the rain-storm is ineffective. Rise of nucleation on October 19

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 205

is broken by a cloud effect. The occurrence of marked sun-spot disturbances,
as announced during the month, may be noted. After October 23 the winter

CHARTS 27, 28, 209.

temperature effect is again reinstated, both for the general march through the
weeks and for many daily variations.

‘2

206 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

1s. In November (charts 29, 30, 31) the cold wave on the 7th has no
counterpart in the nucleation, which is throughout nearly uniformly high. A
curious result is the unexplained minimum on the 11th for clear weather, much
below the rain effect. After November 22 the sweep toward cold weather on
November 27 shows itself in the nucleation. The cold snow on November 29
has but slight, if any, denucleating tendency.

40" ; ‘ Q ; :
Xe, 10 ul 1 re) #4 5 16 Tf 8 49 20 a ae 8%

CHARTS 30, 31, 32.

16. Soin December (charts 31, 32, 33) the precipitation on the 2d and 3d
scarcely reduces the nucleation, and this remains high except during the rains
of December 9 and 13. The clear weather minimum on the 8th is noteworthy.
From December ro to the rain-storm of December 21, the temperature effect is
very sharply marked, as appears particularly in charted data. The same is
true for the remainder of the month. On December 26, 27, 28, precipitation of
snow below the freezing-point actually raises the nucleation.

17. Successive monthly data for 1904.—With the beginning of the new
year, 1go4, and very cold weather (charts 33, 34, 35), there is striking parallel-
ism between the temperature curve and the excessively high nucleation, as far
as the minimum introduced by the thaw and wet snow on the 9th. The con-

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 207

trast between the negligible dry snow effect on January 2 and the wet snow
minimum on January g is to be noted. The maximum reaches its height (the
- diffractions actually surpassing the large green-blue-purple corona) during the
cold blizzard and snow drift of January 3. As the coronal method breaks down
for nucleations exceeding the case for the g-b-p corona, the higher nucleations
are merely suggested in value on the charts. There is a fog effect on January
7. Moreover, the maximum is well sustained while the winds shift from north-
erly to southerly on January 6.

Ah eallbe-caniianl Nee taad 0h
1004) Scientia Ly

CHARTS 33, 34.

After the rains on January 13 and 16 a second march of nucleation into
extremely high maxima (January 18-21) coincides with the sweep of very cold
weather, ultimately to be broken by the wet snow on January 21. During the
remainder of the month there are some suggestive data on the nucleation ac-
companying a strong fog, though but few of the active nuclei are probably
entrapped in the latter. Cold snow effects on January 26 and 29 may be noticed.

18. February (charts 35, 36, 37) begins with a sweep of cold weather and
high nucleation, terminating in the thaw and fog of the 7th. The next cycle
extends to the snow on the 14th, a low minimum, remarkable as being below
freezing. The temperature effect from February 12-14 is obscure. High nuclea-
tion again prevails until the rain and thaw on February 22. The large number
of observations entered are due to a number of subsidiary experiments made
during this interval. There is cold snow on February 15 and 19. Even the
dense snow-storm on February 24 only partially removes the nucleation, as this
falls off gradually into the minimum following the rain of February 29.

208 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

80

| 37

CHARTS 37, 38, 39-

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 209

19. The high nucleation of March, 1904 (charts 37, 38, 39), is in distinct
contrast to the relatively low nucleation of March, 1903. This is true in like
measure for April, May, and even June. In 1904, the winter nucleations are in
all cases very gradually replaced by the low summer nucleations. The sweep
of temperature from March 1~15 is in very full accord with the changes of
nucleation. Remarkably high rain minima occur on March 3, 7, 18, 22, and
the same is true of the snow minima on March 11, 15, 18, and of the Sunday
minima on March 6, 13, 20, 27. On the 14th there is a cloud effect. Toward the
end of the month temperature correspondence ceases and the nucleation remains
high, in spite of weather much above freezing.

2

1904

CHARTS 40, 41, 42, 43.

20. The first period of nucleation, beginning with the rain of April 1
(charts 39, 40, 41) and terminating in the rain of April 7, partakes of the rela-
tively high character for the season already instanced. It is possibly referable
to the corresponding sweep of temperature which approaches freezing on April
4, supposing that the nucleation due to fall of temperature outlasts the latter.
The period from April 10-18 is less pronounced, but high nucleation corresponds
to the moist snows on April 14 and 16. The night observations betray nothing
unusual. The final cycle of the month from April 19 to the very low minimum

210 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

on April 27 and 28 is again high in nucleation and here there is a reversal of
the winter temperature effect, both in the high maximum of April 21 (which
comes too late) and the minimum of April 27 (which is also too late). In rela-_
tion to temperature, both regions have been shifted to the right. The rise of
nucleation during the rain of the 29th may be pointed out. There are many
results during the month which seem to imply a change of nucleation with the
direction of the wind.

CuHaRTS 44, 45, 46, 47, 48.

21. May (charts 42, 43, 44) also begins with a period of high nucleations,
reaching from May 2 to the rain of May 9. The temperature effect is usually
reversed or obscure. The contrast with the low nucleation of the preceding
year is marked. Rain minima as a whole are low. Strong nucleations (as was
the case in April) mark the close of the month.

22. June (charts 44, 45, 46) at last introduces the characteristic summer
nucleation, during the first half of the month. The temperature effect is re-
versed. Sunday nucleations are low on June 5, 12, but not on June 19, which
shows phenomenally high values. Abrupt high maxima occur on June 14 and
16. On the 17th there is change of nucleation with the direction of the wind.
One may note the sharp reduction of nucleation due to the thunder-storm of
June 26. After June 20 the values are as a whole seasonable.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 211

23. Observations during the first half of July (charts 46, 47) were sus-
pended. High rain values occur on July 16, 18, 27, 30. Thus the rain on
July 27 has but partially wiped out the maximum. On the other hand, the
rain minima on July 23, 24, like those of August 4, are excessively low, being
among the lowest nucleations observed (1000-2000 per cub.cm.). Sunday nu-
cleations are not remarkable.

24. The nucleations of August (charts 47, 48) are low at the outset, pre-
ceding the abrupt maximum of August 9. The temperature effect is vague.
During the middle of the month observations were suspended. The maximum
on August 31 suggests the approach of the winter nucleations.

25. September nucleations (chart 48) present no new features and but
few are charted. The observations were suspended from September 13-27,
during my absence in St. Louis.

GENERAL INFERENCES.

26. Efficiency of apparatus.—The experiments as a whole are in the first
place to be regarded as a severe test, amounting almost to a strain of the method
employed. Since both maxima and minima are registered with equal facility
at exceedingly low as well as at high atmospheric temperatures, the temperature
error of the method is not menacing. Rain minima, snow minima, and other
exceedingly low nucleations have all been recognized. The high maxima re-
quire a long apparatus, since the diameters of the corona occupy nearly halt
a meter at the condensation chamber; but the maximum of atmospheric nuclea-
tion does not, except in very rare cases, exhaust the limit of measurable
coronas. Nevertheless a higher pressure difference would make it possible to
confine the measurements to normal coronas exclusively—a desideratum, inas-
much as Chapters VII. and VIII. have shown that after the g-b-p corona is
reached, the data lying in the region of the cusp are estimates. Higher pressure
differences are somewhat less convenient, and in regions remote from cities it
is improbable that the normal coronas will ever be exceeded. Care must be
taken to have the influx pipe far enough from the walls to guard against ex-
halations from the building in summer. If this pipe is sufficiently long and
thin (influx current rapid), the air arrives in the condensation chamber almost
at room temperature and without appreciable loss of nuclei during the passage.

27. Variations of nucleation—Mere inspection of the charts shows the
extreme variability of atmospheric nucleation, though quiescent periods are
also met with. A part of this must be a local effect, even if the changes cor-
respond to the weather. Work in cities where there is so much chance for the
pollution of the atmosphere is to some extent unsatisfactory, but the method
could hardly have been developed in the country. It is probable that even the
small variations of the chart (if they exceed 2000 nuclei per cub. cm.) are real.
If the nuclei were colored, the atmosphere would look like mottled soap with

to
H
to

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

the clear regions usually but by no means always accompanying rain or lying
under clouds.

28. Wind effect——Changes of wind velocity were not observed. The
nucleation consistently follows certain changes in the directions of the winds,
as will appear from a detailed examination of the charts as a whole. The
following table, which is a summary of this kind, shows the distribution of
maxima above 2=60000 with the wind direction.

TABLE II.

Winp DtREcTIONS FOR MAxIMA EXCEEDING ”2=60000.

Time. East. S-East. | South. | S-West. | West. | N-West.| North. | N-East.
Oct., 1902-Jan., 1904....| 0 I Or ian 54 25 14 2

Jan., 1904—-Sept., 1904... I I 11* IO By 26 | II 3

* 8 after March, 2 after May.

By far the greater number of these maxima occur for nearly westerly
winds and but few for easterly winds, if stress is laid on the winter nucleations.
In 1904, where high maxima occurred in the spring and even in the summer,
the distribution is more southerly, but in the main like the preceding. It by
no means follows, however, that nucleation is inherently associated with these
particular winds, inasmuch as the winds observed are merely those prevailing
in Providence. Moreover, the density of population, etc., is far greater towards
the southwest than towards the northwest of the University, whereas the
northwest winds prevail during the period of maxima.

Considered as a whole, therefore, it is improbable that any real variation of
the nucleation with the direction of the winds is in question. What has been
observed is the obviously more frequent occurrence of maxima during the
prevailing winds.

29. Rain effect—The observation next in importance is the occurrence of
pronounced minima during rain, of which the charts contain examples in great
abundance. There is rarely an exception to this rule. It imples a faster
removal of nuclei by precipitation in a saturated atmosphere (the result of fall
of temperature probably) than the supply of nuclei to the same region by
either diffusion, or subsidence, or convection, or other more occult causes.
Whether this deficiency is eventually made up from the lower air strata in
contact with the surface of the earth or from the higher air strata is at the
outset left open.

Rain minima never fall quite down to the zero of nucleation, in cities
rarely below m=1000, and they are themselves quite variable. Thus the
summer minimum is as a rule much lower than the winter minimum.

Minima quite as low as the rain minima are sometimes observed in clear
weather, but they are very rare.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 213

One may note that the tendency of rain to change the normal air potential
from positive to negative values would thus be accompanied by relative ab-
sence of nuclei. In other words minimum nucleation exists here cotempo-
raneously with maximum negative ionization.

30. Snow ejfect.—Wet snow usually acts similarly to rain, but less power-
fully. The effect is particularly noticeable during a thaw. There are cases,
however, in which the nucleation is high with a fresh fall of wet snow. Dry
snow may even increase the nucleation.

The present experiments, moreover, show that ordinary terrestrial dust,
properly so-called, has no bearing on the nucleation; for this is frequently a
maximum immediately after a rain, when the earth is blanketed with water,
or after the earth has been covered with snow or with sleet to the exclusion of
all dissemination of dust as such.

31. Cloud effect.—Another interesting feature are the cloud minima as
seen on October 15, 17, 21, 23, 1902;-March 19, 25, 27, April 1, May 20, August
27, september 27, October 18, 19, 26, December 15, 1903, etc. These observa-
tions are incidental, and a larger number would have been found if systemati-
cally sought for. Usually a higher nucleation is again established after the
cloud train has passed the sky, the phenomenon beginning and ending in periods
of clear weather. At other times, however, the minimum remains, as on
October 17, when many observations were made with this end in view. The
explanation of this result is at hand: the air has moved bodily with the cloud,
the whole constituting a region of deficient nucleation. The nuclei may have
been precipitated by rain elsewhere and the cloud may even have vanished
from the region.

32. Solar nucleating effect absent.—Since the nuclei cannot re-enter a re-
gion by diffusion as quickly as is usually observed, one may be tempted to
infer that solar radiation is the cause by which the nucleation of a deficient
region is re-established. There is no evidence of this in the chart and much
against it. Thus remarkably low minima are frequently maintained through-
out the day in full sunlight. The minimum may be part of a cloud region with
which the day closed, but sunlight is powerless to replenish it. Simular refer-
ences may be made to the notched midday minima which so often occur. By
contrast high nucleation develops in spite of an overcast sky. Finally night
observations show neither increase nor decrease of nucleation, but are usually
normal in character. Hence there is no evidence, so far as these observations
go, that ultra-violet light or other solar radiation has any potency in producing
the nucleations here immediately in question, and cloud effects have therefore
been explained as purely convective. Indeed, compatibly with the final sum-
mary ($37) of mean monthly nucleations, the sun must be regarded rather
in the light of a denucleating agency.

33. Lemperature effect—An important general result is the frequent oc-
currence of maxima of nucleation, cotemporaneously with the sudden fall of
atmospheric temperature in cold weather. So far as the drop of temperature

214 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

is concerned, one would have the conditions for the formation and growth of
water nuclei, remembering that persistent nuclei may be made in an atmos-
phere not too far from saturation, from solutions, or in the presence of a solute.
Without actual accession to their numbers, however, there could be no in-
crement of nuclei in the condensation chamber at about 20° centigrade. In
other words, an actual rise and fall of the number of nuclei per cubic centimeter
must occur in the atmosphere at the place of observation.

This temperature or cold-weather effect is very striking in winter where
both the daily and the periodic effects frequently coincide. In warm weather
the temperature effect is vague or may be even quite reversed. In relation to
this one may note that rains in winter usually correspond to rising temperatures,
but in summer to falling temperatures. This would accentuate the winter
effect, which is otherwise quite independent of rain, but would obscure or even
reverse the summer effect, if 1t exists.

If a polar air current or a current from the upper atmosphere is associated
with the cold waves in question, a part of the region where there is continua]
production of nuclei may be bodily transported to the place of observation.

34. Local effect-—The prevailing westerly winds sweeping over a part of
the city of Providence carry the products of combustion and other local im-
purities along with them. A large part of the nucleation at the place of ob-
servation must therefore be of artificial origin. The local effect should be
greater as the wind velocity is smaller and as the temperature is lower. In this
way the winter temperature effect is in a manner explained, seeing that fue]
will be consumed more rapidly during the cold waves. Observations of wind
velocity are inadequate.

The difficulty with this view is that it does not consistently explain all of
the observed facts. Thus cold waves occur in winter without maximum nuclea-
tion, as, for instance, on November 11, December 8, 1903. The Sunday nuclea-
tions during clear weather in summer are not uniformly low. Maxima during
relatively warm weather are often quite as high as during very cold weather,
the temperature effect is often reversed even in winter, etc. But the most
direct reason for caution will be given in the summary. of monthly nucleations
where it appears that the highest nucleations do not occur in midwinter so far
as cold weather is concerned. Speculation as to the origin of the nuclei is thus
premature. Whether it is the residue of the ionized products of combustion,
or whether ultra-violet light or other radiation at the boundary of the atmos-
phere is the efficient and preponderating source will only be made clear in a
series of correlative observations obtained in a wilderness remote from cities or
on isolated mountains. Work of this character is now actively under way.

SUMMARY AND CONCLUSIONS.

35. Mean daily nucleations—Meanwhile it is expedient to deduce from
the above data, both the daily and the monthly average nucleations or number

ee ee ee

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 215

of nuclei per cubic centimeter. Incidental disturbances will thus be to some
extent removed at least from the latter, so that the sweep of nucleation through-
out the year may be more clearly apparent. The rain effect will be more uni-
formly distributed during the summer months, though it cannot of course be
eliminated.

_The data for the mean daily nucleations are given in Table III, and require
but little explanation. The symbols for the state of the weather have been
abbreviated, f denoting fair, f’ slightly cloudy, c’ partly cloudy, c cloudy, R’
slightly rainy, R rainy, etc. Sn refers to snow, Sn’ to a light snow.

TABLE III.

Mean Daity NuCLEATIONS FROM Oct. 2, 1902, TO NOVEMBER 1, 1904.

|
|
|
|

E 5 s
S : tsa fa 3 : ait
ral A = = Zs || > A = = Ze
1902 |Oct. 2 | c 18 ||1902 |Nov.7 |c 52
gue 45 | 8 |f£N BY
Boat ZI 9 Ile 18

ee aie 12 10 |fe 30

Gi PRC! 19 | Tee} Mf a5

Peale 22 | | D2 Wee 45

Sia ial: 21 | 3h eC Vaese

me ie 16 | 14 |f eels

10 | f 52 || TAG es 37

me |(xC 1) 24 16 | f’ 7

1A) AR | 6 | De 9

Tita le 22 | 18 | ¢ 24

Te RE 12 19 |c 2

Tipe oft 24 20 |f 42

nO) || CE 18 ar |f go

Te 7E et Cs 1g Bae 30

Ts} | © LRM 9 || Dia) (che 22

TOPs |e Ce 6 24 |c! 34

Zin LE 12 25 kReC 34

Tae | 15 200 Re 13

22, | f 16 ENG 30
oomalpcecrt yi 2) Sealer: 27

oy Nie 21 2 f 75

25 | ft 9 30 |fic 36

2 Oni lnieec 10 (Dect 1 |i after R: | 60
7c 31 2 eit 45

Oe me 18 Bip pic 30

ZO L. 21 Au fee 45

30 |f 37 ec 60

un et 30 6 |f 100
Nov. 1 | f 22 7 |Sn go
2 ANGE 24 Sih 75

ga] tf 30 9 |f 100
4\f 30 to /|Sne go

5 peat 27 Te |e 52
6/cR 45 12 |Snc 37

210 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE I1[I1—Continued.

¢ ¢
uw | g = J ae) 20s = s ge
oO 3 ¥ & Seth oO 5 vo I 3x
a A | £& E Ze|| > A = = Zz
1902 |Dec.13 |Snc 75 ||1903 |Feb. 2 |c R 52
| 14 |f 100 |, Bt 4c
TG) yf hic 100 A Re 450
16 |R 36 cee 52
ifs |e 75 | hai 60
18 |f’ 37 || Tomi go
1g |f’ go | 8 |cSnR 34
Zou 45 9 |f 75
eae pane - 33 10 |f 45
22) ee 45 TE fic 45
23| \f go aia ok 30
24 |f go To aelake 45
25 |Snc 36 rat 52
26 |'c’ 75 15 |c Sn’ 45
27 (Gh 100 16 *| Sn 45
28 |f 100 17 *|Sn Bl 60
29) ee ; 100 TS) tc? 83
307 |t 42 a) ti 92
Bie Hit 100 20 If 93
Pm ie 7°.
1903 |Jan. 1 |f TOO Bem ie 30
20a Ch 100 23 If 60
ay Roc 40 24 lice By
A alec 48 2 Seale 60
Ge (et 55 26 |f 62
6 |RSnf 75 27, | hie 39
7 pheno n: 75 28 |R —
Serta 45 Mar. x |\c’ f 19
9 |f | 100 ai 61
10 |f | 100 SealiC 27
rr jc Sn 45 Ae ih on
ey Nai 75 Be aN RY 44
U3) it 100 6 |f 48
14 |f go eit 26
TS) Gate go Seal aR 20
Ole |p go Gy Pike 13
Tai alee 100 10 |c 23
18 |f 100 rine alee 28
TOW Ie go Tot |e 32
20) fe 75 | Denes 32
25r | pier! 4o TA) at 36
22 ure, 45 DS ea inc 84
2cten iba 42 LO! ic 22
24 If 45 7c 44
ZL) piesa 52 18) CB 5
26h 45 ne) Gu 41
27 N.C 90 20 | f' f W. 36
28 |R Fog 45 2x |Ric E. 25
2 c 75 22 |R N. 17
Bonn Rvan 75 230) | ReMi N.E.S 19
yaaa i 60 24 |e S.W. 230
Feb. 1 |¢ 45 Pe Ante W. HO)

* End of the approximate data.

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

Year.

Weather.

Wind.

1903

Qo

Qa
hh

~
°

| FrHRAPHO OO Wo Phe eh

|

eh

~
°

Wo Leyte agils are ey eel top) ‘Ga uG.
2a ee Hh

2a

PHA RRR eR eH ER ehh OO OO

222288
see a2:
no =

ZPDH
gah
ZZ
er

eae
= eg

ABA~LZLZSAZLZLZLZAD SS
me

=
Me eee

Zz aezh an
See.

Breit eee ed cee et ec
Say =
Dn
re

ZZINNnN
mas

TABLE I1[—Continued.

|
| § |
ree |
Roe a | 8
Salis) A
| 32 |/1903 |May17
emesis 18
18 19
31 | 20
39 || 21
26 | 22
2 23
26 | 24
Zi | 2
26 20
42 2
32 28
28 29
21 30
27 31
24 June 1
25 2
24 3
20 4
13 5
10 6
13 7
25 8
36 9
20 10
26 II
— 12
28 15
24 16
14 17
13 18
21 19
20 20
20 21
17 22
13 2
21 2
23 ae
18 20
14 27
28 28
Toy) 29
8 30
20 July I
32 i
30 3
2 4
22 5
15 6
TO | 7

Weather

|
|

wo
wW

al

aad

Faia ap staan XJ ae Q RRO Mme
~
~
ol lhUoA
loan’ ‘Co

ned
o—-

og

—

Haze f

en

Ane
o
°

AA™ARAAS ©
200

Qa

~
QO

~

bia

w

An

Fh Fh
2

sige aim m oe iC Q mrhoono Awe Cha Goaiera
ar]

SEVP ABV SSZAZBZEYV EE AZZL OE EPPADOAA

AZ2RNZ
435

ZZgz222E9
= = 82485

ZAALZP ZV ONAAZS
eee

ae

DDD,

nS

Dn

n

n

Dn

Bo

nA
=

| Wind,

A

a
=

} Forest fires.

ts
=
~

|
|
|
|

| Nucleation
| X I073,

16

22

218 A CONTINUOUS RECORD OF ATMOSPHERIC* NUCLEATION.

TABLE I1I1—Continued.

ri 8 K 8
8 Bo o Gm
alae 3 E gel e| 2 3 = $8
o i I
o ! A 5 5 2 Ciena 5 z ee
1903 |July 8 | f W. 17 ||1903 |Aug.31 | Re N. 8
9 |f N.W 8 Sept. 1 | f’ W.S. 34
1o |fR S. 27 2a ahe W.S. 19
De alla |W 40 Bilee N.W.S 25
re |) — — || 4h S. 24
Te Rec |W. 19 || Eau Cuiee W. 18
14 | f W. 21 olf N. 22
LS | Rf tie 18 || yi fc N. 25
16 | f W. 18 8 | f N. 29
7 ab W. 17 gif Ss. 39
TO) | atarcuks S.W. 13 TONmC S.W. 29
Ola Rec N. 3 Te | W. 30
2OmNit |N.W. 24 r2| W. 50
2x | R’feRaqN.E 44 Tet S.W Ir
22 Ct |S. 12 14/f W. 34
23) UR isiy S.E. 15 || rsaiie W. 39
24 |f N.W. 13 i6|c 8. 10
2h W. 16 || 17) |eCyle S.W. 19
20M aCe W. 7 |l 18 | f’ |N.W. 25
27 \¢ jN.W. 17 || rufa) ne N.E. 42
23) Neb |IN.W. 21 20|/c’c E. 8
20m Guky S.W. 17 2 elon N. 32
30: || £ 5.W. 17 224k W. 30
Bee W. 14 220 S:E: 360
Aug. 1 | f N. 12 24 eRSC. W. 28
2\f N. 6 25. | £ N.W. W. 48
Bee N. 18 26) f S. 37
4/|cR N.E. 12 27 to Cue Clas: 10
iy kere E. 5 28 | f W. 30
6) ‘cu N.E. 8 29\| f W. 39
Pal Gan S.W. 10 301] £ W. 53
S| Meher N.W. 24 Ogie ae it S.W. 34
g|cR — 7 2 \c W.N.E. ir
LOw | W. 160 oe at N.E. 19
iz |/cR S: 20 4 | f S. 32
Ly ie W. 16 eae tes W. 40
a W. 19 | Gi i etaae N.W. S.E 17
14 |f W. 13) eu E. 9
7a ia N. 16 | So |Kexe’ S.E 20
LO) lit N.W.S 22 Oneac E. 5
19 |f S. 28 nite), ||P AR N.E. | Uy
20" |¢c IR: Ss. 20 Tata ele N. 8
aire Wt N.W. 19 12 | Rie N. 23
22 | £ W. 25 Tee |G N. orn
230) ti W. 9 T4) |b N. 45
24} Ret N.W. 32 TiS ie N.E. 40°
25/Re’R_ |N.W. 21 TiO | eter Ss. 33
20- | ¢ N. N.W. ry 7a CE De 27
27 || tic N. 17 TOM | pte W. 20
28 |cR N.E. S.E 22 TON Al placed W. 48
20) Ik Bye 12 20 | f S.W. a0
30 | R E. 5 2m ti N.W. 47

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 219

TABLE I1I—Continued.

u 5 ce K 8
fs go | s ai
s| $ e = gx] g| 3 : z oe
alee = ZAe|| > A E E Zz
|
1903 |Oct.22 | f N.W.S 32 |\|1903 |Dec.13 | Rf S.W. 18
Dota lla Cols: S.W. N.W 26 || pa ih if 77
Py Nan onto N.W: 37 Gn RCAG get |W. 58
eye atic N. 58 16 | f W. 64
26 | W 39 07) pit |W. 80
Ag | at W 46 TO) et N.W. 86
Ae | a N.W 59 LO) |) |W. 82
20) |i W. 51 20 | R Ieee Bic
30 | f W. 35 Biteelee |W. 39
Bp |p W. 36 22) \| Riot |W. 44
Nov. 1 | f W.S.W. 2 oe aak 1S.W. 81
2|f W. 48 an Ele IS.W. 73
Bult W. 43 25 eC |W. 31
ZA oi W. 38 26 |cSne |N.W 2
& | we S.W. 31 24 |eSnic |S.W. 63
6 |cSn N. 35 28 | f N.W. 76
ale W. 41 29 |c Sn S.E. 54
nlnt W. 37 BO lal |N.W. W. 69
9 |f W. 59 Bie alt |W. 72
ROM fet S.W. 42 |
mt |p N:W. N.E. 19 ||1904|Jan.1 | f |W. 48
Teo | cane W. 2 | 2 |-esn N. 79
eae |e S.W. 49 | Beale N. 100
TeAw | Gal: N.W. W. 43 | Ay Nak W. 100
TS || W. 43 | ert W.N.W. 10O
MO Gul N.W. N.E. 30 Gr alee 5. S.W. 97
7 | Re N.W. 43 | 7 | £ Fog W.S. 68
OM (Cee N.W. 28 | 8 |cSn N. 67
TQ) | N.W. 46 | 9 | Sun f’ Sn |/N-W. 2
20) || E N.W. N. 50 | TO! et W. 41
21 | f N. 38 || Tella N.W. 63
22 |¢c N. 37 || Te cee N. 75
23m Cue N.S 46 Tea) fCalke N.E. E. 26
ona etn CArc W. 30 Trae lietg S.W. 66
Pica ti W. 43 iste liek W. 58
26 | f N. 63 TO: Wet Ome: USeWiens 46
ay |fc N. 70 17h W. 37
28 | f N. W. 54 TS | ae N.W. 100
29 |c Sn N. 36 19 |f N.W. 100
BO Hai W. 2 20 | snc N. S.W. 100
DWeerter | N.W. 43 2m \aceon N.E. 41
2 |cSn R’cIN. 61 22 |R N.E. 30
SeleCulc N. 63 23 | Fog S. S.W. W. 59
AL MN OSi N. 55 Mitt W. 38
ana W. 66 | aan W. gl
6 \f W. 39 XS Wai Ke N. E 35
Tae W. 7 OT ab W. 71
Salat W. 42 | 28, et cic” W. 81
OMe Re N. N.E 25 29 | Sn N. or
reve | ai W. 45 ZOn yt N. 69
rete | W. 59 atten ahGe S. 46
12 |fe W. 69 || Feb. 1 |cSn’e_ |N.W. W. 48

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE I1I]—Continued.

3
4 2 g =
$ = o “ct
a a = =
1904|Feb. 2 | fc W.S.S.W.
aah W.
4 |f W.
Cia W.
@ |e Sn’ RZ IN:
Tac S. W.
8 |f W.
Ont N.W. W.
Io |f NW.
11 |fc N.W.
Eo: |hiel N.
3) hae N.
14 |cSn |N.
15 | Snf N.W.
16 | f W.
man ie W.
18 |fe N.W.
19 | con S.W. W.
20 | f N.W.
oT et N.
22 |Rf S.W.
2a i S.W.
24 | Snic N.E. N.
25 eh W.
26 | f N.W.
27 |—
25 ast Cure |S.
ZO Mine |N.E.
Mar. x |'c:Sne EB:
2|\f N.E. S.
Balk os
| Aut N.W.
Sane N.S.
Onlkere? Son Se
a NWeuR! 5:
Sah W.
©. UE W.
TOM at W.
11 |cSn N.W.
12 || N.W.
moe ee N.
14 |fc’ N.W.
L5.)| onc N.
TO lit N.W.
Ti ae N.W.
18 | Sn R! S.E. N.W.
LOy || £ W.
20 | f W.
Droit N.
22 |R S:
23 nend N.W.
24 \f N.W. S.

Nucleation
nX 10-3,

loo)
“SO

as
oO 0c

| Year.

1904

Weather.

a0
~

rhrh ho 0
Q
e
w

Nucleation

Wind.
n X 1073,

a= 3

DAN AZAAZGION
nan =

PEBAZ
anaes

=
p
=
=

sf

mn
n
an

HS

San ahhhs
<a

AP AZAZALZLZLZLZOOPALZAV PAA AAAOA“|L
a2
Ra n

am
nn

Zn
a
mM

2222250002
Maa S un

|

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE I1]—Continued.

S

: : sg | $0 : : 4 ; eonle
8 $ S = are 38 $ $ = ois
S| 4 5 5 Zell < | A = 5 2%
1904|May16|f’ R’f |W. 36 ||\1904 |Julyrg | f 5. 16
7c Siw s: 39 20 |-f N.W. W. 25

ON NC RC N.E. 8 Teh W. 25
TG) || OAS N. oT 22) |ohe IRE E. 8
ZO C1 f S.W. W.S 32 Qn RAC N.E. N. 4
ar | f- N.W. S. 44 24 |cR N. 255

oma a Ss. 16 25 Next Sy 19

2a it S: 34 | 26 (atl £ Ss. 37

24 |\f S.W 29 | 27 leRe tL W. 37

2ineel it N.W 29 | 28 | fc S. 23
2 Ona |fe S. 25 ZOO uc £ W. 28

Deel he S.W 215 | 30 |f N. 2
2a \E W. 49 ura eer S.W. 8

29 |f S. 38 Aug. 1 jc S.W. 22
Bone S.W. 16 uae: N. 15

Sire N.E. 22 enlace N. 15
Juner|cR N.E. 16 AG Neak N.S. 8
mac N. 9 5 | tee S. 14
ulencr N.E. 22 6A (Rc |S: 18

Aeelat N.E. S.W. 19 7 || ae W. 12

See et Ss. 22 8 |Re S.E. S. W 18

6 | c (R) N.E. 6 9 |f N: S. 46

7 |cRf (R)IN. S.W. 22 to |e R’c¢ S.E. E 23

8 | cf SE. E. 5 or |(R)keicealiWe 32

in IRE te N.E. 25 26 | fi W.N.W. 33
LO ||| —— — — et S. 30
Tes aa N. N.E 12 28 | f W.S. 10

1 2p | ak N. 14 29 | f' W.N.W. 2

3 pl eles N.E. 23 Zorn E. 21

14 /f Ss. 49 gira |b iW. S.W. 43

| ba 9. 22 Sept. 1 | f |S. 27

TON teak 5.W. 61 2) cf 1S: W. W. S. 28

ye ue N.E. 26 Bale |S. 30

BS ealet W. 25 AE lnCp rey ah N. W. 21

19 | f N. 65 Se haat N.W. W. 30

20 | f Ss. 40 6 £ [NSE Savers: 35

pte |inient 5.W. 31 Fella |W. S. 31

2 tae fi W. 28 Sullttieeeglnge an Wie Sei 5

AS |e N.E. 27 Co} | oude N.E. 8

Bie at W. 38 1o|cf |N. 17

25 |f (R) S.W. 27 TAN aes ee ERs ONS 26

BO | eahet Ww.” 25 12) £ W.S.W. 22

Da Nae N.W. 32 3h le E. 10
2a eCuc E. 23 —|— | — —

29 |Re S.E. N. 15 27 Nhe |N.E 19

Boni eR: SE 2 28} cf N.E 26
July: |Re 5.5. W. 25 29 |fc...R |S.W. 34
BNA S.W. 38 z0 /f Rf S.W. W.N.W.| 31

Tec eds 29 Octet W. 33

OM Ca S. 23 Oe |W. 34

Teja ate N. 5 yt |N.W. 35
MSwrlet N. 46 4 |f |N. 53

2:22 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

TABLE IlI1—Continued.

L 3. | ‘ oe
2 Bo ll 2 : ai
: | - SB J Ch ces ~ 3 oo
8 & 3 & ex Sots g & oy
= | a = s zei|s | A 5 5 Za
1904 |Oct. 5 | f W.S. 41 ||tg04 |Oct.19 |c N. 10
6 RSt N.W. N. 34 20 |\c S. 28
A MV N.E. 53 2ite aCuReati S. (gale) 28
8 |c S.W. 38 Bo. Win S. 24
gic S. 15 Dey NN IW. S.W. 46
TO! |Ne@re Ss 23 || 24 |f WwW. N.W. 42
Tay | Rec S. N.W. N.E 22 Ale | bai IW. S.W. 47
12 |R’cR_ |N.E I5 || ZOnACeRac S. 377
£30) |C N. 47 Oven |eGicy N.W. 41
Ln els N. 50 28 |f N.E. N.W. 52
Te lf N. 45 20mm S. W. 52
16 | f |N. 68 Boral ti N. 52
yee Variable 59 30 ib N. N.W. gl
18 |f S.W. 48

These data have been constructed in the charts 49 and 50, with the two
years of observation overlying the same abscissas to bring out the contrasts.
The series for October, 1g02-October, 1903, are easily distinguished from the
later series, October, 1903—October, 1904, by the continuous curves. The
nucleations are given in thousands (7~"?). Clear, partly cloudy, and cloudy
weather are indicated on the curves by the usual Weather Bureau symbols—
©,@, @. Rain is shown by r.

Apart from details, the striking difference of time changes of nucleation,
1902-03 and 1903-04, are apparent. In both cases the high winter nucleations
as compared with the summer nucleations are the essential feature; but in 1903
these nucleations fall off almost suddenly and permanently in March, whereas
the change in 1904 beyond March is much more gradual.

36. Mean monthly nucleations——The character of these secular changes
will, however, appear much more clearly in the monthly averages given in
Table IV.

TABLE IV.
Mean Montuty NuCLEATIONS FROM OCTOBER, 1902, TO OCTOBER, 1904.
Year. Month. n X 1073, Year. Month. n X 10-3, Year. Month. nX 10-3,

1902 Oct. | 19.9 1903 Jan. 72 | 1904 Jan. 66.3
Nov. 35.8 Feb. | 55.0 Feb. 60.8
Dec. | 60.4 Mar. | 31.2 Mar. 46.5
April eens April 41.5
} May 17.9 May 30.4
June T4.1 June 26.3
July 18.4 July 23.2
Aug. 15.4 Aug. 22.6
Sept. | 29.4 Sept. 2563
Oct. 30.8 Oct. 40.8

Nov. |

Dec.

223

CLEATION.

A CONTINUOUS RECORD OF ATMOSPHERIC NU

‘S0W d 'NIVY SALONA J ‘SNOIG AVaNNG YaAHLVaM AHL o1 NOILIddy NJ “SLUVHO aAOMY AHL woud
Ta NIAY “WILNAD “AAD Add SANVSNOHY, NI (SALVNIGUQ) SNOILVATOAN ATIVG NVa]yY—'oS-6b sLavHg

£06) Wo;

224 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

These data have been plotted in chart 51, the nucleations (” per cubic
centimeter) being again laid off in thousands. Both the similarity and the diver-
gences of the results of 1902-03 and 1903-04 become apparent at a glance. It
has been possible to connect the successive mean data with a curve almost at
once without resorting to much smoothing, except in the summer observations
of 1903, where the dotted line is drawn to eliminate the probable and excessive
rain effect, as well as the effect due to building operations on the college campus,
etc. It should be noticed that the new data for 1904 (August, September,
October) fall very closely upon the curve for 1903 prolonged. As stated above,
§ 3, the results from October, 1902, to March, 1903, are reduced from the old
scale and are not at once comparable in absolute magnitude with the remaining
data, but the relations are nevertheless well indicated.

80

60

ae a
t| || Pea
Qt Ni Deo. Sam Feb Mar Apr May. June July Quy Sept: Oct: Mov

CHART 51.—MEAN MontHLy NucLEATIONS (ORDINATES) FROM OCTOBER, 1902, TO OCTOBER, 1904,
IN THOUSANDS PER CuB. CENTIM.

37. Occurrence of maxima and minima of nucleation during the winter and
the summer solstices, respectively.—The remarkable result of both curves is un-
mistakable: remembering that the mean nucleations hold for the middle of the
month, the maxima of nucleation both in 1902 and 1903 occurred between the
middle of December and the middle of January, nearer the latter; in other.
words, about at the time of the winter solstice. The minima for summer occur
between the middle of June and the middle of August somewhat after the
summer solstice. The winter maximum is in both cases sharply indicated and

A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 225

its coincidence with December 22 not improbable. The summer minimum is
much more prolonged and uncertain, a result to be anticipated from the marked
rain effect which prevails at this season.

The identification of maxima with the winter solstice and of minima with
the summer solstice is alluring: for in these cases the earth is respectively nearest
and farthest from the sun. At the same time the orbital velocities are re-
spectively greatest and least, so that the path volume of the earth would have
corresponding values. Both causes are qualitatively in harmony with the
observed results, if the nucleation comes in great part from the sun. Quantita-
tively the results are less convincing.

If the data for 1903-04 (which are more nearly absolute) be taken, the
ratio of greatest and least nucleation is about 3, seeing that December 22 corre-
sponds to about ~=66000 and June 22 to m=22000. The case of 1902-03 is
even more accentuated. On the other hand, the greatest and least values of the
radius vector of the earth’s orbit are 1.0168 and .9832, in terms of the mean
radius. The same effect, approximately, is attributable to the differences in
velocity, making only about 6.7 % by which the winter nucleation should ex-
ceed the summer nucleation, for the case of linear distribution. No easily dis-
cernible distribution of nuclei from the sun outward would account for the
three- to four-fold winter nucleation as compared with the summer nucleation,
seeing that the decrement of about 2 % of distance corresponds to an increment
of 100 % of nucleation. Thus the reasonable surmise which makes the density
of the solar output vary as the inverse square of radius, while the path volume
per unit of time varies as the inverse radius, gives a compound law of the inverse
cube of radius, which, however, is quite inadequate.

Unless some occult law of distribution is at the root of these cases, mere
change of distance and velocity does not account for the facts. On the other
hand, if the influence of the sun is atmospheric and productive of dissipation of
nuclei (by breaking them into fragments beyond the lower limit of nuclear
size, for instance), or if there is greater tendency toward supersaturation of
the atmosphere when the nights are longest with the consequent pre-
cipitation of local nuclei, the effect due to greater length of days in the
summer as compared with the winter will much more nearly correspond
to the nucleations observed. The effect postulated is, however, wholly
conjectural.

38. Conclusion.—On the other hand, the two graphs showing the distri-
bution of mean monthly nucleation are not compatible with the postulate of a
purely local or incidental origin of the nuclei. Rains in general would make
the winter nucleation higher than the summer nucleation, while the products
of combustion still further increase the former. But neither of these inci-
dental effects is so distributed as to give rise to the salient maximum in Decem-
ber, reproduced independently and sharply by both curves.

Briefly, then, while local and incidental effects enormously modify the
seasonal variation of atmospheric nucleation, it is not improbable that

226 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

its origin is in part to be traced to a more deep-seated cause. At least
the results obtained are sufficient to warrant the extension of similar ex-
periments over a wide area of territory, including regions remote from_
the habitations of man—researches for which I am now making extensive
preparation.

ar ce r 5 r
i 4 ?
é
4
' .
=
pad
‘
ya
Cpr
7 wy
r
coe
A
>)
‘
x
e
2
~

; HSONTAN CONTRIBUTIONS TO KNOWLEDGE:

PART OF VOLUME XXXIV

¥ THE CAMADIAN ROCKIES
AND SEI

(No. 1692) {

1907 oe J i

2

2

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE

PART OF VOLUME XXXIV

MexcieERS OF THE CANADIAN ROCKIES
AND SELKIRKS

(Smirnsontan ExpepITION OF 1904)

BY

WILLIAM HITTELL SHERZER, Ph.D.

(No. 1692)

CITY OF WASHINGTON
PUBLISHED BY THE SMITHSONIAN INSTITUTION
1907

Commission to whom this Memoir has been referred :

THOMAS CHROWDER CHAMBERLIN
HARRY FIELDING REID
GEORGE PERKINS MERRILL

The Tnickerbocker Press, Hew Work

ADVERTISEMENT.

Doctor Wi.iiaAmM H. SHERzER, Professor of Natural Science at Michigan
State Normal College, has brought together in the present memoir the results
of an expedition undertaken by the Smithsonian Institution among the glaciers
of the Canadian Rockies and Selkirks in the year 1904. The general objects of
the research were to render available a description of some of the most access-
ible glaciers upon the American continent, to investigate to what extent the
known glacial features of other portions of the world are reproduced in these
American representatives, and to ascertain what additional light a study of
similar features here might shed upon glacier formation and upon some of the
unsettled problems of Pleistocene geology.

A systematic survey was made of the Victoria and Wenkchemna glaciers in
Alberta and of the Yoho, Asulkan, and Illecillewaet glaciers in British Columbia,
located about two hundred miles north of the boundary of the United States.
The largest of these is the Yoho Glacier, extending more than three miles below
the névé field, and a mile in width for two-thirds of its length. Doctor Sherzer
investigated various surface features of each of these glaciers, the nature and
cause of ice flow, the temperature of the ice at various depths and its relation
to air temperature, the amount of surface melting and the possible transference
of material from the surface to the lower portion; their forward movement and
the recession and advance of their extremities, and the general structure of glacial
dee:

In summarizing the most important results Doctor Sherzer discusses the
indicated physiographic changes in the region during the Mesozoic and Pleis-
tocene periods; the question of precipitation of snow and rain, and the effect of
climatic cycles on glacial movements, the structure of the ice as to stratification,
shearing, blue bands, ice dykes, glacial granules, and the possible methods of
their development. In discussing the theories of glacial motion the author
expresses his conviction that the nature of the ice movement can be “‘satis-
factorily explained only upon the theory that under certain circumstances and
within certain limits ice is capable of behaving as a plastic body, that is, capable
of yielding continuously to stress, without rupture,” but ‘‘the plasticity of ice,
a crystalline substance, must be thought of as essentially different from that
manifested by such amorphous substances as wax or asphaltum.”’

Doctor Sherzer also discusses the cause of the richness and variety of coloring
of glaciers and glacial lakes.

In accordance with the rule of the Institution this paper has been referred
to a commission consisting of Professor T. C. Chamberlin, of the University
of Chicago; Professor Harry F. Reid, of Johns Hopkins University, and Doctor
George P. Merrill, of the United States National Museum, and upon their
favorable recommendation is published in the series of ‘‘Smithsonian Contribu-

tions to Knowledge.”
RICHARD RATHBUN,
Acting Secretary.
SMITHSONIAN INSTITUTION,
WasuincrTon, D. C., January, 1907.

ee

PREFACE.

THE five glaciers selected for investigation are located in Alberta and
British Columbia, along the line of the Canadian Pacific Railway. They repre-
sent the great snow-ice masses which accumulate, season after season, upon the
higher slopes and within the amphitheaters of the Selkirks and Canadian Rockies,
the slow downward movements of which prevent indefinite accumulation and
bring these great ice bodies to a level where complete melting may occur and
the waters again be put into circulation. The observations here described
were begun by the writer in the summer of 1g02 and continued through the
seasons of 1903, 1904, and 1905; the entire field season of 1904 being devoted
to the surveys and more detailed studies... Camps were established in the
immediate vicinity of the glaciers selected and they were kept under almost
continuous observation during the hours of daylight. Beginning with the nose
of each glacier, surveys around either side to the névé field were made with
plane-table, transit, or compass; the measurements being with a steel tape.
It was found impracticable and unnecessary to traverse the névé areas and
those portions mapped were drawn from field observations and original photo-
graphs together with maps and illustrations from the Canadian Topographic
Survey, and other sources. The writer was ably assisted by Mr. DeForrest
Ross and Mr. Frederick Larmour, to whom he desires to make grateful acknowl-
edgment for intelligent and faithtul service, rendered: often under trying cir-
cumstances. During the latter part of the season of 1905 very efficient assistance
was rendered by Messrs. E. W. Moseley and O. K. Todd.

Being the most accessible glaciers upon the American continent it was
desired to render available as complete a description as time and facilities
would permit and to ascertain to what extent the known glacial features of
other portions of the world are reproduced in these American representatives.
It was hoped that a study of the same features, produced under somewhat

- different conditions, might shed additional light upon their method of formation
and upon some of the unsettled problems of Pleistocene geology. A dispro-
portionate amount of time was devoted to the Victoria Glacier, at the head of
the superbly beautiful Lake Louise Valley, since this glacier is geologically the
most interesting and may well be taken as a type by students of glaciology.
A delightful camp site lies under the lee of the outer massive block moraine and
a still more picturesque one farther up, on a low shoulder of Mt. Whyte, over-

1A preliminary report upon the expedition appeared in May, 1905, in the * Smithsonian Miscellaneous
Collections,” vol. 47, Quarterly Issue, pp. 453 to 496.
v

vi PREFACE.

looking the small lakelets. Students who may carry this report into the field
generally desire an explanation which they can put to the test, upon the spect,
and so an attempt has been made at interpretation of the various phenomena
described. The value of such interpretation will be known only after others
have passed judgment upon the same features and more extended observations 3
are available.
Numerous forest fires in the season of 1904 prevented distant photography,

on account of the smoke or haze, but through the courtesy of the Dominion
Topographic Survey and of the Detroit Publishing Company we are permitted
to reproduce some of their general views, obtained under favorable conditions.
In addition to the views so used the writer is indebted to Captain Eduard Deville,
Surveyor General of Dominion Lands, and his Chief Topographer, Arthur O.
Wheeler, for a series of maps and photographs and much information concerning
the regions under study. To the Director, R. F. Stupart, of the Canadian
Meteorological Service, to the Assistant Director, B. C. Webber, and to Mr.
N. B. Sanson, very grateful acknow ledgment is made for meteorological data
relating to British Columbia and Alberta and for the use of instruments kindly
placed at the disposal of the expedition. Very sincere thanks are hereby ten-
dered also to Prof. Joseph B. Davis, of the University of Michigan, and to Prof.
Elmer A. Lyman, of the Michigan State Normal College, for the use of surveying
instruments. The writer further desires to express his deep gratitude to the
officials and employees of the Canadian Pacific Railway, who permitted the use
of their Swiss guides for the necessary higher climbing and in many ways rendered
very substantial assistance to the expedition. Finally to the packers and
outfitters, Messrs. Robert W. Campbell and George W. Taylor, with their indis-
pensable though often unwilling cayeuses, the writer desires to gratefull
acknowledge the most generous and courteous treatment.
W. H. SHERZER.

THe MicuiGcan State Norma CoLiece,
YpsiLanti, Micu., December, 1906.

ABE EOF CONDENS:

Advertisement
Preface .
List of Mlestrations’

CHAPTER I.
INTRODUCTION.

Geographical Data
-a. Physiographic features
b. Streams ; :
c. Glaciers selected for eredy, ;

Historical Data

Geological Data
a. Stratigraphy
b. Physiographic changes
c. Lakes
d. Alterations in deine

Climatic Data ;
a. Geographic disenbution oe moisture
b. Chinook winds
c. Oscillations in climate

CHAPTER II.
VicTortaA GLACIER.
General Characteristics
Nourishment
Double Tributary

a. Mitre Glacier; the Host
b. Lefroy Glacier; the parasite

Drainage
a. Surface Bp ition
b. Surface drainage
c. Margina] drainage
d. General drainage brook
e. Water temperatures .

Forward Movement

a. Measurements .
b. Frontal changes
c. Shearing

d. Crevasses

PAGE
lil

Xi

WwW &®W Ne

Cmwr An nN

1o
I2

15

19

20

22
22
22

24
24
25
27
28
29

29
29
32
34

Vill CONTENTS.

CHAPTER III.

Victoria GLACIER (continued).

1. Glacial Structure . : : 3 ‘ ‘ f
a. Stratification
b. Dirt zones
c. Granular structure
d. Capillary structure
e. Melting features
7. Blue bands
g. Ice dykes

2.- Surface Features

Superficial débris
Lateral moraines
Medial moraine
Terminal moraine
Dirt bands

Dirt stripes - :
Dust and pebble wells
Débris cones x
Glacial tables .
Surface lakelets

Rock reflection

Top ETS AO STA

3. Former Activity

a. Terminal moraines
b. Lake Louise basin

c. Lake Louise delta

d. Lake Louise Valley
e. Ancient till sheet

CHAPTER IV.
WENKCHEMNA GLACIER.
1. General Characteristics . : : 5 ; i : ; :

2. Piedmont Type

3. Nourishment

4. Drainage ; : : :

5. Moraines

6. Crevasses

7. Movement about the Front. : ; : ; : : ;

8. Former Activity ‘ : : : : , ; : ji
a. Bear-den moraines
b. Moraine Lake .
c. Valley of Ten Peaks .

CONTENTS. 1X

*
CHAPTER V.
Youo GLACIER.
a ; PAGE
eneral Characteristics. ‘ P . : ; : ‘ : : = 70
urishment . 4 3 : : : : : ; ; Laer 72

Distributary . : E ‘ : : ‘ es : 2 , ees

é peace : : 2 : : : : : 4 : : 56

ntal Changes. : ; : : k : : : : . ~ 978

aa. Moraines . : : E : 5 . : ¢ i , ; 78
-b. Plucking action : : : : : ‘ : : . 5 : 79
c. Yoho Valley . ‘ 3 ; ‘ : ; ; ; 3 : : 80
CHAPTER VI.

ILLECILLEWAET GLACIER.

eneral Characteristics . ; : A 7 : : , ; ; - 85

oraines ae : ; : J % i ; ona : ee
Surface débris . : ; : : ; 5 5 : ? é 83

Left lateral moraine . 5 7 : . : ‘ : : : : 83

Terminal moraines . 5 : i A 3 : ; S 3 84

Right lateral moraine : 5 : ; ‘ : : 2 ‘ : 85

Boulder pavement . . F : z : : ol : : ‘ 85

asses : 86

ce Structure . ; ; : 2 3 : ‘ : ; ; ; = 86
ainage 2 ; : : 3 : 87

e', a. Surface and marginal drainage. : . : : : : : ‘ a
> ee b. Terminal drainage . ; : j 5 3 j : ; : , 88
_¢. Temperatures . p , , 2 , 5 ; : ; ; : 89
ward Movement : : : : : : : 3 Eis a 480
rontal Changes. : ; ; : : 5 , : : : Boe 3)

a. Recession data c : 6 : 5 . 9 ; 5 : : gr

6. Ice waves : : : : : : : 5 : : : : 93
former Activity . : 3 : ; a z : : : : OS

a. Rock scorings : : 5 : : ; ; ; : : : 95

_ 6b. Bear-den moraines. ‘ 3 . 5 3 - % : 4 . 96

CONTENTS.

CHAPTER VII.

ASULKAN GLACIER.

General Characteristics

Piedmont Characteristics .

Nourishment

Moraines

Crevasses

Ice Structure .

Drainage

Frontal Changes

Former Activity . : :
a. Development and decadence

b. Bear-den moraines
c. Rate of retreat

CHAPTER VIII.

SUMMARY AND CONCLUSIONS.

Physiographic Changes in the Region
a. Mesozoic peneplain ; e
b. Pre-pleistocene erosion
c. Pleistocene erosion
d. Pleistocene deposition

Precipitation . : :
a. Geographic distribution
b. Climatic cycles
c. Ice waves
Piedmont Type of Glaciers
Parasitic Glacier
Bear-den Moraines
Surface Features . 3
a. Dirt bands, zones, and stripes

b. Differential melting effects .

Ice Structure
a. Stratification

b. Shearing
c. Blue bands
Ice dykes

e. Glacial granules
Theories of Glacial Motion

Color of Ice and Glacial Water .

PAGE

97
98

. 99
100
102
103
105
106
107
107

107
108

109
109
IIo
110
I1I

I1I2
Ii2
I1i2

113
114
114
Tetas;

118
118
119

120
120
121
122
123
125

129

131

a i i

EIST sOF TEEUSTRAGIONS:

Piate I.—Mt. Victoria and Lake Louise : ‘ ; : : Frontispiece
Puate I].—General view of Canadian Rockies from Mt. Balfour
Puate III.—Map of Victoria Glacier

Pirate I1V.—r. General view of Victoria Glacier, looking southward. 2. Lefroy tribu-
tary and Victoria Glacier .

PuaTe V.—1. Débris-covered nose of Victoria Glacier. 2. Névé field of Victoria
Glacier :

Pirate VI.—1. Path of avalanche along Victoria névé. 2. Hanging glacier upon Mt.
Victoria j

Puate VII.—1. Hanging glacier upon Mt. Lefroy. 2. Double névé field of Mitre
Glacier . : ; : : : : : oe A :

Prats VIII.—1. Parasitic Lefroy Glacier being carried by Mitre Glacier. 2. Western

face of Mt. Aberdeen. 3. Surface drainage stream upon Victoria Glacier. 4. Ob-
lique front of Victoria Glacier

Prate IX.—1. First’ stage in formation of a moulin, Lefroy Glacier. 2. Marginal
lakelet on west side of Victoria Glacier. 3. Inner end of abandoned drainage
tunnel, Victoria Glacier. 4. Mouth of abandoned drainage tunnel, looking outward;
Victoria Glacier

_Piate X.—1. Drainage from Victoria Glacier after a day of much melting. 2. Refer-

ence boulder A, Victoria Glacier. 3. Stratified ice front, Victoria Glacier. 4. Front
of Victoria Glacier, showing irregular stratification and shearing .

PiatEe XI.—Time of contact between two dirt zones, Lefroy Glacier. 2. Dirt zones
upon Lefroy Glacier

Puiate XII.—1. Glacier capillaries, Yoho Glacier. 2. Glacier capillaries, Ilecillewaet
Glacier. 3. Stratification on wall of ice tunnel, Victoria Glacier. 4. Blue bands
in Lefroy Glacier :

Pate XIII.—1. Blue bands near nose of Illecillewaet Glacier. 2. Contorted blue
bands, Yoho Glacier. 3. Ice dyke filled with horizontal ice prisms. 4. Crevasse
in Victoria Glacier .

PLatE XIV.—1. Stony till, left lateral moraine, Victoria Glacier. 2. Sharply crested
left lateral moraine, Victoria Glacier

Pirate XV.—1z. Ground moraine, Lefroy Glacier. 2. Right lateral and medial mo-
raines of Victoria Glacier

Prate XVI.—1. Formation of Forbes’s dirt bands, Deville Glacier. 2. Forbes’s dirt
bands, Victoria Glacier

Pratr XVII.—1. Forbes’s dirt bands, Asulkan Glacier. 2. Dust wells, Victoria
Glacier. 3. Small dirt cone, Victoria Glacier. 4. Same cone, dirt veneering
removed

xi

FACING
PAGE

I

6
19

20

21

22

23

24

27,

39

41

47

48

50

X1l LIST OF ILLUSTRATIONS.

Prats XVIII.—1. Glacial table, Victoria Glacier. 2. Dethroned glacial table, Vic-
toria Glacier . :

Pirate XIX.—1. Boulder mound, Wenkchemna Glacier. 2. Surface lakelet, Vic-
toria Glacier .

Piatt XX.—Map of delta, head of Lake Louise
Piate XXI.—The Continental Divide, Canadian Rockies
PLateE XXII.—Map of Wenkchemna Glacier

PiaTe XXIII. 1.—Drainage brook from Wenkchemna Glacier. 2. General view of
eastern end of Wenkchemna Glacier .

PrateE XXIV.—1. Front of Wenkchemna Glacier. 2. Front of glacier showing
forest invasion. 3. Disintegrated blocks of bear-den moraine. 4. Melted area
on north side of surface block of quartzite, Victoria Glacier

PLateE XXV.—General view of Moraine Lake and eastern end of Wenkchemna Glacier
PiatE XXVI.— Map of Yoho Glacier

Pirate XXVII.—1. Ice distributary from Yoho Glacier. 2. Ice plucking upon a
mountain peak, head of Yoho Valley. 3. Yoho Glacier, head of Yoho Valley.
Awe hree-hundred-foot ice arch, Yoho Glacier p :

Prate XXVIII.—1. Knoll and ridges of ground morainic material in front of Yoho
Glacier. 2. Yoho Valley, showing Wapta and Waputik snowfields

PLatE XXIX.—1. Formation of Forbes’s dirt bands from crevasses, Yoho Glacier.
2. Hanging Valley, head of Yoho Glacier .

PLATE XX X.—Map of Illecillewaet Glacier : ; :
Pirate XX XI.—Tongue and moraines of the Ilecillewaet Glacier, August, 1899

Pirate XXXII.—1z. General view from Roger’s Peak. 2. General view of Illecillewaet
Glacier

Pirate XX XIII.—Map of the Selkirk snowfields and glaciers

PratE XXXIV.—1. Beginning of subglacial fluting by pressure melting, Ilecillewaet
Glacier. 2. Subglacial fluting, Illecillewaet Glacier. 3. Roches moutonnées
near nose of Illecillewaet Glacier

Pirate XXXV.—1. Regelation of ice blocks at foot of ice cascade, Illecillewaet Glacier.
2. Stratification in upper part of Illecillewaet Glacier

PLrate XXXVI.—1. Illecillewaet Glacierin 1888. 2. Ilecillewaet Glacier in 1905

Pirate XXXVII.—1. Bear-den moraine made conjointly by the Illecillewaet and
Asulkan Glaciers. 2. Illecillewaet Glacier in 1897

PLateE XX XVIII.—Map of Asulkan Glacier

PLaTteE XXXIX.—1. General view of Asulkan Glacier in 1902. 2. The Asulkan
glaciers and snowfields from Mt. Avalanche j

Pirate XL.—1. Left Asulkan moraine. 2. Nose of Asulkan Glacier, 1904

Pirate XLI.—1. Stratification of Asulkan Glacier. 2. General view of Asulkan Glacier
in 1898.

Pirate XLII.—1. Development of seracs from glacial blocks, Asulkan Glacier. 2. Dis-
rupted quartzite blocks illustrating plucking power of glaciers

FACING
PAGE

56

57
61

62
63

64
68

7°
71

74
78

80

82
84
85
86

88
92

96
97

98

100
I02

108

a ee — SS ee ee ee

ee eee eee ee SA eas nas LA WAINA ANT —*SSTMAONT

*Ao1jaT ‘IN

AUM IN
‘T ALV1d

MAZYAHS—AVIGATMONY OL SNOILLOAAINLNOO NVINOSHLINS

OTT:

—.

fr elensS OF THE CANADIAN ROCKIES
AND SE EKIRAS

(Report of the Smithsonian Expedition of 1904.)

By Wiiii1aM Hitrevy SHERZER, Ph.D.

CHAPTER I.
INTRODUCTION.
I. GEOGRAPHICAL DaTa.

a. Physiographic features—The Canadian Pacific Railway crosses the
Rockies and Selkirks between north latitude 51° and-51° 30’, working its way
up the left bank of the Bow River and its small tributary Bath Creek, to the
Kicking Horse Pass, attaining an altitude of 5,329 feet above sea-level. Upon
the more abrupt western slope of the Rockies the road follows the left bank of
the Kicking Horse River to its junction with the Columbia, crosses this great
waterway of the mountains, and slowly ascends the eastern slope of the Selkirks
along the left bank of the Beaver. The summit of the Selkirks, Rogers’ Pass,
is crossed at an elevation of 4,351 feet, whence there is rapid descent along the
swift-flowing Illecillewaet to the Columbia again, which has encircled the system
to the north, forming the “Big Bend.” These transverse mountain valleys
are lined with most majestic peaks, many of them rising a mile above the valley
floor and furnishing some of the grandest of mountain scenery upon the American
continent. The highest peak in the Rockies, seen from the railway, is Mt.
Temple, with an elevation of 11,627 feet, and in the Selkirks, Mt. Sir Donald,
10,808 feet. Numerous peaks range from 10,000 to 11,000 feet and are believed
to culminate to the northward in latitude 52° to 53°.

The Rockies and Selkirks, together with the Gold and Coast ranges to the
west, make up the Great Cordillera in this part of Canada. North of the inter-
national boundary this great system is much narrower than in the United States,

I

2 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

having a total width of about four hundred miles, and the component ranges
are straighter and more regular. The systems are progressively higher from
the coast eastward, culminating in the Rockies proper, which stand as a lofty
buttress along the western margin of the great central plains. Between the
Coast and the Gold ranges there lies an interior plateau a hundred miles wide
with an average elevation of about 3,500 feet above sea-level. The Gold,
Selkirk, and Rocky systems are separated by the Columbia and the Columbia-
Kootenay valleys, made by the action of water and ice along the strike of the
geological formations, assisted probably by some dislocations of the strata.

The Rocky Mountains, or as formerly called, the Stony Mountains, consist of
an imposing array of parallel ranges with a general trend in this region of north-
northwest to south-southeast, separated by longitudinal valleys and attaining
a total breadth of 40 to 50 miles. Compared with the systems to the west
they are strikingly rugged in character and free from vegetation. Skirting the
eastern border, and a part of them both geologically and structurally, are the
‘‘foot-hills,”” consisting of folded parallel ridges, reaching out 15 to 20 miles and
merging into the ‘“‘plains” at an elevation of about 3,300 feet.

b. Streams.—The main streams occupy the longitudinal valleys for a
portion of their course, leaving the mountains by the transverse valleys, which
extend into the foot-hills. According to Dawson the base-level of the streams
upon leaving the mountains to the eastward is about 4,360 feet, while to the
west it is about 2,450 feet above sea-level. Upon the eastern slope of the Great
Continental Divide the waters are gathered into the Saskatchewan and reach
the Atlantic Ocean by way of Hudson Bay; while those to the west drain into
the Columbia River and work their way to the Pacific Ocean. As pointed out
by Dawson, the actual water parting does not correspond entirely with the
highest crest line of the mountains, but lies to the eastward, in which direction
it seems to be moving. Between the international boundary of north latitude
49° and 52°, the Rocky Mountains are sharply separated from the Selkirks to
the west by the Columbia-Kootenay Valley, which maintains a considerable
breadth and a remarkably straight course through more than three degrees of
latitude. This valley is filled with drift materials to a considerable depth and
is undergoing but little erosion, the river simply cutting tortuous channels
through the loose deposits. The eastern side of the valley is generally steep
and escarpment-like, while the western is rounded and wooded. The Columbia
starts within a mile and a half of its southward-flowing tributary, the Kootenay,
and moves northwestward in a great sweep as though intent upon capturing
the drainage of the region, before starting for the sea. In this great fold it en-
closes and sharply limits the less rugged, but picturesque, Selkirk System,
with its subdued outlines and forested slopes. Some of the eastern ranges are
continuous and have the same general trend as those of the Rockies, but, in
general, there is less regularity and continuity in the arrangement of crests
and peaks, and they do not attain as great a height. The drainage is all into
the Columbia River, and the streams are unable to develop any considerable

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 3)

size. Being so largely glacier fed here, as well as in the Rockies, the streams
maintain themselves during the summer months, but reach their highest stage
in the late spring, or early summer, from the melting of the snows, when the
Columbia may rise 30 to 40 feet above its usual level. The streams are generally
turbid with glacial sediment that gives them a milky, or yellowish, appearance,
changing to green and, upon the loss of the sediment, to a blue color wherever the
water is of considerable depth. The lakes of the region owe their origin mainly to
former glacial action, consisting either of rock-basins, or of depressions in the
glacial or fluvio-glacial deposits. Certain ones have been dammed back by
morainic material deposited either beneath or at the extremities of glaciers of
greater extent than at present. Those lakes which receive glacial sediment,
or which are shallow, have a greenish cast, while those free from sediment
and of moderate to considerable depth are rich blue.

c. Glaciers selected for study.—The glaciers selected for study lie close to the
main crests of the systems above described, between north latitude 51° and 52°,
and west longitude 116° and 117° 30’, from 160 to 200 miles north of the inter-
national boundary between the United States and Canada. They are but a
few of a series available for study, those being selected which are most easily
reached by well established trails. They are at such low altitudes that one may
comfortably ride almost to the nose of each and none require climbing except
to reach the névé regions. The two most easterly of the glaciers here discussed,
the Victoria and Wenkchemna, lie east of the Great Divide in Alberta, the other
three are west in British Columbia.

2. HistoricaL DATa.

The establishment of the international boundary to the south, along the
4oth parallel, and the opening of the railway in 1885 called for geographic,
geologic, and topographic work which was started by the various Dominion
departments concerned and is still in progress. Dr. George M. Dawson began
his work in 1874 along the boundary and extended it northward to include the
region pierced by the railway, where he was assisted by R. G. McConnell.
Topographic work of a preliminary nature, along the line of the railway, was
begun in 1886 by J. J. McArthur. Photographic methods were introduced into
the survey in 1889 by Director Deville and the accompanying triangulation of
the “railway belt” placed in charge of W. S. Drewry, D. L. S. The same
year Mr. St. Cyr made a survey along the upper Columbia, between the Selkirks
and Rockies, and in 1896 he and McArthur continued the work from Revelstoke
down the Columbia and Arrow Lakes, with the view.of connecting the surveys
of the railway belt with those of the boundary commission.!. Two topographic
maps, upon a scale of two miles to the inch, were issued in 1902 by the Depart-
ment of the Interior, under the direction of James White. geographer. These
are the Banff and Lake Louise sheets and are issued by the department at

1 The reports of the work of McArthur, Drewry, and St. Cyr will be found in the Annual Reports of the
Canadian Dept. of the Interior for 1886, 1888 1889, 1890 1891, 1892, and 1893.

4 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

Ottawa. The topographic work of the mountains is now in the hands of Mr.
Arthur O. Wheeler and there is being issued an enlarged map (scale 5,000 feet
to an inch) of a section of the mountains lying between the railway and the
Great Divide and extending from the Kicking Horse to the Vermilion Pass.
This map includes completely the regions surrounding the Victoria, Horseshoe,
and Wenkchemna: glaciers, with corrected elevations, essentially the same
territory covered by Wilcox’s map, 1896, on the scale of an inch and a half to
the mile. Based upon work done during the seasons of tgo1 and 1902, there
will be issued with vol. 1 of Wheeler’s Selkirk Range an admirable piece of
mountain mapping, extending from the Columbia to the Columbia, across the
Selkirks along the line of the railway.

The opening of the railway and the wonderful attractions of the region
brought in a body of non-professional explorers and mountaineers, among the
first of whom was the Rev. W. S. Green, Carrigaline, Ireland. He spent the
working season of 1888 in the Selkirks, using Glacier House as a base, and gathered
material for an interesting volume, Among the Selkirk Glaciers, Macmillan &
Co., 1890. The map accompanying the volume, originally published in the
Proceedings of the Royal Geographical Society, vol. x1, 1889, was the first attempt
at detailed mapping in the Selkirks. One year earlier than Green, in 1887,
Messrs. George and William Vaux, Jr., of Philadelphia, visited Glacier House,
secured a valuable collection of photographs and began a series of observations
upon the glaciers to which frequent reference will be made in the later chapters
of this report. During the closing decade of the last century, and the opening
decade of the new, there has been much work done in the region of an exploratory
and mountaineering character. There should be mentioned especially the names
of Wilcox, Fay, Parker, Collie, Stutfield, Allen, Habel, Topham, Thompson,
Huber, Sulzer, Noyes, and the English ladies Benham, Tuzo, and Berens.
Besides three superbly illustrated and attractively written volumes by Wilcox,
Wheeler, and, conjointly, by Stutfield and Collie,! there have been prepared
a number of descriptive papers for the scientific societies and magazines. A
bibliography of the region, full but not complete, will be found in Appalachia,
vol. x, 1903, pp. 179 to 186. The Canadian artist, F. M. Bell-Smith, of Toronto,
has spent many seasons in the mountains and, based upon the various maps,
photographs, and original sketches, has prepared relief maps of the best known
areas of the Rockies and Selkirks. Copies of these maps are placed in the hotels
operated by the railroad.

It is not likely that these mountain valleys ever supported anything more
than a scant Indian population, owing to the scarcity of fish, game, and available
pasture. Providing food, en route, has always been a precarious matter for
exploring parties. Aside from the marmot and rock-rabbit and an occasional
porcupine, there is a strange and impressive feeling of desertion. The few

1 Camping in the Canadian Rockies, Wilcox. G. P. Putnam’s Sons, N. Y., 1896. The Rockies of Canada,
Wilcox. Putnams, 1903. Climbs and Explorations in the Canadian Rockies, Stutfield and Collie. Longmans,
Green & Co., N. Y., 1903. The Selkirk Range, Wheeler. Government Printing Bureau, Ottawa, 1905.

Ee

GLACIERS OF THE CANADIAN RO@KIES AND SELKIRKS. 5

birds that one meets seem awed into silence by the grandeur of their surround-
ings. The mountain Crees had possession of the region at the coming of the

white traders and trappers, but within rather recent time have been assimilated

by the Stoneys, a tribe of Assiniboines, from the plains to the east.

3. GEOLOGICAL DATA.

a. Stratigraphy. The first work of a geological nature in this region was by
Dr. Hector in 1858 to 1860, as a member of the Palliser expedition, his observa-
tions being confined mainly to the Rockies and the region to the east. A
geological map and numerous sections were prepared to accompany a paper
presented to the Geological Society of London, in advance of the publication
of the results of the expedition.! For detailed knowledge of the geology of the
Rockies and Selkirks we are indebted mainly to Dr. George M. Dawson and his
assistant R. G. McConnell, the former of whom began his work in 1874, as geol-
ogist of the boundary commission. The Bow River region was entered in 1881
and in the Annual Report of the Geological Survey for 1885 there was published
a preliminary report upon the geology of the Rockies lying between the boundary
and north latitude 51° 30’. The report was accompanied by a large scale
geological map, which was followed the next year with a geological section by
McConnell, approximately along the line of the 51st parallel of latitude. Work
was extended westward into the Selkirks and, at the Washington meeting of the
Geological Society of America, Dr. Dawson, in 1890, presented the results of his
observations amongst these ranges.2 The present Geological Survey, under the
directorship of Robert Bell, is still at work upon the detailed study of portions
of the region.

The Selkirks and Rockies consist of an enormous complex of sedimentary
strata, 50,000 to 60,000 feet in thickness, underlain by crystalline rock. In
age they range from the Archean to the Laramie, at the close of which the
final stages of upheaval were accomplished. The rock strata graduate in age
from the west, eastward, and were folded and faulted by pressure from the
west, by which they were forced against the resistant layers underlying the
“great plains.’’ The crystalline rocks of the series, of presumable Archzan
age, consist of gray gneisses, passing into schists, and occur only along the
western margin of the Selkirks, where they constitute the Shuswap series. No
trace of them has yet been discovered in the Rockies. Overlying the series occurs
the Nisconlith, with an estimated thickness of 15,000 feet, consisting of dark
colored argillite-schists and phyllites, showing various stages of alteration from
true argillites to micaceous schists. Interbedded layers of dark limestone and
quartzite are seen in certain sections. Although the beds yielded no fossils
they were referred to the Cambrian by Dawson, because of their relation to the

1“*On the Geology of the Country between Lake Superior and the Pacific Ocean,” James Hector,
M.D., 1861, Quart. Journal Geol. Society, vol. xvu, pp. 388 to 445.

2‘*Note on the Geological Structure of the Selkirk Range,” Geol. Soc. of Amer., vol. 2, 1891, pp. 165 to
176. An extract from this paper is given in Wheeler’s Selkirk Range, vol. 1, pp. 405 to 409.

6 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

crystallines and their supposed equivalency with strata of known Cambrian ‘ij
age in the Rockies to the eastward. The crest of the Selkirks, the region in *
which occur our snowfields and glaciers, consists of the folded Selkirk series, oa
with an estimated thickness of 25,000 feet and believed to be of Cambro- Silurian: ;
age. The strata have been forced into a synclinal fold, which terminates to

the eastward by a thrust fault, produced by the western half of a sharp anticline

being thrust upward with reference to the eastern half. The rocks consist of —

eray schists and quartzites, passing into grits and conglomerates which weather

to pale yellowish or brownish colors. The latter are often more or less schistose

from pressure and other metamorphic agencies, silvery mica, or sericite, being —

developed. At times the strata are wrinkled and contorted.

Passing into the Rockies, to the eastward, we find them made up very
largely of the representatives of the Nisconlith and Selkirk series just described,
but known in the report of Dawson as the Bow River and Castle Mountain series. _
In the western part of the Rockies, adjacent to the Columbia, the upper and
younger of the two is continued from the Selkirks, showing crumpling and folding, ;
with metamorphism, but without faulting. The rocks are dipping eastward
and have their “ strike’? parallel with the mountain ranges. Along the center
of the range the folds are broad and sweeping, while eastward, for about 2 ie
miles, there is a succession of thrust faults, running parallel with the ranges, .
the maximum vertical displacement being estimated at 15,000 feet. McConnell
made out seven of these faults, giving rise to a series of massive mountain
blocks resting in succession upon one another and forming escarpments to the —
east and relatively gentle slopes to the west. It was to this type of mountain
that Leslie Stephen applied the suggestive term “‘ writing-desk.”’ ,

The ranges making up the central portion of the Rockies are of the Castle —
Mountain series (Selkirk series) and of Cambro-Silurian age. They are more —
regular and depart less from the horizontal than the strata to the east and
west. Mt. Stephen, on the line of the railway, gives a 5,000-foot section of the
series, one shaly band being remarkably rich in Cambrian trilobites. The total
thickness of the series is estimated at 10,000 feet, as compared with 25,000
feet in the Selkirks, and consists of limestone and dolomite, with calcareous iz 5
schists and shales. These rocks give the steep-sided, massive, block-like cliffs, _
typically shown in Castle Mountain, which has furnished the name for the —
series. These rocks extend lengthwise of the central ranges to the Yoho ~
Glacier at the north and the Victoria and Wenkchemna glaciers to the south.
At the base of Mt. Stephen. at the head of Lake Louise, and in the Bow River,
there has been brought up from below by an anticlinal fold the “Bow River —
Group,”’ or Nisconlith series of the Selkirks. This is of Cambrian age, estimated _
at 10,000 feet in thickness, and consists of quartzites and Conse with —
dark gray, purplish, and greenish argillites.

b. Physiographic changes. When viewed from a high elevation the rough — 3
ridges and jagged peaks appear to blend, as far as the eye can reach, into a great
plateau with a notably even sky-line (pl. m1), giving the appearance of an Z

Sis aurp-Ays Jo raioeiwyo uaagq ~*Aaarng orydevasodoy, uetpeuey ayy

‘uetdauad paqoasstp puv payyttdn uv so
Way Aq ‘tobi ‘paydeis0} Yq "3SaAqy10U-ysa.M SULOO| ‘(jaazy 1EL‘01) AMOJ[V “AP Jo Wwuins wor; SaLypooy ue

2.) JO MOIA [RIAUdy

jo Asajanoo ay) Y al pu 1apaeyA\ “O J

IT peoup

MAZUAHS—ANGATMONM OL SNOILAALIALNOO NVINOSHLIWS

Tl GLv id

~

ee eee ae S| ee

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 7

uplifted peneplain, similar to that observed by Gilbert farther north in Alaska.
The upheaval of such a mass, by lateral pressure, would give rise to troughs and
open gaping crevices parallel with the previously formed mountain folds and
into these the drainage streams would, for the most part, be diverted. Deepened
by stream action, widened by atmospheric agencies, and still further modified
by Pleistocene glaciers, we have the longitudinal valleys of the Rockies and
Selkirks. The transverse valleys, noted by Dawson as extending into the
foot-hills and due to “‘causes not now apparent,’’ probably mark the location
of drainage streams developed while the peneplain was being formed and antedate
the final upheaval of the region. That these valleys were occupied by extensive
glaciers, presumably in Pleistocene time, is everywhere evidenced by the
morainic accumulations, rounded rock contours, glaciated surfaces, extensive
plucking, truncation of mountain spurs, amphitheaters, rock basins, and hanging
valleys. The valleys generally were filled with ice to a depth of 2,500 to 4,000
feet during the maximum period of glaciation, the height, as pointed out by
Wilcox, rarely falling below 7,000 feet above sea-level. Either because the
mountains were so completely enveloped in ice and snow, or because of the
nature of the final retreat, extensive terminal moraines were not formed in
the main valleys. Ground-morainic deposits, however, hundreds of feet
thick, occur in places favorable for their lodgment beneath the ice. Near
Banff, in the Bow and Cascade valleys, Wilcox discovered evidence of two
distinct till-sheets, the older highly charged with pebbles, with little clav, the
younger consisting mainly of very hard clay.? In the extension of the Bow
Valley to the eastward of the mountains, McConnell and Dawson found three
separate till-sheets. The lowest and oldest of these appeared to have been en-
tirely derived from the mountains and to pass eastward gradually into the
“Saskatchewan gravels” of the plains. For this formation Dawson suggested
the term ‘‘Albertan,’’? which he regarded as of pre-Kansan Age.

The upper two boulder-clays contained rock fragments of both eastern and
western origin, each variety preponderating in the direction of its origin, showing
a commingling of the deposits of the Cordilleran and Hudson Bay ice sheets. The
middle of the three till-sheets Dawson correlated with the Kansan and the
upper with the Iowan (p.509). In the light of our present knowledge the upper
would be referred to the J/lino1an, which succeeded the Kansan in the upper
Mississippi Valley and was of much wider extent than the Iowan. The corre-
lation of these sheets with those observed at Banff has not yet been made.

c. Lakes. In the very interesting paper above referred to, Wilcox recog-
nizes four types of lakes in those portions of the Canadian Rockies which came
under his observation:

First—Lakes lying in depressions of the valley drift, often in chains of

1 Harriman Alaska Expedition, vol. 111, Glaciers and Glaciation, Gilbert, p. 183.

2‘*A Certain Type of Lake Formation in the Canadian Rocky Mountains,” Jour. of Geol., vol. v11,
1899, Pp. 249. :

3‘*Note on the Glacial Deposits of Southwestern Alberta,” Jour. of Geol., vol. 111, 1895, p. 510. See
also note on page 384, vol. 111, Geology, by Chamberlin and Salisbury.

8 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

three, or four, and especially numerous near the summits of the mountain
passes. These lakes show no regularity in form, or location, are usually shallow,
and frequently have neither inlet nor outlet.
Second—Lakes dammed by terminal moraines. In the same class may be
included those dammed by alluvial cones, or deltas formed in preéxisting lakes.
Third—Lakes lying in rock basins, excavated by former glacial action.
Fourth—A special type of lake, of which Lake Louise is an example, formed
just within the mouth of a tributary valley. These lakes are leaf-shaped and _
from three to ten times longer than they are wide. They owe their existence ©
to the presence of a ridge of ground-morainic material, thrown across the mouth —
of the tributary valley from the up-stream side and curving gently out into the
trunk valley. These ridges have apparently been formed beneath the ice,
when the valleys carried glaciers, by the joint action of the tributary and trunk
glaciers. The lakes may be shallow and so filled with silt that they are reduced _
to swamps, or, where the ice was especially active, as in the case of Lake Louise,
the depth may still be surprisingly great. As recognized by Wilcox, these
lakes may present a combination of the rock-basin and morainic-dam types.
d. Alterations in drainage. In a region of the character above described,
exposed for countless ages to the effects of weather, water, and ice, marked
alterations in the drainage would be expected, such as reversals, stream capture,
and the migration of divides. A study of the Columbia-Kootenay Valley has —
shown that the upper 200 miles of the Columbia flowed at one time to the south, a
instead of encircling the Selkirks to the north as at present. This must neces-
sarily have been the case so long as that portion of the valley known as the
‘‘Big Bend” was in possession of the ice. The withdrawal of the glaciers into
the tributary valleys would permit the present northward flow from the Columbia —
Lakes while the Kootenay, flowing in the opposite direction in the valley to
the east, enters the main valley, approaches within one and one-half miles of
the head-waters of the Columbia, but completely skirts the Selkirks to the south
before joining it. Upon the opposite side of the Rockies attention has been
called by Dawson, McConnell, and Ogilvie to changes in the Bow and its tribu-
taries. The long, slender Lake Minnewanka Valley, near Banff, was evidently
the course of a prepleistocene valley that was occupied and modified by an ice
stream during the maximum period of glaciation of the region. Dawson con-
sidered this to mark the former course of the Bow, which was deflected to the
southeastward, along the strike of the soft Cretaceous shales, when the lake
valley was occupied by ice.1_ From barometric observations made by Dawson, _
his assistant McConnell noted that the Ghost River opposite the Devil’s Gap, —
the mouth of the Lake Minnewanka Valley, is considerably higher than the valley —
floor and concluded that the Ghost River turned and entered the mountains —
through this valley, joining thus the Cascade and with it forming a tributary _

1 Annual Report of Canadian Geological Survey for 1885, ‘‘ Preliminary Report on the Physical and
Geological Features of that Portion of the Rocky Mountains between latitudes 49° and 51° 30’,”” 1886,
p. 141B.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 9

of the Bow.! The topographic map of the region, issued in 1902, shows that
the river opposite the Gap is fully too feet higher than the level of Lake Minne-
wanka and about 4oo feet higher than the present level of the Bow in the
vicinity of Banff. According to this view of McConnell the ice-filled Minne-
wanka Valley compelled the Ghost River to find for itself a new course across
the foot-hills to the eastward, which it deepened sufficiently, assisted probably
by the ice, to prevent the return of the river into its former course upon the
withdrawal of the glacier from the valley. In 1904 Dr. I. H. Ogilvie examined
the region and reached the conclusion that the upper Bow and Minnewanka
valleys were formerly continuous and that the Bow Valley below Banff was
occupied by a stream which cut back into the soft shales until it tapped the
upper Bow and effected its capture. She concluded? that this had been ac-
complished in prepleistocene time and that the Lake Minnewanka Valley was
occupied by a glacier, fed by hanging glaciers, which moved westward, rather
than to the east, deepening the western end of the valley and forming certain
morainic deposits about the western end of the lake. The lake itself and its
southwesterly drainage would then date from the withdrawal of the ice from
the Cascade and Minnewanka valleys. Dr. Ogilvie holds that the drainage in
the Lake Minnewanka Valley before the advent of the glaciers was eastward.
The bed of the Bow River at Banff is approximately 4,500 feet above tide,
while that of the Ghost River, opposite the Devil’s Gap, is not far from 4,900
feet. The present Bow, in the same distance as that from Banff to this portion
of the Ghost River, drops 300 feet, so that the bed of the Bow at Banff is some
700 feet lower than it should be in order to have the upper Bow leave the
mountains by the Devil’s Gap and thence by the lower Ghost River. All
will grant that this is too much cutting to expect of the Bow in postpleistocene
time and that Dr. Dawson’s theory of the diversion of the Bow by the Pleistocene
glaciers is untenable. Noting that the Ghost River has also been deepening
its bed, with a much steeper gradient and presumably for as long a time as the
Bow, the above 700 feet must represent the excess of cutting by the upper Bow,
when compared with the Ghost, since its capture by the lower Bow, upon Dr.
Ogilvie’s hypothesis. We have no knowledge of the depth of the drift deposits
opposite the Devil’s Gap, but there is no reason to think that they would be
any deeper there than in the valley of the Bow. The explanation that lies
nearest at hand is that of McConnell, viz., that the Ghost River upon reaching
the foot-hills made a sharp turn and reéntered the mountains by the Devil’s
Gap, but was diverted eastward when the Minnewanka Valley became ice-
filled. According to this hypothesis the valley of the Ghost from the Gap
down should show less maturity than that above the Gap, except so far as it
may have been modified by ice action. During the period of maximum glacia-
tion the ice movement in the Minnewanka Valley must have been eastward,

1 Annual Report of Canadian Geological Survey for 1886. Report on the Geological Features of a
Portion of the Rocky Mountains, 1887, R. G. McConnell, p: 9D.

2**Geological Notes on the Vicinity of Banff, Alberta,” Jour. of Geol., vol. x11, No. 5, 1904, pp. 408
to 414.

Io GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

any tendency towards a westerly movement being checked by the ice in the
Cascade and Bow valleys. The westerly movement noted by Dr. Ogilvie was
simply a minor episode toward the close of the Pleistocene glaciation. _

4. Criimatic DATA.

a. Geographic distribution of moisture. The climatic conditions of this
section of country are peculiarly dependent upon the physiographic features
above outlined, combined with its proximity to the Pacific. The centers of
the areas of low pressure commonly move in from the ocean to the northward
of the region, give rise to westerly winds which convey the warmth and moisture
of the Pacific currents across British Columbia and Alberta. At Nanaimo,
upon Vancouver Island, separated from the mainland by 30 miles of strait,
the precipitation records available show a rain- and snowfall combined ot 41.36
inches, only 5 per cent. of which falls as snow. Opposite, upon the mainland,
the total precipitation at Vancouver rises to 63.06 inches, with 4 per cent.
falling as snow. This increase is due to the Coast Ranges, having here a north-
west-southeast trend, which compel the westerly winds to ascend their westerly
slopes, by which rise the air is cooled and its capacity for holding moisture
thereby diminished. At Agassiz, in the lower Frazer Valley, the precipitation
is but slightly less, although it is located some 70 miles from the Strait of Georgia,
up the broad open valley, and about 50 feet above sea-level. Records are
available here since 1890, with the exception of the years 1891 and 1899, and
give for the 14-year series an average precipitation of 62.02 inches, 6 per cent.
of which falls as snow. Over the broad interior plateau which lies between the
Coast and Gold ranges, the temperature is colder and the precipitation relatively
slight. At Kamloops, with an elevation of 1,160 feet and in latitude 50° 41’,
the combined rain- and snowfall averages but 10.66 inches. Passing eastward
the air currents impinge upon the westerly slopes of the Gold Range, are again
compelled to ascend to still higher altitudes, with the attendant loss of moisture.
In consequence, the station of Griffin Lake, located in this range at an elevation
of 1,511 feet, and go miles east of Kamloops, receives an average precipitation
of 34.37 inches, or over three times as much as the latter place. Crossing the
Columbia a still higher barrier is encountered in the impressive Selkirk system
of ranges, and a correspondingly increased amount of moisture extracted from
the still laden air currents. The records in the Selkirks are unfortunately

meager, but they indicate a precipitation almost as great as that of Vancouver —

and Agassiz, and considerably in excess of Nanaimo and Victoria out in the
Pacific. The station of Glacier House is located just west of the main crest
of the Selkirks, in latitude 51° 16’ and at an elevation above tide of 4,093 feet.
The average total precipitation here is 56.68 inches, of which 77 per cent. falls
as snow. If the entire amount were precipitated as snow, as is practically the
case upon the peaks and elevated névé fields, this would represent an average
fall of over 47 feet. This heavy snowfall in the Selkirks and Gold Range has

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. cr

necessitated the erection of numerous snow-sheds over the tracks of the Canadian
Pacific Railway, to guard against the frightful avalanches which come crashing
down the mountain slopes.

Although Donald is located but a few miles east of Glacier House, it is upon
the lee slope of the Selkirks and in the Columbia Valley some 1,500 feet lower.
The precipitation here drops to 25.39 inches. The still higher Rockies im-
mediately follow, but the Selkirks have proven greedy and there is relatively
little left in the way of moisture. Full precipitation data in the region of the
crest ranges of the system are wanting. The snowfall at Field, however,
averages about 27 feet, at practically the same elevation as that of Glacier
House. If we assume that the same ratio holds here, between snow and rain,
as at the latter place the precipitation at Field, lying just west of the main
crest, would be about 42 inches. Passing the continental. crest the currents
are drawn to lower levels, they become warmed by their descent, and their
capacity for retaining moisture increases. At Banff the precipitation for the
13 complete years available averages 20.14 inches, of which 39 per cent. falls
assnow. Beyond the foot-hills, at Calgary, the precipitation is reduced to 16.64
inches, of which 28 per cent. is snow. The following table furnishes a summary
of the climatic data of special interest in connection with this report.

TABLE I.

CiimaTic Data, FROM RECORDS OF THE CANADIAN METEOROLOGICAL SERVICE.

| ea [eae ew EER
E € g 83 en es 28 5 Es + 5 | 88
Stations, ar- i a Rohe as § 5 s § § 3 B82 |e 3 33 Remarks concern-
tanged from West o vu on| og | as pe +8 O8 |88lae ing location.
pease. 3 e Beales nee) oe 3 Ba |§slo8
2 Pa S| sa | 8 a8 ge B35 |o8|e2
w 8 aa) 24 sé LOO 73) z3 Sea |Sal 3°
yn A AS|He | a3 Ge | wAs <a [Ave|a8
Nanaimo 4g°ro’ | 123°57’ |-30| x17/| 48.7°| 90.3°| —3.9°| 41.36] 5) 5|East side of Vancouver
| | Island.
Vancouver Ager Wera3zo5f 10| 136] 48.0°| 89.0° 6.0° | 63.06] 4| 4) West slope of Coast
| Ranges.
Agassiz 49°14’ | 21°31’ | 70] 52| 47.7°| 103.0°| —13.0° | 62.02] 6|14|In open valley of Fra-
zer River.
Kamloops 50°41’ | 120°20' | 270\1160| 47.1° r01.0° | —27.0° | 10.66] 24/10 |Interior plateau.
Griffin Lake | 50°58’ | 118°31’ | goo|1srz| 44.6°| 110.0°| —23.0° | 34.37| 37|10|Gold Range.
Glacier House| 51°16’ | 117°30’ | 460 4093| 36.9°| 89.0°} —21.0° | 56.68] 77| 5|Selkirk System.
Donald 51°29’ | It7°TI' | 470|2580| 37-4°| 97.0°| —45.0° | 25.39| 51 |10|Columbia Valley be-
| tween Selkirks and
Rockies.
Banff 51°10 115°35’ | 550|4542| 35-3°| 88.79] —48.8° |20.14 | 39|13|East of Continental
| Divide about 15 miles.
Calgary gto? 4 114°2’ | 600/3389| 37-2°| 95.0°| —49.4° | 16.64) 28|19|Just beyond the foot-
hills of the Rockies

Quite in contrast with the still higher mountains of this same system to
the south in the United States, the following factors conspire to yield the neces-
sary meteorological conditions for extensive perennial snowfields and glaciers;—

I2 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

the narrowness of the svstem, its proximity to the Pacific, the higher latitude, _
the arrangement of the Cordilleran ranges with respect to height, and the position _
of the region with reference to the ordinary paths of the great cyclonic areas.
If these areas of low pressure commonly crossed the continent so that the path
of their centres lay well within the United States, the prevailing winds would be
easterly over this region. Such winds would be colder in the winter and warmer
in the summer than those which now prevail and capable of supplying relatively
little precipitation. There can be but little doubt that any great shifting of
the paths of the cyclonic areas, either to the north or the south, would lead to —
a great reduction in the size of the glaciers of this region, and perhaps to their
complete extinction. ‘-
b. Chinook winds. One of the most interesting and important climatic
factors of this section is connected with the prevailing westerly winds and the _
north to south trend of the main mountain ranges, giving rise to the so-called
‘“‘Chinooks.”” These winds must have been long familiar to the aborigines
and voyageurs, but were first noted by Mackenzie about the year 1790 in the
region of the Athabasca and Saskatchewan rivers, bringing clear, mild weather —
in the winter and spring.‘ He ascribed their warmth, very naturally, to the —
nearness of the warm currents of the Pacific, their progress over the snowfields
being assumed to be too rapid to permit of their being cooled. The same
winds are known in Montana and still farther south in Colorado, where they
are known as “‘zephyrs”’ and “snow eaters.’’ They also occur in South America
with the passage of westerly winds over the Andes, having been described by
Bishop in the vicinitv of San Juan, Argentine Republic, under the name —
‘““zonda.”’? Here they were supposed to derive their warmth from volcanic
sources. The chief characteristics of these, and similar winds are their
warmth and dryness and consequent accompaniment of bright sky. They
are most conspicuous in the winter and spring, but occur also in the summer and
fall. Standing upon the Victoria Glacier, in midsummer, opposite the nose of
Mt. Lefroy, one frequently notes gusts of balmy air sweeping down from the
elevated snowfields and puzzles over the source of the warmth. The following
graphic description of these winds will serve to get before the reader their —
climatic importance.

‘The extreme severity of the winters in certain parts of our northwestern
states, among the Rocky Mountains and along their eastern base, is much
tempered by the prevalence of a mild westerly wind, locally called the chinook.
Its name is derived from that of the tribe of Chinook Indians, living near Puget’s _
Sound. It is said first to have been applied by the early Hudson Bay trappers
and voyageurs, who, meeting the wind while travelling towards the Pacific ~
coast, and finding it particularly strong and warm as they approached the lands —
of this particular tribe, called it the chinook wind. “f

“It is described as a soft, balmy wind, varying in velocity from a gentle

1 Voyages on the River St. Lawrence and through the (Gani of North America to the Frozen and Pacific
Oceans in the Years 1789 and 1793, London, 1801, p. 138
2A Thousand Mile Walk across South America. BoEtoer 1869. N.H. Bishop.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 13

breeze to a steady gale. Though its temperature rarely exceeds 50°F, yet coming
as it does when one is accustomed to a low temperature, it seems warm by con-
trast with the preceding weather. The thermometer often rises from below
the zero point to 40° or 45° in a few hours, and the maximum temperatures of
the winter months in the Rocky Mountain region nearly always are coincident
with the occurrence of a chinook.

“The sky is usually clear while the warm wind blows, though observers
often note a few leaden-colored clouds of a kind seen only during the chinook.
These clouds are described as pancake-shaped, with peculiarly smooth, rounded
edges, and stand apparently motionless, high above the mountain ranges.

“The continuance of a chinook is as uncertain as its coming. It may
last a few hours or for several days. With a change of wind the temperature
falls rapidly, and winter weather once more sets in.

‘“The chinook wind possesses to a remarkable degree the power of melting
snow, for it is not only warm, but appears to be dry. Although a foot or more
of snow may lie on the ground at the beginning of a chinook, it disappears
within a very few hours, often seeming rather to evaporate than to melt. For
this reason the chinook is most welcome to the cattlemen on the plains of
Montana and Wyoming. In fact, without it, stock-raising would be almost
impossible, as the dried grasses of the plains, on which the cattle subsist, would
otherwise be buried the greater part of the winter. To a few, however, this
wind, instead of being hailed with delight as a break in the cold of the winter,
is a source of much discomfort.”’!

The scientific explanation of this type of wind was first given, in part, by
the American meteorologist Espy,? and later completed by Helmholz, Tyndall,
and Hann. Simply stated and adapted to the region under discussion, this
explanation may be of interest to many into whose hands this report may fall.
The presence of an area of low barometric pressure to the north gives rise to a
system of rotary air currents, moving counter-clockwise about the center and
constituting a great ‘whirl.’”’ The interposition of mountain barriers, such as
those already described, with a north to south trend, compels the winds, which are
westerly in the southern portion of the cyclonic area, to mount these obstructions,
from the crests and through the passes of which the air is again drawn to lower
levels through the suction of the great rotating mass. As the air rises upon the
windward slope of the mountain range, it experiences less pressure from the
surrounding air, is permitted to expand, and is, in consequence, cooled at the
rate of 1° F. for each 180 feet of ascent. This cooling reduces the ability of the
‘air to hold moisture and when the dew-point is exceeded precipitation must
result. These facts account for the relatively large precipitation upon the
western slopes of the Coast, Gold, and Selkirk ranges. When vapor is condensed
there is liberated the so-called latent heat, which disappeared when the water
was originally evaporated. It so results that while the air is being cooled by its
own expansion, the consequent condensation of its moisture is supplying it
with heat, and the actual cooling experienced after condensation begins may be
1° for every 300 to 500 feet of ascent. The amount of heat thus supplied to the

1H. M. Ballou: ‘The Chinook Wind,” American Meteorological Journal, IX, 1892-93, pp. 541-547.
2 Fourth Meteorological Report, Washington, 1857, p. 147.

14 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

ascending currents will obviously depend upon the amount of moisture con-
densed and the rate at which this condensation takes place. The air will reach
the mountain crest at a higher temperature than if no moisture had been
condensed; it is capable of retaining more of its moisture in consequence, to be
carried to the leeward of the mountain range. In being drawn down the lee-
ward slope of the mountain barrier the air is compressed, as it descends, owing
to the greater weight of superincumbent air, and is still further warmed at the
rate of 1° F. for every 180 feet of descent. As it is warmed its capacity for hold-
ing moisture is thereby increased and it becomes, relatively, more and more
dry, although it may actually possess considerable moisture. Such a warm,
thirsty wind is our chinook.

The temperature of the air about the peaks and mountain passes differs
less in the winter and spring from that of the valleys and lower plateaus, so that
the chinook is a more conspicuous feature during these seasons. It is also
during these seasons that the cyclonic disturbances are the most pronounced.
In the summer and fall the temperature about the peaks and passes is generally
sufficiently below that of the lower levels so that, although the heating effect,
due to the condensation of vapor and the compression of the air in its descent,
may be actually as great, it becomes much less perceptible. During the most
favorable season for the chinooks we may, theoretically, account for a sudden
rise in temperature of 20° to 25° F., but occasionally it is much greater than this,
sometimes amounting to 50°. Mr. E. B. Garriott mentions a rise of 43° in 15
minutes occurring at Ft. Assiniboine, Montana, Jan. 19, 1892.! In order to
account for such a phenomenon we must postulate a correspondingly high tem-
perature about the crests of the mountains, brought about by excessive and
rapid condensation upon the windward slope, or by some other agency. As
is well known the temperature about the crests of mountains is often considerably
higher than that in the adjoining valley, giving rise to what are known as “‘in-
versions of temperature.’’ Since the establishment of the meteorological station
upon Sulphur Mountain, October, 1903, with an elevation of 7,459 feet, there
have been some 300 such inversions noted up to the close of June, 1906. The
following list of some of the most pronounced cases, with dates, is taken from
data kindly supplied by Mr. N. B. Sanson, of the Banff station. The upper
station is 2,917 feet above the lower and 1? miles to the south-southwest. Of
the 206 instances sent, 163, or 79 per cent. of them, were noted in the morning;
26, or 13 per cent., in the evening, and 17, or 8 per cent.,at noon. Of this number
IoI, Or 49 per cent., occurred during the winter months, giving the most pro-
nounced cases of temperature inversion. The spring and fall were each repre-
sented by 27 cases, or 13 per cent., while the remaining 51, or 25 per cent., were
woes during the summer months.

1 Monthly Weather Review, 1892, p. 23.

ee

4 a full discussion cannot be here introduced, there still remains the question of
its distribution in time. Here again our data are far too scant for satisfactory

generalization, but, so far as the evidence goes, it is in harmony with the theory

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 15

TABLE II.
INVERSIONS OF TEMPERATURE, BETWEEN THE BANFF AND SULPHUR
MounrTaAIN STATIONS.

Data supplied by the Canadian Meteorological Service.
(Elevation Banff Station, 4,542 feet; Sulphur Mt., 7,459 feet.)

Date. Time. Banff Sulphur Difference.
Station. Mt. Station.

Nov. 17, 1903 6.00 A.M. —22.6° F. —14.0° F. 8.6°
Jan. 5, 1904 6.00 A.M. — 7.4° 16.0° 2 aRAO

ee atlas 6.00 P.M. — 6.6° 4.0 10.6°
Jan. 7d 6.00 P.M. 0" 20.0° 21.6°
Mch. By g 6.00 A.M. —19.1° 4.0° 2igs
Aug. aay 6.00 A.M. 38.5° 44.0° mee
Sept. aa 6.00 A.M. 33-8° 40.0° 8.2°
Oct. ast ar 6.00 A.M. 26.6° 35.8° 9.2°
Nov. BAC 6.00 A.M. rad 6.8° 8.2°
Dec. BAG os 6.00 P.M. — 5.6° Banc Sevorre
Jan. —_ 29, 1905 6.00 A.M. TE 43.0° ROSE
Jan. BOsmmias 6.00 A.M. —22.2° 1.8° 24.0°
Feb. Bs Viaeics 6.00 A.M. = Gey? 4.2° T3e7 0
Feb. Sa 6.00 A.M. 10.20 10.5° 20.7°
Feb. ores Se 6.00 A.M. —14.0° 36.2° 50.2°
Aug. 23 6.00 A.M. 30.4° 36.0° Sea
Sept. Zs 6.00 A.M. 33.8° 46.0° 12.2°
Nov. TOs 6.00 A.M. 26.6° 32.2° Gs0e
Dec. Ter as 6.00 A.M. — 9.5° 2-30 r1.8°
Dec. yin (CS 6.00 A.M. — 6.2° 4.0° 10.2°
Apr. 22, 1906 6.00 A.M. 242° 35-8° TE.6°
tne) a4. <6 ; 6.00 A.M. 35.2° 52.0° 16.8°

Geologically the chinooks must exert an influence, the accumulated effect of

which may be considerable. They remove much snow from the eastern slopes

of the mountains, which might otherwise be available for glacier formation,
raising the lower limit of the snow-line and rendering it more irregular. That
portion of the melted snow which is not evaporated will course down the moun-
tain slopes as water, accomplishing a different work than if it had remained
solid. The alternate periods of thawing and freezing during the winter and

spring will accelerate the disintegration of rock upon the eastern slopes of the

-tanges. By this means glaciers lying to the east of, the mountain crests will
receive a heavier load of rock débris from neighboring cliffs, talus deposits will
_be larger than otherwise, and soil formation go on at a more rapid rate. At
Innsbruck, in the Alps, where the foehn winds blow upon an average 42.6
days in the year, the mean annual temperature is raised 1.08° F., or the equiva-
lent effect of one degree of latitude.
-¢. Oscillations in climate. The problem of precipitation has been discussed
thus far with reference to its geographic distribution over the region, and although

_ that precipitation occurs in cycles, there being a series of years in which the total

16 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

amount is in excess of the normal, although it may fall below for some particular
year of the series. This damp phase of the cycle is followed by a series of
years during which the total amount of precipitated moisture is less than the

normal for this number of years, although it may be in excess for some

particular year. Based upon the meteorological data from 321 stations
distributed over Europe, Asia, Africa, Australia, North and South America,
Brickner discovered the length of the cvcle to be 35.5 vears,! the
dry phase averaging slightly longer than the damp one. At neighboring
stations the cycles may partially overlap and, in the case of coast regions, the
phases may be completely reversed, as recognized by Brtickner. Coincident
with the damp phase there is a general reduction in temperature, an increase
in the level of lakes and rivers, a rise in the ground-water level, a halting or
advance of the front of glaciers. The following table shows at a glance such
data as are available along these lines, compiled from publications of Heim,
Richter, Brickner, and Hess. The glacial data refer mainly to the Alpine
region. Constructed for any particular region the figures would necessarily
differ more or less from those given. So far as the United States is concerned
the dry phase through which we have just passed appears to have closed and
we are entering upon another series of damp years. In the Canadian region
under study the damp phase seems to have started about three or four years
earlier than in the Great Lake region and the preceding dry phase about as long

before.
ANB, WNL

PreRrIopIc OscILLATIONS OF CLIMATE, WITH THEIR EFFECTS UPON LAKE LEVELS
AND ALPINE GLACIERS.

ll Tl A i Nl ee le

le
PRECIPITATION. TEMPERATURE. LaKkE LEVELs. GLACIERS.
Damp. Dry. | Cool. Warm. High. | Low. Advancing. | Retreating.
1591-1600 ceeeteate 1595-1610
1601-1610
1611-1635 1611-1635 1630
1635-1645
1646-1665 1645-1665 1677-1081
| 1665-1690 |
1691-1715 | 1691-1705 | 1710-1716 |
| 1716-1735 1706-1735 1720
1736-1755 | 1731-1745 | 1740 1735
| 1756-1770 | 1746-1755 1760 1750-1767
Tid eo} 1756-1790 | || 1777-1780 | 1760-1786
1781-1805 | | | 1791-1805 | | 1798-1800 | 1800-1812
1806-1825 || 1806-1820 | 1820 . 1811-1822 j
1826-1840 | 1821-1835 | 1835 * 1822-1844
1841-1855 | 1836-1850 | ote a 1850 1840-1855
| 1856-1870 | 1851-1870 | 1865 1855-1875
1871-1885 | || 1871-1885 1880 1875-1893
1886- 1886— | 1892 1894-
| | :

' Klima-Schwankungen seit 1700,

Edward Brickner, Wien, 1890, pp. 133-193.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 17

ietous factors conspire to prevent the movements of glaciers from
Bc being exactly coincident with the corresponding climatic phases. For the
_ Swiss glaciers Heim found that, ordinarily, an advance began from 3 to 6 years
after the opening of a damp-cool phase and reached its Tae in 4 to Io
years, but in the case of the longest glaciers the maximum position might not
: be reached until the close of the cycle itself. These facts partially explain the
anomalous behavior often noticed in neighboring glaciers. When the Illecille-
waet Glacier was first visited by the Messrs. Vaux, in 1887, it was standing
close up against a small moraine, which it had just formed, or more probably
assisted in forming during this period of halt. Since 1887 this glacier has been
- in constant retreat at an average rate of 33.2 feet per annum. In r1gos the
_ retreat was found to have been reduced to 2 feet and the inference is that the
glacier is preparing to advance. So far as we may judge from a study of this
one glacier, the best known of the series, a damp-cool phase of the precipitation
_ cycle closed in the early 80’s, and was followed by a dry-warm phase, which lasted
_ for 16 to 18 years, ending somewhere near the close of the century. The follow-
ing quotation fron. Dawson furnishes confirmatory evidence of the existence
_ of the preceding wet phase of the cycle, which was itself preceded by a dry phase. !
. “Bvidence of a remarkable character has been found, which tends to show
that a somewhat rapid increase in the total annual précipitation, has taken
place during late years, and deserves to be recorded here. The evidence referred
to is that afforded by the abnormal height of small lakes, without outlets, occur-
ting in regions characterized by moraine hills. These serve as natural gauges,
_ but instead of measuring the actual rainfall give-a result, dependent on this and
the counteracting effect of evaporation. The abnormal character of the rise
_of the water in these lakes is shown by the facts that it has killed a belt of trees,
some of large size, and at least fifty years in age, along parts of the margins of
_ somé of these lakelets. Both the Douglas fir and the yellow pine—the latter,
- never naturally growing even in damp ‘Soil, —have been found in numbers thus
_ killed. The condition of the trees shows that they have been killed within a
_ few years, and their size indicates that the waters of the lakes in question have
‘ not been for any considerable time during a period of 50 years or more, at the present
high level. These observations were made in both 1883 and 1884. The lakelets
_ observed to be so affected were numerous and scattered over a belt of country
along the western part of the range | Rockies| for a length of about 140 miles.’

_ Looking to the records of the Canadian Meteorological Service for still
further evidence of the periodicity of the climate of the region, we find that only
three of the stations have records sufficiently continuous and reaching far enough
back to be of help. These stations are Agassiz, Banff, and Calgary and their
averages are more nearly the real, but still unknown, normal. Average annual
temperatures and precipitation for the mountain stations, based upon observa-
tions made between 1880 and 1897, will certainly be found later to be too high
for the temperature and too low for the precipitation, while those averages
based upon observations taken since 1897 will prove to yield too high a pre-
tation and too low a temperature. In table 1v we place side by side the

1 Geological Survey of Canada jor 1885, p. 32 B.

18 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

precipitation data for the above three stations, so far as such data are obtainable.
In the first column for each station is given the total precipitation, in inches,
the snowfall being reduced to rain upon the supposition that 10 inches of snow
are equal to one of rain. In the second column is given, for each year, the actual
deficiency, or excess, when the amount is compared with the normal, or average
for the entire series of vears. In the third column the actual precipitation
appears as a percentage of the normal, while in the fourth there is given the
accumulated excess, or deficiency, to or from the beginning of the year 1897.
So far as we may be permitted to draw conclusions from all available data,
the break between the damp and the dry phases of the precipitation cycle in
this region occurred about 1897 in the Rockies and 1898 in the Selkirks and we
may venture to predict that for another decade the precipitation will be in excess -
of the true normal for the various mountain stations. An examination of the
Agassiz data, from the lower Frazer Valley, shows that while Banff and Calgary

were deficient, this station was accumulating an excess and that since 1897 —

there has been a marked deficiency. The inference is that we have here an
example of one of Brtickner’s exceptional coast regions, in which, although the
precipitation is also in cycles, the crests of the great waves correspond with
the troughs over the general surface of the earth. It is very unfortunate that
the data from Vancouver, Nanaimo, and Victoria are not full enough to show .
whether or not they are included in this region.

TABLE IV.
PRECIPITATION DaTA, BY YEARS, FOR CaLGARyY, BANFF. AND AGASSIZ.

CALGARY. | BANFF. AGASSIZ.
— — | :
ss ae sy
Year. | : se : be i eg
mL Ss B3| § 33 g Bs
| 3 uf aa | oS 3 we &3 | ss 3 wD fi | a9
oes og | ga | 3a] 3 6g ga |g 3 88 $¢ |3u
I eats aig) || Ser 1 een pees 26 gE g5)) eos 25 BE | 85
| #2 | de | G2 gs| a | fe | 88 jes] de | fe | Ee las
ea | As | acd. | iS) es As Bac) |e Stl emmeney as as |g
1885 | 12.91 | 3-737 Nee g : "i
1886 | x1r.32| —s.32| 68 ha 3) 2
1887 15.69| — | oS i 2
| 15-69 0.95] 94 | 8 | = S
1888 | 37-42 | +0.77| 105 g 5 5
ae [eee eee sl ones | s g
1890 75-47 | —Z-E7] 93 |B | 19.54 |) — 0-01.07.) 5 tls 150430 eres oo) eon
1891 | ley | TOn4s, —3.66| 82 fe 3
1892 7-9t| —8.73| 47 |-% | 3 | 67-78| +5.76 3
| & oc 7-7 5-7 109g Et
1893 II .05 | —5-59| 66 z | 10.88 | —9.26|} 54 = 76:95 |; +14.93) 124 §
1894 11.63 | —5.or/, 7o | 8 | | 5 78.01 | +15.99] 126 | &
1895 | 16.00} —o.64| 96] < | < 54-50 —7-52| 88
1896 16.05 —o.59| 96 | | 15.86 —4.28| 79 68.25 +6.23] 110
1897 | 20.58 | +3.94) 124 | ¢ | 23.40 | +3.26| 116 | |; 63-43 +1.41| 102 :
1898 | = | 20.58 +0.44| 102 | 2 5 Oe 3 70 tS
1899 | 26.15 | +9.51| 157 | S | 26.34 +6.20] 131 | 5
1900 | 17.57 | +0.93| 106 g | 23.29 +3.-15) 116 | g 72.00 +9.98) 116 a
IQOL 225 +5.67| 134 & | 19.27 —o.87| 96 | 2 52.98 —9RiOA ro eS
1902 34-57 | +17.93| 208 3 | 30-59 | +10.45| 152 | 2 54.68 7 -oaie oom 3
2908 22.7 alias Care, tae | S | 24.82 +4.68| 123 | 3 57-74 —4.28) ‘93 "| cue
es 10.82 | —5.82| 65 | z 14.80 —5.34| 73 zs 54.67 =7-35| 88 3
AO 29 era S2 i232 Com 8 16.00 —4.14| 79 2 60.59 —1.43| 98 4
| | < < 5
|
Normal | 16.64 | | | 20.14 || 62-02 <

i

UTIONS TO KNOWLEDGE—SHERZER.

2
3

SMITHSONIAN CONTRIE

4000 feet

oO
2°
°
o

H. Sherzer, July, 1904. Field assistant

Surveyed and drawn by W.
ss and Frederick Larmour. Adjacent regions based upon maps of A. O. Wheeler and D. W. Wilcox.

se Valley, Canadian Rockies.

Lake Loui

O:

a Glacier,
orrest R

F

lap of Victori

N

GLACIERS OF THE CANADIAN ROCKTES AND SELKIRKS. 19

CHAPTER II.
VICTORIA GLACIER.

1. GENERAL CHARACTERISTICS.

THE VICTORIA GLACIER originates at Abbot’s Pass, upon the crest of the
Great Continental Divide which separates British Columbia from Alberta,
flows due north for a mile between the precipitous walls of Mts. Victoria and
Lefroy, makes an abrupt turn to the northeast and pursues a straight course
for another two miles before wasting away in the Lake Louise Valley. The
collecting area at the Pass is much restricted and narrows down to 600 to 700
feet, where the rocky cliffs upon either side begin to frown at each other from
beneath the snow. These cliffs become higher and separate gradually, allowing
the glacier to broaden to about one-third of a mile as it approaches the bend in
its course. Rounding the nose of Mt. Lefroy the glacier receives its double
tributary from the southeast, attains its greatest breadth of one-half mile,
and for the last mile narrows regularly to its débris-covered nose. In this
lower third of its length it lies between Mt. Aberdeen (elevation 10,340 feet)
and Castle Crags upon the east and Mt. Whyte (9,776 feet) upon the west. A
general view of the lower two-thirds of the glacier, with the tributary entire,
is shown in plate Iv, figure 1, while plate Iv, figure 2, shows the upper portion
and gives a lengthwise view of the tributary.

A Watkin mountain aneroid was carried to the crest of the Pass, July 22,
1904, and gave an elevation, when corrected, of 9,370 feet above sea-level. The
more accurate methods of the Canadian Topographic Survey gave Wheeler
9,540 feet elevation for this same Pass, from which the descent through the
so-called ‘“‘Death-Trap” is rapid, requiring the cutting of steps in the snow
when it is hard from freezing. Owing to the north-south direction of this part
of the valley and the height of the bounding cliffs, the sun has little direct
power and the glacier is permanently covered with snow which assumes a granu-
lar form, owing to the partial melting of the flakes, and constitutes the névé.
This condition of the snow causes it to resemble granular tapioca and may be
seen in the snows which, in more southern latitudes, linger until late in the
spring. The névé-covered portion of the Victoria reaches an altitude of about

. 7,500 feet, or about 2,000 feet below the Pass, when, as the glacier turns to the

northeast, the ice makes its appearance through the snow. The line of separa-
tion between ice and snow, as seen at the surface, is irregular and uneven,
shifting with the season and from year to year. Winters of scanty snowfall,
followed by bright warm summers, will send the névé line up the glacier; while
winters of heavy snowfall and cool summers will cause this line to move toward
the nose. Plate v, figure 2, shows the upper third of the glacier, leading to the
Pass through the “trap,” and the névé line in the foreground, as it appeared in
July, 1904. Rounding Lefroy, the glacier descends rather abruptly some 4oo
to 500 feet, owing apparently to a sudden change in the inclination of its bed,

20 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

giving rise to a series of transverse crevasses. The lower one and one-half
miles of the Victoria presents a remarkably even surface slope, suggestive of a
correspondingly even bed, so that it may be ascended with entire safety by the |
most inexperienced. The nose reaches an altitude of about 6,000 feet, so that
the average slope is at the rate of about 1,200 feet ‘to the mile. For the lower
half the surface slope is but 650 to 700 feet to the mile, or at the average rate
of. 7° to 8°.
In the névé region the surface of the glacier is concave, owing to the accu-
mulation of snow along the base of the cliffs, this being permitted by t
relatively small amount of heat radiation and reflection. For a short distance
opposite Lefroy the cross sections of the ice stream have a horizontal surface
line, while in the lower portion the surface is flatly convex, owing to the lateral
melting caused by the rock walls and, to a greater or less extent, by the late:
drainage streams. (See the cross-section of the Victoria along the line of
steel plates, page 30.) It does not seem probable that the glacier attat
any considerable thickness, the thickest portion, apparently, lying opposite
tributary where the ice may be 500 to 600 feet in depth. The nose is round
completely veneered by rock débris so as to conceal the ice, and perched up
above the valley floor upon an old moraine which it has partially overridden,
but has not had the strength to push aside (plate v, figure 1). The front here
is steep, the angle being about 38°, and about go feet in height, with a series
gradually rising crests from the medial and right lateral moraines. Back from
this nose some 2,000 feet, there is exposed a steep ice wall, 35° to so°, which
attains a height of 125 feet, and continues for about 800 feet. This now appears S
as the side of the glacier, but the position and form of the older moraines show
that the front has been gradually swinging around into this oblique node ns
the cause of which is apparent from an inspection of the map. The eastern
side of the glacier is much better protected by the right lateral moraine ané by
the much broader and closely placed medial. This portion of the front is
prevented from melting, while the less well protected western half has been
rapidly receding. Between this oblique ice wall and the real nose other smaller

faces are developing and being enlarged with each season’s melting. | ti
ei

a

2. NOURISHMENT.

The main glacier is nourished in four ways, which may be separately recog =
nized, as follows: 2

a. By the moisture directly precipitated into the valley between M Ss.
Victoria and Lefroy. The great bulk of this is in the form of snow, wh
probably amounts to about 25 feet per annum. The lesser amount in the
of hail, rain, dew, fog, and frost would also contribute to the substance of
glacier. . ae

b. The funnel-like form of the valley, with its opening to the north, en
it to catch and retain large quantities of snow drifted southward by the 1
winds, as well as that which collects in the lee of Victoria when a west wi d is

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE IV.

Mitre. Lefroy. Victoria.

Fic, 1.—General view of Victoria Glacier looking southward.
Copyrighted, 1902, by the Detroit Photographic Co.

Mitre. Lefroy. Abbot's Pass. Victoria.

Fic. 2,—Lefroy Tributary. Victoria Glacier.
Copyrighted, 1902, by the Detroit Photographic Co.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZE

PLATE V.

Aberdeen. Lefroy.

Fic, 1.—Débris-covered nose of the Victoria Glacier. July, Igo4.

Lefioy. Abbot's Pass. Victoria,

Fic, 2.—Néve field of Victoria Glacier, looking southward through ‘‘death trap.” July, 1904.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE VI.

Fic, 1.—Path of an avalanche along Victoria névé. Photographed by De Forrest Ross, July,
1904. Note ice levees along the margin of the path.

Fic. 2,—Hanging glacier upon Mt. Victoria as seen from summit of Mt. Aberdeen (elevation 10,340 feet). Photo-
graphed, 1903, by Arthur O, Wheeler.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 21

blowing, or in the lee of Lefroy with an east wind. Thus because of its shape,
position, and depth the upper valley is able to capture much more snow than it
would ordinarily be entitled to.

c. Upon the opposing faces of Victoria (elevation 11,355 feet) and Lefroy
(11,220 feet) large beds of snow accumulate during the fall, winter, and early
spring. During the late spring and early summer much of this snow is pre-
cipitated into the valley as avalanches, with the roar of thunder and the blast
of a tornado. It is the danger from this source that has suggested the name
“Death-Trap,’’ for the narrower portion of the valley, although no fatalities
have yet occurred here. These avalanches may shoot directly across the
valley, or they may turn and move along it lengthwise as shown in plate v1,
figure1. They must bring down numerous rock fragments, which are distributed
completely across this portion of the glacier and incorporated into its névé.

d. A considerable portion of the snow which clings to the eastern shoulder
of Mt. Victoria is compacted into stratified ice, and over an area of about one
square mile there arises a true glacier perched up on the mountainside, with a
slope that appears too steep to give it a foothold. Such a glacier is known
as a “‘cliff glacier,” or “hanging glacier’’ (plate vi, figure 2). It moves down
the slope, probably with considerable velocity, and is avalanched into the
valley upon the back of the Victoria Glacier, filling the air with ice dust. At
the crest of the precipice the ice has an estimated thickness of 200 to 300 feet,
from which great blocks, sometimes as large as a city square of buildings, are
detached and fall vertically 1,200 to 1,500 feet. At certain places where the
falls are more frequent there are built up débris cones of coarse granules upon
the western margin of the glacier. This avalanched ice spreads over the glacier
and is incorporated into its body along with more or less ground-morainic material
which has been manufactured between the hanging glacier and its bed. When
the weather is cool and’cloudy these ice avalanches are infrequent, but upon
a warm bright day, with much melting and more rapid forward movement of
the ice, they occur every few minutes from some portion of the long front. On
August 25, 1903, during the mid-portion of the day avalanches were noted as
follows:

10:39 A.M. moderate. 12:29 P.M. moderate.
10:43 slight. 12:49 moderate.
II:05 slight. H2eG2 slight.
rs27 slight. rsd moderate.
11:28 heavy. 1:18 moderate to heavy.
11:40 slight. 1:35 moderate.
12:20 P.M. _ slight.
227 moderate. E52 moderate.
2:10 moderate.

Owing to its western and southern exposure the opposite face of Lefroy
does not support a hanging glacier, although there is a suitable collecting area

22 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

for the snow. That which is not precipitated into the valley as avalanches
melts away in the course of the summer and this water, along with that from
Mt. Victoria, forms slender cascades, which are partially absorbed by the
névé and, in part, work their way to the bed of the valley and are incorporated — 4
into the subglacial drainage. In these various ways the upper Victoria receives
the precipitation from about two square miles of collecting area. Through
pressure, rain, and surface melting, due to the intense solar action, as well as
the chinook winds, this mass of granular snow is compacted into a very fine
granular ice. Powerful winds sweep over the bare peaks and ridges and spread
over the névé more or less fine matter, organic as well as inorganic. This
material is concentrated at the surface, when sufficient melting has taken place,
and gives a rather sharp line of demarcation between the older snow and that _
which falls freshly upon it. In consequence of the melting and the presence
of the foreign matter, the névé acquires a characteristic stratification, which in
the case of the Victoria persists to its nose. Owing to the avalanches of snow
and ice from Mts. Victoria and Lefroy these strata are rendered more or less
irregular. The great weight of this snow and underlying ice forces the entire
mass valleyward and thus prevents indefinite accumulation.

3. DovusiLe TRIBUTARY.

a. Mitre Glacier; the host. Uponeither side of the small peak known as the
Mitre (elevation 9,470 feet), there descend two steep snow slopes, which give rise
to two névé-covered streams of ice (see plate vir, figure 2). That to the right is
intersected by a great crevasse, caused by the glacier drawing away from the
snow and ice which adhere to the rocky slope, and forming what is known as —
the ‘‘bergschrund.’’ This schrund renders this stream impracticable, but the —
other may be safely ascended with a guide, to the Mitre Pass leading over —
into Paradise Valley. Here from an elevation of 8,480 feet the descent is
very rapid for about 1,200 feet, when the two streams unite into a single glacier,
for which the name Mitre, first proposed by Allen for the entire tributary, may
best be retained. For a very short distance the glacier is permanently covered 4
with névé, but in midsummer this soon disappears and discloses a very weak and
poorly defined medial moraine (plate vu, figure 2). It flows lazily down
the straight valley, one and one-third miles, between Mts. Lefroy and Aberdeen,
attains an average width of one-third to one-half a mile, and joins the Victoria —
with a breadth of 3,200 feet. Z

b. Lefroy Glacier; the parasite. Upon the eastern and northern slope of ~
Mt. Lefroy, because of its exposure and other favorable conditions, there has
arisen another hanging glacier similar to although smaller than that just de-_
scribed upon Mt. Victoria. Plate 1v and plate vit, figure 1, give views of this
elevated glacier, clinging to the steep mountain slope, the latter view being
taken from the summit of Mt. Aberdeen, looking westward and from an elevation —
of 10,340 feet. From the steep, vertical ice face great blocks are avalanched
2,000 feet into the valley, much of the ice being ground into dust and shot az

’

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE VIL.

Fic. 1.—Hanging glacier upon Mt, Lefroy as seen from summit of Mt. Aberdeen (elevation 10,340 feet). Photo-
graphed, 1903, by Arthur O. Wheeler.

Aberdeen, Mitre Pass. Mitre. Little Mitre. Letroy.

Fic, 2.—Double névé field of Mitre Glacier, July, 1904. Faulted névé strata are seen in left tributary and a bergschrund in the right.
In foreground a snow-filled crevasse ; beyond which lies the névé line.

:
z

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 23

out beyond the base of the cliff. Upon the western or Lefroy side of the upper
half of the Mitre Glacier, there is thus heaped up a mass of pulverized ice, which
is compacted by freezing into strata. This mass constitutes a new glacier,
since the structure of the original one must have been destroyed by the plunge,
with the exception of the granules and their fragments. Such a glacier is said
to be “reconstructed,” or “‘regenerated.’’ Furthermore, this new glacier
rests upon the back of the Mitre; is nourished differently; is of a different form;
has its own distinct set of strata, unconformable with those of the Mitre; moves
across the valley instead of lengthwise of it, and is accomplishing a totally
different geological work. This glacier, for which the term Lefroy is appro-
priate, is one of the best known examples of what Forbes termed a “‘ parasitic
glacier,” ! far better, indeed, than the type itself. It is parasitic in the sense
that it is carried by its host, the Mitre, and in that it is nourished entirely by
snow and ice which would be otherwise available for the host.

Just what is the structural relation of the Lefroy to the Mitre, the parasite
to its host, can only be conjectured, since the contact was not observed and the
plane of separation may not be at all distinct. There is, however, a very
evident, deep-seated motion down the valley and an equally evident
superficial motion across from the base of Lefroy to the foot of Aber-
deen. The result of the latter motion is to carry most of the ground-morainic
material, such as clay, sand, bruised and scratched pebbles and boulders, which
has been manufactured beneath the hanging glacier of Lefroy, entirely across
the valley and dump it in a great heap upon the eastern or Aberdeen side
of the Mitre (plate vit, figure 1, and plate xv, figure 1). The front of the
parasitic Lefroy being parallel with the szde of the Mitre, some of the ground-
morainic deposit is arranged in ridges parallel with the side of the latter, in
which form it is being dealt out to the Victoria. Until the above stated relations
were discovered it was a serious puzzle to ascertain how a glacier could get its
ground moraine upon its own back and arrange it in ridges parallel with its side
(see plate xv, figure 1). That the material could not have come from Mt. Aberdeen
was evident from the fact that it does not support a hanging glacier upon its
western face, as shown in plate vu, figure 2, although there is a buried mass of
stagnant ice upon the northern shoulder. Avalanches of snow and the ordinary
processes of weathering have brought down considerable angular material from
Aberdeen which covers most of the ground-morainic deposit from the opposite
side of the valley. While this ground-morainic material is being moved east-
northeast by the Lefroy for a distance of 1,800 to 1,900 feet, it is also being carried
north-northwest for a distance of about 3,800 feet by the underlying Mitre and
the resultant motion is somewhat east of north. This will be made clear from
an inspection of the map, plate 11.

Opposite the large débris cone seen in plate rv, figure 2, upon the western
side of the Lefroy Glacier, there is a marked depression in the surface of the
ice and also across the entire tributary where it joins the Victoria, giving good

1 Travels through the Alps of Savoy, James D. Forbes, Edinburgh, 1845, p. 201. ‘i

24 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

exposures of the outcropping edges of the strata, to be noted later. These
depressions furnish confirmatory evidence that the general movement of the
superficial layers is across and not parallel with the valley. The position of the
left lateral moraine from the débris cone, above noted, to the Victoria, shows,
however, that along the base of the Lefroy cliff the movement is normal and
due to the Mitre, although the strata belong to the Lefroy. In consequence
of this a relatively small amount of morainic material is captured from the
Lefroy and delivered to the medial moraine of the Victoria at the nose of Mt.
Lefroy. This double tributary joins the Victoria at an elevation of about
6,670 feet, having an average surface slope of 1,360 feet to the mile, and is at
once compressed to about 600 feet, as compared with 3,200, or as 54 is to I.
There being no corresponding increase in the height of the ice, the inference
is that the tributary delivers relatively little ice to the Victoria and that its
movement is correspondingly slow.

4. DRAINAGE.

a. Surjace ablation. The drainage supply of the glacier originates from
the rainfall over the glacier and adjacent mountain slopes, from the melting of
the snow and ice in the region of the hanging glaciers, and from the general
melting of the glacier itself and its tributaries. No definite data are available
concerning the rainfall over the glacier and adjacent slopes. Owing to its greater
altitude it would be much less than at Field and Banff and would practically
all fall during June, July, August, and September. After heavy showers the
streams from the mountain slopes are in many cases highly charged with
sediment, those originating from the melting of snow being generally clear.

The temperature of the ice during the summer was found to be either
just at the freezing point, or so near it that any addition of heat was
sufficient to start the process of melting. In the abandoned drainage tunnel,
to which reference will be made later, holes were bored into the ice wall, 140
feet back from the entrance, and a standard mininium thermometer inserted
its full length. Owing to the course of the tunnel the point of observation
was estimated to be 70 feet from the foot of the oblique ice wall and about 17
feet from the actual ice face (plate vii, figure 4). During the week from
July 31 to August 7, 1904, the readings were 31.8° F., 31.6°, 31.8°, 31.9°, 31.7°,
and 32°. The maximum temperature of the air in the tunnel during the week
ranged from 31.4° to 33.0° F. Owing to the warmth of the body and that of
the candle used, it was found impracticable to get the temperature of the air
at the same time that the temperature of the ice was taken.

In the rarified atmosphere at these high altitudes the midsummer sun
strikes with surprising force and the surface ice, so near its melting point, is at
once converted into water without changing its temperature. In the case of
scores of observations made upon the surface streams of the series of glaciers
the temperature was almost uniformly 32° F. In rare cases it was found to be
a small fraction of a degree above. In order to secure some data for an estimate

“TILDB[Y) BLIOJITA uodn WeI14S advurerp dOBJING—'"t ae) Aq

&

‘toot ‘An

"BLIOJOLA 9Yy JO ‘AoayjaT ayy Aq 24 24} SSO1B palates
[ ‘UsapPIAqY “J JO sovVyj UIDISAaA\—'S “OLY puv Sorjay “JA uo Ja1orys St
“SULeIOW punts

JOBJINS WOT; J

uey 94) Yjeoueq painjovynueu

UL S}SISUOD UVapIOGY “IJ JO aseq je UOl}e(NUu

“NOOB OUTeIO AY “Aa1Ov[H aaqtpy AQ pattivo Suraq saepy AosjaT DIVISBIBY—'1 “OY

“Aorja'T “211IN “ugapsaqy

IMA ALVId “NAZAAHS—AVDGATMONY OL SNOILOFINLNOD NVINOSHLINS

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 25

of the rate at which surface ablation was taking place over the lower portion
of the Victoria, accurate elevations were taken upon the series of steel plates
used in determining the forward movement. The results are shown in table
V, page 31, giving the total melting from July 13 to August 4, 1904, for a
period of 22 days of midsummer. The maximum lowering of the ice surface
occurred at plate No. 13, nearly two-thirds of the way across when measured
from the west side, and amounted to 3.8 feet, or a daily average-of 2.076 inches.
This plate was located on a portion of the glacier least protected by rock débris.
Although this lowering of the surface is due, in the main, to the sun’s action and
general effect of the atmosphere, usually above the freezing point in summer,
there are other minor agencies which would tend to give the same result. One of
these is the rain, 1.506 inches of which fell during the period of the observations.
Other agencies are subglacial melting and subglacial erosion, longitudinal
stretching, or a lateral spreading of the ice, all of which would tend to lower
the upper surface. It should be noted, further, that in the 22 days this plate
moved forward 60 inches and with the average surface slope ot 7° to 8°, the
plate would be lowered by this agency alone about one-third of an inch daily.
Making this correction the actual ablation, from sun, atmosphere, and rain,
would amount to about 1.74 inches daily and for the two main months of July
and August would give a total of about 9 feet. Observations upon the lower
Lefroy showed that the ice surrounding certain morainic heaps had been
lowered during the season by about this amount. No glacial tables of this
height can be found, however, because of the undercutting effect of the sun’s
rays and the consequent destruction of their pedestals. The broad medial
depression lying just west of the medial moraine (see plate Iv, figure 2, and cross
section page 30) has been produced by the greater surface melting and this
has been permitted by the thinner covering of rock débris, the ice of this portion
of the glacier coming from the Lefroy side of the upper Victoria. This depres-
sion continues down the glacier for 2,200 feet, where it thins out, apparently
from surface melting. With a forward motion here of about 64.5 feet annually,
it would require 34 years for the ice to move from the line of plates to the oblique
ice face, during which time some 306 feet of ice might be melted away. If
the rate of melting and rate of forward movement remained constant, this
number would represent the approximate thickness of the ice beneath plate 13.
It is very probable that the rate of melting becomes less, owing to the concen-
tration of rock débris at the surface, but it is also very probable that the rate
of forward movement becomes also less as the ice diminishes in thickness.
The work with the spirit level indicated that this plate was originally 393 feet
above the lower margin of the ice in this depression, the difference of 87 feet
representing the rise of the valley floor in this distance. If this latter figure is
_ approximately correct the inclination of the bed is much less than that of the
surface of the ice itself
b. Surface drainage. Over the entire névé area the water from the melting
snow, as well as that from the rainfall, is absorbed into the body of the glacier

26 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS,

and refrozen, binding the granules together into an ice conglomerate. Where
the ice itself is exposed and crevasses absent, the melted ice and rainfall are
concentrated into more or less well defined channels, which persist from season
to season. In the early morning the glacier is impressively quiet, the ice is dry,
and many of the small pools are frozen over with a thin layer of ice. When the
summer sun enters the valley the exposed ice becomes moist, small trickles of
water unite into rills, that grow larger and larger from the union of innumerable
others, and these form still larger brooks which empty their clear, ice-cold waters
into the main drainage channels and, after a day of rapid melting, we have here
roaring torrents. These streams slowly cut their way into the solid ice by
mechanical erosion, assisted by the rock fragments which the water is able to
move along, and by melting. The water being apparently at the melting point
of the ice, 32° F., it is incapable of imparting heat to that over which it flows
and the melting must arise from the conversion of its kinetic energy into heat.
Such heat would be imparted to the ice, rendered latent in the process of lique-
faction and the temperature of the water would not be sensibly raised. The
question arose in the mind of the writer while studying these ice streams
whether water at 32° is capable of dissolving ice at the same temperature, as
it might dissolve rock salt over which it was flowing. So far as he has been able
to learn the question has not been investigated, but if water does have any such
effect upon ice under these conditions, it would help to explain the formation
of these ice channels.

In the upper part of their courses these stream beds are generally free from
débris and quite straight, but as the bed is broadened, boulders, too large for
the stream to handle, slide in from the surface and the stream is compelled to
go around. In this way a system of meanders is formed, as shown in plate
Vill, figure 3, and the ice banks are rendered steep and, here and there, undercut
by the rushing water. In the lower portions of the course the bed may contain
considerable rock débris, but this has simply slid down from the surface and
nowhere suggests an aggrading action of the stream. When the stream channel
is contracted for any reason, the level of the water is raised, its velocity increased
in consequence, and an ice basin cut out upon the down-stream side, filled with
more quiet water. This is suggestive of the manner in which the Lake Louise
rock basin, to be later described, may have originated when the entire valley
was ice-filled.

In portions of the glacier intersected by crevasses it is obvious that surface
streams, of any considerable size, cannot develop. The water escapes by an
englacial or subglacial tunnel, to reappear at or near the nose. When a stream
encounters such a crevasse, from which there is drainage beneath, it forms a
small cascade and begins to cut a channel in the vertical face of the crevasse
wall. If the velocity and volume of the water are sufficient, a corresponding
channel may be produced in the opposite wall. As the lips of the crevasse
are subsequently brought together by movements of the ice body, and the
crevasse is healed, this small vertical channel persists and still furnishes an

PLATE 1X.

Fic. 2.—Marginal lakelet west side Victoria Glacier.

Fic. 1.—First stage in formation of a moulin, Lefroy Glacier.

i
) 3.—Inner end abandoned drainage tunnel, Victoria Glacier, July, 1904,

Se

Fic. 4.—Mouth abandoned drainage tunnel, Victoria Glacier, looking outward. July, 1904.

i
‘a -

GLACIERS OF THE CANADIAN ROCKIES AND, SELKIRKS. 27

escape for the surface stream. ‘This is well shown in plate 1x, figure 1, the small
stream entering the nearly healed crevasse from the right. As the ice moves
valleyward the drainage area of the stream may be enlarged, the amount of
water correspondingly increased, and our glacial well, or ‘“‘moulin,’’ may be
indefinitely enlarged both in diameter and depth. Since the water is introduced
from above and escapes below, they are more like wells, turned wrong-side up.
They may be found in various stages of development, the younger in the region
of the open crevasses, the mature examples in the lower course of the glacier,
where they have been carried by the general forward movement of the ice,
and into which the surface torrent plunges with a sullen roar to unseen
depths.

The main drainage stream of the Victoria Glacier starts near the nose of
Mt. Lefroy, to the west of the medial moraine, and passes somewhat obliquely
downward to a moulin, opposite the oblique ice face (plate Iv, figure 1, and
plate 111). This stream drains that portion of the glacier over which the
surface melting is the greatest. A second drainage stream originates near the
above, but in the depression between the medial moraine and the Lefroy tribu-
tary. This stream collects the surface waters from the tributary and extends
for one-quarter mile down the deep depression between the medial and right
lateral moraines and disappears in a system of crevasses that cut this portion of
the glacier.

Two short, but rather deep, drainage channels occur upon the western side
of the glacier, lying upon the inner side of the left lateral moraine. One of
these has incised the ice to a depth of 18 to 20 feet. Since these streams have
probably occupied their present sites for many years it is rather remarkable that
they have not completely cut through the ice to the rocky bed beneath. From
the fact that they have not done so we are forced to conclude, either that the
rate of cutting is surprisingly slow, or else that the glacier thickens in such a
way that the bottoms of the streams are elevated with reference to the base
of the glacier. It is possible that the longitudinal compression to which the
glacier is subjected in its lower half may be sufficient to secure this result.

c. Marginal drainage. Upon the eastern side of the lower Victoria, along
the base of Mt. Aberdeen and Castle Crags, there is no visible marginal drainage
at present, but water may be heard trickling amongst the rocky débris. Just
at the head of this depression, however, where the tributary joins the main
Victoria, there is evidence of earlier drainage here, in the form of an abandoned
lake bed, with a length of 500 to 600 feet. A small gravel delta was formed
at the head, while the rest of the bed is filled with silt. A still smaller lakelet
existed for a short time at the nose of Mt. Lefroy, in the depression between
this side of the tributary and the Victoria. Upon the western side of the glacier
there occurs a similar marginal lakelet, between the glacier and Mt. Whyte,
fed by a mountain stream and a discharge stream from the glacier itself (plate 1x,
figure 2), which cascades over the lateral moraine from a dozen different places.
The lakelet has an elevation of 6,554 feet, is largely filled with fine silt, and

28 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

contains a number of low islands, supporting some vegetation. The outlet
stream is marginal for about 500 feet and then enters the side of the glacier.
d. General drainage brook. The subglacial and englacial streams are all
united into a single stream which emerges from the base of the ice at an elevation
of 6,127 feet and back about 1,000 feet from the real nose, wpon the west side.
This stream cascades over the coarse blocks of the terminal moraine, receives
a small tributary from Mt. Whyte, and rushes headlong to the lake, one mile
distant, dropping some 450 feet. In comparatively recent time the dis-
charge was through a tunnel which is being rapidly destroyed by melting.
plate vit, figure 4, shows the entrance to this tunnel as it appeared in 1904 and
plate 1x, figure 4, furnishes an interior view looking out and down the valley.
At this time the opening was 12 feet high by 7 feet broad and the tunnel could
be entered for a distance of 160 feet, when it appeared to have been clogged by
ground-morainic material. The opening narrowed rapidly toward the inner
end (plate 1x, figure 3) and the severe melting of 1905 showed that it connected
with an englacial passage Jeading towards the main moulin. This difference
in the size indicates that when the ceiling of the portion of the tunnel shown in
plate rx, figure 4, was being fluted by torrential waters, the bed of the stream
was at a considerably higher level than that there shown and that the stream

worked its* way down from an englacial to a subglacial position. The amount

of water discharged through this tunnel was probably no greater than that
at the present time through the present exit. The floor of the tunnel, at its
entrance in 1904, had an elevation of 6,192 feet, or 65 feet above the present
place of discharge.

During periods of minimum melting, and always in the early morning,
the amount of water discharged from the glacier is relatively small.. Late
in the afternoon and evening of a day of rapid melting it gushes forth with
great power and volume (plate x, figure 1). It was impracticable to
measure the amount of discharge at this exit, but measurements were made
at the delta where the stream enters the lake. So little additional water was
being received at the time from the adjacent mountain slopes that the results
secured represented approximately the flow from the glacier itself. An accurate
cross-section of the stream was secured by taking the level of the bed for each
foot, establishing a gauge, and determining the velocity from surface floats.
A calculation of the flow was made, after a week of minimum melting, by averag-
ing the flow at 9:00 A.M. and at 6:00 p.m. During this period there were 0.671
inch of rainfall near the nose of the glacier. A similar determination was
made after a week of maximum melting, with 0.030 inch of rainfall. These
results gave 73 and 93 cubic feet per second for the average flow. At the
time of the minimum flow the water at the exit from the glacier was found to
possess 0.230 oz. of sediment to the cubic foot and 0.506 oz. during the time of
maximum discharge. This is enough to make the water decidedly turbid.
The total amount of sediment carried out daily during the maximum discharge
period was estimated to be about six tons, anu one-third this amount for the

i A

et

ere ae ee

wt Ral Sine fee

vaca ..

Ta at a et a i a Oe

i
a
‘a

“ULDIVUL adt WOIF Jody xIs-AjuaAas { C061 ‘Ez ysnsny

‘poor ‘Aqnf *Surrvays puv uoivoyiyesys avpndsairt SurMoys ‘d19v]H BUojJIA Jo JuOI—'t OT ‘punoisai0} aypprw ayy ul y Japrnoq eouaiajay ) BIZOJOIA ‘JuUOoIT IOI paytyesyGS—'E *

‘AB ‘A SapavyoD ‘sorg Aq ‘g6gr ‘of [nf

paydrisoioy gy ‘OOL oy ul peppequta {os “Tater BLLOJILA, V Japjnoq dIUDIDJaY—*S “O1A

aut yonut jo Avp v Alaiye JaIOV[H BLIOJIA Ww “OI

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 29

minimum period of flow. During the spring and fall the flow is very greatly
reduced and must be very scant, or nothing, in the winter. Mr. Robert Camp-
bell informs me that he has seen water in the stream, however, beneath the
snow and ice of winter.

e. Water temperatures. The temperature of the water at the main exit
was found to vary from 32.0° F. to 32.4,° at various times of the day. Near the
site of the camp, just outside of the older of the two great block moraines, and
some 2,000 feet from the exit, a series of observations was made upon the temper-
ature of the water in the west branch of the drainage brook. The observations
were made between July 2 and 27, 1904, and in the early morning, near midday,
and in the evening, but not at any stated hours. Most of them were taken
between 7 and 8 a.M.; 12 and 2 p.m., and 8 and g P.m. Of 56 observations,
those for the morning averaged 35.4°, for midday 41.4°, and for the evening
35.2°. A small amount of drainage was received from Mt. Whyte, which must
have materially affected the temperatures. Upon July 18, 1904, simultaneous
_ observations were made at the glacier, camp, and at the delta, one mile below.
The maximum temperature for the day was 51.7° F. and the minimum 34.8°.
The results were as follows:

Time. Glacier. Camp. Delta.
9:00 A.M. 32.4 35.2 35.8°
6:00 P.M. QonA BArOe 36.0°

By the time the water has moved across Lake Louise to the foot, its tem-
perature has been raised some 6° to 12°, the temperature ranging from 42° to
48°, during the summer months. The temperature of the Bow River at Laggan,
into which Lake Louise is drained, was found to be 54.9° F., Aug. 15, 1904.

5. Forwarp MovEeMENT.

a. Measurements. Long before the attention of scientists was directed
to glaciers as suitable subjects for investigation, the Swiss peasants had dis-
covered that they possessed a forward, down-valley movement and that they
slowly transported boulders and other objects left upon their surface. In 1841
the Bishop of Annency, M. Rendu, published his remarkable work, Théorie
des Glaciers de la Savoie (edited by George Forbes, 1874, translated by
Alfred Wills), in which he shows a surprising insight into the laws of
glacial motion. ‘‘ Between the Mer de Glace and a river,”’ he writes, ‘‘there is
a resemblance so complete that it is impossible to find in the latter a circum-
stance which does not exist in the former” (page 85). Since this was written,
in 1839, glacial investigation has done much to justify, if not to actually verify
the generalization. Streams of ice and streams of water have many character-
istics in common, as well as important differences. All are agreed that the
cause of the movement is, in both cases, the force of gravity, acting upon the
mass itself, or some other mass in contact with it. As to the nature of the ice
movement, those whese opinions should carry the greatest weight are not yet

a
ay¥ 8 =
c co
dees = S
e 2 eno 2 &
ee San = oe 17
Se Z “1 ¢ 3 4 5 6 Toes 9 0 WW e is ite a E
1 he Le Ik Medial Moraine.
3 b Bee Elevakion 6555 Fear.
2
ere) irae ee
a Sc Hovizontal and Vertrcal Scale. a
@ ee =i ao ae
se Scale of Feev. sie
et Assumed Valley FLooe—— ~~

30 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

inagreement. So far as the present investigation is concerned the phenomena
observed can be satisfactorily explained only upon the theory that under certain
circumstances, and within certain limits, ice is capable of behaving as a plastic
body, that is, capable of yielding continuously to stress, without rupture.
In the discussion of this property of ice in the closing chapter of this report
it is pointed out that the ‘‘plasticity”’ of ice, a crystalline substance, must be
thought of as essentially different from that manifested by such amorphous
substances as wax or asphaltum.

That the Victoria Glacier is flowing valleyward is capable of direct demon-
stration. From range lines across the glacier, the Messrs. Vaux marked the
position of a certain large boulder July 26, 1899. This was on the portion of the
glacier opposite the tributary and back some 6,800 feet from the nose. By July
24, 1900, this boulder was found to have moved forward 147 feet, while a second
“near the terminal moraine’’ was found to have moved 115 feet.! So far as

may be inferred from the phenomenon of the ‘‘ dirt bands,” to be later described, 7
the former boulder was not in the locus of maximum surface movement, but —
some 360 feet to the west. In order to gather more definite data concerning —

the movement of the lower Victoria a line of 18 steel plates was set by means of a
transit, the plates being placed as nearly as convenient at an average distance of

100 feet. The line of plates was back 3,600 feet from the nose of the glacier (see
map) and was marked by setting the instrument over an established point upon
a large block on the shoulder of Mt. Whyte and sighting across to the sharp
edge of an easily recognizable cavity in the face of Mt. Aberdeen. The plates were
of the style successfully used by the Vaux brothers upon the Illecillewaet Glacier

and were 6X6 } inches, with a g-inch piece of } inch gas pipe screwed into the _

center. They were given two coats of brilliant red paint, were numbered, and
had the actual reference line marked in white, after the short piece of pipe had
been driven into the ice. An arrow was marked upon each for purposes of
orienting, in case they should become turned. In general, where the melting
was greatest, it was found that the pipe did not sink into the ice and retain
its vertical position, but allowed the plate to drop to one side. The end of

‘ Proceedings of the Academy of Natural Sciences of Philadelphia, Mch., 1901, p. 213. ‘Observations
Made in 1900 on Glaciers in British Columbia.”

an

ee ee ee

tomer a = ae

ee

x
“

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 31

the pipe, however, marked the spot upon the ice over which it had stood, so
that its original position could be readily restored (see cross-section). It so
happened that the next ten days in the valley were cool and cloudy, with small
amounts of rain almost daily. This period was followed by ten days, and more, of
bright warm weather, with considerable surface melting. After each of these periods
the instrument was set in its former position and the original line again established,
but upon ice which had moved down into this position. | Measurements were made
from this new line to the various plates and the amount of forward movement thus
determined. The full data gathered from observations upon this set of plates
are shown in table v. The relative vertical position of the plates is shown in
the cross-section, as well as their relation to the surface features of the glacier.

TABLE V.

_ OBSERVATIONS UPON THE SERIES OF STEEL PLATES, SET ACROSS THE VICTORIA
, GLACIER, JULY 9, 1904.

(Total Distance across Glacier along line of plates 2,167 feet.)

|

a = a Wes a) oe rom ai
o © - 3 aio Om i OF 3 !
ee | |e |e (2 | *2.)e3 | 82] ee] 2 | 2
D = s - 3 an a
Poca ee Jee olek | ts)e" | S| es | 8° ie
G 43) HO a ce OL m .
Sree eece cen a (seo | Obs 8 laa) ee] el cae R
a oe 20 20 ga | #4 |S Bal oO a Su os or 5} emarks.
Smelassiay | ae SS So | 29 | 85 Bos|/ eG [ssa] ws | 88] ae
“3 ce Sore eo 3a Bee | SIMO) Bia. |S on 2 |) eet SOM
Ss Se aug aos ws es |per SES) o43¢ | Se og 3 Ba
8 2d = oa HS OQ Bo ao |\Oaa Odo| bao | S&o5 Ea) a a 2
nO Ag Rar Bar As | O38 |auH aeal dhs | aos <5 Ow Aa
Feet Feet | Feet Feet. | Inches.,| Inches. | Inches. |Inches.| Feet. |Inches.| Feet. | Feet
ne 63.4| 6585.60] 6585.47 0.13 0.068 | -0.787| 1.181] 0.020 © 21 | 0.006 0.18 62 | Crest of left lateral

ys moraine.
ae 163.2| 6578.20 | 6576.11 | 2.09 1.116 | 0.394| 0.000] 0.197 3-00 | 0.084 2.56 100 | Motion irregular.

3- 262.8| 6577.74 | 6576.02 1.72 0.936 0.000] 0.000} 0.000 ayarsvejia|| le totsteroiaatl|iMereleaei sia 137 No summer motion.
Plate lost in r905.

1.174 2.953] 2.953| 0.295] 15.70] 0.445 | 13.54 160 | Motion for warm pe-
| riod retarded.

Se 443-3| 6582.64 | 6579.89 | 2.75 1.386 | 7.874] 10.630] 0.925 | 26.10 | 0.744 | 22.63 200 | Western slope of low
crest.

1.231 | 10.433] 17.126] 1.378 | 33-80 | 0.959 | 29.17 245 | Western slope of low
crest. 5

Crest of low divide.

ne
n
©

4. | 342.7) 6575.26 | 6573.04

6. 549-9| 6595-27] 6592.70) 2.57

nb
x
un

Te 642.1} 6601.22 | 6599.24 1.98 0.866 | 11.811] 22.835] 1.732 | 47.60 1.350 | 41.06

8. 741.4| 6595.36 | 6593.06 | 2.30 | 1.009 | 12.402/ 27.362] 1.988 | 56.50 | 1.603 | 48.76 292 | Eastern slope of low
crest,

9- 839.8) 6589.06 | 6586.26 2.80 | recoil 18.504| 26.378| 2.244 | 51.50 | 1.460 | 44.41 302 | Eastern slope of low
crest.

1.292 | 18.110| 25.984] 2.205 | 72.10 | 2.045 | 62.20 306 | Eastern slope of low
crest.

Ir, |1037.6| 6579.59 | 6576.65 | 2.94 1,273 | 19.882] 33.661| 2.677 | 76.30 | 2.165 | 65.85 306 | Maximum motion for
year.

£2, |1222.0| 6574.49 | 6570.86 | 3.63 | 1.724 | 18.701] 22.638| 2.067 | 74.20] 2.105 | 64.03 287 Lowest plate.

Io. 938.3| 6585.02 | 6582.15 2.87

IZ. |1320.5| 6580.91 | 6577.11 | 3.80, | 1.735 | 18.701] 36.024| 2.736 | 74.80 | 2.122 64.54 276 Riese suomien abs
ation.

14. |1419.7| 6593.45 | 6590.15 | 3.30 1.491 | 16.929| 33.071] 2.500] 71.00 | 2.014 61.26 262 ean slope to me-
ial.

15. | 1519.1} 6613.69 | 6611.45 2.24 0.959 | 10.827| 31.693] 2.126 | 67.40] 1.912 58.16 252 eoaree slope to me-
ial.
16. |1618.0| 6629.03 | 6627.25 1.78 0.689 | 14.370| 31-299] 2.283 | 63.30; 1.795 54.60/ 236 Cra slope to me-
ial.
Bldr. | 1704.0] 6647.90 | 6646.46 | 1.44 | 0.535 11.024) 29.528| 2.028 | 58.50 | 1.660 50.49 220 | Boulder on medial
| moraine.
I7. ‘1791.6| 6653.14 | 6652.03 pene 0.374 | 11.614| 25.787| 1.870 | 55.50] 1.574 47.88 188 | Crest of medial
: moraine,
18. | 2083.5| 6687.12 | 6687.15 | —0.03 | ..... =—91646)|| 4.1 34)|| osleieese 0.17 | 0.005 0.15 80 | Crest of right lateral
| moraine. .

In column 7 is given the total forward movement for the cool period, July 9 to
_ Ig, from which the average daily movement can be seen at a glance by shifting

32 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

the decimal point one place to the left. The greatest movement was shown by
plate 11, almost 20 inches, and this was the plate nearest the centre. From —
this point the motion diminished very gradually toward the west, but less rapidly —
eastward, toward the crest of the medial moraine, and then fell off very suddenly. —
The plates in the crests of the two lateral moraines had moved up-stream, sup- as
posedly from the settling of the débris due to ice melting beneath. Plate 3
seemed to be located upoa a stagnant portion of ice, showing no movement for
either the cool or the warm period, and then could not be found in 1905, appar- _
ently because of the side cutting of a surface stream. For the warm period, —
July 20 to 29, the results are given in the next column for ready comparison.
The motion is seen to be greater, almost twice as great for plate 14 and almost
three times as great for plate 15. Both plates upon the lateral moraines had _
moved ahead, although that upon the right lateral had not yet regained its —
original position. Plate 2 showed no movement whatever for the warm period, -
and number 4 showed no increase of movement. In column 9g there is given the
average daily motion for these 20 days of midsummer, which may be regarded as _
typical of the season. For this period the greatest forward movement was shown _
by plate 13, having a daily average of 2.736 inches. It is located some 287 feet —
east of the centre of the glacier, and, as previously pointed out, showed the —
maximum ablation. Number 3 showed nomovement whatever. On September
5, 1905, the plates were again located, all being readily found except plate 3, 4
and their distances from the original line determined. These results are listed _
in column 10, from which have been calculated the average daily motion and the
motion for a year. In all cases there was a down-stream movement indicated, ~ ;
although very slight for the two lateral moraines. The greatest movement was ~
shown by plate 11, the one nearest the centre of the glacier, amounting to a total —
of 76.3 feet for the entire period of 423 days, or a daily average of 2.165 inches.
This represents a yearly motion of about 66 feet. The central position of the
plate of maximum movement for the year was to be expected from the very
straight course of this part of the glacier. Its daily motion for the year is about _
81 per cent. of its midsummer motion, which means that for the greater part —
of the year the movement is fairly uniform. The table suggests that during the
season of maximum motion there are cross-currents set up in the ice, and, with —
reference to the body of the ice itself, not the bed, even back currents. During —
the year, however, the impulse is steadily and regularly forward. Studies upon —
the dirt bands of Forbes, to be later discussed, indicate that as we approach the
steeper ice slope opposite Lefroy, the motion is more rapid than that given in the —
table.
b. Frontal changes. The rate of melting about the nose and side of the glacier,
in connection with the rate of forward movement of the ice, determines the
behavior of the front. When these two factors are balanced the glacier appears |
to halt, and, if carrying débris, begins to build a terminal moraine. If either fy
the rate ot forward movement, or the rate_of recession from melting, is

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 33

“fact that the ice of the glacier is continually moving forward. Owing to the rock
_ veneer which completely covers the nose of Victoria, the amount of melting
even during midsummer is very small. The last episode here was one of advance,
Bie glacier having extended itself, some decades ago, into a forest of spruce and
fir and checked its own advance by mounting a heavy moraine of rock fragments
which it was incapable of pushing aside (plate v, figure 1). The cut stumps and
broken trunks which lie about the nose, some of them entirely out of reach of
e present glacier, appear to have been produced by an avalanche from between
_Mt. Aberdeen and Castle Crags, which encircled the nose when it stood some-
what farther back than at present. Trees now growing in the path of this
- avalanche are 28.3 inches in circumference and by calculation should be 130
years old. In order to determine how the nose was behaving, three accurate
measurements were made with a steel tape between definite points upon coarse
blocks of the old moraine and upon others that seemed rather firmly embedded
in the frontal slope. Between July 9 and September 13, 1904, in all 66 days,
each of the latter blocks had settled back approximately an inch, presumably
owing to the wastage of the ice beneath from melting. Confirmatory evidence
that such melting was in progress was furnished by a small clear stream of water at
32°, which escaped through the rocks just west of the nose. Between September
13, 1904, and September 2, 1905, when measurements were again made, this
small recession was partially made up, but the blocks still lacked .36 in. to .72
in. of regaining their former position. With the front so delicately poised it is
evident that a very small additional impulse from behind would inaugurate an
_ advance.
At a point 2,000 feet up from the real nose, at about the middle of the oblique
ice front already noted, there lies a large red quartzite boulder, which was used
_ by the Messrs. Vaux as a reference block. This is the largest of the three blocks
in the middle foreground of plate x, figure 3, as it appeared in August, 1903.
_ This boulder was observed protruding from the ice, a little over half-way up the
_ face, in the midsummer of 1898, by Prof. Charles E. Fay. In this position it was
_ photographed by him, and also a week later, when it had fallen. Plate x, figure
_ 2, shows the boulder in position in the ice. In 1899, July 26, this boulder was
found by the Messrs Vaux to be 20 feet from the ice front. How much of this
_ 20 feet was due to recession and how much to the rolling or bounding of the block
in falling, cannot now be determined. On July 24, 1900, the boulder was found
to be 26 feet from the ice, indicating a recession of 6 feet for the year. August
23, 1903, the block was found by the writer to be 76 feet from the ice foot, giving
an average recession since 1899 of 14 feet. The following July the block was
marked with bright red paint, so that it could be readily located by others: ‘‘A.
Ice foot 74.5 ft. 7/23/’04. Sr.’’ The elevation of a line upon the face, deter-
‘mined by spirit level from Lake Louise as a base, was recorded as 6,264 feet!
a above sea-level. When compared with the distance noted above for the previous

1 This elevation was based upon 5,675 feet for the height of I Lake Louise above sea-level. The corrected
ation as now given by the Canadian Topographic Survey is 5,670 feet, or five feet lower,

34 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

year this would indicate an advance, whereas the glacier was actually in retreat.
The necessity of taking the measurements as nearly as possible at the same cor-
responding time in the season becomes evident. In 1904 measurements were als O
made August 4 and September 13, giving distances of 76 and 86.45 feet respec
ively. If we assume a uniform recession between these last two dates we have
a daily amount of 0.26 feet and for the interval between August 4 and August 23,
at which date the measurement was made in 1903, an additional recession of 4.94
feet. This amount, added to that of August 4, gives 80.94 feet, and the approxi-
mate recession from August 23, 1903, to August 23, 1904, was 5 feet. In 1905 the
measurements were made September 2, and gave a distance of 106.8 feet fron m2
the reference block to the ice foot. This means a recession of about 25 feet for
the year 1904-5. For the series of six years 1899 to 1905 the total amount |
recession observed at this point was 86.8 feet, or an average of 14.5 feet annual r.
The following summary for this boulder may be given:

1898. Fell from ice early in August.
1899. 20 feet from ice foot.
1899-1900. Ice receded 6 feet.
Igoo—1g03. Average recession of 19 feet.
1903-1904. Ice receded 5 feet.
1904-1905. Ice receded 25 feet.
1899 to1go5. Average recession 14.5 feet.

About 375 feet nearer the nose a second block was selected for reference and
upon July 23, 1904, marked ‘‘B. Toice38.5ft. 7/23/’04. Sr.” Between this)
date and August 4, 1904, the recession amounted to 3.9 feet and up to Septembe 7
13 equalled 11.5 feet. The distance from the block to the ice foot was agai n
measured September 2, 1905, and amounted to 63.2 feet, indicating a recession
between September 13, 1904, and the latter date of 24.7 feet. Calculated, as
above, for the year September 2, 1904, to September 2, 1905, the recession w
approximately 15 feet. Since the exposure of the ice face opposite blocks A ar
B is so nearly uniform, we may assume safely that the rate of melting upon the
oblique face is substantially the same at the two points of reference. T!
diminished recession of the ice at B would then indicate that the forward m
ment of the layers here must be greater than at A. From data already ci
it is seen that the forward movement at the nose is insignificant and it app
that the main current of ice, as it approaches the nose, is deflected to the west
ward and that the oblique ice wall is in reality part of the front. |

c. Shearing. The steeply inclined ice front, having a slope of 46°, n
reference block B, shows a succession of ice strata, more or less well defined, wl
dip back into the glacier at a rather steep angle. At the mouth of the abandone
drainage tunnel (plate vit, figure 4, and plate xu, figure 3), these strata in I -JO4
had a dip of 26°, which is below the actualdip. Between these strata there is d
sand, a little fine gravel, and, occasionally, a cobble-stone, but the amount of:
eign matter is small and inconspicuous. A few consecutive days’ visits to

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 35

art of the glacier, in early July, showed that a differential movement between
jacent strata seemed to be taking place (plate x, figure 4), the upper layers being
ushed beyond the lower. There was not enough foreign matter in the layers to
plain the phenomenon by differential es In the case of the South Point
d other Greenland glaciers Prof. T. C. Chamberlin observed a jutting of the
upper stratum which was apparently ae to the more rapid melting of the under
ayer, Owing to its heavier load of dark-colored débris and consequent more rapid
sorption of the sun’s heat. Upon the same glacier, however, he found very
clusive evidence that the upper stratum may be pushed bodily over the lower,
iving rise to a shearing action between the adjacent strata.!. In August, 1903,
f. I. C. Russell found a similar phenomenon on one of the small glaciers
ited on the Three Sisters, Oregon. In this case he thought the evidence con-
clusive that the jutting of the upper stratum was due to differential melting?
In order to ascertain whether or not this was a similar case of shearing, a place
was selected 50 to 52 feet above the base of the ice and heavy spikes driven into
the ice until their heads were flush with the surface. Three were placed in the
base of the upper stratum, about three feet thick, and three corresponding ones
in the upper part of the subjacent layer, which had a thickness of about two feet.
July 21 the upper layer projected eond the lower 19.7 inches at the place selected
or observation. Two days later it was evident that the melting was greater
on the upper layer, in spite of which it now projected 24.4 inches beyond the
lower. The spikes were now visited regularly for 1o days, July 25 to August 3,
th e amount of melting measured, as well as the amount of projection of the two
layers, and the spikes reset. These measurements were necessarily rough, but
they showed each day that the melting was greater upon the upper stratum,
the average amount for each spike being 1.76 inches, while that for the lower
tratum was 1.53 inches, or nearly } inch less. Some sand and fine gravel,
shed down from above, daily accumulated in the lee of the projecting upper
er and gave the appearance of a concentration of dirt in the upper part of the
er stratum. When this dirt was small in amount it was observed that melting
was accelerated; when greater in amount, that the melting was retarded. The
upper stratum continued to gain slowly, but irregularly; reached a maximum
f 26.6 inches and closed at 25.6 inches, or about 6 inches more than at the begin-
g of the observations. The results are tabulated below for inspection. Time
did not permit the verification of the results at other points where the same
thing appeared to be taking place, but there seemed to be no question that the
Jer layer was moving bodily over the lower. This movement represents a
learing of the body of the glacier, the shearing-plane lying between the adjacent
rata, but not a shearing of the ice itself. Knowing how readily iron absorbs
sat it may be supposed that six-inch spikes might induce melting sufficiently
ender their use unsatisfactory. Lying in the ice horizontally there was con-
rable melting about the outer half of the spike, allowi ing it to sag and slide

J rournat of Geology, vol. 111, eo p. Gs
: 2 Glacier Cornices,”’ peat of Geology, vol. XI, 1903, p. 783.

36 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

forward, but they did not seem to penetrate the ice any by such action. But
even if such an effect was produced to an appreciable extent, it would hay
been more pronounced upon the upper row of spikes, which received more sun
heat owing to their more exposed position, and the actual melting upon the up
layer would have been still greater than is indicated in the table. Altho

TABLE VI.

SHEARING OBSERVATIONS, VICTORIA ICE FRONT.

Date. Upper stratum. Lower stratum. Projechionwos Temperature.
= — = upper layer
p beyond lower. -
July 23 Melting about spikes. Melting about spikes. | Max. Min.
| —————— — al — — — — | i
July 25 2.75 in. 2.00 in. 2.50 in. T.50 in, 2.25 in. 2.25 in. 25.2 in. | 60.7°F. 35.0° F) Cloudy. ne
Sores ae a hare a re a > > Sa
“26 | 2.00 2.06 2.25 1.25 2.25 2.25 25.0 | 55.2° 40.0° Warm. =
= = — | : = == Panel
e 27 2.37 2.56 2.37 1.25 3.37 2.25 Z 26.6 68.4° 38.8° Warm and bright.
“ 28 3.25 2.56 2.50 2.75 2.25 2.38 25.6 69.7° — 45.4° Very warm.
ss 29 1.62 1.56 1.38 1.19 I.50- 1.87 24.6 | 74.0° 45.5° Cloudy.
“30 | 0.87 0.69 1.25 Wfofoun aiolas 1.06 25.0 | 74.0° 36.29 | Cool and cloudy. _
a 31 1.75 T.19 I.00 \Eortos 1.06 1.44 ‘26.4 cna 50.4° 38.0° Mostly cloudy. —
August 1 | 1.50 1.63 2.00 | x.63 1.38 I.50 26.2 ~ Bright.
+ 2 2.31 2.00 2.38 I.50 1.50 1.75 25.0 Bright. -
se 3 2.00 2.12 1.50 1.38 1.62 1.56 25.6 Bright and warm,
1o days’ obs.| Average melting 1.76 in. Average melting 1.53 in. | Increase 5.9 in. i

= Sea fe | since July 21. |

than the actual, it is interesting to compare the maximum average daily effect
here observed with the maximum melting observed upon the surface of the
glacier, where least protected by débris. See table v, column 6, plate 13, page 31.
d. Crevasses. The general forward movement of the ice, and its inability
adjust itself to inequalities in its bed, give rise to systems of cracks, or crevass¢
These show that the limit of tensional strain, without rupture, has been exceed
in this part of the ice. They occur in all parts of the glacier from the bergschrund
to the very nose, and when insecurely covered with snow, they form the greates
menace to glacial exploration. The inexperienced cannot be too strong
cautioned against the danger arising from these concealed traps, against which
judgment of best trained Swiss guides is sometimes pitted in vain. In pas
from one portion of its bed to a sufficiently steeper slope, as that opposite the n
of Mt. Lefroy, v-shaped cracks in the ice occur, extending directly across th
glacier. They penetrate to considerable depths into the ice, as measured i
feet, but their depth, compared with the total thickness of the ice, is proba
small, unless the change in the inclination of the bed is very abrupt, when th
may reach the bed of the glacier. When these transverse crevasses have an ea:
west trend, the sun’s rays strike the northern lip of the crevasse more stron
than the southern and, in the*course of the season, it becomes more round
In passing down the slope the crevasse walls come together and the crevass
healed, except for the slight depression caused by the greater melting upo

M GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 37
northern crevasse wall. It is into this depression, which becomes convex down-
‘stream owing to the more rapid central movement, that fine débris may collect
and give rise to the “dirt bands” of Forbes, to be presently described. As
_ the glacier rounds Lefroy and enters a broader portion of its valley, it has a
chance to spread laterally, and longtitudinal and somewhat radiating crevasses
are opened which may intersect those having the transverse position. If the
_ glacier is again contracted these crevasses will also be closed, and if any depression
is left, it will slope down-stream and not have a tendency to collect débris.

_ The more rapid movement of the middle portion of the glacier, when compared
with the sides, which are retarded by the friction of the valley walls, induces
_tensional strains between the central and marginal masses. In consequence,
along the sides, there is opened up a characteristic system of marginal crevasses
at right angles to the resultant strain. These extend inward and upward,
making, theoretically, with the sides angles of about 45°. The difference between
, the central and marginal flow must reach a certain value, and be sufficiently .
abrupt, otherwise the ice seems capable of yielding without rupture. In this
“way we may account for the absence of marginal crevasses over the lower west
side of the Victoria. The very sudden change in movement, shown in table v,
between the margin and the ice of the near-by medial moraine, plates 18 and 17,
is evidently responsible for the series of marginal crevasses that are seen between
the line of plates and the tributary (see plate 1). From their absence upon this
‘side, farther down, we infer that the ice beneath the medial moraine becomes
more sluggish as the main flow is deflected westward. Opposite Mt. Lefroy
‘conditions are favorable for their formation and they are well represented upon
ther side. Opposite the tributary they do not occur, as the marginal ice is
sufficiently yielding. Upon the tributary itself these crevasses are well repre-
sented, except over the collecting area for the Lefroy. After their formation
their i inner ends may be swung around until they assume a transverse, or even
_Teversed, position, as seen Suen the Aberdeen side of the Lefroy. Here we find
one series, pu eine: N. 51° E. and making angles of about 66° with the margin
but ranging from 52° to 86°; and a second series, many of them nearly closed, and
apparently older than the sneaetling having an average direction of N. 95° E. and
making with the side angles of about 111°

E ig The size of many crevasses in the Sane and their contents of fresh snow show
that they may persist through a series of seasons. Sometimes they become
partially filled with water which may melt out cavities in their walls and give
tise to the most exquisite ice grottoes, a peep into which is worth miles of travel.
‘The closing of crevasses sometimes confines pools of water, often under hydro-
static, or ice pressure, and as the surface of the ice is lowered by melting, the water
denly bursts forth with geyser-like action. The compression of air enclosed
avities, or brought in by surface streams, often gives rise to a bubbling at the
ace and a faint hissing, or chirping sound—the “‘sighing”’ of the glacier.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. =

Ww
oo

CHAPTER nie

ee mk

VICTORIA GLACIER (Continued).

1. GLACIAL STRUCTURE.

a. Stratification. In this chapter there is set off for description a number of
features, especially well shown upon the Victoria and its tributary the Lefroy,
but which were more or less well represented upon the other glaciers also and are
characteristic of glaciers in general. Among the first of these is the stratification
the origin of which in the névé has been given on page 22. It is conceivable that a
stratification in the basal layers might arise exceptionally through the operation
of differential stresses in the body of an unstratified glacier. As pointed out by
Chamberlin in the case of the massive Greenland glaciers shearing-planes may thus
arise leading to a concentration of débris. The lower stratum over which the
shearing takes place may be protected from the shearing thrust, may be more
heavily charged with débris, or may be more rigid because of its temperature
and water content.!. In the case of the Canadian glaciers studied it seems
probable that the strata are depositional, in very large part, at least. Conditions
most favorable for the formation of shearing-planes would seem to be found in the
case of the Ilecillewaet Glacier, owing to the body of ice and its rapid descent
from its reservoir. The depositional stratification is almost completely obliterated *
by the ice cascade and none other has arisen to take its place.

The stratification of the Victoria continues throughout the glacier’s extent,
and is seen at the oblique front, in the drainage tunnels and channels, in the
moulins, and upon the walls of the crevasses. The line of demarcation between
adjacent strata is usually only a soiled streak, but sometimes there is sand, gravel,
and an occasional cobble-stone. The strata vary in thickness from 12 inches to
to or 12 feet, as seen upon the Lefroy. . This thickness would indicate that 9 to 110
feet of loose snow had taken part in their formation. The average thickness of
the Victoria strata 1s not. too great to suppose that they may represent the
accumulated and compacted snow fall of the year. Those of unusual thickness
are to beascribed toavalanches. About the mouth of the abandoned drainage tun-
nel in 1904 the stratification of the ice was well displayed (plate viu, figure 4, and
plate xu, figure 3) as previously referred to. Threestrata here averaged 26 inches,
the full thickness of the lower one not being seen. The uppermost layer was
wedge-shaped and thickened from 13 inches to 81 inches. The strata all dipped
back into the body of the glacier at an average angle of 26°, as measured upon the —
tunnel walls, but this was less than the actual angle when measured at right —
angles to the strike of the layers. The irregularities shown in the strata here, —
as well as in the oblique ice face, are probably due to the partial nourishment of ©
the glacier by means of avalanches of snow and ice. Upon the regenerated Lefroy -
Glacier the strata are massive, 6 to 12 feet in thickness, having been produced
entirely from the avalanches from Mt. Lefroy. These strata all dip towards the

Oe ee ew?

1 See Geology, vol. 1, Chamberlin and Salisbury, p. 303. “

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER PLATE XI.

Fic. 1.—Line of contact between two ‘‘dirt zones,” Lefroy Glacier. These zones repre-

sent outcropping edges of depositional strata,

>

Aberdeen. Mitre.

Fic. 2,—‘ Dirt zones” upon Lefroy Glacier, frequently confused with ‘‘dirt bands” of Forbes, Compare figure 2,
plate XV1.

Aer erty

SOE IN

"4

ee RET et)

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 39

region of accumulation, directly beneath the front of the hanging glacier. In
the lower part of the glacier this dip averages 22°, ranging from 12° to 26°,
while farther up-stream the dip is more gentle, only 5° to 10°, as well seen
in the crevasse walls. The Mitre Glacier, near the junction of its two feeding
streams, is crevassed and faulted and displays a very regular stratification
(plate vil, figure 2).

b. Durt zones. Upon a moderately steep slope, such as is found upon the
lower Lefroy, the outcropping edges of the strata, somewhat differently charged
with débris, give rise to broad contrasting zones which pass evenly and symmetri-
cally around the slope. As generally seen these bands are convex in the direction
of flow, but irregularities in the surface slope of the ice, or in the angles at which
the strata come to the surface, may make them concave down-stream for portions,
at least, of their course (plate x1, figure 2). The upper edge of one zone upon the
Lefroy contrasts very strongly with the adjacent layer, as shown in plate rv,
figure 2. It was the abnormal position of this line, first seen from the Devil’s
Thumb, that furnished the clue needed to decipher the relation of the Lefroy to the
Mitre Glacier. A nearer view of this zone line, and two adjacent ones, is shown
in plate vu, figure 1, and a still nearer view in plate x1, figure r. Because of the
irregularity and small size of the strata, as well as the débris covering, the phe-
nomenon is not well seen upon the Victoria. At the place where it should show
the best it is, furthermore, obscured by the dirt bands of Forbes, with which
the zones are often confused. These two features are so different in origin and
significance, yet often so similar in appearance, that they should be sharply
separated in the field and in descriptions of glaciers. Plate tv, figure 2 shows
the dirt zones, upon the Lefroy, at the left, and the dirt bands, upon the Victoria,
in the middle foreground.

c. Granular structure. A lump of ice from the body of a stratum, which has
not yet begun to show any signs of melting, is compact, firm, brittle, without cleav-
age, and beautifully blue by transmitted light. It appears quite homogeneous,
except for the presence of air spaces, which may be sparingly and irregularly
scattered through the ice, or they may be arranged in seams, to be presently
described. Under the polariscope, in thin slices, the ice is seen to be crystalline
in structure and made up of closely pressed polyhedrons, ranging in size from
hazel nuts to goose eggs. These polyhedrons are the so-called glacial granules,
that may be traced back to the névé, growing smaller and smaller, upon an
average, as we recede. from the nose. They fit tightly together, interlocking
perfectly, have curved rather than plane faces, and show no spaces nor signs of
any cementing material between the individual granules. There seemed to be a
correspondence between the size of the glacier and the size of the granules seen
about the nose, the largest granules being observed in the Illecillewaet and Yoho
glaciers, in the case of the latter ranging from o.2 inch to 2.75 inches and averaging
about one inch. From the fact that such granules occur in no other form of ice,
that they may be traced back to the névé, becoming smaller and smaller and
more numerous, the inference is reasonable that, in some way, these granules

40 ; GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

must be derived from those pellets which constitute the typical névé. The ques-
tion as to how the granules are developed at once arises, but cannot be yet an-
swered with certainty. (For a fuller discussion of this subject see page 127). That
the larger are not produced by the simple freezing together of a certain number _
of the smaller pellets is shown by the fact that each mature granule is crystallo- _
graphically homogeneous. Those who have written most recently upon the sub-
ject hold the view that the granules are permitted to grow by a process of partial
melting and refreezing, the larger thus appropriating to themselves the water
derived from the melting of the smaller. Mugge holds that this melting takes
place at the outer limits of the individual granules because of the constant read-
justment of pressures within the body of the glacier,! and in this change of the
granules he sees the cause of glacial motion. Chamberlin believes that a similar
change occurs because of differentialstresses upon the granules undergoing constant
adjustment, assisted by whatever heat energy may be conducted into the glacier
from above.? Drygalski recently argues in favor of a melting of the granule
by pressure both internally and at its outer surfaces, by which some granules
may be completely liquified and subsequently refrozen.2 Upon this action
he bases his theory of glacial motion and the orientation of the granules
about the nose, as brought out in his Greenland report in 1897 cited below.

Experiments of Hagenbach-Bischoff in 1883 showed that when two ice crystals,
having differently directed axes, are pressed together they unite without melting
into a single crystal, ‘‘the larger eating up the smaller.’ The union differs from -
the regelation of Tyndall in that there is a rearrangement of the molecules by
which the resultant crystal is crystallographically and optically homogeneous. To
distinguish it from the ae . granular growth due to melting and refreezing
it is spoken of as a ‘‘dry union.”” This principle applied to the glacier would lead
to a continual reduction in the number of granules and a corresponding increase in
their size, as pointed out by Hagenbach-Bischoff, Heim, and Emden. It will be
shown later (page 128) that this theory of granular growth seems to the writer
to best explain the remarkably perfect preservation of the often delicate lamine
and blue bands seen about the nose and sides of the glacier. Combined with
the special type of plasticity exhibited by ice crystals this method of perfect
dry welding may explain the absence of noticeable distortion of the ice granules,
which, as urged by Chamberlin, should be observed in the direction of flow if the
glacier moves because of its viscosity.

In order to determine whether or not there was any tendency towards the

orientation of the granules in the basal layers about the nose, thin slabs of ice ~

Sie ee et ee ae

~
Ce

eee a ee oe ee Te

1“ Weitere Vomtcee tiber die Tyarislationsfahigkeit des ieee nebst Remenameen aber die Bedeutung
der Structur des grénlandischen Inlandeises,”’ Neues Jahrbuch fir Min., Geol., und Pal., 1900, Bd. 11,
S. 87 zu 98. 4

2 “Recent Glacial Studies in Greenland,” Presidential Address before the Geological Society of America, —
Bull. Geol. Soc., vol. 6, 1895, p. 211; ‘“‘A Contribution to the Theory of Glacial Motion,” Decennial Publica-
tions of the University of Chicago, vol. Ix, 1904, pp. 10 and 11; Geology, by Chamberlin and Salisbury —
vol. I, 1904, pp. 299 to 306.

3 “ Ueber die Structur des grénlandischen Inlandeises und ihre Bedeutung fiir die Theorie der Gletscher-
bewegung,” Neues Jahrbuch fur Min., Geol., und Pal., 1900, Bd. 1., S. 71 zu 86.

ray

*LOLNV[D pajyetouadar v jo afdwexa joajaed

Ssuymo ‘

Vo ‘apsuv ysry ye eieays Jaioelg Aosyjaq Jo javd aaddn ut uaas spueq antq— ts

"BIBS YAM 19}}R] JO Ayrutoyuooun
a}OU puv spueq antq puv sajnuvsd pviov]s aatasqG ‘wo OS oj yno untade}y
1993S ‘hobr ‘Aynf ‘ta1ovpyH vitojzorA ‘fauuny aor Jo [JVM uodn uoyLoyyVAIG—"E ‘ory

‘OZIs ul peonpoat *OZIS ul peoupal YOu
PON “AsTOVpH) JoRMal[Hal[T ‘payenypyur ‘saeypidva ssvwpg—s ‘ow ‘sapnueLis [ees Sururpyno ‘aielsy oyo

. KX ‘sateppides raivl9q—"r ‘oy

‘
- -
"
~
eS % 7
he Z
z
‘
i
‘
s

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 41

__weresawed out in various directions and melted down to thin slices by rubbing them
over the face of a warm saw-blade. Examining these sections with the polariscope,
_ it was found that in the case of those cut horizontally from the glacier, from
_ +$ to 4 of the granules remained dark when revolved. In the case of sections
E cut vertically, either across or lengthwise of the glacier, only an occasional granule
was found to show this phenomenon. From this it appears that there is a ten-
dency towards the orientation of the granules near the lower portions of the
Victoria, Yoho, and Illecillewaet glaciers, a considerable percentage of the
_ granules having their main optic axes in a vertical position. The same phe-
_ nomenon was observed by Drygalski in the case of the Greenland glaciers.!
a d. Capillary structure. When glacial ice is subjected to a moderate melting
__ temperature for a sufficient length of time there is developed a network of capillary
tubes, at the junctions of three or more granules. These tubes are approximately
_ circular in cross-section and from 0.008 inch to 0.04 inch in diameter. Their
_ walls reflect the light strongly and give the appearance of silver threads, more or
less perfectly outlining the granules. From the ease with which liquids course
through the tubes one infers that they are free from or contain but little air.
_ From beneath the margin of the Yoho Glacier it was possible to get some of
them upon the camera-plate, although, many of them being out of focus, they
all appear disconnected (plate x11, figure 1). By making a strong solution of po-
tassium permanganate and placing it in a basin hollowed in the ice, the capillaries
were in a few minutes beautifully infiltrated, the red solution contrasting strongly
with the rich blue ice (plate x11, figure 2). Upon the faces of crevasse walls, and
_ in the drainage tunnels, where the sides are smoothed by melting, these tubes may
¥ be seen in longitudinal section, forming a pattern by which the irregular
granules are outlined. These are the tubes which Agassiz and Forbes found in
_ the Alpine glaciers, but which Huxley and Tyndall did not discover. Agassiz
_ was in error in supposing the entire body of the glacier to be permeated with such
__asystem of capillary tubes and Huxley in denying that any part of it was.
e. Melting features. As melting proceeds the capillaries become larger; ir-
regular, ‘‘crinkly”’ spaces are opened between the faces of adjoining granules, and
the delicate network is gradually obliterated, as shown in portions of plate x1,
_ figure 2. With this increased reflecting surface the ice loses its deep blue color,
_ becomes whiter, and when the granules are small it assumes somewhat the
appearance of névé. A slight pressure now, or a sharp blow, will cause the ice
_ to crumble into its component granules. These granules are shown, but rather
_ indistinctly, in plate xu, figure 3. While still in position, as well as after they
have fallen apart, the granules are seen to be covered completely with delicate
_ parallel ridges and rows of fine points winding over the surface and having no
definite direction with reference to the crystal. The ridges and rows of points
are about 0.04 inch distant, but show some variation, and form a complicated
_ pattern that is different for each granule, suggesting more strongly than any-
_ thing else the ridges seen upon the inside of one’s finger- -tips. This phenomenon

1 Gronland-Expedition der Gesellschaft fur Frdeunde zu Berlin, 1891-93, Bd. 1, 1897, S. 494.

42 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.
was noted by Drygalski in the granules of the Greenland glaciers and described
briefly upon page 488 of his report cited. It had been previously observed
and described by Emden in his paper Uber das Gletscherkorn, p. 22, figure
5, and designated as melting water curves. While neighboring granules were
in position, no correspondence could be made out between the ridges and
furrows of adjoining faces. An attempt was made to take impressions of the
markings but no suitable material was at hand. The wall preparation “‘alabas-
tine’’ reproduced perfectly the finger markings, but refused to work with a wet
ice surface. The “‘stripes of Forel’’ are delicate ridges, passing around the
granules at right angles to the main optic axis, and evidently connected with
the intimate crystalline structure of the crystal. They mark the edges of the
very fine plates of which each ice crystal is composed, placed together with their
flat faces perpendicular to the main optic axis. The ridges here described are
entirely different and do not suggest to the writer any possible explanation.
They are certainly due to the manner of surface melting but it is far from appar-
ent what could give rise to such a pattern. In the prisms of lake ice Emden
found both the melting curves and Forel’s striping present, with an intermediate
type of ribbing, and concluded that all three were due to one and the same cause
and independent of the structure of the crystal (p. 24).

Granules that have been well acted upon by the sun show a system of very ,

flat, circular disks, all with their planes parallel and at right angles to the main
optic axis. These were first observed and figured by Agassiz in his Systeme
Glaciaire, 1847 (plate v1, figures 7 and 10) and described also in his Geological
Sketches, vol. 1, p. 275. They were believed by him to be air bubbles, flattened
by pressure, although observed to lie differently in adjoining granules. These
are now known as ‘‘Tyndall’s melting figures,’ described in his Glaciers of the
Alps, Ed. 1896, pp. 353 to 361. . They represent “‘vacuous space,” left in the
ice by the contraction of the water when changed from its solid to its liquid
condition, the melting planes coinciding with the crystalline plates, of which
the granule appears to be composed. They are thus serviceable in enabling one
to determine the direction of the main axis of each granule, but there were not
enough of them seen at one time about the nose of the glaciers to settle the
question of the orientation of the granules.

j. Blue bands. Many observations were made upon the blue bands, of
which the strata are generally composed, with the hope of shedding some light
upon their position and direction in the ice and their relations to the strata. In
general, they were found well developed about the nose and along the sides of
the glaciers, well up toward the névé region. The lower Victoria has too much
débris covering to enable them to be well seen at the surface, but in the tunnels
and moulins and along the walls of the surface streams they are to be seen in a
good state of development. At the mouth of the tunnel they were found to
average 0.59 inch to 0.75 inch and to dip back into the glacier at an average angle
of 9°, while the average slope of the strata was 26°. This unconformity of the
laminz and strata is well shown in plate xu, figure 3, although the laminz

feet GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 43

become indistinct in proportion as the granules separate. In the moulins,
opposite the oblique ice face the average inclination up-stream was found to be
30°. Under the medial moraine, near the nose of Mt. Lefroy, the bands were
longitudinal, vertical near the centre but radiating, fan-like, upon either side;
the outer ones inclining as much as 45°. Although for over 65 years the subject
of study, we are not much nearer an explanation of this common glacial feature
than when first observed in 1814 by Brewster. The idea of Forbes that they
represent ice-filled crevasses, or shearing-planes, has been generally abandoned.
The early view of Agassiz, that they represent the original lamination of the névé
snow, successively compacted by rain or melting, and then frozen (Geological
Sketches, p. 247), has been revived by Reid! and Hess.2 Crammer accepts
this same view of the origin of these bands, and argues further that they repre-
sent shearing-planes along which the motion of the glacier proceeds. In his prize
essay, Uber das Gletscherkorn, p. 37, Emden advances the theory that these blue
bands were formed by the overflow from glacial brooks, infiltrated and frozen.
The view of Tyndall, that these blue bands result from pressure and, when
formed, are at right angles to it, had received very general acceptance. In
the former view the lamination is to be regarded as an organic part of the glacier;
in the latter, the banding is of secondary origin, and might not be present at all,
under certain circumstances. Tyndall’s theory is set forth clearly in his Glaciers
of the Alps, chapter 31, and is summarized thus: ‘The ice of the glacier must
undoubtedly be liquified to some extent by the tremendous pressure to which it
is here subjected. Surfaces of discontinuity will in all probability be formed,
which facilitate the escape of the imprisoned air. The small quantity of water
produced will be partly imbibed by the adjacent porous ice, and will be refrozen
when relieved from the pressure. This action, associated with that ascribed to
pressure in the last section, appears to me to furnish a complete physical expla-
nation of the laminated structure of glacier-ice.”’

The Lefroy Glacier, being a regenerated and at the same time a parasitic one,
moving in a different direction from its host, furnishes an opportunity for testing
our two theories. In plunging 2,000 feet into the valley all traces of the original
stratification and lamination of the névé must bedestroyed. Since the avalanches
of snow and ice occur only, or mainly, during a few months of the year, it may
be safely granted that layers of this material will be spread out, more or less
unevenly, about the base of the cliff, alternating probably with layers of snow
which falls directly into the valley, or is in part drifted there. The result of this
action will be to restore the stratification seen in the hanging glacier at the crest
of the precipice. It cannot be assumed, however, that anything like the
original lamination of the ice can be reproduced. Possibly around the margin
of the area covered by the avalanches, there might be built up a succession of

1“ The Relation of the Blue Veins of Glaciers to the Stratification,’’ Comptes Rendus IX. Congrés Geol.
Internat. de Vienne, 1903, pp. 703 to 706.

2 Die Gletscher, 1904, p. 175.

3 Eis- und Gletscherstudien. Neues Jahrbuch fir Min., Geol., und Pal., xvi. Beilage-Band, 1904,
pp. 105 and 106.

44 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

fine layers, but such a deposit would be of limited thickness since it would very
soon be pushed outward and beyond the reach of the snow dust. The bulk of
the avalanched ice would come down in great heaps, which could show neither =
original nor acquired lamination. Granting that some of the avalanched snow
and ice would become finely stratified, we would expect it to alternate with much _
more that was not, and with frequent layers, produced by the direct snowfallinto
the valley, showing the typical névé lamination. Furthermore, the position of © 4
the Lefroy upon the Mitre is such that these lamine along the sides of the former, ~
as well as over the surface, should run across the valley and should be entirely
conformable with the strata. 4

Upon the other hand if the banding is of secondary origin and the result of © 4
pressure against the valley walls, it should be entirely similar in adjacent strata
of the same character, exactly as found in ordinary glaciers not formed as is the _
Lefroy, should be found near the sides of the valley and parallel with them,
and should show an utter disregard for the position of the Lefroy strata. In |
ascending the Lefroy to apply our test we find a beautifully perfect and typical —
banding upon the Aberdeen side, well shown upon the crevasse walls, under the |
lateral moraine. The inclination of the blue bands is very steep, ranging from
72° to go® and averaging 83°, as they dip downwards and into the body of the
glacier. These bands are continuous across the gently inclined strata and cut
them at a high angle. Plate xu, figuré 4, shows the perfectly developed bands,
the margin of the glacier lying to the right, but does not give the desired view
of the strata. Toward the centre of the Lefroy these bands become obscure
at the surface, or disappear entirely, but are found again upon the Lefroy side,
between the collecting region and the nose of Mt. Lefroy. So far as this feature
is concerned the Mitre and Lefroy seem to be a unit and the evidence isallin
favor of the pressure theory. i

Near the nose of the Illecillewaet Glacier the blue band structure is very a
perfectly shown about the sides, as seen in plate -x1m, figure 1. There is no a C
lateral pressure upon either side and the bands conform with the valley floor. —
Furthermore, they would be conformable with the strata, providing the latter
were present, but these have been destroyed, presumably at the ice cascade —
farther up the slope. It may be maintained that at such a cascade it is only the
superficial layers that are disrupted and that their fragments are destroyed —
by melting, while the basal layers are preserved intact. This is undoubtedly —
true, at times, but in the case of the Ilecillewaet, the stratification, well seen _
above the cascade, has been destroyed to the very base and it is difficult to —
believe that the much more delicate lamination could possibly have escaped
destruction at the same time. Beneath this same glacier boulders are seen flut-
ing the under surface, as the ice is pressed against them and melted; this is shown
in figures 1 and 2, plate xxxiv. If the banding were simply the original névé —
stratification the edges would be cut off squarely. Upon examining the ice
which has been pressed against a boulder there may be seen a set of bands curv-
ing about the stone, as though they had been there produced.

|ONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER.

PLATE XIII.

Fic, 2.—Contorted blue bands, Yoho Glacier. Supposed to indicate differential ice
ue bands giving rise to ‘‘dirt stripes,” near nose of Tllecillewaet Glacier. Bands flowage.
would here be conformable with strata if latter were present.

3.—Ice dyke filled with two tiers of horizontal ice prisms meeting
towards center,

Fic. 4,—Crevasse in Victoria Glacier, showing how superficial débris may
attain an englacial or subglacial position,
,
, .
}
“

ae

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 45

Wherever observed, the phenomenon of blue bands suggested the structure

‘in rocks known as schistosity, rather than stratification, the bands thinning out
and overlapping. It is possible that they may still be due to pressure and yet
_ the ice may not have become liquid, as Tyndall supposed in order to account
for the scarcity of air bubbles in the blue bands, when compared with the whitish
vesicular ice in which they are embedded. There may be serious doubts as to
_ whether the pressure has been sufficient to produce liquefaction in all such cases
__ where the bands occur, and there is no reason for thinking that the crystalline
condition of the ice would be essentially different in refreezing. We may account

that differential movements have occurred in the ice mass since the bands were
_ formed. A double set might be induced without the complete obliteration of the
first. It is quite possible that, in the case of a glacier of the simplest supposable
type, having a very even bed and without the restraint of rocky walls or lateral
_ moraines, the original lamination of the névé would be preserved to the nose
E and give rise to acertain type of “blue band.’ It would seem that such a type,
_ however, could be distinguished from the more common variety and that it
- would lose in distinctness towards the nose. The writer had come to the con-
clusion that the diverse views held by investigators concerning the origin of
_ these bands were due to the fact that two very similar structures had been
studied under the same name, when his attention was attracted to the following
_ paragraphs written by Agassiz when glacial study was still in its infancy:

“Undoubtedly, in both these instances, we have two kinds of blue bands,
namely: those formed primitively in a horizontal position, indicating seams of
_ stratification, and those which have arisen subsequently in connection with the
movement of the whole mass. . . . With these facts before us, 1t seems to

me plain that the primitive blue bands arise with the stratification of the snow

in the very first formation of the glacier, while the secondary blue bands are

formed subsequently, in consequence of the onward progress of the glacier and

the pressure to which it is subjected. The secondary blue bands intersect the

planes of stratification at every possible angle, and may therefore seem identical
_ with the stratification in some places, while in others they cut it at right angles.”
_ Geological Sketches, vol. 1, pp. 260 and 261.

_ In this report the writer uses the term lamine by which to refer to these
_ “primitive blue bands ”’ arising in the névé, and blue bands for the similar, but
essentially different, structure resulting, apparently, from pressure, or from some
_ other possible agency.

_g. Ice dykes. These were well developed upon the lower Lefroy in the early
_ part of the summer, but became somewhat obscured as the season advanced.
_ They were found sparingly upon the Wenkchemna, but were not observed upon
_ the other glaciers studied. They consisted of gashes in the body of the ice,
_ apparently former crevasses, from two to fifteen inches across, which were filled
_ with columnar ice crystals. The columns varied in diameter from } to 1 inch
and stood at right angles to the crevasse walls, having thus an approximately

46 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

horizontal position. Very commonly the inner ends of the columns met and _
interlocked at the centre, but sometimes they were simply attached by their —
bases to the crevasse walls and left a space at the centre. Plate x11, figure 3,
will give some idea of the appearance of these dykes, although the individual —
ice crystals could not be made to show in a general view. As a rule the columns ‘3
were straight, but sometimes curved and geniculated. The dykes were some-
times many feet in length, occasionally cutting across the walls of crevasses s
presumably younger in age. In certain cases similar columns were found fillin:
elliptical cavities in the ice, the crystals meeting at the centre. Such structur
as these were observed by Agassiz upon the Aar Glacier and described by him
in 1847 under the name of ‘glace d’eau.”’! Although their origin was not under- a
stood he clearly saw that they resulted from the freezing of water in cavities in the —
ice. The ellipsoidal cavities with their radially arranged columns were figured
upon plate vi of his atlas and described under the name “‘étoile de glacier,”
or ‘‘Gletscherstern” (p. 187). These structures probably arise from the rearinaa ;
of water-filled crevasses, moulins, and smaller cavities, the cooling surfaces being .
the walls of the cavity, instead of the atmosphere. When a lake surface freezes _
similar columns of ice are formed, with their main axes at right angles to the -
cooling surface, and, hence, ordinarily vertical. In the case of these dykes the
columns also take a position at right angles to the surface of refrigeration, —
but these surfaces now being vertical the columns assume a horizontal position.? 5.
If the freezing is complete, the columns meet at the centre, the growth of the
columns proceeding at about the same rate, inward from the sides. Should the
water be drained off before the freezing is complete a space will be left at
the centre. They are probably formed in the early part of the season, while the
body of the glacier still retains some of its winter’s temperature and after t he
melting has proceeded far enough to supply the necessary water. After be:
once formed they would persist through many seasons, although their uppet
surfaces might be obscured by various agencies. Somewhat similar dykes wer
sparingly observed upon the western side of the Lefroy but filled with grant le
ice, instead of the ice columns. Obviously these have had an entirely diffe
history. The most plausible explanation is that they represent crevas
which were filled with the granular ice avalanched from the hanging glac
upon Mt. Lefroy.

2. SURFACE FEATURES.

a. Superficial débris. The narrow valley through which flows the upper
third of the Victoria Glacier, permits the avalanches of snow and ice to distribu
rock débris over the entire surface. The most of this material is derived from
the Mt. Victoria side, fr om 1 which the avalanches may shoot completely acr¢

Gan elles Ett ae et Exper riences sur les Glaciers Ae 237 Premiére Partie, p. 185, plate
figures 14, 15, et 16.

2 While this report is going through the press the author has been enabled to study the valuable

of Crammer referred to upon page 43. Under the head of Leisten he describes similar structures (p

and ascribes to them the origin here given.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XIV.

Fic. 1.—Stony till, left lateral moraine, Victoria Glacier. Manufactured beneath

hanging glacier upon Mt. Victoria and carried down with avalanches.

Whyte. Devil's Thumb. Bow Valley. Lake Louise.

Fic, 2.—Sharply crested left lateral moraine, Victoria Glacier. Moraine consists of a core of ice over which is spreac
) >

a relatively thin covering of clay, sand, and roc k fragments.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 47

the valley. This being the region of accumulation, rather than melting, the
rock débris is almost completely enveloped in snow and remains temporarily
covered (plate v, figure 2). As the névé is pushed beyond the snow-line upon
the glacier, surface melting begins and the rock fragments begin to make their
appearance at the surface. As this action continues the rock rubbish is con-
centrated more and more, forming an almost complete veneering over the lower
third of the glacier, completely obscuring the ice except where it has been incised
by the drainage streams. The most of this material is sharp and angular, con-
sisting of irregular fragments of quartzite, sandstone, limestone, dolomite, and
quartz and argillaceous schists; in the main of Cambrian age. The ice lying im-
mediately to the west of the medial moraine has come from the Lefroy side of the
valley and being less well covered with débris has experienced more surface
melting. This depression thus formed, shown in the cross-section along the line of
plates (page 30), determined the position of the main drainage stream, previously
described. The effect of this débris, in general, is to retard surface ablation
and recession about the nose, so that the glacier attains a lower altitude than
would otherwise be possible for it under the present climatic conditions. So
far as we may judge from the ice front, the walls of the tunnels, crevasses, mou-
lins, and drainage streams, the Victoria is not carrying much englacial material.
A portion of this is in the position originally deposited in the névé and a portion
has worked down from the surface by means of the crevasses, as shown in plate
xu, figure 4.

b. Lateral moraines. Along the margins of the upper Victoria and Lefroy
conditions are especially favorable for the reception of rock detritus, both from
the action of the ice avalanches and from the various weathering agencies that
are operating upon the overtowering cliffs. Material derived from the cliff walls
will ordinarily be sharp and angular, but may rarely show a single glaciated
face, produced when in its original position during an earlier stage of glaciation.
Most of this has been pried loose by the water in the seams and joints expanding
in the process of freezing. The material carried by the hanging glaciers is almost,
if not entirely, subglacial and has been subjected to severe abrading action
between the ice and its rocky bed. Boulders, cobbles, and pebbles have had their
corners and edges partially rounded, have had their faces bruised, gouged, and
irregularly scratched, and are embedded in glacial sand and clay, of a bluish gray
color. This ground-morainic material, mixed indiscriminately with that from
the cliffs, is heaped up along the névé margins, embedded in snow and ice.
Moved slowly along, very slowly compared with the central portions of the névé,
the quantity is augmented and by the time the snow-line is reached there is
formed a thick band of this débris covering the margin of the ice. Protected
from the action of sun and rain more effectually than the general surface
of the glacier, in spite of its débris covering, the ice beneath melts less rapidly
and the marginal material is gradually elevated, with reference to the general
surface. About the sides of this marginal ice ridge the débris slides and rolls
down, allowing the less well protected ice above to melt into a sharp crested

48 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

ice ridge, with a veneering of rock rubbish, the whole looking like a great 1
road embankment, as seen in plate xiv, figure 2. The ordinary visitor
scarcely prepared to admit the existence of the ice core, which constitute
reality, the main bulk of the ridge (see plate x1, figure 1, from the Asu
Glacier). In this way are formed the lateral moraines. Should the glacier
pletely disappear from the valley by melting it is obvious that the ]
moraine would be gently set down along the side of the valley, forming a ri
but of insignificant proportions compared with its apparent balls ae
glacier.
Upon the western margin of the Victouty the glacier’s left, Sipoe
trance of the tributary, there occurs a considerable mass of angular débris, co f
uted from the Mt. Victoria side of the valley. Most of it is arranged in three
four somewhat poorly defined ridges, parallel with the margin of the glacie i
sudden contraction occurs here in the breadth of the glacier (see plate m1), a1
there is continued a prominent, sharp-crested ridge for one-quarter mile, marl
the margin of the glacier and losing gradually in height (plate x1v, figure 2). TI
portion of the left lateral consists almost entirely of ground-morainic material
derived from the hanging glacier upon Mt. Victoria (plate xrv, figure 1). Soaked
with water after heavy rains, mud flows occur, upon the surface of which cobbles |
and small boulders are slowly moved down the marginal slopes, thus reducing the
covering of the ice core and permitting further melting. Along the base of M
Whyte there are found two small moranic ridges, consisting mostly of angula
material, from which the ice has withdrawn rather recently. They appear sollte e
the continuation of the two outer ridges which farther up-stream rest upon
ice itself.
The right lateral of the lower Victoria is derived entirely from the right late te
of the double tributary, already described. It consists at first of two high
sharply crested ridges, mainly of ground moraine, which can be traced arout
into the great accumulation dumped at the base of Mt. Aberdeen by the pa:
Lefroy Glacier (plate vu, figure 1; plate xv, figurer). The angular materi
been derived mainly from Mt. Aberdeen, while the ground moraine comes f
the hanging glacier of Lefroy, as previously described. The inner of e
morainic ridges is being destroyed by sliding and mud flows into the dey
between itand the near-by medial moraine. In places it has become
that only with the greatest difficulty can one maintain a foothold upon i cl
About -2,000 feet back from the nose, an outer third ridge makes its ap
(plate xv, figure 2), and together the three pass around and over t
separating into minor ridges and mingling with those of the medial and
moraines (plate tv, figure 1). The lower portion of this moraine has th
ance of composure and comparative stability, giving support to moss
alpine plants, shrubs, and evergreens. One Lyall’s larch was noted 8 fee
and 2 inches in diameter at the base.
Since the upper Victoria receives relatively little material from Lef
right lateral above the tributary is rather meagre, and inconspicuo

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHE

PLATE XV.
Aberdeen. Mitre.

Fic, 1,—Ground-morainic material manufactured beneath hanging glacier upon Mt. Lefroy
and carried across Mitre Glacier by parasitic Lefroy Glacier. Beginning of

ridges seen below in figure 2.

Lefroy. Victoria.

Fic, 2.—Right lateral and medial moraines of Victoria Glacier. Longitudinal ridges in lateral are well shown, the

work of the parasitic Lefroy Glacier,

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 49

previously pointed out, a small amount of ground moraine escapes from being
carried across the valley and moves down in the left lateral of Lefroy. In
addition to this there is a large detrital cone, with its base resting upon the
ice, and slowly dealing out morainic material as the ice moves down the valley
(plate 1v, figure 2). The covering of the general surface of the lower Victoria with
rock débris prevents a great amount of differential melting, so that the lateral and
medial moraines attain no great height.

¢. Medial moraine. Owing to the stream-like nature of the flow, the left
lateral of the Lefroy and the right lateral of the upper Victoria unite at the nose
of Mt. Lefroy into a single medial moraine. This is at first a poorly defined
ridge, but it becomes higher and broader as it moves across the valley from
which emerges the tributary and serves as a divide for the two main drainage
systems (plate rv, figure 1). Owing to the small volume of ice delivered to the
Victoria by the double tributary, the medial moraine lies close to the right lateral,
being separated at first by a deep depression, shown in plate xv, figure 2, which
gradually disappears below as the two moraines merge. The western slope of the
medial becomes long and gradual in the lower part. The entire length of the
moraine is about 7,500feet. Toward the nose it broadens as shown upon the map
and in plate rv, figure 1 and becomes poorly defined, implying a sluggish condition
of the ice upon which it rests. Its crevassed condition in the neighborhood of the
line of plates was described upon page 37. Owing to the source of the material
above noted the moraine contains a certain amount of ground-morainic material,
but the bulk of it is angular and consists of quartzites, sandstone, schists,
dolomite, and limestone. Some of the blocks show alge, tracks, lingulas, and
bryozoan-like stems. It has practically all been derived from Mt. Lefroy.

d. Terminal moraine. Although the front of the ice at the nose is in a con-
dition of halt, the ice is practically stagnant and no frontal moraine has yet been
formed (plate v, figure 1). Along the oblique ice front the retreat has been gradual
enough to distribute the superficial and englacial rock débris somewhat uniformly
over the valley floor and there has thus been formed no prominent ridge, as shown
in plate vir, figure 4. The apparent heaps seen at the right, alongside the face, still
contain a core of ice, which will eventually melt and allow the rock to settle upon
the valley floor. A small ridge, from 100 to 125 feet back from the ice, indicates
a somewhat recent short period of halt, perhaps but one or two decades ago.
It is quite probable that this halt was contemporaneous with that of the Illecille-
waet, which closed in 1887. Between the oblique front and the nose conditions

_ have been favorable for the formation of a somewhat poorly defined terminal
moraine, 7. ¢., the front has been in a condition of halt while the ice was moving
forward and dumping its load of angular débris. Two of the ridges that pass
across the glacier, just back from the nose, extend off the ice upon the terminal
moraine, without interruption, testifying still further to the sluggish condition
of the ice about the nose. The medial moraine has introduced some ground-
morainic material into the mass which has furnished a foothold for vegetation.
Spruce and larch are climbing up the slope, the largest of the former showing

5° GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

70 rings of growth and of the latter 77 rings. It is over this morainic heap that 4
the drainage brook from the glacier cascades. - The swift stream and its load of —
hard angular sediment have a perceptible rounding effect upon the corners, —
edges, and faces of even the hardest quartzites. This effect was most strikingly | 4
shown in quartzite boulders lying in the bed of a glacial stream coming foi
the Asulkan ridge. ‘

e. Dirt bands. Under this term there was described by Forbes, in 184gem
(Travels through the Alps of Savoy, p. 162), a superficial feature of certain —
glaciers which is of much interest and, possibly, of much importance. It is F
found in those glaciers which change their slope sufficiently to give rise to a
distinct system of transverse crevasses, not necessarily to a cascade or ice-fall.
The phenomenon was not understood by Forbes himself and, by various writers —
since, has been confused with the dirt zones, described upon page 39, which are -
the outcropping edges of variously marked strata. It is to the keen observation —
and shrewd interpretation of Tyndall that we are indebted for the true explana- _
tion? of the feature. The Victoria and the Lefroy glaciers furnish an excellent —
opportunity for the study of dirt bands under very simple conditions, as |
well as the dirt zones for comparison. The two types of structure may be gotten f
upon the same photographic plate and are well shown in plate tv, figure 2. In
very simple form the dirt bands may be seen cutting across the dirt zones upon %
the lower Lefroy, owing to the abnormal position of the latter. Under ordinary
conditions the two would be more or less conformable and possibly difficult —
to separate.

The typical dirt bands are of sucha nature that they can be seen most strik-
ingly at a distance of a half-mile or more from the glacier and at a considerable
elevation above it. When once seen, however, it is possible to locate them in a
very general way while upon the surface of the glacier itself. In the summer of
1904, from the summit of the Devil’s Thumb, which overlooks the Victoria Glacier —
from a height of 8,000 feet, there could be counted 23 soiled streaks passing across
the glacier. Beginning near the crest of the ice slope opposite the nose of Mt. —
Lefroy, the bands were narrow, straight, and extended nearly across the glacie
They showed so dimly that there was uncertainty in regard to the count, until
they had been gone over a number of times. Upon the face of the slope they
became more distinct, curved so as to be convex down-stream, and correspondingly
shortened. A few of them could be traced around into the transverse crevasses
which had not been completely closed. Beyond the foot of the ice slope the
bands became still better defined, especially upon the southern, or up-strea

circumferences to hyperbolas. Towards the lower end of the series the ba
Dee much shor ter, the arms extending into and blending with the su

1 Agassiz, acces Sketches, vol. 1, pp. 244 and 254; Russell, Glaciers of North America, p. 43; ]
Die Gletscher, Pp. 169.

2 Glaciers of the Alps, pt. 11, chapter 26.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XVI.

Fic. 1.—Formation of Forbes's ‘‘ dirt bands,” Deville Glacier, Selkirks. From summit of Mt. Fox (10,572 feet),
looking eastward. Photographed, 1902, by Arthur O. Wheeler.

hic, 2.—Forbes’s ‘‘ dirt bands,” Victoria Glacier. Photographed from the Lefroy Glacier, July, 1904. Often confused
with ‘‘ dirt zones,” Compare figure 2, plate XI.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 51

of the ice. Finally the bands mingled with the superficial rock covering of the
_ glacier and weré lost. Standing upon an individual band the dark color seems
to be imparted by the fine dust and sand and not by the coarser débris. In
_ September, 1905, owing to the excessive melting of the summer, the bands stood
. _ out with unusual clearness, so that they were photographed from the side of the
_ Lefroy Glacier, as shown in plate xvr, figure 2. By signaling to an assistant,
_ the well defined up-stream margins of 19 of the bands were located by erecting
small cairns of rock, and their distances apart, in the line of their apices, were
: later measured (see map, plate 111). The results were as follows, beginning near
7 - the foot of the ice slope. The average interval between the bands is 97 feet.

we

Band No. 1--____ Band No. 1o0-____
iis==>159 feet _ => 100 feet
NO; Ges ae No. r1-<==~_
Wee =174 feet “Tt =- 86 feet
NOs Seer NO Nin2<=-—aaee
i -— 126 feet Tt 88 feet
BeuINO Ny grea ara Ee Nosig=—— ie
peer 24 feet = -57 feet
Han NOS Soma ote » UNosi4===-ea
pes --1 13) feet > DE===-81 feet
Now O== <<a No. 15<-=2__
Bie Tog eet __ Lo 66 feet
BEEN O My mcm o i SNosi6== see
_>>=+100 feet peas = 384 feet
No. 8 Nowi7=-=se0
= ——oaiteet ~ > S== 83 feet
NOnO=—=su No. 18===22_
= ~75 feet __ a= 45 feet
NOM i0=-— am me NON TO noi

For reasons to be given later the writer believes that the intervals between
_ these bands mark the annual progress of the ice down the slope, as conjectured
by Tyndall, and offers the following explanation of the phenomenon. As the ice
_ of the glacier is pushed over the crest of the ridge in its bed, which is responsible
3 _ for its change in surface slope, there is formed successively a series of transverse
crevasses, as explained upon page 36 of this report. The distance between
Biese crevasses will be determined mainly by the thickness of the ice and the
* change i in its angle of slope. Since the glacier is moving forward in winter as
_ well as summer, although at a less rate, these crevasses must originate at all
seasons of the year. Those which have been formed in the late fall, or winter,
_ upon passing down the slope will be perfectly healed, since their lips have ex-
_ perienced practically no melting from the sun’s action. The opposite crevasse
_ walls come slowly together, refreeze, and leave no visible scar in the ice. Those
crevasses, however, which have formed in the late spring and summer have their
- lips much rounded by the sun’s rays. If the glacier is moving northward as in
the case of the Victoria and Lefroy, the northern, or down-stream lip of the
-erevasse will receive the maximum effect, the southern comparatively little.
Should the glacier be moving southward, the northern lip of the crevasse would
‘still be the one most strongly acted upon by the sun, but in this case it would be
the up-stream side. Glaciers flowing east or west, and having their transverse
-Crevasses in an approximately north-south position, would have their crevasse

52 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

walls affected more evenly, unless surrounding mountain cliffs interfered. In
the healing of such crevasses there would be left a depression, representing the —

stin’s action upon the lips of the crevasse, not simply for one season but through —
a series, and in this depression the wind-blown dust would collect and the fine
débris would be washed by rain and melting ice from the adjacent portions of the —
glacier, rendering it lighter by. contrast. Owing to the more rapid central —
movement of the ice the bands, at first nearly straight, will begin to curve down- _
stream and become more and more sharply bent, their apices marking the locus is
of maximum surface motion. Between them will lie swellings, or ridges, having —

the same general form of the depressions, from which much of the finer dirt has

been removed. These ridges and intervening depressions may be very inconspicu- _

ous, as upon the Victoria, or they may become very prominent, as shown upon the

Deville Glacier in the Selkirks, forming what Forbes termed ‘“‘ wrinkles” (plate xv1, F
figure 1). They mark that portion of the ice which passed the crest of the —

slope in the late fall and winter and appear as ridges, partly because of the severe
compression to which the ice is subjected and mainly because the adjacent ice
has been lowered by melting. Owing to the more rapid movement of the ice

down the slope the bands will be farther apart and less well defined, than after the

more gentle slope below has been reached and the ice is subjected to longitudinal
compression. Upon this more gentle slope they have a better chance to catch
and retain the fine débris. Since the sun’s action was more powerful at thé
center of the crevasse, the depression is greater at the apex of the band and
persists after that of the extremities has been finally lost by surface melting.
In consequence the bands become shorter and shorter and lastly disappear,
when ablation has reduced the surface to a general slope and the fine débris is
redistributed. Very often it must happen that instead of a single crevasse
being formed during the season of melting there would be formed a series of
them. Upon a steep slope of the Asulkan they seem to be formed in pairs as
shown in plate xvit, figure 1, in which it is seen that the ridge of ice separating
two adjacent crevasses is acted upon from either side and lowered, assisting in

the formation of the depression. The crevasses that are forming the depres- —

sions, preparatory to the reception of the dirt, may be traced around to the
almost healed crevasses at the left, while between them are seen traces of crevasses
that have healed with practically no marginal melting. These are presumably

those which opened and closed soon enough to éscape the rounding action. —

If the surface slope is too great the depression produced in the ice may not be
sufficient to retain enough dust to bring out the series distinctly, as is the case

with the Asulkan just noted. Study figure 1, plate xxrx, from the Yoho Glacier.

That the method of formation of these dirt bands is essentially as outlined
above admits of no doubt. The question as to whether they are produced
annually, or at irregular intervals, needs to be investigated. The average inter-

val of those bands originally described by Forbes upon the Mer-de-Glace was
711 feet. Opposite his station D the interval was 667 feet (Travels through the —
Alps of Savoy, p.165). Ina postscript to his volume, p. 420, he gives the move-—

¢

*
os
et

a des

a

“<4
a

c

4
ig

a

“2409 “SUNpaUL aovyins

B91 MOYS 0} APIs AUO WOIY paaowar Surtaauaa Ap YA ‘€ aimdy Ut UMOYs aud aIUBG—"f ‘OI spivjaa

Suqep Ayoor ary Huenb yuaroyns up satoxps viuoyor, ‘auc Wp jpeuwg—'é ‘ong

“‘Suypaut saye

saqyuenb yews uy “Sunpaut [eMuaraytp Jo yfnsar ay} ‘Ma1orpH viAoyOIA ‘sTTaa Isn(q|—"z ‘Oly

‘fobr ‘ysudny
“NLOV][L) URY[USy Jo uonod daaqys

c ‘spueq WIp,, Ssaqiogq jo uoneuioq~—I ‘ong

«

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 53

1 eater, Bee wording closely with the inter male of tie. ‘dirt bande of the

lac Although the size of the intervals in this series differs plainly to the
“still Forbes states that the difference for any one interval is probably not
feats of the mean. From the same point of view as that used by Forbes in
842, Tyndall counted upon the Mer-de-Glace, 17 years later, exactly the same
yer of bands and remarked: ‘‘The entire series of bands which I observed
the exception of one or two, must have been the successors of those
ed by Professor Forbes; and my finding the same number after an interval

: . Germans. these are due, in part, to the filling of transverse crevasses,
n during the fall, with snow and then its later compression into a white
ar ice. Since, in general, these crevasses would be those which had been
ted upon by the summer sun, they would be the counterpart of the dirt bands
u er discussion. Sévé found the average interval for these white seams upon
36ium Glacier, in Norway, to be 218 feet and that this represented also the
age forward annual movement.' So far as the Victoria Glacier is concerned
have not sufficient data at hand to settle the question of the annual character
the dirt bands. At the line of plates, about a third of a mile below, the
ximum annual movement of the ice was found to be 65.85 feet. The average
ual interval for the lower half of the series is 76.56 feet, which is about what
uld be expected in the way of annual ice movement, when compared with the
ve. We should also expect the movement to increase as we approached
ie crest of the ice slope. So that the actual and relative spacing of the bands
strongly suggests their annual character. If due to some “regularly
urrent cause,” as Tyndall suggests, this cause must recur with the seasons.
We are, however, not entirely without evidence that the intervals between
dirt bands indicate approximately the annual movement of the ice. As
ited out upon page 30, the Messrs. Vaux marked the location of a large
ulder upon this portion of the glacier July 26, 1899. From range lines, one
ear later, they determined that the boulder had moved forward 147 feet. In
ember, 1905, this boulder’was found opposite the 9th band of the series
upon page 5r. In 1899 it should have lain opposite the 3rd band and,
motion there had been the same as it was in 1904-5, it should have moved
1899-1900: ‘the distance of 126 feet. The previous year it should have moved
4 feet. My field notes say that the second and third bands were indistinct,
hat there is strong probability that the three intervals between one and four
lay not have been properly distributed. The average for the three is 153 feet,
ich agrees very well with the actual observed motion of the boulder. If the
. band intervals are an approximate indication of the annual ice movement

1 Quoted from Heim’s Gletsche rkunde, p- I40.

54 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

the Vaux boulder had moved downward in 1905 from its original position some
676 feet, or at an average rate of about 113 feet per annum.

fj. Dirt stripes. Somewhat closely related to the dirt bands just described,
so far as their method of formation is concerned, are the fine streaks of dirt seen
along the margins of most glaciers, sufficiently free from surface débris. They
may be found, however, anywhere upon the glacier that the blue bands are
well developed, reach the surface at a fairly steep angle and are being subjected
to surface melting. The blue bands, being composed of relatively firm, compact
ice, are more resistant of the sun’s action, than the vesicular ice in which
they are embedded and project as delicate ridges, separated by narrow
furrows. Into these furrows the wind-blown dust settles and is washed
from the adjoining ridges, forming narrow, parallel dirt streaks, or stripes.
When well developed, as upon the Lefroy, the glacier has the appearance of
having been swept with a coarse wire broom; the strokes having all been long,
regular and parallel. The dirt stripes mark the position of the vesicular bands
in the ice and the lighter streaks between the position of the blue bands them-
selves. In this way the banding is clearly shown at the surface, whereas, other-
wise, it might be obscure. Views of these stripes have already been shown in plate
xu, figure 4 and plate x1, figures 1, 2. Sometimes they run down the face of a
crevasse wall (plate xm, figure 4), as though they might be something more
than a superficial feature, but a little chipping of the ice shows plainly that they
are not. After they have once been formed the dirt stripes will absorb the
sun’s heat and still further emphasize the small furrows. Running, in general,
lengthwise of the glacier these furrows become the sites of minute rills which
have a tendency to clear away the fine dirt, as fast is it collects. For this
reason, as well as because of the nature of the banding itself, the individual
stripes are not continuous for any considerable distance. They are sometimes
so closely placed that 10 stripes may be counted within the distance of an inch,
but are usually considerably coarser.

g. Dust and pebble wells. Where small pebbles, or patches of fine dirt,
often black from the presence of organic matter,! are thinly distributed over
the surface of the ice, heat is absorbed and the ice immediately beneath is melted
more rapidly than the surrounding ice. Cavities are thus formed with vertical
walls, which for a time retain the water. They sink into the ice for a few inches,
until protected from the direct rays of the sun by their own walls, when further
melting would be delayed until the general surface was lowered sufficiently
to allow the sun to again reach the foreign matter at the bottom. Such wells
are shown in plate xvu, figure 2. Although their depth at any one time is seldom
greater than a finger’s length, still in the course of the season their total length
would be g to 10 feet upon the Victoria and Lefroy. A thin film of water often
freezes at night over the surface and then thaws out promptly when again
exposed to the sun. After thus freezing the water is at times drawn into the

1A sample collected from the Illecillewaet in 1903 contained 14 per cent. of organic matter, enough so
that when set away moist in a warm room it soon became offensive.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 55

glacier, by means of the capillaries developed between the granules, leaving
the well free from water, but with its ice cover. Where pebbles, or small
dirt patches, are abundant, as shown in the last figure, the ice between the
adjoining wells is melted more rapidly by the sun than it would ordinarily be,
forms minute pinnacles and appears whitish and spongy. In this way the lower-
ing of the general surface of the glacier by ablation is accelerated. By keeping
itself thus at the bottom of a small well the dirt of these small patches is pre-
- vented from being blown away, or washed away, and thus it is possible that the
same well may persist through, not only a season, but a succession of seasons.
Should the well, however, collect additional dirt, beyond a certain limit, this
excess would then protect the bottom of the well from further melting, the
adjoining ice would soon be lowered below the bottom of the well and the well
would be literally turned wrong-side-out. Where one has a few days to spare
about the same glacier an interesting experiment would be to sift dirt into a
group of typical wells, filling them to varying depths, and observing the result.
Such an experiment may easily be performed upon a snow bank of sufficient
depth, when it is being strongly acted upon by the spring sun. It would prepare
the way for a clear understanding of the next three surface features to be
described.

h. Débris cones. When the amount of dirt, sand, gravel, or rock débris,
is sufficient to protect the surface of the ice from melting, or to even partially
protect it, over a limited area, the surrounding ice surface will be lowered
more rapidly than that beneath the protecting material and the débris will
begin to be elevated, with reference to the neighboring surface. The loose
débris will slide, or roll down about the side, exposing the edges and corners to
the melting action of the sun, allowing still more sliding of the débris and still
further melting. The ice core will finally assume the form of a ridge, cone, or
mound, with its thin veneering of foreign matter, as in the case of the lateral and
medial moraines already described. The companion figures 3 and 4, plate xv,
show the structure of a small gravel vone, only 15 to 16 inches in height; figure 3,
as it was found upon the ice, figure 4, after the gravel upon one side had been
washed off to show the ice core. It is seen what a thin covering will suffice to
bring about the result. Depending upon the nature of the covering they are
known as dirt, sand and gravel cones, and boulder mounds, and they may vary
in height from a few inches to many feet. In plate xix, figure 1 is shown
a mound upon the Wenkchemna Glacier, estimated to be 80 feet high. This
pile of rock rubbish was either dumped in a heap by an avalanche, or collected
in the bottom of a lakelet, as described by Russell for the Malaspina in Alaska.!
Cones of all types, varying in height, from a few inches to 12 or 15 feet are
to be found upon the Victoria in the region of maximum melting. They may
persist from one season to another, but there is a limit to the height to which
any particular cone may attain. As the height of the cone grows the lateral
surface is increased, over which the débris must be spread in order to suffi-

1 Gletns of North America, p. 115.

56 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

ciently protect the ice. When this covering becomes too thin, or when it is
blown off, or washed down the steep slopes by heavy rains, the ice core becomes
exposed and rapid melting ensues, resulting in the destruction of the feature.
In case the surface covering is distributed about the base and the pure ice core
exposed, the melting does not cease when the general level is reached but con-
tinues more rapidly than the surrounding ice to which has been transferred the
débris. Instead of a cone, we may now get a basin-shaped depression, which is
gradually extended laterally by melting, and into this depression the original
material may again slide and be collected at the centre until there is sufficient
to prevent further melting. An interesting and instructive experiment, in
connection with that suggested upon the dirt wells, is to wash down the gravel,
sand or dirt, from a collection of small cones, mark the location, and watch the
changes from day to day.

1. Glacial tables. In the case of a single rock fragment, of sufficient size,
resting upon the ice over which surface melting occurs, protection is afforded
the ice immediately beneath. As the result of the more rapid melting of the
surrounding ice the rock is relatively elevated upon a pedestal of ice and there
results what is termed a “glacial table’’; as seen in plate xvitl, figure 1. As the
rock is elevated a short and narrow ridge of ice lying to the north of the pedestal
(observe the shadow in the figure) is protected from the noonday sun, so that

viewed from the east or west the pedestal is unsymmetrical. This lack of

symmetry is further emphasized by the undercutting action of the rays of the
noonday sun upon the southern side. Some observations were made with a
view of discovering the lower limit of the rock fragments that were capable
of furnishing the protection necessary to form tables. The following were
found forming low tables, or starting to form them. It is obvious that the
color and nature of the rock would both have their influence in determining
the effect upon the ice.

Dark gray limestone, 12x 12x 4.5 inches.

ce ce ae ae

13% 9X 3
Light =~ S EL Xs Ook Geog
Reddish: quartzite “160 mes ar5
Rusted limestone, Oa Aree re
Dark limestone, 8 x4x 2.5 to 3 inches.

In the case of the last specimen the thicker end was found to be protecting,
while the thinner was inducing melting. Owing to the undercutting action
of the sun’s rays blocks of this size can form only low tables. Larger blocks
may rise to a height of three to five or six feet upon the Victoria, the latter
heights being unusual. They may persist from one season to another but there

is a limit to the height which any particular table may attain, determined mainly.

by the size and shape of the rock. As the rays undercut, mainly upon the
_ southern side, the block begins to lean to the south and finally topples off in
that direction (plate xvii, figure 2). The remnant of the pedestal is removed by

Cl es i

SHERZER. PLATE XVIII.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE

Fic. 1.—Glacial table, Victoria Glacier, looking southwest. Observe undercutting action of
rays upon south side and shadow cast by the rock upon north side, with

resultant ridge of ice.

Kc, 2.—Dethroned glacial table, Victoria Glacier, looking northeast. The boulder fallen

to the south by undercutting action of sun’s rays.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER PLATE XIX.

Fic, 1.—Boulder mound, Wenkchemna Glacier, illustrating the protective effect of rocky debris, Estimated to be
eighty feet in height.

Whyte. Devil’s Thumb. Bow Valley. Fairview.

Fic. z.—Surface lakelet, Victoria Glacier, resulting from lack of debris protection. Enlargement towards right is still

in progress by melting, but has practically ceased towards the left, owing to the rocky cover,

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 57

melting, the block settles into its position of equilibrium and the making of a
glacial table begins anew. In exceptional cases the undercutting of the pedestal
may be done by a surface stream. In the case of 25 tables selected at random,
it was found that the longer of the horizontal axes of the pedestals had an average
magnetic bearing of N. 36° W., or 11° W. of true north. With a larger, or a
different, series, the average would probably be more nearly the true north.

y. Surface lakelets. Upon the middle portion of the Victoria, western side,
where the ice is presumably quite stagnant, there occurs a series of surface
lakelets, the crater-like basins of which have been hollowed in the ice. The
largest of this series ts somewhat elliptical in form, 200 feet long by 100 feet
broad (plate x1x, figure 2), and filled with deep blue water in which miniature
ice-bergs may be seen floating about. The southern and eastern banks of the
lakelet are from 12 to 20 feet high and under cut, apparently by the melting
action of the lake water. The northern and western banks have been acted
upon more strongly by the sun, causing them to recede and the débris to slide
down until the margins of the lake are filled and the ice banks veneered sufficiently
to retard melting (plate x1x, figure 2). These banks are as steep as the débris
can stand and from 25 to 30 feet in height. From the still steeper ice walls
the gravel and small boulders are splashing into the water with a sound suggestive
of considerable depth. The lake has no visible outlet and persists from season
to season. Several similar lakelets, but smaller, occur in the same vicinity,
some having their sides completely veneered with rock débris, which has checked
melting and allowed the lakelet to become almost dry.

These lakelets may have originated in marginal crevasses and been enlarged
and shaped by melting, or they may have originated by surface melting over
certain limited areas less well protected by débris covering. In the preceding
discussion of débris cones it was shown how miniature basins might originate.
In a stagnant portion of the glacier it is possible that the basins of such lakelets
might arise in a similar manner from such a mound as that figured from the
Wenkchemna Glacier (plate x1x, figure 1). The rock débris rolling or sliding to
the base would leave the cone sufficiently bare to permit rapid melting to a depth
at which the marginal débris would begin to slide back again. The accumula-
tion of the débris in the basin would check further melting at the center, while
the surrounding ice has lost, in proportion, a part of its débris. As observed by
Russell upon the Malaspina, the surrounding ice would be lowered until the basin
disappeared and what had been the centre of the basin would become the crest
of a boulder mound. If conditions remained favorable, 7.e., sufficient thickness
of stagnant ice and continuous surface ablation, the sides of the mound would
become more and more steep, as it gained in height and the time would come
when the bulk of the débris would slide or roll to the base, the ice core would
be removed to the general level and a new basin would be started. For the
larger lakelets the complete cycle would probably have to be reckoned in decades
and centuries.

k. Rock reflection (?). <A final surface feature remains to be described, al-

58 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

though a satisfactory explanation has not been found. Observations were begun
upon the Victoria in 1904 before the winter snow had disappeared from the lower
half of the glacier, which is ordinarily bare in the summer. It was noted that as
the surface boulders protruded through the snow a larger percentage of them
showed melted areas upon their northern sides, the shape and size of which sus-
tained a certain relation to the breadth, height, shape, and possibly position of the
boulders themselves. Over the melted area the snow was removed, in whole
or part, to the soiled surface of the glacier, so that the feature shows with
much clearness in the photographs secured. The phenomenon was seen
upon the Lefroy, as well as the Victoria, and sparingly upon the Wenkchemna
in midsummer, near the névé lines. The same thing was also seen upon the
snow of an avalanche which had descended from Mt. Whyte and carried along
some small rock fragments, which were scattered over the surface. The block
shown in plate xxtv, figure 4 is a gray quartzite standing ro inches high and is 29
inches broad. The melted area has the same length as the rock and has a corre-
spondence in outline. The farther right hand corner of the rock is somewhat
lower than the general surface and the corresponding corner of the melted area
is seen to be rounded and incompletely melted. Boulders showing the phenom-
enon were not hard to find upon the Victoria, but were very abundant. The
north-south axes of ten of the areas, selected at random, gave an average
magnetic reading of N. 25.5° W., with less range than was shown in the case of
the glacial tables. The magnetic declination of the region, as obtained by the
Canadian Topographic Survey, is N. 25°5’ E., so that these areas are oriented
with reference to the noonday sun, and might have been used for determining
approximately the meridian and the magnetic declination. The natural infer-
ence is that the phenomenon is due to the reflection of heat from the surface
of the boulders, this action being at a maximum when the sun is upon the meridian.

3. Former ACTIVITY.

a. Terminal moraines. Between the present poorly defined terminal moraine,
described upon page 49, and Lake Louise there occurs a series of ancient
terminal moraines which furnish evidence of the glacier’s former extent and
greater activity. The first two of these moraines are remarkable in that they
consist of massive blocks of quartzite and sandrock, tumultuously heaped
together and without the usual filling of gravel, sand, and clay. The position of
these is shown upon the map, plate m1. The spaces between the great blocks
enable man, or animals, to creep in between and under them and they form
an ideal home for the marmots. For moraines of a somewhat similar appear-
ance, although probably different history, in the Mount Ktaadn region Prof.
R. S. Tarr has used the expressive term “‘bear-den moraine.”.! The inner of
these two moraines extends obliquely across the valley from the present nose,
being partially overridden by the glacier and along the side of the valley nearly
parallel with the oblique front. It consists of what were originally massive

1 Glaciation of Mount Ktaadn, Maine. Bulletin Geol. Soc. of Amer., vol. x1, 1900, plate 37.

OTS ee OS Fe eee ee eee

eee er

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 59

blocks of quartzite, sandstone, and schist tumultuously heaped into a ridge 30 to 4o
feet high. The sandstone and schist have undergone considerable disintegration,
in place, forming more or less soil which supports a growth of moss, shrubs,
spruce, and fir. The rings of growth, in the spruce and fir of the Lake Louise
Valley, were found to average 0.884 mm., and, in the case of the averages for
individual trees, to range fromo.51 mm.to1.26mm. As the tree matures the new
rings are excessively thin; owing to the reduction in the relative amount of leaf

surface, the scant precipitation, the short growing season, and the lack of direct

sun light in a deep valley with a north-south trend. The largest tree found upon
this moraine gave a circumference of 221 cm., at a distance of 50 cm. from the

_ base, and should be approximately 4oo years old. The material of this moraine

is arranged in two main heaps, between which the glacial brook passes, that
upon the eastern side having come from Mt. Lefroy and that upon the west
from Victoria. After the formation of the moraine the glacier retreated up
the valley to a point greater than that occupied by the present nose.

From 300 to 800 feet farther down the valley there occurs the second of this
type of moraines, the blocks consisting very largely of quartzite, lichen-covered
and moss-grown, but not disintegrated. As in the case of the preceding moraine
the blocks are disposed in two heaps, upon either side of the glacial brook,
the bulk of it lying to the west, where it forms an oblique ridge 700 to 800 feet
long, with a maximum breadth of 300 feet and a height of 70 to 80 feet. Made
up of such coarse blocks and with no filling of sand, gravel, or clay the whole
presents a very imposing mass and impresses one with the possible importance of
glaciers as geological agents. The largest block seen had split in falling, measured
31 X 25 x15 feet, and was estimated to weigh 970 tons. The blocks are generally
sharp and angular and have not been subjected to stream or iceaction. They were
carried either upon the ice or within it and show almost no signs of ice abrasion.
An occasional single face is glaciated but in such a way as to show that this was
done when, in its original position, it formed the face of the cliff. Owing to the
lack of soil the growth of shrubs and trees is scant. The largest spruce seen
upon the moraine itself was estimated to be 450 years old, while another just
beyond the outer edge was estimated at 580 years. Upon the side of Mt. Whyte
just at the line of plates, there is a large collection of similar blocks, and appar-
ently of equal age, which appear to have become stranded here while the others
were undergoing transportation. It should be noted that the present Victoria
is entirely incapable of making such a moraine now, no matter how prolonged
the halt. The present terminal moraine, and the oldest of the series to be
described, are essentially alike but very different from these great block, or
bear-den moraines. The cliffs which contributed the bulk of the material,
so far as we may judge from their favorable situation and the location of the
blocks, have a north-northwest trend and the blocks fell from them to the
eastward.

Down the valley, a distance of one-quarter of a mile, there occurs a double
detrital cone, derived from the opposite mountain slopes of Mt. Whyte and Castle

60 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

Crags. The slide from the latter slope is still in process of formation and consists
of freshly broken angular fragments. That from Mt. Whyte is older and covered
with spruce, fir, and willow. This material overlies, and almost conceals, a
triple moraine, composed of boulders, gravel, and clay. The inner and higher
ridge of the three curves regularly across the valley and in one place about 60
feet of it have been cut away by the drainage brook. It here has a breadth of
too feet, a height of about 20 feet, and consists largely of a yellowish, stony
till. Between it and the outer bear-den moraine the drainage stream has filled
in with sand and gravel, forming a nearly level flat. Lower down the slope
and 4oo feet distant, there is another morainic ridge, approximately parallel
with the first, but made up more largely of boulders while 350 feet farther
on there is a third ridge containing a higher percentage of yellowish clay.

b. Lake Louise basin. This lake is roughly elliptical, with a major axis of 1}
miles and a width of 1 to 3 of a mile, placed with its longer axis parallel with the
valley (see plate 1 and plate xtv, figure 2). The chief irregularity in the outline
is due to the presence of a rock slide from Mt. Fairview and to the delta deposits
at the head. The level of the lake is given by the Topographic Survey as 5,670
feet above sea level. This level, however, fluctuates in the course of the
season and from year to year, being some 15 inches higher in the spring, as a
result of the rapid melting of the snow. Four determinations upon the inflow
at the head gave an average of 80 cubic feet per second. Two measurements
_ upon the outflow, through a rectangular orifice at the dam, gave an average
of 88 cubic feet per second, the lake receiving a small additional flow from
Mirror Lake and Lake Agnes. In midsummer, owing to the presence of the
glacial sediment, the lake has a superb green color, but during the winter this
has a chance to settle to the bottom and in the spring the color is more of a
blue, the natural color of pure water.

From the studies of Mr. W. D. Wilcox, reported in the paper previously cited
(page 7), the basin of the lake is seen to be a U-shaped trough, with a maxi-
mum depth of 230 feet just beyond the centre. This shows that the valley
was excavated by ice and that, in all probability, the bottom of the lake is a
glacially excavated rock-basin, similar to that of Lake Agnes, just west but
at a higher level. Bedrock is found upon all sides of the lake, except about the
foot, where there is the ground-morainic dam previously described (page 8)
filling the valley to an unknown depth. At the head of the lake the valley walls
are much contracted, being but 570 feet apart. They consist of a very firm
quartzite, in the main, with a little slaty schist, dipping up the valley at an angle
of ro° tors°. This feature of the valley must have greatly contracted the ancient
glacier passing through the gateway, caused it to thicken correspondingly
and to vigorously gouge out its bed until it had a chance to again expand laterally.
In this way we may account for the presence of the rock-basin, but should
expect the deepest part of it to be somewhat nearer the head of the lake than is
shown in Wilcox’s contour map. It is not at all improbable, however, that the
deepest part of the rock-basin is really so located and that there has been a

‘ih simi iia area il a Oat te oe Bol ee

: ~ emai

ae

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XX.

a

Map of delta, head of Lake Louise, by W. H. Sherzer. Plane-table survey, July, 1904. Field assistants De
Forrest Ross and Frederick Larmour.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 61

filling of the lake bed here with glacial sediment and débris from the Mt. Fair-
view slide, to the extent, at least, of 50 to 60 feet.

c. Lake Louise delta. After the ancient Victoria permanently retreated from
the head of the lake, there began the formation, at the head of Lake Louise, of
a gravel, sand, and silt delta, which is a minimum measure of the amount of
glacial erosion that has since been taking place in the upper part of the valley.
This delta extends into the lake 400 feet, with an average breadth of about 300
feet, as shown upon the map, plate xx, that was prepared with a plane-table in
July, 1904. A shelf of fine sediment borders that portion above water and
then drops off rapidly, forming a layer of unknown thickness over the bottom
of the lake. Much of the delta is elevated but 10 to 12 inches above the
lake level and is under water during high water stages of the lake. Portions
of it are flooded during midsummer after periods of excessive melting. In this
way the delta is gradually growing in height. Low, broad levees line the
main stream. This comparatively small portion, which projects into the lake,
is simply part of a very much more extensive deposit of glacial silt, sand, and
gravel, which reaches up the valley for a distance of $ to 4} of a mile and has a
breadth of 500 to 600 feet. Over this very gradual slope the glacial drainage
courses in three rapid, turbid streams, which unite into a single channel, 40 to
50 feet broad and one to two feet deep. Grasses and moss cover the lower,
flat portion of the deposit, close set willow and spruce the central part, and
isolated willows the upper gravelly region. The rock-slide from Mt. Fairview
has encroached upon the lake as well as the morainic dam at the foot of the
lake, upon which stands the chalet. Between this dam and the range of moun-
tains in the background lies the Bow River Valley, which supported a great trunk
glacier, leading eastward from the mountains (plate x1v, figure 2).

d. Lake Louise Valley. When the mountains were in process of making it
is very probable that this valley began as a structural feature either as a trough
between mountain folds, or from the opening of a joint in the rock strata.
Before the coming of the glaciers it is probable that ages of stream action, aided
by atmospheric weathering, deepened and broadened the original feature into a
V-shaped valley. With the advent of the perennial snowfields a new geological
agent entered, deepening the bed and broadening out the base so that the cross-
section of the valley, particularly the lower half, assumed the characteristic
U-shape of glacially excavated, or glacially modified valleys. An inspection of
the general views of the valley, such as plate 1, shows a lower portion with steeply
inclined, nearly vertical walls, while the upper portion has more flaring walls,
the arms of a truncated v. This upper portion was glaciated to a height of about
9,000 feet above sea level, or some 3,000 feet above the valley floor, the walls being
smoothed and fluted and the spurs evenly truncated; see the shoulder of Mt.
Fairview at the right in plate xrx, figure 2. It is not improbable that these
more gentle, higher slopes represent portions of the original pre-glacial valley,
produced mainly by the action of weather and running water, and not very
materially modified by the ice. These slopes produced until they intersect may

62 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

be assumed to represent approximately the original valley before the ice invasion.
The lower portion, with the very steep walls and broad base, probably represents
that portion of the valley which was profoundly affected by the great ice stream
that so nearly filled the valley. The destruction of the strata necessary to
secure such a result was very probably accomplished by the disrupting power of
the ice, known as “‘ plucking.”

e. Ancient till sheet. The rock débris was, in great part, delivered to the
Bow Glacier and carried beyond the mountains. The finer portions, consisting
of gravel, sand, and clay, with some cobbles and boulders, secured lodgment in
certain favorable places, particularly near the mouth of the valley, and com-
pacted by the great weight of ice formed the ground moraine, or sheet of till.
From 75 feet to too feet of this deposit have been exposed between the foot of
the lake and the Bow River, but the maximum thickness was very probably
much greater. It forms an irregular sheet, of unknown extent and thickness,

reaching up to and under the modern glacier, mantling the actual rock bottom _ _ |

of the valley. It is entirely without stratification and the rock fragments are
largely bruised and scratched. The color, as seen in the exposed sections, is a
brownish-yellow, as the iron of the clay is being slowly oxydized. The more
deeply buried beds would, undoubtedly, show more of the bluish-gray, which
characterizes this material when it is fresh.

CHAPTER IV.
WENKCHEMNA GLACIER.
1. GENERAL CHARACTERISTICS.

NESTLING close in behind the northern base of that grand array of peaks for
which the Canadian Geographic Board has recently adopted the name Wenk-
chemna Group lies the Wenkchemna Glacier. The name is of Sioux origin, w7k-
chemna signifying ten, and was given the glacier by Mr. S. E.S. Allen, in allusion to
its relation to the series of ten peaks upon which he bestowed the Indian numerals.
These peaks, the highest of which has been rechristened Mt. Deltaform, with
their high connecting ridges constitute here the great Continental Divide (plate
xx1). Between them and Mt. Temple lies a broad valley originally called “‘ Deso-
lation Valley’? by Wilcox, but now known as the Valley of the Ten Peaks.
The glacier occupies the southern half of the upper third of this valley, and
faces north, while the valley itself slopes eastward and then northeastward.
In a direct line it lies only about six miles south-southeast from the Victoria, -
and may be reached by crossing the Mitre Pass, encircling the Horseshoe Glacier
at the head of Paradise Valley and entering the Valley of the Ten Peaks by the
Wastach Pass. The ordinary way of reaching the glacier, however, is from
the chalet at the foot of Lake Louise, over a good trail to Moraine Lake,
where a summer camp is maintained by the Canadian Pacific Railway. From

*Aaeains yeotydeisodo 7, uvipsueg “opaaq Ay

‘Oo nyITy Aq ‘(j@9F gzg‘r1) ajdmay ay jo ywums wor ‘Cob ur paydeisojoyg ‘syvag uay, ay} jo Ad][VA ‘saryooy uvipeuRD ‘apa [vjuaunuog ayy,

*191IO¥[4) BUUTAYIHUA MA "WIRAS 3: 1eiqy “OYRT aUIeIOIW

“yenjdan “WAOpeEIICT "LON ‘9 *ON “SON by ON “EON * ‘ON "ART

[XX GL V1d “YWAZYAHS—AOGATIMONN OL SNOILNAIMLNOD NVINOSHLINS

gt = NSS ee nO Poona y Po ® ee, ae eas

oP OO ee ey ee See.

*XOOTIM, “M 'C pue Tapaaq AA *O “V Jo sdevur uodn paseq uoisar
jusoe(py ‘“MouIVT Youeparq puw ssoy IsertoT eC syurysisse ppatq ‘robr “ysnSny “taztays “H “M Aq umvap pure pakaaing “sarppoxy uvipeury ‘sea Way Jo Kawa ‘atryy vuumorypyuay jo dey

gyead yo dn0s18 PUM>USHUSy

“33-S22'11
w40jez}aq

43.9490 ge PRG é “Wb L918
‘ = ‘ enj\de
ZON aii is ert 6, _ yen3den

ee
sg

a

e % 7 } “ 5
Ro Sage
ae

as “he
esa
v pea

POY

PL,

ON De

Yps

er

WUMAYVINUIM

go00e

‘IIXX ALV1Id YWAZUAHS—AIGATMONYM OL SNOILOSINILNOO NVINOSHLIWNS

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 63

this camp, at the foot of the lake, the trail is rough for horses, but practicable to the
front of the glacier, about two miles distant.

Quite in contrast with the Victoria, and with mountain glaciers in general,
_ the Wenkchemna presents noteworthy peculiarities of form, in that it is very
broad for its length and has a remarkable amount of frontage. Its breadth
is about three miles, while its length is from one-half to one mile, the frontage
amounting to something over three miles. The area of the glacier is estimated
at. about two square miles. It lies mainly between 7,500 feet and 6,400 feet
above sea level, the easternmost nose attaining the latter elevation, or about
400 feet higher than the Victoria.

2. PIEDMONT TYPE.

The peculiarities above noted inform are dependent upon the very unusual
method of formation. Instead of there being a trunk stream, to which the
minor ice streams are tributary, the entire glacier results from the amalgamation
of twelve, more or less, independent ice streams, each with its own feeding ground,
which lie side by side. There is no propriety in speaking of these streams as
tributaries; but since they are all nourished from the same general source, the
snow which falls upon the eastern slopes of the Ten Peaks, since they coéxist,
are tolerant of one another’s presence, and maintain their own identity and
independent velocity from névé to nose, they may be spoken of as ‘‘commensal
streams,” to borrow an adjective from the biologists. The form of the front,
the position of the medial moraines, and the névé areas enable us to differentiate
these streams as shown upon the map, and less well in plate xx1. Owing to the
general slope of the valley floor, the commensal streams are deflected eastward,
their natural course being northward. However, it is quite apparent that they
interfere with one another's movements. The easternmost stream is relatively
very small, terminating back some 3,000 feet from the nose of its neighbor,
where it is forming a terminal moraine. Its neighbor to the west is narrow,
but in conjunction with streams three and four, counting from the east, it reaches
the general front and together they form a broad rounded nose. Number five
spreads out fan-like at its lower end, and in consequence six and seven, in their
lower third, are deflected rather sharply to the north. Streams seven and eight,
from Mt. Deltaform, are exhibiting the greatest amount of relative activity.
In the western part of the glacier the ice streams are turned eastward by the
tremendous accumulation of morainic blocks which they are unable to push
ahead or override. In consequence, number nine, which has only a limited
collecting area, is considerably compressed, being forced laterally against its
sturdier neighbor to the east. It is not likely that any one of these streams
could exist by itself as an independent glacier, since it would flatten out, be
more thinly clad with débris, waste more rapidly from surface and lateral melting,
and disappear soon after leaving the shadow of the mountains.

This form of glacier is known as the “‘ piedmont type,”’ and, so far as the writer
is aware, only one other example, the Malaspina, of Alaska, has thus far been

64 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

described. This has a breadth of 70 miles and an average length of 20 to 25
miles, with approximately 1,500 square miles of area. It is made up of four great
commensal glaciers, with an innumerable number of smaller ones. Farther
west there lies the Bering Glacier, known to be of the same type, but not yet
visited and described. It is quite likely that this variety of glacier is more
common than has been recognized, since in addition to the Wenkchemna there
is the Horseshoe Glacier at the head of the adjoining Paradise Valley, with some
16 commensal streams, and the Asulkan in the Selkirks (plate xxx1x, figure 2)
which represents a piedmont glacier in process of disintegration into its compo-
nent streams. If the ordinary Alpine glacier, with its tributaries, is compared
to a river, the body of a piedmont glacier should be thought of as an ice lake,
of greater or less magnitude.

3. NOURISHMENT.

Each of the component streams of the Wenkchemna may be traced back to a
more or less well defined and fairly distinct patch of névé. This may be no
more than a cone of avalanched snow, or it may be a strip of permanent snow
field filling a couloir in the mountain side, or between two adjacent peaks.
The highest point of the Divide here is Mt. Deltaform, with an elevation of 11,225
feet, somewhat lower than Victoria. To the west, and closely connected with
Deltaform, is Neptuak (Allen’s No. 9) with an elevation of 8,767 feet. East-
ward from Deltaform there occur in order No. 7 (10,648 feet), No. 6 (10,520
feet), No. 5 (10,018 feet) and No. 4 (10,028 feet). The northern face of this array
of peaks is very abrupt, furnishing only a meager collecting area for the snow.
The snowfall is probably not materially different from that at the head of the
Lake Louise Valley, where it is estimated as about 25 feet annually. That which
clings to the steep slopes during the winter is largely avalanched upon the glacier
below in the spring and early summer. The most of that which remains is melted
and but little survives the warm season. The snow accumulates along the
northern base of the range, where it is protected from the noonday sun, allowing
it to become converted into névé and compacted into ice. Owing to the increased
altitude of the glacier’s surface toward the west there is a greater deposit along
the base of Deltaform and Neptuak (plate xx1). This rather meager supply of
snow could support a glacier of such dimensions only because of certain favorable
conditions. Being in the lee of some 3,000 feet of nearly vertical cliff, its névé
field is sheltered from the noonday sun, while that portion which is exposed is
almost completely veneered with a protective covering of rock débris. The
form of glacier, with the ice streams lying side by side, reduces the lateral melting
to a minimum, while the slope of the valley floor, upon which the glacier rests,
is sufficiently gentle to allow the ice to remain in a very sluggish condition.

4. DRAINAGE.

Because of the conditions just outlined the amount of ablation is reduced
to a minimum and the surface drainage streams are correspondingly small and

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XXIII.

Fic. 1.—Drainage brook from Wenkchemna Glacier, August, Igoq. Free from sediment

and having a summer temperature of 35° to 36° F.

No, 2. No. 3. No. 4. No. 5-

Looking southward, The very complete

Fic. 2.—General view of eastern end of Wenkchemna Glacier, August, 1904.
covering of rock débris and irregular surface are w ell shown,

+

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 65

inconspicuous. Some near the center have cut their way a few feet into the ice.
None of them reach the margin of the glacier, but find their way to the base
through crevasses, or moulins. Opposite peak No. 7 a small stream of pure
water from the valley enters the side of the glacier. The drainage from the Wenk-
chemna Lake, which collects the waters from the base of Mt. Hungabee, reaches the
glacier through the great accumulation of morainic blocks (plate xxtv, figure 3).
No surface lakelets were observed upon the glacier, although numerous depres-
sions occur, suggestive of the former sites of lakelets. The water is probably
lacking now because of the small amount of surface melting. There was practi-
cally no marginal drainage observed. At the east end, between the side of the
mountain and the glacier, there occurs a small lakelet and opposite Mt. Delta-
form, at the front, there is a very shallow lakelet, or marsh, of insignificant
proportions.

The various subglacial drainage streams are collected beneath the glacier into
a single stream which gushes from the extreme eastern compound nose and cascades
over the coarse blocks of the frontal moraine in several channels. These form a
single broad drainage brook (plate xxl, figure 1), about too feet across, shallow
and rapid, which enters Moraine Lake one-half mile below, dropping 210 feet
in the distance. The volume could not be measured, but was estimated at about

go cubic feet per second, being somewhat in excess of the volume of the Victoria

drainage brook. The volume did not fluctuate during the day nearly so much as
is usual for glacial brooks, suggesting that the supply is not so dependent upon
immediate melting of the ice. This is further indicated by the temperature of
the water, which remained during the middle week in August very steadily
at 35.6° F., and rarely varying more than 0.2° to 0.3°, no matter at what time
of the day taken. September 8, 1905, it was still 35.6° at 12:30 p.M. A sur-
prising feature of the brook is its remarkable freedom from glacial sediment,
the water issuing from the glacier perfectly pure. Flowing over coarse gravel
and boulders it acquires no sediment upon the way to the lake and has formed
not even the suggestion of a delta at the head of Moraine Lake (plate xxv).
This indicates that the subglacial erosion is practically nothing and has remained
so for centuries, testifying to the sluggish condition of the glacier. In flowing the
half mile to the lake the temperature of the water in August was raised from
35.6° to 36° or 37°. The lake is about a mile long, has an elevation of 6,190
feet, is apparently shallow, and filled with the purest water of an intense blue
color. Passing the length of the lake the temperature in August is raised to
about 44°F. The freedom of the water from sediment permits it to exhibit its
natural color, a simple glance at which, as far away as the color could be dis-
tinguished, enables one to safely predict the absence of sediment from the brook
emptying into the lake. Should the glacier become active and start to erode its
bed the color of the lake would change to some shade of green.

5. MOoRAINES.

Except for the comparatively narrow strip of névé along the base of

66 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

the cliff, well seen in plate xx1, the entire upper surface of the glacier is
veneered with angular rock débris, effectually preventing surface melting, as
already shown. This material is derived from the Wenkchemna group of peaks,
through the agency of avalanches and the ordinary processes of weathering.
With the entire breadth of the glacier spread out along the base of the cliff
all portions receive their quota, leaving no portion of the ice exposed to the
sun. The débris, at first, is covered with the snow, but it is concentrated by
melting until the amount is sufficient to prevent further loss of ice at the surface
when the action ceases. No ground-morainic material was observed upon the
surface, in contrast with the Victoria, and this is accounted for by the absence
of hanging glaciers. What might be mistaken for such upon the northern face
of Mt. Deltaform, and upon either side of peaks 4 and 5 (plate xxt), are simply
the continuous névé fields of the commensal streams. This method of acquiring
its load leads to a somewhat irregular distribution of the rock débris, resulting in
hummocks and depressions, especially towards the northeastern corner (plates x XI
and xxv). These irregularities of surface are also shown in plate xx11I, figure 2.
It was from this portion of the glacier that the view for plate x1x, figure 1, was
taken. This irregularity of surface renders travelling across the glacier laborious
and somewhat dangerous, except near the névé line. The almost complete
concealment of the ice by débris renders this glacier a poor one for the study
of ice structure and the usual surface features. The third ice stream, coming
from between peaks 5 and 6, in the vicinity of the névé shows stratification
and dirt zones to advantage, the strata ranging from five to ten feet in thickness.
Low glacial tables occur here and the phenomenon described upon page 58 of
this report, as possibly due to reflection of heat from the surface boulders.

The line of junction between neighboring ice streams is roughly indicated
upon the surface by ridges of rock débris, somewhat low and poorly defined
near the névé, but gaining in height and distinctness in their course across
the glacier. These ridges are the lateral moraines of the individual ice streams
and they are especially well defined over the eastern third of the glacier. Upon
either side of the third stream these ridges are double for a considerable dis-
tance. Toward the western end of the glacier these moraines are neither so
well defined nor so continuous and, owing to the deflection of the streams to
the eastward, by the ancient moraine, swing around into a position almost
parallel with the frontal. By the blending of these lateral moraines upon
adjacent streams the single ridges resulting become the medial moraines of
the piedmont glacier, and the outermost laterals of the two marginal streams
become the laterals of the unified glacier. At the eastern end of the glacier the
first stream is so short that the right lateral of the second ice stream constitutes
the right lateral of the glacier asa whole. If the first stream ever extended to the
front then this lateral moraine was originally a medial. A deep depression, snow-
filled in tg904, separates it from the double débris cone and from the mountain
spur shown upon the map. At the western side of the glacier, owing to the
deflection eastward of the ice streams, there is no distinction to be made between

Ay ee yee fee SES Sarai ras

=e

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 67

the lateral and the frontal moraine. They end in a peculiar series of short
closely pressed ridges, slightly concave outward.

Since a glacier of this kind cannot be said to have an end, the term frontal
may be more appropriately applied to the rock débris that is being dumped
along the united extremities of the individual ice streams. Were the glacier
to begin a uniform retreat from its present position, there would be left a ridge
of angular rock débris, over three miles in length, marking the shape and present
position of the front. Inside of this would be left upon the valley floor the débris
which now mantles the surface of the ice, or is contained within. Because of
the very slow advance, to be noted below, the frontal morainic material over
the eastern half is being very slowly urged forward, giving a steep and unstable
frontal slope, but not so steep that it can not be climbed at almost any point.
At only one point, nearly opposite peak No. 7, is there any ice showing and
here the débris cover is partially lacking. Should the ice front actually halt a
frontal moraine would form very slowly, in spite of the amount of débris carried,
because of the sluggish condition of the ice. Toward the western side the front
becomes less steep and high and finally merges into the névé and snow bank

which mantles the col between Neptuak and Hungabee.

6. CREVASSES.

The glacier is remarkably free from crevasses in the lower part and about the
sides. In the case of the commensal streams, measurements would probably
show that the sides were moving forward at about the same rate as the centers,
so that there is lacking that differential movement that gives rise to marginal
crevasses. The mutual pressure from the sides is sufficient to prevent the
opening of radial crevasses along the front. The absence of prominent transverse
crevasses indicates that the bed is of even slope and the motion slow enough
to allow the ice to yield, without rupture, to most of the inequalities that do
exist. The absence of crevasses in this case is quite as instructive as their pres-
ence would be. Upon the steeper portions of the névé slopes there occur nu-
merous transverse breaks of the nature of bergschrunds, caused by the upper mass

_ clinging, for the time being, to the rocky wall while the lower portion draws

away from it. If kept under inspection these schrunds would be found to close
up, as they work their way down the slope and to open again at a higher level.

7. MovEMENT ABOUT THE FRONT.

In a little booklet prepared for the Canadian Pacific Railway by Messrs.

- George and William Vaux, and entitled Glaciers, attention was first called

to the evidence that this glacier is advancing into the adjoining forest. No

‘data were at hand for determining the amount of this forward movement, or

whether it is still in progress. Dead trunks of forest trees, from which the
bark and branches have fallen, are seen projecting from near the frontal slope
(plate xxiv, figure 1). Some of these trees were probably killed by a forest fire

68 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

which swept through the valley 70 to 80 years ago. Other trees in similar posi-
tion, but also dead, still retain their bark and boughs, but show no signs of fire.
It is likely that these trees were killed, and more or less displaced, by the ad-
vance of the ice front (plate xxrv, figure 2), since which time the ice has advanced
less than a dozen feet. This is still further evidence of the almost stagnant
condition of the glacier. Only at one point, near the center, were there any
trees which have been recently cut by rolling blocks from the frontal slope.
This is taking place about the nose of the stream coming from Mt. Deltaform.

In order to gather some definite data concerning the frontal movements,
a series of eight sets of reference blocks was established along the eastern half
of the front, beginning at a point just east of the drainage brook. Between
certain marked points upon boulders that had rolled forward and others firmly
embedded in the frontal slope, accurate measurements were made with a steel
tape. From August 9 to September 12, 1904, an interval of 34 days, it was
found that there was no perceptible movement at the station east of the drainage
brook. Passing westward along the front, and up the valley, the data indicated
that there had been a wastage of the ice, causing the blocks to settle back 1.2
inches and 0.7 inch. The next two stations showed an advance of 1.9 and 1.3
inches, while the next two gave a retreat of 1.0 and 4.6 inches respectively.
At the upper station, where the trees had been freshly cut, the advance for
the 34 days amounted to 11.8 inches. One year later, September 8, 1905,
measurements were again made between the series of blocks and at all of the
stations (the upper block at station D could not be located because of disturb-
ance) there was a small advance indicated, varying from 1.7 inches to 20.4
inches. The least movement was about the extremities of the easternmost
streams and the greatest was towards the center.

A summary of the measurements is given below.

Stations. Movement for 34 days, Aug. 9 to Movement for 361 days, Sept. 12,
Sept. 12, 1904. 1904, to Sept. 8, 1905.
A o.o inches 1.7 inches.
A =Tae Aue 2.4 i
B =O/n7= aes 4eSti ics
Cc LQ) est 12.0)
D ito erg Missing.
E = Tis Sie oe 3185)
F =4°/6) aes ZONA ass
G DI Ors oe : .

DS .cb

These figures indicate that the component glaciers are as independent in their
movements as in their structure, and that some may be stationary, or in retreat,
while others lying alongside are advancing. The question of the frontal be-
havior of a piedmont glacier is thus seen to be complicated in ptoportion to
the complexity of its structure. Measurements made at single stations can give
only very incomplete data concerning the glacier as a whole. There should be
at least one such measurement for each commensal stream.

8. FormMeR ACTIVITY.

a. Bear-den moraines. Along the western front of the Wenkchemna, for a

*bFobr ‘ys

My ‘a1oe[y

s

Bio} A ‘ayizyvnb Kvas yo YoorTq aovjins jo apis y74ow ayy uodn ware poyayy—'h ‘org “ID1OB[D, vuUaYyo

“UDA ( Jure1ouL Uodp-1vag qustoue jo syxpo|q auojspuvs pajerdaquistc] —'€ ‘OI

”

> ‘ahi,

at

“So61 ‘1 aquiaydag “UOISLAUL JSaIO} JUIIII SULIMOYS ‘TIIOVIL) BULTAYIYUIAA JO JUOIY—'S ‘OT

‘robr ‘jsnsny

“JSo1OJ 91} uodn JUIUIYIVOIDUD S}t OUTMOYS "TILED PUMAYOYUdAA JO JUOTY—'I “OMY

“AIXX ALV Id “MAZUAHS—IOGAIMONN OL SNOILOAIM. LNOD NVINOSHLINS

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 69

distance of over a mile, there occurs a tremendous accumulation of huge mo-
rainic blocks of a red and brown sandstone. The blocks are much disintegrated
by the weather, and falling apart, but each roughly indicates its former size and
_ shape (plate xx1v, figure 3). Near the upper end of the valley the moraine is a half
mile across, extending from the glacier to the foot of Eiffel Peak and almost
completely surrounding Wenkchemna Lake, as shown upon the map. Toward
the east the moraine becomes narrower and there may be distinguished an
older portion, partially soil-covered and forested. These latter blocks are more
completely coated with lichens and have plainly the appearance of greater age.
There is thus evidence that the moraine was formed at two different periods,
the older portion surrounding the lake and extending eastward, and that a con-
siderable interval, as expressed in years, separated the two periods. Opposite
peak No. 7 there occurs at the front an accumulation of coarse blocks, evi-
dently part of the younger of the two ancient moraines, which the present glacier
has been able to partially override. Farther west the glacier has been unequal
to the task of either pushing the blocks ahead or of overriding them, and the
streams have been deflected eastward, as previously noted. When the forma-
tion of the older of the two moraines began the glacier reached across the valley
and deposition started. The glacier had so little depth, owing to the meager
supply of névé, that it was unable to heap the blocks into a great ridge. Ap-
parently the rather thin edge of ice was pressed against the moraine and there
melted by pressure, and the blocks deposited along the southern margin until a
belt a quarter mile in breadth was formed. A considerable period intervened
during which time the eastern portion of the glacier had shrunken to much
smaller proportions than it had formerly held and than it has at present. In
a manner still to be accounted for, the glacier a second time became loaded
with very coarse blocks and started to advance, possibly because of the protection
afforded the surface by this load. Encountering the former moraine, however,
it was unable to go over, or around, and it used its energy in a vain attempt
to push the obstacles ahead. The pressure exerted caused the melting of the ice
and the blocks were deposited in another broad, continuous belt, parallel with
the first. As though it had learned wisdom by the experience, the glacier now
rather calmly turns eastward and passes around the obstruction which it has
built in its own path. These two moraines, lying side by side, are of the same
type as the two described in the Lake Louise Valley, and, as far as may be judged
roughly from their general appearance, of approximately the same age. The
cliff from which the material was obtained has a west-northwest trend and the
blocks were dropped from it to the eastward.

b. Moraine Lake. ‘This lake has had a different history from that of Lake
Louise in that it is apparently not a rock-basin and so attains no great depth.
It is, however, like Lake Louise in that it has a morainic dam across its foot,
although in the case of Moraine Lake the dam is of a different type. This
consists of a sharply defined heap of rock débris about 400 feet long, placed
at right angles to the main axis of the valley. The ridge increases in height

7° GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

rather gradually toward the west and attains a height of about 70 feet, ending
abruptly, as steeply as the débris will stand and with no trace of any continuance
across the valley. This is so unusual a feature for a terminal moraine that many
are disposed to consider the mass as a rock slide from the adjoining mountain
face. There are several varieties of rock represented in the heap and time did
not permit an examination of the adjoining face to see whether they might have
had such a source. The strata in the region, however, are so nearly horizontal
that, whether the feature is a moraine or a slide, the same rocks probably occur in
the adjoining mountain as farther up the valley. Standing upon the highest
crest, the ridge is seen to be double, the outer one somewhat convex down stream,
while from the western end there passes a short spur down the valley. The
writer is disposed to accept the view of Wilcox, who gave the name to the lake,
that we have here a moraine. It is, however, not of the bear-den type found
farther up the valley, but very much older than the most ancient of the two.
Its general lack of vegetation may be due to the scarcity of suitable soil, although
it does support a sparse growth of timber. The unusual features of the mass,
considered as a moraine, will be understood when the unusual nature of the glacier
that formed it is considered. This represents the position of the front of the
easternmost ice stream, of the ancient piedmont Wenkchemna during a pro-
longed period of the halt. This moraine originally abutted against a wall of ice
at the west end, the side of the adjacent ice stream, which probably extended
far down the valley and may have been engaged in making a correlative moraine.
A relatively small amount of the débris was dragged for a short distance down
stream by this neighbor, forming the spur above noted. When this ice wall
melted away finally the débris rolled down and assumed the “‘angle of repose.”
As has been pointed out the present easternmost stream is short, compared with
its neighbor, and were it to make a sufficiently prolonged halt there might be
produced, upon a smaller scale, this identical feature.

c. Valley of Ten Peaks. The time that could be devoted to this glacier did
not permit of an examination of the valley from the lake to the Bow River, or of
the interesting Consolation Valley, which still supports a glacier at its head.
Observations of only a general nature from the elevated trail around Mt. Temple
could be made. There is evidence that the entire valley was occupied by a great
ice stream, a tributary of the trunk glacier that filled the Bow Valley, the glacier
then being of the Alpine type. The lower half of the valley was altered by the
ice into the characteristic U-shape, while the upper half retained its flaring side
walls from pre-pleistocene time. In making the bend from its east-southeast
course to the northeast, the glacier pressed hard against the western face of
Mt. Babel, while upon the opposite, or concave, side there was deposited a high
ridge of ground-morainic material which swings around in a very regular curve
from the Eiffel to Mt. Temple. From the northeastern shoulder of Mt. Temple
there extends into the Bow Valley, curving gently down stream, a spur of ground-
morainic material, identical with that described for the Lake Louise Valley upon
page 8. This was deposited beneath the ice and along the line of junction of the

gO) oye v SOOUT i acy oy jo Asayanoo oq} ysnoiyy a1 peystqud “LOTOB[E) BUMMIYO Ud AL jo Ayiuiaayxa UlO]SBO pue ae'T JUIVIO JY jo MOLA [Btouar)

SE oS er i/o
. . ‘;
x

ining

“AXX ALV Id YWAZAAHS—AOVGATMONY O.L SNOILLOGIN.LNOD NVINOSH.LINS

il
1
t

“MOULLET AIO PIT PUB SSOY JSIIIOT a] spuvisisse platy “Fobr YsnSny ‘zoztayS “EH *A\ Aq uMvap pur pakaaing ‘sarpoy uvrpeurg ‘ATA OYOR JO pvay ‘sd19¥PH (vide) oyox yo dey

BMIODIAN MOF 41915 e)/5

oo Ee]
39935 000+ o00e 0002 000!
3|e3s

IAXX ALVI1d “AMAZAAHS—AOGATMONY OL SNOILAGINLNOOD NVINOSHLINS

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. VL

ancient Wenkchemna and Bow glaciers. Looking down the valley towards the
Bow, Babel Mountain is on the right-and Mt. Temple upon the left, rising above
the high morainic ridge mentioned. From the shoulder of Mt. Temple there ex-
tends to the center of the view the morainic ridge reaching out into the Bow
Valley, gradually losing in height and breadth. As pointed out by Wilcox, this
deposit probably mantles a rock spur which escaped destruction by the ice.
During the height of glaciation a tributary glacier moved in northwestward from
Consolation Valley and joined the Wenkchemna at a level 400 to 500 feet above
the present valley floor, forming a “‘hanging-valley.’”’ From the ‘‘Tower of
Babel” there curves across the mouth of the valley what appears to be a morainic
ridge, of the same nature and origin as that just described. The height of this
hanging-valley above that of the Ten Peaks is believed by some to measure the
differential erosion between the ancient Wenkchemna and that of the tributary
which occupied this valley.

CHAPTER V.
YOHO GLACIER.
I. GENERAL CHARACTERISTICS.

Tuts glacier, the largest and most northerly situated of the series studied
constitutes a tongue of ice from the great Waputik snow-ice field which mantles
the Continental Divide to the north of the railway. Its nose lies in latitude 51°
34’, at the head of the picturesque Yoho Valley, and is most conveniently reached
from Field, via Emerald Lake. The day’s ride, over’a fairly good trail, up this
ice-cut valley, with its hanging glaciers and plunging cataracts, is an experience
never to be forgotten. The return trip to Field should be made over the Burgess
Pass. During the summer the Canadian Pacific Railway maintains a camp at
Laughing Falls, some four miles from the glacier. The glacier was first made
known through the descriptions and photographs of Jean Habel,' secured in
1897, and each summer since it has been visited by gradually increasing numbers
of tourists and students. The original name was derived from the Wapta River,
another name for the Kicking Horse, the name ‘“‘wapta’’ itself meaning river
in the Stoney Indian language. The name Yoho since approved by the Canadian
Geographic Board is the Indian exclamation of surprise and wonderment.

As one emerges from the forest and comes suddenly face to face with the
glacier, plunging at him from above, he is greatly impressed with its size and appar-
ent power. Its freedom from surface débris better enables it to meet the popular
idea of what a glacier should look like,—the Victoria and Wenkchemna, having
been somewhat disappointing (plate xxvul, figure 3) in this respect. The Yoho
has the general form of a gauntlet mitten, extending in a south-southeast direc-
tion, with the thumb upon the eastern side of the valley and partly surrounding
a great rock embossment (see plates xxvi and xxviul, figure 2). Independently of

1“ The North Fork of the Wapta,’”’ Appalachia, Vol. vii, No. 4, r898, pp. 327-336.

72 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

its névé it is three miles in length and for its upper two-thirds nearly one mile
in breadth. In rounding the rock embossment noted it narrows to a half
mile and then tapers regularly to a blunt nose. The mean elevation of the
névé, according to Wheeler, is 8,400 feet, and the nose descends to an altitude
of 5,670 feet. By noting the temperature of boiling water, Habel determined
this elevation at 5,680 feet. The nose of the Yoho is thus 330 feet lower than
that of the Victoria, and 730 feet lower than the Wenkchemna, in spite of the
lack of débris covering and the southerly exposure. This is due, without doubt,
to the greater precipitation and the greater size of the collecting area, by which
a much larger body of ice is amassed. The mean average slope from the névé
to the nose is about goo feet to the mile; the main part of the descent, however,
is in the lower half. The general inclination of the ice about the front is 20° to
25°. Upon the western side the glacier presses more or less firmly against the
valley wall, except for a short distance, where the ice is steep and not to be ascended
without cutting steps. By crossing the drainage stream, which is not a simple
proposition unless one is mounted, one may easily ascend the glacier without
the cutting of steps, the ice slope being very gentle. Skirting the crevasses and
crossing back to the west side of the glacier, the névé may be easily and safely
reached. Since its discovery by Habel, the glacier has maintained a great archway
of ice at its lower extremity, which spans 250 feet of space, and is estimated to be
70 feet high, from which escapes the drainage. Owing probably to its southern ex-
posure there is formed a cavern beneath the arch, as seen in plate xxvul, figure 4,
extending back into the ice 100 to 200 feet, but not forming a subglacial tunnel.
Toward the close of the summer season, the arch has become so weakened by
melting and the formation of a transverse crevasse (plate xxvul, figure 3) that
the entire structure collapses and lies a heap of azure ruins. The blocks of ice
are melted down to a size that the stream can push, roll, or float, some head being
obtained for the stream by the damming action of the ice débris. Finally it is
all removed and the making of a new archway is started.

The actual nose of the glacier lies to the east of the archway and rests upon
limestone bedrock, with only a sprinkling of ground-morainic material. Upon
either side, and for some distance beyond the nose of the glacier, there is bed-
rock exposed, which has been smoothed in places by previous ice action and
in other places roughened by plucking. Upon the western margin of the drain-
age brook there are shale strata upon edge which have been thus roughened.
Excessively thin laminz alternate of an intense red and yellow color. There are
no reliable data for estimating the thickness of the ice, but it seems to be consid-
erable. In the case of the Victoria and Wenkchemna glaciers the most conspicu-
ous geological work being done is transportation, in the case of the Yoho we
have very plainly a great engine of erosion.

2. NOURISHMENT.

The collecting area of the Yoho is triangular in outline and includes the region
between Mt. Collie (10,315 feet), Mt. Baker (10,441 feet), and Mt. Gordon (10, 336

i El

i ta

ad

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 73

feet), the sides of which triangle are approximately 3x 4x5 miles (plate xxv,
figure 2). The area is located upon the western slope of the Great Divide, the
crest of which extends in a curve from Mt. Baker to Mt.Gordon. The collecting
area is estimated at about 63 square miles. Upon the western slope of the Divide
it presumably receives more precipitation than falls in the Lake Louise Valley
and that of the Ten Peaks. From the meager data available at Field we calcu-
lated that the precipitation may amount there to 42 inches per annum, page 11.

‘To the north it would be somewhat less and may be assumed to be 4o inches.

Over the névé area the great bulk of this would fall as snow, but that which was

precipitated as rain would be absorbed at once and rendered available for the

glacier, representing about 331 feet of snowfalleach year. This amount over the
collecting area, the region in which the snow is manufactured into glacial ice, would
represent some 224 million cubic yards of snow, or about 24,396,000 cubic yards
of ice, available each year for the Yoho. Ifourassumed data are approximately
correct, this must represent the amount of ice to be disposed of annually by
melting and evaporation. Converted into water, this volume of ice would pro-
duce 22,372,000 cubic yards of water, or 604,032,000 cubic feet of the same.
Distributed over the months May to September inclusive, durng which time
the melting is most active, this would give an average flow of about 46 cubic
feet per second. During midsummer the flow is probably four or five times this
amount, due largely to the fact that the actual area drained is much larger than
the single névé field, and that the melting is now at a maximum. The névé
coming in from the eastern, or Mt. Gordon side, as well as that from the western
or Mt. Collie side, has already been compacted into glacial ice before reaching
the main flow from Mt. Baker. This ice is incorporated into the Yoho névé
with whatever débris it may be carrying. The absence of overhanging cliffs
about the névé area, quite in contrast with the Victoria and Wenkchemna
glaciers, prevents the névé snow from becoming charged with rock débris.
The glaciers from the slopes of Gordon and Collie, as well as the main stream
from Mt. Baker, are carrying only subglacial material, of which we have
evidence later. To this is to be ascribed the freedom of the glacier from surface
débris. This condition of the ice, combined with its southern exposure to the
sun, is unfavorable to the maintenance of a glacier at low altitudes. This is
entirely obviated, however, by the greater bulk of ice available when this
glacier is compared with the two previously described.

3. DISTRIBUTARY.

Except for the névé-covered glaciers above noted from Mts. Gordon and
Collie, the Yoho has no tributaries; but instead, what has been termed a dis-
tributary, to assist it in getting rid of its ice supply. A considerable volume of
ice is deflected around to the eastern side of the rock embossment and is there pre-
maturely melted. (See plates xxvi, xxvull, figure 2, and plate xxvu1, figure 3.)
This tongue of ice forms a very pretty little glacier, one-half mile long by one-
quarter mile broad and tapering down to a blunt nose (plate xxvil, figure 1).

74 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

The surface is soiled with wind-blown dirt but it carries little débris. At an
earlier stage it brought down from above the left lateral morainic material for the
Yoho, and its ice extends still well under this ancient moraine. The axis of the
glacier is curved as it is forced around the rock embossment, which it hugs closely
and is still engaged in fluting and polishing (plate xxvur, figure 1). The upper
crest of the rock embossment has an elevation of 6,960 feet, while that of the
nose is 6,320 feet, or 650 feet above that of the main glacial stream. The average
slope would be at the rate of about 1,300 feet to the mile. It is evidently in retreat
although no data are available for determining the rate. In the upper portion
of its course, it appears to descend over a steep step in its bed and is much
crevassed. These crevasses completely heal, however, or are destroyed to their
bases by melting, leaving the lower half exceptionally smooth. Over this portion
of the small glacier there are developed three very pretty drainage systems, two
of them marginal and the third central. The central drainage, which is collected
into a trunk stream (plate xxvit, figure 1) from a network of small tributaries,
has cut a longitudinal channel in the ice, and continues to a point just east of the
nose. Habel’s map of the Yoho Glacier shows this tongue of ice continuous
around the rock embossment, forming of it a rock island, or so-called “‘nunatak.”’
Such a position it originally held, but not less than 200 years ago, so far as we
may judge from the size of trees growing in the valley. At a still earlier stage
of glaciation the entire embossment was overridden by the ice, much of it being
disrupted, wherever the ice could get a satisfactory grip upon the strata. The
resistant portions were planed down, rounded, and fluted.

4. MoraIneEs.

Because of the lack of overtowering cliffs, above noted, the general surface of -

the Yoho is practically free from coarse rock débris, in striking contrast with the
Victoria and Wenkchemna. For the same reason also the lateral moraines
are poorly developed and almost absent in the lower half mile. Upon the western
margin, just before reaching a broad glaciated valley, originally carryiug a
tributary, the right lateral moraine begins to make its appearance and develops
across the mouth of the valley into a well defined ridge of stony till, or ground-
morainic matter. This ridge is continuous up the slope to the line of junction
of the main Yoho with the glacier from the eastern slopes of Mt. Collie, where
the latter is seen to be delivering this material from its under side to the surface
of the Yoho. This ground moraine has been produced between the Collie
Glacier and its bed, frozen into the ice, and urged down the slope.

The left lateral moraine begins along the southern side of the rock emboss-
ment, down in the valley, asa double ridge from which the ice has been withdrawn.
It extends around the embossment, on the west side, for a distance of some
4,500 feet, developing into a prominent, high, sharp-crested ridge at the head
and curving across to the eastward. This consists, also, almost entirely of ground-
morainic matter, which must have been derived from the basal layers of the ice,
which became stranded at the head and about the west side of the rock emboss-

esse he

oe

Dea Niet

*Pobr ‘ys C
eB *po61 Ysnsny ‘ssoy jso110y act Aq paydea 404g ‘“Aa[eA OYOA JO pray ‘a1R[L) OYOR—'E “OY

-0j}0Ud "Io ALY Oyo ayy sansst Ory WOdy “IDOI (

Pain L—'P “oly

“ADONIS OYO A WO Areynqtiystp a[—"I “OLY

‘AaypeA OYOR Jo pray ‘vad urejunou ev uodn suryontd ,, a9J—"s “Oly

“IIAXX ALW1d “MA ZYAHS—-ANGATMONY OWL SNOLLAAIYLNOOD NVINOSH.LINS

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 75
ment. In the valley occupied by the distributary described, and from 600 to
700 feet down from its nose, there begins a second lateral which curves around
upon the débris-covered ice and continues for two miles along the shoulder of
Mt. Gordon. This portion of Gordon has an elevation of 9,510 feet, but its
cliffs do not overhang and contribute only a moderate amount of material to the
moraine. A considerable portion of it consists of ground moraine which may
be traced to the glacier from Mt. Gordon. This sustains the same relation to the
Yoho as does the Collie Glacier upon the opposite side of the valley. Up to
heights of 30 to 40 feet patches of morainic material can be seen upon the valley
wall, left there when the surface of the glacier stood at a higher level. The rocks
in the moraine are largely limestone, light and dark, yellow and mottled; some
pieces being odlitic.

An older lateral moraine, upon the eastern side of the valley slope, may be
traced for some 2,000 feet up the valley from the nose. The ridge is some 200 to
300 feet from the margin of the ice, is but three to four feet high and inconspicu-
ous, but it has heaped promiscuously over it a mass of broken tree trunks very
evidently brought down by an avalanche and heaped against the side of the ice
when it stood here. The distance from the margin of the ice to the ridge, at one
point, was found to be 260 feet, along the slope. The wood is somewhat decayed
and gives some appearance of age. A photograph was taken when the trunks
were covered with a light fall of snow, which had melted from the surrounding
rock, rendering them much more conspicuous than they would otherwise be
with their dark surroundings. Growing in the path of the avalanche trees
were found, the largest of which gave 25, 28, and 47 rings respectively. It is
likely that the avalanche occurred between 1850 and 1860, since which time
the glacier has been retreating down the slope at the average rate of 5 to 6 feet
per annum.

With so little rock débris carried upon and within the glacier it would require
a very prolonged halt of the front in order to build up a terminal moraine of any

considerable proportions. About 200 feet from the present nose, at the end of

the bedrock upon which it rests, there swings in from the side a weakly developed
double ridge, low and inconspicuous. It may be the correlative of the lateral
moraine above noted, carrying the avalanched timber, or it may mark a still
more recent halt in the general retreat. Within recent time the glacier has
deposited very little ground moraine, the conditions not being favorable for its
lodgment. The lowermost stratum shows upon the western side of the drainage
stream, beneath the archway, and is seen to be charged with débris. Much of
this is delivered to the stream and swept away by the swift current, the remainder
being spread thinly over the valley floor.

5. CREVASSES.

Opposite the head of the rock embossment described, the glacier and its dis-
tributary plunge over a steep step in their beds, of which the embossment itself
is probably a portion which the glacier was unable to reduce to the general level

76 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

of its bed. In making this rapid descent the ice is crevassed both transversely
and longitudinally. The irregular blocks of ice formed melt into sharp points,
or steeples, to which the term ‘‘seracs” is applied. The transverse crevasses
have been noted which open just behind the archway at the nose, allowing the
arch to collapse toward the close of the melting season. The absence of pressure
in front allows the arch to drop forward, faster than the ice can yield to the ten-
sional strain, and the crevasse is the result. Along both margins, for nearly the
entire three miles,the normal lateral crevasses, described for the Victoria, occur.
They extend inwards and upwards for varying distances, are irregularly spaced
and become less numerous toward the névé, where some of them are snow-filled
and snow-covered. In passing the ice cascade the ice is too much shattered to
permit the formation and preservation of the dirt bands described upon page
52 of this report. However, at the crest of one of the minor slopes the phenome-
non may be seen, as shown in plate xxIx, figure 1, where the depressions for
three bands are plainly marked out. These mark the sites of former crevasses,
and, if rightly interpreted, the distances between them show the approximate
annual motion at this portion of the glacier.

6. IczE STRUCTURE.

In both the main glacier and the distributary the stratification of the ice
is poorly preserved, possibly because of its destruction in passing the cascade.
General views, as wellas detailed ones, give almost no trace of the strata. In plate
Xxvil, figure 4, one stratum, relatively much charged with débris, forms the base
of the arch, but does not appear upon the opposite side. This basal stratum
where seen is 2 to 5 feet in thickness. Its upper surface may represent a shearing-
plane, the body of the stratum being held more rigidly by its content of débris
while the superincumbent ice is forced over it.1 In places where the strata are
still preserved, the dividing planes show poorly, and it is to be noted that this
may arise because of the paucity of foreign material concentrated at the upper
surfaces of the strata. In the case of this particular glacier the size of the névé

field precludes any but the finest dust from reaching the general surface, and -

with so few peaks uncovered in the region the supply of dust must be meager.
No opportunity was afforded for observing the structure of the névé itself. As
pointed out by Reid, in the paper cited upon page 43, the basal layers of a
glacier may be able to pass a cascade without suffering destruction, while the
upper strata may be destroyed and in large part melted away. This may be the
cause of the poor development of strata in the case of the Yoho, and the absence
of the dirt zones, which should show especially well over the smooth lower half
of the distributary.

In spite of the almost complete obliteration of the strata in the upper part,
the blue bands are shown in great perfection where the ice presses against the
west valley wall. The edges run parallel with the margin and the bands dip

1“ The Influence of Débris on the Flow of Glaciers,’’ 1. C. Russell, Journal of Geology, vol. 111, p. 823,
1895.

le a Me

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 77

down irito the ice at angles of 42°, 48°, 53°, 54°, 56°, and 61°. Upon the walls
of the crevasses they may be seen to curve around into a position parallel with the
valley floor. At the surface the position and approximate thickness of the bands
are indicated by the dirt stripes. Differential movements of the ice, after the
formation of the bands, have given rise to curved, twisted, and contorted patterns
in numerous places towards the center (plate xitt, figure 2).

The fine development of glacial granules and capillaries in the Yoho Glacier
_ has been already noted upon pages 39 and 41. They here attain the largest size
of any seen in the series of glaciers studied and appear to have about the same
amount of orientation near the nose.

7. DRAINAGE.

Owing to the crevassed condition of the main glacier, there is little oppor-
_ tunity for the development of surface drainage streams, the water soon making its
way to the bottomof the bed. Inthe upper portion where the crevasses are not so
_ numerous toward the center, there seems to be too little melting to call for much
surface drainage. The drainage upon the distributary has already been referred
to. There enters its side a strong flow of water from the hanging-valley to the
east (plate xxrx, figure 2), derived from the glaciers lying between Mt. Balfour and
Mt. Gordon. Opposite the head of the rock embossment there is a short strip
of marginal drainage as shown in the map, but the stream is small and the flow
weak. Upon the opposite side of the valley two streams with a brisk flow
enter the side of the Yoho from the broad, glaciated valley noted, while a third
flows down the northern slope from the glacier upon Mt. Collie. Marginal or
surface lakelets were nowhere observed. Augmented with the flow from the
hanging valley, there rushes from beneath the nose of the distributary a torrent
of slightly turbid water, which flows for 4,000 feet over the débris-strewn floor
and enters the side of the Yoho. At the upper end of the line of avalanched
wood described, this stream has cut a gorge, 40 to 50 feet deep, across a ridge
of limestone strata. The gorge extends beneath the present margin of the ice
and, in all probability, has been cut very largely under subglacial conditions,
this part of the valley being under ice when the distributary completely encircled
the rock embossment. This stream flows for 1,600 feet beneath the ice of the
lower Yoho and contributes the bulk of the water which issues from the cavern
at the nose, the North Fork of the Kicking Horse.! This stream, although
shallow at first, is rapid and has a breadth of 240 feet, spreading over the gravel
flat with a network of channels. About one-quarter mile from the exit the chan-
nels are collected into a single one, forming a river of very respectable size,
considering its youth. The water is somewhat turbid, but much less so than that
which ordinarily issues from the Victoria. This is because the drainage from the
glacier itself is so largely diluted with that from the adjoining valleys, derived
in considerable part from the simple melting of snow and carrying a minimum
of sediment. Owing to the volume and velocity of the stream, much of the rock

1 Marked Yoho River upon the latest maps of the CanadianiTopographic Survey.

78 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

débris is carried and rolled down the valley. The channel beds are lined with
coarse rounded boulders, making the fording of the stream, afoot or mounted,
somewhat difficult, especially after a day of rapid melting. In the early morning
the volume and velocity are somewhat reduced.

The dilution of the Yoho drainage, with that from the adjoining valleys,
raises the temperature, as was noted in the case of the Wenkchemna drainage
brook. The last two weeks of August the temperature ranged from 33.8° F. to
35.2° and gave an average of 34.7°. Opposite the Takakkaw Falls, about five
miles from the nose, the stream has descended 770 feet and its August temperature
has been raised from three to four degrees. At Field the temperature of the
Kicking Horse, August 23, 1904, was found to be 44.2°F. at 5:15 p. M. On
August 30, 1905, at 6:30 A. M. it was 39.6°.

8. FRONTAL CHANGES.

In August, t901, reference marks about the nose of the Yoho were indepen-
dently established by Miss Vaux and Mr. H. W. Du Bois, from which it was deter-
mined in r904 that the glacier had receded, in the three years, a distance of 111
feet (August 16, 1901 to August 18, 1904), or at the average rate of 37 feet a
year. Measured to the block of ice which had until very recently constituted
the nose, the distance was 92.1 feet, of which 23 feet was for the year 1903-4,
reducing the average to about 31 feet for the three years. Between August 18,
tgo4, and August 31, 1905, the retreat was found to have been but 9 feet. The
average annual retreat for the four years 1901 to 1905 has been 30 feet. Ata
second station upon the western side of the drainage stream the retreat from
‘August 17, 1904, to August 31, 1905, was 4.6 feet. From these meager data
it seems that the Yoho is having its retreat checked.

g. Former AcTIVITY.

a. Moraines. Lack of time prevented any careful survey of the entire Yoho
Valley from the glacier to where the valley joins that of the main Kicking Horse.
No coarse moraines of the type described for the Victoria and Wenkchemna
were seen and their absence is easily accounted for by noting the absence of steep
cliffs about the glacier and its névé fields, plate xxviu1, figure 2. In passing up the
valley the trail crosses two steep ridges, densely covered with vegetation, but
these appear to be of the nature of mountain spurs, or rock slides, from the
western side of the valley. About 1,000 feet from the present nose there is an
interesting display of modified ground moraine, lying mainly upon the eastern
side of the stream (plate xxviu, figure 1). The structures consist of low
knolls and crescentic ridges, connected with the weak lateral moraines by faint
ridges. Six series may be made out, concentrically placed and with their con-
vexities directed down stream, diminishing in height and distinctness toward
the glacier. The ridges vary in height from one to twelve feet, the longest
being in the form of a semicircumference with a radius of twenty feet. The
ridges possess the smooth, rounded outlines of drumlins, but lack their profile

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XXVIII.

Fic. 1.—Knolls and ridges of ground-morainic material in front of Yoho Glacier. Suggestive of

frontal moraines of latest ice sheets in North America and Europe

Collie. Baker. Gordon. Balfour. Waputik snowfield.

Ses
: SSS
Tak
- ~

-

teat
“a

~ 4

ai

é

b

Fic. 2.—General view looking up Yoho Valley, showing Wapta and Waputik snowfields. Photographed, 1904, by
Arthur O. Wheeler, from summit of Mt. Wapta (9,960 feet). Looking north-northwest.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XXIX.

Fic. 2.—Hanging Valley, head of Yoho, lying between Mts. Balfour and Gordon.
August, 1904.

a Re er

an poy

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 79

and arrangement. They are like drumlins in that they consist of ground-morainic
material, with but a thin dressing of gravel and sand. They differ essentially
from drumlins in that their longer axes are parallel with the former ice margin,
instead of at right angles to it. If composed of stratified sand and gravel they
would be kames. In short they have the structure of drumlins and the form
and position of kames. They so much resemble in miniature the knolls and
ridges formed by the last great ice sheet in its retreat from the United States
and Germany that their origin becomes of especial interest. They have evidently
been formed by the glacier, with its nose consisting of a series of lobes, plowing
into the ground moraine previously deposited upon the valley floor. A retreat
occurred and then an advance, falling a little short of the first position, by which
a series of mounds and curved ridges was pushed up. This was repeated.
at least a half dozen times, each advance being somewhat weaker and falling a
little short of the preceding. Finally the glacier advanced over the entire series,
but overrode them so lightly that instead of being destroyed they were simply
smoothed and rounded. In melting back the last time, either from the ice itself,
or from the subglacial drainage, or from both sources combined, a thin layer of
sand and gravel was deposited over the structures. Had the ice lobes been larger
the ridges would have appeared more nearly straight, as we find them in the case
of the Pleistocene deposits.

b. Plucking action. About the nose of the glacier, as has been already
noted, and upon the eastern side, toward the rock embossment, there are numer-
ous illustrations of the bodily disruption of the rock strata, to which the term
plucking is applied. Conditions are most favorable for this action when the
strata are thin bedded and jointed, when the strike of the strata is transverse
to the glacier and the dip is down stream. Under these conditions the rock layers
are ripped off and the bed lowered with relative rapidity, the rock fragments
being pressed into the ice and moved forward to assist in further work of a like
nature. In this way portions of the bed are much roughened by ice action,
instead of being smoothed. The edges and corners of the strata which were able,
at the last, to resist the action of the ice will be found to be rounded more or less.
A mountain spur, lying between the nose of the glacier and Mt. Balfour was
overridden by the ice and experienced this action upon an extensive scale. The
mountain is made up of curved, concentric strata, the upper layers being of a
dark limestone. Upon the southern side these layers dip to the southwest at an
angle of about 30°. In passing over the peak from north to south many feet of
strata have been removed, those able to resist the action forming a succession of
steps upon the steeply inclined slope. One only of these steps is shown in plate
XXvIil, figure 2, behind the hard crest of which the loose fragments have collected,
while upon either side they have been swept clean by the ice. This furnishes an
illustration of what is known as a ‘“‘knob and tail’? phenomenon. If the com-
bined height of the successive steps were ascertained we should have a figure
representing the minimum amount of this plucking action upon the southern

80 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

slope, if not over the crest of the peak. It is well to note that this was done
with a relatively thin sheet of ice, while in the valley bed, with some 3,000 feet
additional of ice thickness, the result, under equally favorable position of the
strata, would have been correspondingly greater. It seems very probable that
the peculiar form of the peak Trolltinder (9,414 feet), just south of Balfour
(plate xxvitl, figure 2), is due to similar plucking action.

c. Yoho Valley. There is abundant evidence that the entire valley, from
the Kicking Horse at Field, was occupied by an immense ice stream, seventeen
miles in length, which served as a tributary to the ancient Kicking Horse Glacier,
of Pleistocene time. It, in turn, received short tributaries from the adjoiming
valleys and mountain slopes. The valley was filled with ice to a depth of 1,500
to 3,000 feet above the valley floor, by which the lower portion was transformed
into the characteristic U-shape, seen best from below. When viewed froma height
as in plate xxvul, figure 2, the more flaring walls of the upper portion become the
more conspicuous. The valley seems to have had the same general history as
that given for the Lake Louise district, page 61. Being a longitudinal valley
of the Rocky System, it was originally a trough between mountain folds, or a
great crevasse, which collecting the drainage of the region was cut into a V-
shaped valley by the joint action of running water and the weather. With the
coming of the glaciers, the valley was occupied by ice and the lower one-third
to one-half deepened and broadened, while the upper portion, as high as the ice
could operate, was simply smoothed and subdued. Spurs were cut off and
faces exposed to the action of the ice were grooved and fluted, polished or
scratched.

A series of typical hanging-valleys occur along the Yoho beginning with that
of the distributary, the floor of which is not yet uncovered. This ice stream has
not been able to lower its bed as rapidly as has the main Yoho, and when melted
back to the head of the rock embossment there will be exposed a side floor at a
higher level than the main floor. However hanging-valleys, in general, may
arise, this one seems certainly due to the differential effect of the two streams upon

their respective beds. To the right of the distributary there extends a hanging- .

valley to the northeastward between Mts. Gordon and Balfour, still occupied
by two glaciers, which appear to have built conjointly a double frontal moraine
(plate xxx, figure 2). This valley has a double floor, of which time did not permit
anexamination. From the photographs taken there appears to be a lake, occupy-
ing a rock-basin upon the lower level. About 34 miles down from the nose of the
Yoho the Twin Falls drop into the valley from the floor of a hanging-valley
coming in from the west. The falls are 310 feet in height, but their crest (6,500
feet) is 1,050 feet above the floor of the main valley opposite. Five miles down
from the glacier are seen the Takakkaw Falls, in the center of plate xxvu1l, figure
2, the crest of which is 1,200 feet above the valley floor. The valley floor from the
glacier to these falls descends about 770 feet, or at the average rate of 154 feet
to the mile. These figures, based upon data supplied by Wheeler, indicate that
the main valley has been lowered from 1,000 to 1,200 feet more than the tributary

YA4ON D2aUuday

“43-809'01
pleuogJis “3

“33 166'6
Meey Jeuiusa,

‘MAZUAHS—ADGATMONN OL SNOILOAGIALNOD NVINOSHLINS

‘XXX GALVId

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 81

valleys. With the effect in mind of relatively thin ice sheets upon the neighbor-
ing peaks, the writer is quite prepared to admit the sufficiency of glaciers to pro-
duce hanging-valleys, when the ice is deep, concentrated, and operates for a long
period over stratified formations.

CHAPTER VI:
ILLECILLEWAET GLACIER.
t. GENERAL CHARACTERISTICS.

Passinc from the Rockies westward to the Selkirks,we find much evidence
of the increasing precipitation; one of the first to which our attention is unpleas-
antly called is the tantalizing number of snow-sheds which obstruct our view.
The mountains are much more completely forested than we found them in the
Rockies and nearly everywhere the valley slopes are scarred with avalanche
tracks. From extensive snow-fields (plate xxxu, figure 1) hundreds of tongues of
ice descend to much lower altitudes than is possible in the Rockies, with their
slighter snowfall. The largest of these ice tongues to be seen from the railway
is the so-called ‘‘Great Glacier,” or Ilecillewaet,! the glacier that gives rise to the
“rushing water.” Owing to the ease with which it may be reached from the
station it has been visited by more people than any other glacier in the two
Americas, although, so far as known, it was not seen by the eye of white man
until the year 1883. In that year it was discovered by Major Rogers, who
was in search of the railway pass which now bears his name. It was originally
named Agassiz Glacier by Ernest Ingersoll,? but this name has since been
transferred to one of the commensal streams of the great Malaspina, in Alaska.

The glacier lies just to the south of Mt. Sir Donald (10,808 feet), between it
and Glacier Crest, and as a great tongue of ice spills over the rim of the extensive
collecting basin enclosed between Mt. Sir Donald, Mt. Macoun (9,988 feet), Mt. Fox
(10,572 feet), and Mt. Lookout (8,219 feet). See maps, plate xxxand xxx. The
glacier flows to the northwest, 1s but 14 miles in length, and in this distance tapers
from a mile in breadth to a sharply pointed nose. The axis of the glacier is slightly
curved, with its convexity turned toward the southwest. Lying in a broad valley
with this exposure, and with no covering of débris, the glacier receives the full
effect of the noonday and afternoon sun. In spite of this the nose attains the
altitude of 4,800 feet, or 870 feet lower than the Yoho. Since the collecting
areas are very similar in size, this difference must be due mainly to the differences
in the amount of snowfall received by the two regions. ‘The latitude of the nose
of the Illecillewaet is 51° 15’, being nearly a third of a degree farther south
than the Yoho. From the névé line, with an elevation of about 7,500 feet,
the glacier descends 2,700 feet to the nose, or at the rate of about 2 2,000 00 feet to the

1 This name is pronounced as noes it were rected ‘Tlly- silly: -wet, with ine stress upon Pe middle
syllable.
2‘*The Rocky Mountains as Seen from the Canadian Pacific Railway.’’ Science, vol. vil., 1886, p. 243.

oO
bo

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

mile. The greater part of this drop is in the upper half; the glacier descending
from the rim of the basin in a steep cascade, by which the ice is shattered and
its original structure destroyed (plate xxxu, figure 2). In its short length the
glacier receives no tributaries, but instead has a series of short distributary noses
perched high up along the eastern line of cliffs leading to Perley Rock (plate xxx),
with elevations ranging from 6,450 feet to 7,000 feet. No data exist for estimates
upon the thickness of the glacier, but the greatest thickness of ice probably
occurs below the crest of the cascade and may amount to several hundred feet.
The marginal ice is very steep upon the eastern side and for a quarter of a
mile back from the nose upon the western side, so that it can not be easily
ascended. Toward the nose the general inclination is about 30° to 35°, dimin-
ishing to 20° and less, so that one may mount the glacier from the nose for a short
distance. The névé may be reached by a rather rough climb, either around to
the east by Perley Rock, or by ascending to the depression between the glacier

and the steep left lateral moraine, keeping a sharp lookout at first for rolling

rock from the eastern face of the moraine. About the main nose, upon the
eastern side, there occur a number of minor noses, shown in plate xxx and plate
xxx, figure 2; and also a sharply defined, trough-like depression in the surface
of the ice, from 200 to 300 feet across and tapering up stream for a considera-
ble distance. This depression appears in all the photographs taken since 1887
(plates xxxvi, Xxxvu, figure 2), since which date about 400 feet of the floor
immediately beneath it have been uncovered without disclosing any cause for
the depression. In plate xxxm, figure 2, it appears to be continued up the glacier
to the cascade and may have its origin there in some obstruction of the bed by
which the ice is diverted to either side and left thinner in the lee. Just to the
left of the nose there has been uncovered, since 1898, a mass of bedrock for a
distance of 400 feet, its more rapid radiation of heat accelerating the melting of
the ice resting upon it. The rock consists of a brownish, schistose conglomerate,
furnishing an interesting display of glacial features’ to be described in another
section of the chapter (page 95). This is the only bedrock observed in the
floor of the valley from the glacier to the station.

2. NOURISHMENT.

Meteorological data at the station of Glacier House are, unfortunately, very
meager. The average precipitation for the five years available amounts to 56.68
inches, of which 43.7 inches (36 feet and 5 inches), or about 77 per cent. of the
whole, fell as snow.! Over the névé region practically this entire amount
would be available for glacier formation and, as snow, would represent about
47 feet. The retangular snow-field extending from Mt. Sir Donald to Mts. McCoun
and Fox is about five miles long, by two miles broad, and hence contains about
ten square miles of collecting area (plate xxx). Of this area about two-thirds,
or six to seven square miles, drains northward and feeds the Mlecilliwaet,
while the remainder moves southward and nourishes the Geikie Glacier, which

1 The Selkirk Range, A. O. Wheeler. vol. 1, p. 414.

el ie

i ie. ee

|

|

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XXXI.

a
S
=
£
x
es

Plate No./

Plate No. 2
Plate No. 3
Plate No. F
Plate No.5
Plate No.6
Plate No.7

Position of plat:

, Suly 3/st., 1899

Position of plates, Sept. 5th, 1899

Average Slope of surface at this point 22°0'

GLA CLE

BHaLE OS
Alder bushes? Sp.
Re Poh OfLor
Stunted scattered

evergreens, 3ft.high
average 29 years ok

oe S
Ala ig e

oP)
Deep depression In surtace of glacier

yes,

39-8.

aes

1D.
Y
34)

Lala’
SD:
a°8S8
Bao ss &
} Alder bushes '
3 PORE
Beg FI BiG

aay ete VA EE
eS Ga
AS aa

” - Y. E—E
SA SAS ~ £250 jy,
Alger. bushes Pye SE line 334 eS
"3 ERTS old: qT Sorder oF ice
PES OOS 7 jin 1888

“BESS
& OSs

SHALL I
ROO EAS
Si ee

<>
POSITIONS OF MARKED ROCKS. o

A .—Large polished boulder on top of moraine, (Aug. ’90.)
#.—Triangular block. (Vaux, 1898.)

C.—Round boulder. (Vaux, 1898.)

D.—Double rock. (Vaux, 1898.)

£.—Large rock in border moraine. (Vaux, 1887.)
#—Large angular block. (Not marked.)
G.—Pyramidal block. (Not marked.)

-—Tongue of ice.

/.—Large bottom rock. (Not marked.)

¥.—Large bottom rock. (Not marked.)

&.—Large bottom rock. (Not marked.)

£.—Large bottom rock. (Not marked.)
M.—Small rounded boulder, (C. O. E. 1895.)
N.—Large bottom rock. (Not marked.)

O.—Large boulder. (Not marked.)

&.—Two boulders. (Oct. ’g5.)

Q.—Long flat boulder. (Ice 4th Aug. end of snout, ’19.)
&.—Polished boulder of green quartzite. (Not marked.)
S.—Long, flat boulder. (16 ft. from nearest ice, ’90.)
7.—Rocks tarred by W. S. Green, Aug. 13, 1888.
U.—Egg-shaped boulder, (Striped with paint.)
I’.—Marked rock.

W.—Large rock, point of view of test pictures.

Oc es Alder bushes Be o.
Re : 3B 7 OL
Oe Oat. Dea aR :

VSI ad We ES
ty Alder bushes & ran

A, s 4 PEED

xs
The points on border of ice marked by a star were determined trig- “> Ses 31 =
Baamcatcallye a = ¥ e ex Alger bushes witha few evergreens Bras
EON a ae on Ce Se

100 0 100 200 300 400 500 feet
fi ee ee

Tongue and Moraines of the Illecillewaet Glacier, August, 1899. Published through the courtesy of George and William S. Vaux, Jr. From the
Proceedings of the Philadelphia Academy of Natural Sciences, 1899.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS 83

flows westward between the Asulkan and Dawson ranges. The collecting
basins for the Illecillewaet and Yoho glaciers are almost equal in size, but the
estimated precipitation over the former is 41 per cent. greater than over the lat-
ter, giving a correspondingly greater volume of ice to be disposed of. This ena-
bles the Ilecillewaet to attain a much lower altitude, as previously pointed out.

The mean elevation of the surface of the névé lies between 8,000 and 8,500
feet above sea level; ranging from 7,500 tu 9,500 feet, not essentially different
from the Yoho névé. In the central portion the névé is much crevassed, from
which one may infer that the thickness of the snow and ice is not great in the ba-
sin. Thesurface is covered with parallel ridges and furrows, probably result-
ing from the rippling action of the wind while the snow was being deposited.
These ridges when frozen render the walking somewhat difficult and treacherous.

3. Moraines.

a. Surjace débris. Because of the wide extent of the névé field and the
absence of precipitous cliffs about the margins, there is very little opportunity for
the névé to acquire any rock débris. There is no evidence that any considerable
quantity is gathered from the bed and carried subglacially, or englacially. It
may be that the basal layers in the névé region are sluggish, or even stagnant,
and that only the upper layers are being pressed out over the rim of the basin.
The result of this lack of débris in the névé is that the general surface of the
glacier, as in the case of the Yoho, is unblemished with rock fragments, but is
somewhat soiled from wind-blown dust concentrated over its surface by melting.
Some dust wells occur sparingly and a few poor examples of glacial tables.

b. Left lateral moraine. Along the western margin of the glacier fed by
the ordinary atmospheric agencies of rock decay operating upon the cliffs of
Glacial Crest and Mt. Lookout, there has been built up a high, sharp-
crested, left lateral moraine. The angular rock fragments are supplied
mainly by two prominent débris cones, which have formed upon the eastern
face of Glacier Crest and which rest with their bases upon the margin of the ice,
the forward movement of which has distorted the cones downstream, plate xxxm.
In the summer rocks may be seen coming down these slopes, one block starting
others and these still others, until a regular cannonading is in progress. As one

tock collides against another with terrific force, small clouds of dust arise and

we have simulated not only the roar but the smoke of battle. John Muir has
given us a graphic description of the streaks of fire to be seen when these ava-
lanches occur at night. In the way previously described for the Victoria (page
47) this material arranges itself in the form of a sharp-crested ridge perhaps
100 feet above the margin of the ice and 150 feet above the valley floor. The
melting of the ice core upon the inner slope makes it steep and unsteady, while
the outer has settled into a condition of more stable equilibrium and 1s being
slowly covered with vegetation. Upon the inner slope occasional slides of the
rock veneering occur, by which the ice core is temporarily exposed. The materials
from the upper part and sides of the scar produced roll and slide down, collecting

.

84 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

at the bottom and growing slowly upward until the ice is again completely
covered. In this way the material is being more thinly spread over the morainic a
ice core and its melting accelerated. The great height which the moraine has
been able to attain in its lower part is due to the fact that the glacier is unpro-
tected by débris over its general surface, and the differential melting i is so much ‘
greater than it would be if the surface of the ice were protected as in the case of ;
the Victoria.
The material of which this moraine is composed is principally quartzite and
a silvery (sericitic?) schist, with a binding of glacial sand and clay. Occasionally
boulders are scratched, and are very generally bruised, with their edges and
corners rounded from rough treatment they have received upon the débris cones
and the moraine. About one-third of the distance along this moraine there occur
two elongated depressions, exactly in the crest of the moraine. The larger and
better defined of the two is 125 by 50 feet and 6 to 8 feet in depth (plate xxxu, f
figure 2). They are seen from a distance most plainly when the sun’s rays — .
strike somewhat obliquely, permitting the sides to cast shadows into the bottoms 4
of the depressions. Attention was first called in print to these depressions by
the Vaux Brothers in 1900,! showing very plainly as they do in their early
photographs. Before these depressions were visited it was thought by the writer
that they might represent the sites of drained lakelets, similar to those found
upon the Victoria, but the explanation suggested by the above investigators |
seems to be the correct one. The moraine appears to have been formed of two”
ridges, laterally welded together, and these depressions appear to be spaces
left where the two ridges did not quite meet. Owing to the sliding of the débris
upon the inner slope of the moraine the basins are being obliterated and will
eventually completely disappear. ‘
Including the débris cones this lower portion of the left lateral is some 4,000
feet in length, rises to less and less height above the general ice level, and gives —
out for a short distance, where the ice abuts directly against the aha eee |
of Glacier Crest. Up to a height of 40 to 50 feet patches of morainic matter have _
lodged upon the rock shelves. One-quarter mile beyond, a crested ridge again
makes its appearance composed of materials derived from Mt. Lookout, just to
the southeast. The moraine is largely made up of quartzite and schistose boul- a .
ders bound together with sand and clay and supporting a sparse growth of
mosses and Alpine plants. The ridge rises to a height of 35 to 40 feet-above
the ice, curves around to the eastward, becomes reduced in height, and disap-
pears under the névé, which is strewn with rock fragments derived from the
cliffs of Mt. Lookout. eae
c. Terminal moraines. From the lower extremity of the left lateral flere a
curve around into the valley two lower ridges, of the nature of terminal moraines. |
The inner and younger of these forms an inconspicuous ridge, from 6 to ro feet a
high, and passes into the terminal moraine at which the glacier was found to be —

\

‘ The G reat Gi ucier of the Tikieiliwacn Geore and William S. Vaux, Jr., Appalachia, vol. 1x,
1900, p. 164.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XXXII.

Sir Donald. Illecillewaet Glacier. Asulkan Glacier. Mt. Bonney and glaciers

Fic. 1.—General view of peaks and snowhelds from Rogers Peak (10,536 feet),
Selkirks, looking southeastward. Photographed in r902, by Arthur O. Wheeler.

Illecillewaet névé. Asulkan névé,

Fic, 2,—General view of Illecillewaet Glacier from summit of Mt. Eagle, elevation 9,353 feet, distance two to three
miles. Photographed in 1903 by C. F, Johnson.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 85

standing in 1887. From their photographs taken in this year the Messrs. Vaux
have established the position of the ice front with reference to a very large
boulder resting upon this moraine. This terminal ridge swings around to the
north and connects with the right lateral, which is of greater age and, in the lower
part at least, has lost its ice core. See plates xxx and XxxI.

d. Rightlateral moraine. The ice has withdrawn from this moraine a dis-
tance of 400 to 500 feet, leaving a somewhat subdued boulder slope and a low
‘ridge. This becomes higher and steeper as we approach the quartzite cliff which
intercepts it about one-half mile back from the nose. Here the moraine is
double, an older one lying just outside and parallel with it. Forest trees have
taken possession of the crest and outer slope. The rocks in the right lateral
are similar to those in the left and are found to be in the same condition of
being rounded and bruised. Upon the rocky ledges, which carry the distribu-
tary noses referred to, there is spread out more or less morainic material, some
of which has been assorted by running water. These ledges of quartzite have
been much glaciated, plucked and extend up toward Perley Rock (7,898 feet).
For about a quarter of a mile there extends an upper double moraine to the
southeastward, where it disappears under the snow. The material consists of
rounded boulders of quartzite and chloritic schist, with a filling of glacial sand.
An inspection of the map shows that there is a correspondence in the arrange-
ment and position of the lateral moraines; there being in both cases, a higher and
a lower portion, separated by quartzite ledges, carrying only a sprinkling of
morainic material. Since the cascade in the glacial stream lies between these
exposures of quartzite it is probable that the ledges are continuous beneath
the glacier; that they have proven too hard for the glacier to remove, and so it
is compelled to cascade over them.

e. Boulder pavement. Between the terminal moraine and the present nose
of the glacier there has been uncovered since 1887 a broad boulder belt, about
500 feet across. This consists of ground-morainic material in large part, with
the rock fragments which were carried englacially, or supraglacially, and de-
posited as the ice front receded. These boulders have been overridden by the
ice so lightly that they have not been disturbed, and yet a number of them were
glaciated while in their present position, forming what is known as a “boulder
pavement.” About the present margin of the ice, boulders are being continually
uncovered which are being subjected to the same action. The ice presses against
the up-stream face of the boulder, and, either because of the warmth of the stone,
ormore probably because the melting point of the ice isreduced by the pressure, or
because of both these agencies, an inverted trough, or fluting, is produced upon
the under surface of the ice, having the form of the rock. In plate xxxiv we have
these flutings shown in different stages of formation; in the last case (figure 2)
the stone was estimated to lie 70 feet back from the ice margin and was under
probably 50 feet of ice. Photographs taken some years ago of the “‘ice grotto”’
show that it was a feature of this kind produced by an unusually large rock.
If the pressure were sufficient, the ice would settle in promptly upon the lee side

86 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

of the boulder and it would be glaciated not only upon the stoss and upper sur-
face, but upon the lee side as well. A certain relation must exist between the extent
of lee-side glaciation and the thickness of the ice, which, if known, would give
some data for estimating the maximum thickness of certain Pleistocene sheets.

4. CREVASSES.

The crevassed condition of much of the névé, especially that in the main
direction of flow, has been noted. Faultings and dislocations occur, disclosing
the stratified nature of the névé and subjacent ice. This crevassing is due,
apparently, to irregularities in the bed, rather than to differential motion, and
indicates that the ice here attains no great thickness. About the margin of the
névé field there occur, here and there, breaks where the névé has withdrawn
from a portion still clinging to the rocky wall. These are the bergschrunds,
described upon the Victoria and Wenkchemmna glaciers (pages 22 and 67). Ona
line between Perley Rock and the western end of Mt. Lookout, there opens up a
series of transverse crevasses as the ice begins its descent into the valley and its
velocity is accelerated (plate xxxt1, figure 2). The ice is unable to yield to the ten-
sional strain and forms long V-shaped gashes at right angles to the stress. These
become convex down stream because of the more rapid central movement of the
ice. Conditions are here favorable for the formation of Forbes’ dirt bands (page
50), but the ice soon plunges over the quartzite ledges, is shattered in every
direction and all structure lost. Upon the crest of the cascade a network of
crevasses opens, dividing the ice into irregular angular blocks. These become
melted upon all sides and assume the form of pinnacles and steeples,—seracs,—
displaying beautifully the stratified structure of the ice (plate xxxv, figure 2).
Reaching the bottom of the cascade these blocks are jumbled together, many of
them completely melted and the remainder frozen together into a great ice con-
glomerate (plate xxxv, figure 1). As pointed out upon page 44, it is quite
conceivable that some of the basal layers might be able to descend the slope
without having their structure destroyed. F é

The rapid central movement of the ice, due to the high average slope of the
bed, gives rise to a very complete system of marginal crevasses, extending in-
ward and upward, and showing conspicuously when the glacier is seen from a
height. In the lower third the ice does not feel the restraint of the valley walls,
spreads laterally because of its own weight, and thereare opened longtitudinal and
radial crevasses, some of them extending to the margin of the ice. As the
surface is continually lowered by melting, only the bottoms of some of the
shallower crevasses remain, and these appear simply as short gashes in the other-
wise smooth surface.

5. Ice STRUCTURE.

There is evidence upon the glacier’s left, back a short distance from the nose,
that the stratification in the basal portion of the glacier is not completely de-
stroyed in passing the cascade. Traces of the stratification may be seen dipping

wy a ope eee, oe a

:
-
3
4
2

ee

Shade

a ine dalla on

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER,. PLATE XXXIII.

Map of the Selkirk snowfields and glaciers, by Arthur O. Wheeler. Reproduced through courtesy of Canadian Topographic Survey,
Department of the Interior. Approximate scale t inch = 11%4 miles. Contour interval 100 feet. Reference datum mean sea-level.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER PLATE XXXIV.

Fic. 1.—Beginning of subglacial fluting by pressure-melting, Ilecillewaet Glacier,

August, 1905.

Fic. 2.—Subglacial fluting by pressure-mellting, Ilecillewaet Glacier, August, 1903. Photo-
5 & bY I 5 g 903
graphed beneath the glacier and at an estimated distance of 70 feet from the
rock responsible for the fluting.

Fic, 3.—Roches moutonnées near nose of Ilecillewaet Glacier. Motion of ice from
right to left.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 87

inward towards the center, this portion of the ice possibly having received less
severe treatment than that nearer the center of the channel. In the case of the
Yoho the question arose (page 76) as to whether the stratification was obscure
because of its destruction by a similar cascade, or because of its original weak
development. In the case of the Illecillewaet there is sufficient bare peak and
rocky crest exposed to supply the broad névé field with successive layers of wind-
transported dust and a very perfect stratification results from the concentration of
this dirt at the surface of suécessive deposits (plate xxxv, figure2). Thealmost
complete lack of stratification about the nose, where it should be well displayed,
along with the dirt zones, must in this case be ascribed to the cascade. The dust,
originally concentrated between the strata, is brought to the lower margin of the
ice, where it collects and drips as black mud (plate xxxtv, figure 1) over the val-
ley floor. - The color is due to the presence of organic matter, of which there is
enough present to render the material offensive, when set away damp in a warm
room. Four determinations of the organic matter present in material collected
iN 1903 gave 16.75, 11.25, 10.68, and 17.23 per cent., or an average of 13.98 per
cent.

It seems impossible that the coarse stratification of the ice could be so com-
pletely destroyed and the finer lamination preserved so perfectly and continuously
as we should have to suppose if we referred the blue bands to the original lami-
nation of the névé. As pointed out upon page 44 and as shown in plate x1m1, figure
1, the blue bands, with the superficial dirt stripes, are very clearly shown about the
nose of this glacier, from 15 to 36 being counted within the distance of a foot.
They are approximately parallel with the valley floor and would probably con-
form with the strata, providing the latter were present. They dip inward, in
general, about the nose at angles of 3° to 8°, but in places are inclined outward
by this amount. As soon as the ice begins to experience pressure from the
moraines, or the valley walls, the blue bands become more and more steeply
inclined, beginning with angles of 8° to 16° and increasing up the valley to 70°
to 75°. The relation of bands in the ice to the stones which are fluting the
under surface (plate xxxiv,) has been discussed upon page 44.

The glacial granules, with the melting phenomenon described in connection
with the Victoria Glacier, are well shown about the nose. In size they stand
next to those of the Yoho and range from the size of hickory nuts to that of hen’s
eggs. They are limited largely to the blue bands, or the white seams that
intervene, but in cases are seen to cut across from one to the other. As the
granules assume distinctness the blue bands become more and more obscured.
Between the granules there is developed, under suitable melting conditions,
a very perfect and beautiful network of capillaries described upon page 41.

6. DRAINAGE.

a. Surface and marginal drainage. Upon the nights of September 7-8 and
8-9, 1904, the minimum temperature of the ice was measured by inserting a ther-
mometer to the depth of twelve inches in the face of a crevasse near the nose.

88 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

The two readings were 32.0° F and 31.9°, indicating that this portion of the glacier
was very near to its melting temperature. Before liquefication, however, can
occur a large amount of heat must be rendered latent, when water is produced with
the same temperature as that possessed by the ice just before melting. The
heat necessary for this conversion is supplied in the main directly from the sun

but in part by that of the atmosphere, rain, friction and pressure of the ice
against the valley floor and sides, and whatever heat may be reflected, radiated,
or conducted, from the same source. Almost wfthout exception the surface
drainage was found to have a temperature of 32°, varying but a very small
fraction of a degree from this. The surface ablation is rapid during the summer,
owing to the exposure of the ice to the sun and the absence of protective débris.

As would be expected from the greatly crevassed condition of the ice, no
surface streams of any size can develop, either over the general surface or along
the margins. Small streams flow directly to the ice margins and cut channels,
in a few cases, to the depth of a foot. This water may be absorbed at once by
the loose materials covering this portion of the bed, or it may collect and give
rise to scant marginal or subglacial drainage streams. These flows soon cease
when the sun drops behind Glacier Crest and throws the glacier into shadow.
In 1904, for a distance of some 500 feet, a small drainage stream was found
between the left lateral and the adjoining ice slope. Since the ice of the glacier
is here continuous with the morainic ice core, this stream was a surface rather
than a marginal one. From the lower end of this moraine there occurred also
two small flows of water, apparently coming from englacial channels in the
moraine. After a six hours’ rain, September 8, 1904, during which 0.58 of an
inch of rain fell, the water of these streams was rendered muddy, while the
turbidity of the main glacial flow was not perceptibly affected.

b. Terminal drainage. Upon the eastern side of the glacier, drainage
streams leave the ice margin in the neighborhood of the elevated distributary
noses, as shown upon the map. The ice here rests upon bedrock and the streams
have sought the lowest depression, there being three such which are draining this
portion of the glacier. The westernmost of these breaks into a network of streams
upon emerging from the ice, and cascading over the rocky ledges again enters
the side of the glacier, to reappear at the nose. The two other streams have
cut gorges so to 60 feet deep across the hard strata, showing that they must
have been at work for considerable periods, probably as subglacial streams.
However, the high velocity of the water and the sharp glacial sediment enable
it to work very effectively, even upon quartzite. These two cascade over the
ledges and unite some 1,600 feet down into a single stream, which receives
tributaries from the slopes of Sir Donald. Just outside of the right lateral
moraine the stream divides into numerous branches, which reunite upon the
gentle gravelly slope and form what is known as‘a “‘braided stream.’ LEast-
ward of the upper, right lateral moraine, between it and Perley Rock, there is a de-
serted gorge, similar to those now being formed, partially filled with gravel and
morainic matter. The drainage has been deflected westward to a lower level.

aE ee

robs

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XXXV

Fic. 1.—Kegelation of ice blocks at foot of ice cascade, [lecillewaet Glacier, September, 1904.

Fic. 2.—Stratification in upper part of Ilecillewaet Glacier.

Copyrighted, 1902, by the Detroit Photographic Co.

ee

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 89

In 1904 there issued from the main nose of the glacier five drainage brooks,
the main one lying to the west of the nose and receiving the drainage from the
left lateral moraine. This streamfollows the inner curve of the terminal moraine
and cuts across it at a point opposite the nose. The other four streams form a
network over the boulder pavement, unite into a single stream, which makes its
way across the terminal moraine and joins the main flow. About 1,200 feet
from the nose of the glacier (1904) this drainage brook from the lower and west
side of the glacier unites with the strong flow from the eastern side of the valley
and together they form the Illecillewaet, or ‘‘ Rushing Water,” a tributary of the
Columbia. The average flow from the glacier is not much different from that
of the Victoria and Wenkchemna, but considerably less than that from the
Yoho, which collects the drainage from a larger territory. Based upon the
estimated size of the névé area and the average annual precipitation, the average
flow for the months May to September, inclusive, should be about 65 cubic feet
per second. Owing to the somewhat larger drainage area and the more rapid
midsummer melting the flow seems greater than this in July and August. The
water, as it issues from beneath the nose, is only slightly turbid compared with
the Victoria, indicating a small amount of subglacial erosion. The color becomes
slightly green in the combined streams and still more so after it has received the
Asulkan drainage farther down the valley. With the loss of sediment, which is
gradually stranded here and there, the water assumes a bluish tinge, except
where lashed into foam.

c. Temperatures. During the last week in August and first week in Sep-
tember, both in 1904 and 1905, some 30 observations were made on the tempera-
ture of the drainage at the nose. The average temperature, taken at various
times of the day, was found to be 33.1° F., with a range from 32.4° to 33.8°.
That from the eastern side of the valley, taken just under the bridge on the
trail, gave an average of 39.9°. One-third of a mile down the valley, at the lower
bridge across the stream, the average temperature of the combined streams
was 38.3°, ranging from 34.9° to 40.6°._ Where the Illecillewaet passes beneath
the railway, having received the Asulkan brook, four observations upon the
temperature gave an average of 39.7 °.

7. FORWARD MOVEMENT.

As early as 1888 observations were made by Rev. W. S. Green to determine
the forward movement of the glacier. On August 13, he set three poles in the
ice by boring holes with an auger, the distance from the nose not being given.'
These were visited upon the 25th of the same month and were found to have
fallen, owing to surface melting. The holes were, however, found and the
poles reset for measurement. The distances moved in the twelve days are
recorded as follows: ‘‘No. 1 pole, near moraine, 7 feet; No. 2, further out, ro feet;
center of glacier, 20 feet.’’ For the middle of the glacier this gives an average
daily motion of 20 inches. About the margin of the ice Green, at the same time,

1 Among the Selkirk Glaciers, 1890, p. 218.

go GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

tarred a number of boulders in closest proximity to thé margin, which rocks —
could still be indentified in 1905. (See rocks marked T, plate xxxt.) ms
In 1899 George and William Vaux set a line of 8 steel plates across the glacier, _
some 1,400 to 1,500 feet back from the nose, their line lying somewhat obliquely 4
to the main axis of the glacier (see map). The average surface slope was given —
as 22°, and the distance across the glacier along the line of plates was 1,720 feet.
A base line was laid out along the higher portion of the right lateral moraine,
229.5 feet in length, and the plates located by triangulation, July 31, 1899. Bear-
ings were taken upon the plates August 11,1899,and September 5, 1899, the lat- _
ter work being done by Messrs. H. B. Muckleston and C. E. Cartwright, of
the Canadian Pacific Railway. One year later (August 6, 1900) the plates were
again located by the Messrs. Vaux and their forward movement for the 372
days determined. The following table is based upon their data published in
Appalachia, vol. 1x, 1900, p. 160, and the Proceedings of the Academy of Natural ia
Sciences of Philadelphia, March, rgot, p. 215.

TABLE VII,

OBSERVATIONS UPON THE LINE OF STEEL PLATES SET ACROSS THE
ILLECILLEWAET GLACIER, JULY 31, 1899.
(Total Distance across Glacier along Line of Plates 1,720 feet.)

July 31 to Aug. rz, 1899. Aug. 11 to Sept. 5, 1899. Soe seco, July 31, Poee Auug.i6;
Distance ee =
from

Plate. East Total motion, | Average Total motion, Average Average Total motion, Average

edge. 11 days. daily motion. 25 days. daily motion, | midsummer 372 days. daily motion.
motion, 36 ds.

Feet. Feet. | Inches. Feet. Inches, Inches. Feet. Inches.

£ 205 3.54 3.86 2.62 1.26 2.506 88.6 2.86

a 500 3183 3.64 8.67 4.16 3.90 I24.0 4.00

3 605 6.25 6.82 S75 4.20 5.51 139.7 4.51

4 750 6.21 6.77 — | aa On 181.0 5.84

5 845 5.96 6.50 Tee AL 5.62 6.06 188.0 6.07

6 980 6.37 O96)" | ecginza 6.62 6.79 197.0 6.36

7 1040 5.00 5.45 aes: 6.88 6.16 158.5 Sigh

8 1310 5.50 6.00 —— — 6.00 170.0 5.48

An inspection of the above table shows that the maximum movement, for
both the summer and the entire year, lies well to the west of the axis of the
glacier. The greatest average {daily movement was made by plate 6, which
lies 120 feet to the west of the center, while plates 7 and 8 show only slightly less
movement. This is in harmony with what is known concerning the flow of
glaciers on a curve, the maximum movement taking place not at the center,
as in the case of the very straight Victoria, but lying between the center and the
convex side. The average daily movement of plate 6 for the year is 94 per cent —
of its summer motion, as compared with 81 per cent for the corresponding plate |
upon the Victoria. For some reason, not easily explained from the data at hand s

— we

—

ae ere ee

EE —— ee ee, eae TP

~e.LUmT!DmU,CUmOUOC eee eee

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. gt

the mean summer motion of the two most easterly plates was less than their
yearly average. It is to be noted that the maximum summer movement of 6.96
inches (July 31 to August 11), 1899, is but about one-third of the maximum
movement observed by Green in 1888 (August 13 to 25). The only way to
reconcile the two results is to suppose that Green’s measurements were
made farther up the slope towards the cascade, where the movement is un-
doubtedly much greater than towards the nose. Messrs. Vaux placed a
ninth plate upon the nose of the glacier and had it under observation from
August 1 to August 20, 1899. The average daily horizontal motion for the
first two intervals between measurements was 5.9 inches and 5.0 inches. A
crevasse then formed, detaching the block carrying the plate, and the subse-
quent apparent motion was 2.8 inches and 2.7 inches daily.

8. FRONTAL CHANGES.

a. Recession data. Owing to the easy accessibility of the glacier and its at-
tractiveness to the ordinary visitor, we have more data from which to determine

_ the frontal behavior of the IIlecillewaet than any of the other Canadian glaciers.

As has been noted, from the photograph taken in 1887 by the Messrs. Vaux the
position of the ice at that time, with reference to a large boulder, was determined
and in 1898 marked conspicuously. In 1888 the margin of the ice was marked by
Green and the glacier was photographed by Notman & Son, of Montreal (plate
XXXvI, figurer). Reference blocks were marked in 1890 and 1895 by interested
visitors. A visit was paid to the glacier September 3, 1897, by Albrecht Penck,
of Vienna, and a sketch made of the tongue of the glacier and its relation to the
lower moraines. This was published in the Zeitschrift des Deutschen und
Osterreichischen Alpenvereins, Jahrgang 1898, Band xxtx, s. 55, under the title
“Der Ilecillewaetgletscher 1m Selkirkgebirge.”’ The height of a number of points
was determined by an aneroid and four reference blocks established and located
upon the map. These blocks were left to be marked by a railroad employee,
but were apparently neglected and in tg04 could not be identified with absolute
certainty, owing to the changes in the ice margin. Based upon the railway
elevation at the station, Penck determined the elevation of the nose in 1897 as
4,793 feet (1,461 meters). The foot-bridge, just beyond the modern terminal
moraine he gives an elevation of 640 feet above that of the station, or above sea
level 4,760 feet, 1 and this he uses as his datum for elevations about the glacier.
The nose of the glacier at this time lay 33 feet above the floor of the bridge, and
the crest of the adjoining lateral just opposite -was 131 feet above the valley
floor at the nose.

In the year 1898 a number of reference blocks and range lines were established
by the Messrs. Vaux and have since done excellent service in measuring the
frontal movements. In August, 1899, they made a very detailed survey of the
nose and adjoining region and prepared a large scale map which is of the greatest

1 The correction of the railroad levels reduces this elevation by 27 feet, giving the bridge-floor 4,73 3
feet.

g2 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

value in the determination of changes in progress! (see plate Xxx1). They!

to detecuitne the marginal changes.

There is reason for thinking that the glacier in 1887 was just completin,
rather prolonged halt at the younger of the frontal moraines described. T!
it had not recently extended much beyond is proven by the size of the a
bushes growing about the outer slope. That it had made a rather prolo

the moraine, which, with the small amount of débris carried by the gla
would require a considerable time in building. From the early photograp
is seen that the glacier was much bulkier and broader at this stage and the s.
about the nose much steeper, enabling the glacier to maintain well its positi
the moraine (plate xxxv1, figure 1). During the year 1887 to 1888 it had b

to and as indicated by the rocks blotched with tar by Green. The retreat be:
somewhat gradually and attained its maximum between 1890 and 1900, avera
for these ten years about 53 feet perannum. The average for the opening lus
of the century is 19.6 feet, the retreat being reduced until it amounted to
two feet for the year 1904-5. For the 18 years from 1887 to 1905, the horiz
retreat from the Vaux reference block was 597.5 feet, or at an average
rate of 33.2 feet. It should be noted, however, that this measurement is
in a line with the main axis of the glacier. The available data concerning
glacier are given in summarized form below. The measurements were ta

variously, most of them between the middle of August and the middle of S
tember, so that the retreat assigned to some years, ser belong in part: to
preceding, or the following year.

Recession Data of the Nose of the Illecillewaet Glacier.

1887-1888. 10 to 15 feet.
1888-1890. Average rate about 23 feet
1890-1898. Average rate of 56 feet.
1898-1899. 16 feet.

1899-1900. 64 feet.

Igoo-1g0r. 15 feet.

1901-1902. 48 feet.

1902-1903. 22 feet.

1903-1904. 11 feet.

1904-1905. 2 feet.

1905-1906. 84 feet.

‘The Great Glacier of the Illicilliwaet,” George and William S. Vaux, Jr., Appalachia, vol.
TQ0O, p. 156. 3 ae:

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XXXVI

Fic. 1.—Illecillewaet Glacier in 1888. Photographed by Notman & Son, Montreal.

Fic. 2.—Tllecillewaet Glacier in 1905, from approximately the same view-point as figure I,

o>,

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 93

Upon the face of the bedrock exposed near the nose a mark was established
September 16, 1903, immediately beneath the nearly vertical side of the ice,
the height of which was estimated as 60 feet. August 24, 1905, it was found
that the ice had withdrawn laterally 2.4 feet from the face. Passing around from
the nose eastward, three stations were established along the margin of the ice.
A large boulder was found just emerging from the ice, the first week in September,
1904, and marked ‘‘ Face emerging, Sep., ’04.’’ Upon the 24th of August, 1905,
it was found that the ice had retreated here 14 feet. Farther along a medium-
sized boulder had been marked in 1903, ‘‘15 ft. to ice. IxX-16-03.’’ By Sep-
tember 1, 1904, a retreat of 12.5 feet had occurred here, while at the upper
station the boulder ‘‘27 ft. to ice. 1xX—16-03,’’ measured September 3, 1904, 27
feet, and August 25, 1905, 27.1 feet. These data indicate that the margins of

- the ice have been receding as we approach the nose, more rapidly upon the eastern

side, but that farther up along the margin there has been no change for the last
two years and, very probably, for a considerably longer time The two views on
plate xxxvi, taken from almost identically the same view-point, the former in
1888 and the latter in 1905, furnish a good opportunity for noting the changes
produced in the glacier in the 17 years. It seems almost possible to recognize
the individual trees standing to the right of the center, but the lower half of the
glacier is unrecognizable. A stadia and trignometric survey of the Hlecillewaet
and Asulkan glacier tongues was made in 1906 by the Messrs. Vaux and a
report made to the Philadelphia Academy of Sciences. Some additional data
concerning the movements of the steel plates upon the [lecillewaet were collected
and appear upon plate xxx of this report.

b. Icewaves. Incomparing their photographs made in 1898 and 1899 from
a certain large boulder, just west of the trail, Messrs. Vaux noted an apparent
thickening in the ice just beneath the névé line. By drawing a delicate line be-
tween corresponding points in figures 1 and 2, plate xxxvi, that may be recog-
nized in the upper névé region, it is seen that the ice margin along the sky-line
stands slightly higher in the 1905 view. The difference is slight, however, and can
represent but a few feet. When the Notman view of 1888 is compared with a
second, which was made in 1897, and here reproduced in plate xxxvut, figure 2,
the heaping of the ice line beneath the névé line is still more plainly seen. There
is thus evidence that a wave, or impulse, derived from an increased precipitation
over the névé region, travels the length of the glacier and gives rise to a halt,
or an advance, of the front; followed by a depression which permits of a retreat.
Such a depression appears to have been at the edge of the névé line in 1887 or 8,
while the glacier about the lower extremity was experiencing the effect of a pre-
vious impulse. The retreat of the glacier was greatest between 1890 and 1900,
and if we assume that it culminated at about the middle of the decade, it required
about 8 or g years for this trough of the wave to reach the nose, or at the average
rate of 800 to 850 feet per annum. Since in 1905 the appearance of the sky-line
along the névé corresponds so nearly with that seen in 1888, we may assume that
the crest of the wave was in this position at a date only a little later than the

94 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

mean of the two. This would bring it to about 1898 or 9, when it was especially
noted by the Messrs. Vaux. The gradual reduction in the rate of retreat observed
during the past three seasons would indicate that this impulse is making itself
felt about the nose and that either a halt, or an advance, is about to be in-
augurated. If we are correct in the inference that the névé line was marked by
a trough, or low stage in the height of the ice, about 1887 or 8 and that a return
to this condition has been reached about 1905 or 6 with a crest, or high-stage
condition of the ice between, an interesting relation is at once established with
Briickner’s climatic cycle (page 16). The time between the appearance of these
troughs for the passage of one-half of the wave is 18 years, and we may venture
to predict that the present relative depression will be followed by the passage
of another crest requiring about the same number of years. The nose has been
in retreat from 1888 to 1906, some 18 years, and we should expect another period
of halt or advance to soon set in. Such a condition was to be anticipated from
the marked reduction in the rate of retreat from tg02 to Ig05, but the very
remarkable recession of 84 feet determined by the Messrs. Vaux for the year
tg05—1906, leaves the matter in doubt.

The relation of the glacial movements to the precipitation cycles becomes
a matter of much interest and here, as above, with our meager data, we can only
point out possibilities, which will either stand or fall, when the next half-century’s
observations have been collected. From our available meteorological records
there was a deficiency of precipitation over the mountains from 1885 to 1896;
how much before 1885 this condition existed we have no means of knowing.
Since 1897 there seems to have been an excess over the normal amount, but it
was at this time that the crest of the wave made its appearance at the névé line.
Then instead of continuing to increase, as we might expect, it gave way to a
trough. The inference is, and it is only an imference, that this wave represents
the gush of ice from the collecting basin due to the excess deposited during the
phase of the cycle which antedated 1885, and probably to be correlated with the
excess in the Rockies, as recorded so strikingly in the evidence of higher lake levels,
described by Dawson (page 17). The approaching trough shown about the névé line
in 1905 must then be ascribed to diminished precipitation received over the collect-
ing region from 1885 to 1896, being then 9 years delayed from the close of the phase
which gave rise to it. The crest of the wave from the basin was delayed some
17 to 18 years from the close of the preceding phase. In a paper read before
the International Geographic Congress, at its Washington meeting (Proceedings,
1904, p. 487), Doctor Reid gave a discussion of ‘‘ reservoir lag,”’ in which he demon-
strates mathematically that the thickening of ice in the collecting basin does not
keep pace with the variation in precipitation, but lags behind it. In the case of
large glaciers this lag amounts to about one-fourth of the period of the variation,
and the ice in the basin should attain its maximum thickness, only about the time
that the annual supply has settled back to the normal amount and is ready
to diminish. After the maximum ice thickness has been attained toward the
center, time is still required for the impulse to reach its maximum at the crest of

— = =

eS etl eee eS. CCU

ee a ee ee Se a ee

os

—_-”-

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 95

the basin, the amount of which will differ with the local conditions. In this way
We may account for the delay in the arrival of the crest at the névé line in the
years 1897-9.

9. Former Activity.

a. Rock scorings. The former work of the glacier is shown in great
beauty and variety upon the mass of bedrock now being gradually uncovered
near the nose. The hard rock features of Pleistocene glaciation are all here for
study by those interested, many of them indicating the direction of ice move-
ment and hence of practical value in the field.!. Excellent examples of the so-
called roches moutonnées occur, groups of which in the distance often resemble
crouching sheep (plate xxxiv, figure 3). In the specimen figured, the ice moved
from right to left across this projection of bedrock, the up-stream, or stoss side being
rounded and smoothed, while the down-stream, or lee side, was affected slightly,
or not at all. Portions of the rock were polished, as the ice was rubbed vigorously
across, and where the ice held rock fragments against it, systems of approximately
parallel scratches were produced, some so fine that they must have been made by
sand grains. At the last stage of the disappearance of the ice from this particular
roche moutonnée, a small clump of rock fragments was gently dropped upon the
upper surface in insecure position. An inspection of this and the adjoining rock
in the figure, shows a system of parallel joints, dipping down-stream at a steep
angle. From the lee side of the central roche moutonnée it is apparent that an
entire block was pried loose by the ice and that a little more vigorous action at the
joint, just beginning to open, would have removed bodily nearly the entire block.
This action is known as “plucking,” already described in connection with the
Yoho (page 79), by which the work done in a few days may exceed the erosion
of years. Places may be seen upon the surface where a rock engaged in produc-
ing a shallow groove has made a succession of jumps and given rise to a series
of short parallel curves, more or less closely placed, with their concavities
directed down-stream. These are the ‘“‘chatter-marks,”’ the production of
which may be illustrated by pushing a dry finger over a polished surface. In
other cases rocks embedded in the under side of the ice have been suddenly
brought into action, producing a crescentic gouge, with its convexity directed
in the direction of flow.2 The bedrock here being a schistose conglomerate
with rather coarse, hard masses embedded in a softer matrix, there have been
produced the “knob and tail,” or “knob and trail’? phenomena, so useful often
in determining the direction of ice flow in the case of Pleistocene glaciers. In
one case examined there appeared a dark colored knob of harder material,
which the ice was unable to cut away as rapidly as the surrounding schist.
The projecting knob had partially protected the softer material in its lee,

1A most valuable paper by Chamberlin upon the effect of ice upon rock will be found in the Seventh
Annual Report of the Director of the U. S. Geol. Surv., 1888, page 155.

2 See paper by Gilbert read before the Cordilleran Section of the Geolugical Society of America at its
1905 Winter meeting. ‘‘ Crescentic Gouges on Glaciated Surfaces,” Bulletin Geol. Soc. of Amer., vol. 17,

PP- 393-3 14-

96 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

forming an elongated tail, or trail, extending from the knob in the direction
of ice motion. In some cases a small quartz vein cuts across the surface in such
a way as to protect in its lee a strip of the softer rock. In front of the knobs
there is cut out, as a rule, a frontal groove lying at the base and curving around
laterally into two others, one upon either side, forming the lateral grooves. In
places where the ice acted with greater vigor, owing to the concentration of its
action, or where differences existed in the structure, or hardness, of the rock
there were cut out basins and U-shaped troughs, representing, in miniature, lake
basins and glaciated valleys. One basin, with perfectly smoothed sides and
bottom, had a length of 15 feet, a breadth of 6 feet, and a depth of 6 to 8 inches
below its lower rim. The greatest depth was located one-third of its length
from the upper end, indicating where the gouging action had been greatest. One
of the troughs was 11 to 12 feet across and 4 to 5 feet in depth.

b. Bear-den moraines. Some 800 to goo feet below the terminal moraine of
1887, or about 1,400 feet from the nose of the ice in 1904, there occurs a moraine
of the same general type as that described under this head in connection with
the Victoria. This consists of very massive blocks of quartzite, arranged in
a north to south ridge across the valley, having a breadth of about 400 feet and
a height above the general valley floor of 20 to 4o feet. The largest block
observed was measured by Messrs. Moseley and Todd and its dimensions, above
ground, were found to be about 107.5 by 28 by 11 feet, from which it was esti-
mated to weigh about 2,000 tons. A portion of this ridge is seenin plate XxxvI,
figure r, taken from one of the blocks of themoraine itself, looking toward the glacier
up the valley. The blocks are blackened with lichens, more or less moss-covered,
and carry enough soil to support considerable vegetation of alarger size. A spruce
growing upon the moraine had been cut and with a circumference of 128 centi-
meters gave 243 rings of growth. A hemlock, also upon the moraine, with a
circumference of 320 centimeters (50 centimeters from the base), was calculated
to be 447 years of age. This estimate was based upon the average breadth of the
annual rings of growth measured in the Illecillewaet and adjoining Asulkan
valleys. This average breadth was found to be 1.140 millimeters, as compared
with 0.884 millimeter in the Lake Louise Valley.

From the outer edge of this moraine, 1,500 feet down the valley measured
along the stream, there begins another similar but larger moraine of the same
type. Starting from the spur of Glacier Crest which separates the Ilecillewaet
and Asulkan valleys, the ridge swings out across the valley bearing N. 8° W.,
and then swings around to N. 15° W. It is 200 to 300 feet across and some 50 to
60 feet above the valley floor, somewhat steeper toward the glacier. The blocks
are very coarse quartzites and schists, blackened with lichens, and presenting
angular outlines. The largest block noted was estimated to weigh 1,250 tons.
The usual filling of a moraine, gravel, sand, and clay, is practically absent. Upon
the eastern side, for a portion of its length, it is covered by a mass of broken
tree trunks which were swept from the side of Mt. Eagle by an avalanche (plate
XXXVI, figure 1) some decadesago. Enough soil has accumulated about the rocks

SMITHSONIAN CONTRIBUTIONS FO KNOWLEDGE—SHERZER. PLATE XXXVII.

Fic, 1.—‘‘Bear-den moraine” made conjointly by the Lllecillewaet and Asulkan Glaciers.
Strewn with timber avalanched from the right-hand mountain slope.

Fic. 2,—Illecillewaet Glacier in 1897. Photographed by Notman & Son, Montreal. Compare with plate XXXVI,

Ispay AA “O “VY Ulorfy suoljeaaly
NOWILT YIAapai~ puv ssoy jsorio. oq sjuURjsisse pjatqg fobr1 YsnSny ‘sazieysS “H “AA Aq uMEIp pue pakaaing *utaqsKg yarpjas ‘Kay[eA uryxpNsy ‘raeyDH uvypnsy jo dew

4 €E1'6

FBWOILL
gse4 UeYINSY

o
a
%
om
S
=
9
+
co
>

,
pat LOYTNSP MOLL UD 15.
ghprad 4S 3224 000€ 0002

ITIAXXX ULW1d ‘MAZYAHS—AVGAIMONN OL SNOILONALALNOO NVINOSHLIWS

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 97

to support a growth of raspberries, blueberries, etc., and also a few spruce 8 to
12 inches in diameter. The bulk of the material lies to the west of the glacial
brook and was derived from the eastern side of Glacier Crest and Mt. Lookout,
the cliffs of which have a northwest-southeast trend. The shape of the moraine
and the way in which the blocks have been deposited indicate, as noted by Prof.
Penck at the time of his visit, that the moraine was built conjointly by the former
Ilecillewaet and Asulkan glaciers. The blocks contributed by the Asulkan
came from the western side of Glacier Crest and the Asulkan Ridge, and are
much less in amount than those derived from the eastern side and transported
by the Ilecillewaet. The largest tree found growing inside of this moraine
was calculated to have been 520 years old when it died and from the condition
of its wood and bark to have been dead about 30 years.

CHAPTER VII.
ASULKAN GLACIER.

1. GENERAL CHARACTERISTICS.

Ly1nG at the head of the Asulkan Valley, upon the opposite side of Glacier
Crest from the Illecillewaet Glacier (see plates xxx and xxx), is located the
Asulkan Glacier. Its broad expanse of snowfield extends in a semicircle from
Asulkan Ridge, past Leda, Pollux, and Castor to the northern extremity of the
Dome, faces to the northward, and under the sunlight is of dazzling whiteness.
The name is of Cree Indian origin and is generally said to mean “goat,” but Iam
assured that it really means “‘bridge.’’ The nose of the glacier lies about three
miles from the station, reached by a picturesque and easy trail, except in the
upper part, where the trail becomessteep. The glacier itself may be safely visited
and studied without a guide, but no one should venture upon the névé unattended,
as it is very treacherously crevassed. ‘This glacier is the smallest and the most
southern and western of the series here reported upon, its nose lying in longitude
117° 28’, west and latitude 51° 13’, north.

The glacier consists of three streams, two of which are closely united and the
third separated from the other two except in the névé region where they are
all united. The length of this third stream, measured from the Asulkan Pass, is
about two miles, of which the first mile is névé and the lower mile is ordinarily
free from snow during the summer season (plate xxxrx, figure 1). The breadth
of the dissipator is about 1,800 feet in the upper part, but about the middle of its
course it makes an abrupt bend from the north to the northeast and tapers
gradually to a sharp nose. The eastern margin curves around gradually to the
nose, while the western side is curiously straight, cutting diagonally across
what appears to be the natural course of the glacier. There seems no apparent
reason for this abrupt bend in the glacier and for the remarkably straight western
margin of the ice, but the explanation will appear in what follows. This peculiar
contouring of this stream gives it the general form of a bear’s paw—a polar bear

98 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

—in which the straight margin represents the sole. From near the heel of this
foot there extends southward a long, slender ridge of glaciated rock, carrying
more or less morainic matter, which separates this eastern ice stream from the
double ice mass immediately to the west (plate xxxrx, figure 1). Judging from
the line of crevasses and faulting across the névé, there lies another similar ridge,
parallel with the first and about one-quarter mile to one-half mile to the west,
which separates this mass into two streams, each having its own separate nose, as
shown upon the map. This ridge is apparently the continuation of the line of
bedrock exposed along the right-hand margin of the westernmost ice stream.
The névé line upon this glacier is about 7,000 feet above sea-level and the main
portion of the névé lies between this altitude and 8,000 feet. From the Asulkan
Pass (7,710 feet) to the nose of the easternmost stream the descent is 2,110 feet,
or at the rate of 1,055 feet to the mile. The altitude of the nose is 5,600 feet,
or some 800 feet higher than that of the Iecillewaet, due apparently to the smaller
volume of ice in the Asulkan and its dissipation at three separate points. . The
altitudes of the two higher noses to the west are about 6,000 feet, or the same as
the Victoria. So far as may be judged from the crevasses and faultings, the
ice responds fully to the irregularities in its bed which indicates that it is
relatively thin. The surface slope of the western and middle streams is very

steep; that of the easternmost, or main, stream is much more gentle, amounting

in places to not more than 6°. Toward the nose the inclination becomes 25°
and then drops off to but a few degrees, so that it may be readily ascended.
Upon either side of the stream the marginal slopes are steep for a few hundred
feet back from the nose.

2. PIEDMONT CHARACTERISTICS.

If the reader has covered Chapter IV of this report he will have recognized
already the piedmont character of the Asulkan, which consists of three com-
mensal streams. The glacier is of peculiar interest because it is an illustration
of a piedmont glacier in its senile condition. It has reached its second childhood
and now illustrates the disintegration of a piedmont glacier into the component
streams, the union of which in its youth gave rise to the glacier itself. Every
glacier of this type begins with the independent development of a system of
Alpine glaciers, codrdinate in importance, which coalesce laterally into a single
ice mass. The length of the glacier is determined by the length of the separate
streams composing it and its breadth by the number of streams and their com-
bined breadth. In the final stages of dissolution, which must come sooner or
later in its life history, the piedmont glacier shrivels back into the original Alpine
components. The eastern tributary has already separated sufficiently so that
it may be regarded as an independent glacier. The other two have separated
for a distance of about one-fourth of a mile, but the separation will not be com-
plete until the ridge of rock above noted has appeared at the surface of the ice.
The middle stream covered the ridge of rock, now exposed between it and the
eastern stream, and sent its nose down the valley as far as the drainage brook

PAs oer he ell

SMITHSONIAN CONTRIBUTIONS TO KNOWLI DGE—SHERZER, PLATE XXXIX.
Mt. Donkin and Asulkan Pass.

—=

Leda Pollux, Castor.

Fic. 1.—General view of Asulkan Glacier in 1902. Copyrighted, 1902, by the Detroit Photographic Co.

Donkin. Castor and Pollux. Dome. Bonney.

Fic, 2.—The Asulkan glaciers and snowfields from Avalanche Mt. (elevation 9,387

feet), showing a decadent piedmont glacier. Photographed in
1go1 by Arthur O, Wheeler.

ie

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 99

shown upon the map. There it formed a series of terminal moraines upon its
eastern side, the eastern component standing at about the same level and forming a
similar series. The sudden bend noted in the eastern component, one-half mile
back from the nose, resulted from its pressure against the side of the middle
stream which it was unable to force aside. Conjointly they formed a straight
medial moraine from the bend to the nose. Upon the more rapid retreat of the
middle stream and its disappearance from this part of its bed, this moraine
became the left lateral of the easternmost stream (plate xxx1x, figure r), and was
of such a massive character that it has continued to deflect the ice from its
natural course.

In plate xxxrx, figure 1, we have shown nearly the entire eastern and middle
streams of the Asulkan, and a portion of the névé of the western. A distant view
of the entire glacier is given in plate xxx1x, figure 2, taken by Wheeler from the
summit of Avalanche Peak (9,387 feet) in 1901. The Dome may be recognized from
its contour and from it there is seen to be a broad ridge extending valleyward and
marking the western limit of the present Asulkan Glacier. To the right of this
ridge, along the eastern slopes of Mts. Afton (8,423 feet) and Abbott (7,710 feet),
four marked depressions occur, each containing small-sized glaciers. The
contour of the rocky slopes separating these amphitheaters, or cirques, as they
are termed, proves that at an earlier stage of glaciation these streams coalesced
laterally and united with the present Asulkan, forming a grand, hanging, piedmont
glacier, extending from the Asulkan Ridge to Mt. Abbott with at least nine or
ten main commensals. Previous to this stage they had united with others from

the head and opposite side of the valley into a grand Alpine glacier, which became

a tributary of the ancient Illecillewaet trunk glacier in Pleistocene time.

3. NOURISHMENT.

The névé field of the present Asulkan is arranged in the form of a semicircular
belt, extending from the Asulkan Ridge upon the east around to the Dome upon
the west, having a length of about three and a half miles and an average breadth
of perhaps three-fourths of a mile. The area of this field is somewhere between
two and a half and three square miles, or less than half that of the Ilecillewaet névé
field. The amount of precipitation over this field cannot be essentially differ-
ent from that given for the neighboring glacier (page 82). From an elevation
of about 7,000 feet the névé snows reach up to the crests of the bounding ridges,
in many places, attaining an elevation of 9,000 feet. It is possible to pick out, ina
general way, the névé fields by which the separate ice streams are nourished.
The eastern stream receives its supply from Asulkan Ridge (9,100 feet) and from
the Pass (7,710 feet), the former moving westward down the oblique slope and
delivering its supply of ice and subglacial débris to the right side of the mainstream
flowing from the Pass. A still less amount is received from the opposite side
from the snow that accumulates upon the northern slope of the unnamed peak
(8,700 feet +) lying to the east of Leda. The Pass and upper névé are reached
by means of the right lateral moraine. Judging from the course of the transverse

100 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

crevasses, which lie at right angles to the main direction of flow, the névé that
accumulates between this minor peak and Leda (9,133 feet) moves northward
and nourishes the middle ice stream. The oblique course is taken probably
because of the continuation southward of the ridge of rock previously noted
as separating the middle and eastern streams. The presence of a similar ridge
beneath the ice deflects the névé snow and ice from the northern slopes of Castor
(9,108 feet) and Pollux (9,176 feet), to which is added that from the Dome
(9,029 feet), and thus is obtained the supply for the double-nosed western ice
stream. The middle stream is least well supplied at the present time with ice
from the névé field and has receded farthest. It seems very probable that
when the ice was thicker over the névé there was relatively less of it deflected
to the western stream by the subglacial ridge and this fact permitted the middle
stream to maintain the same length as the now better nourished eastern. The
superficial layers from Castor and Pollux could move directly across those which
the configuration of the bed deflected northward, but as the general elevation of
the névé was lowered a relatively greater and greater percentage of ice was
deflected to the western stream and the nose of the middle stream retreated
steadily some 2,200 feet up the slope, to an elevation at present 4oo feet higher
than the nose of the eastern stream. Were the ice of all three streams now
concentrated into a single one it is probable that the nose would attain as low
an altitude as that of the Illecillewaet.

4. MORAINES:

Owing to the absence of high, overtowering cliffs, such as we find in the
case of the Wenkchemna and Victoria glaciers, the névé fields of the Asulkan
receive very little rock débris over their surfaces. In consequence, the ice itself,
except along the margins, is quite free from rock fragments. As in the case
of the neighboring [llecillewaet collecting basin, conditions are favorable for
receiving wind-blown dust from peaks and ridges towering above the snow.
This dust is distributed somewhat evenly over the snow and, when concentrated
by melting, gives rise to the stratification and imparts a soiled appearance to the
ice about the lower margins.

The right lateral moraine of the eastern ice stream makes its appearance just
east of the nose and south of the stream from Asulkan Ridge. It rises at once into
a conspicuous, sharp-crested ridge, extending south-southwestward, and bending
abruptly to the south-southeast, attaining the length of a mile before it dips
under the névé snow. The lower portion of the moraine seems entirely free
from ice, the outer slope carrying fir and spruce 50 to 60 years of age. The inner
slope is more steep and has younger vegetation, indicating that the ice has within
a few decades withdrawn from the moraine. The crest rises to a height of 60 to
70 feet above the valley floor upon which the glacier rests. The rocks consist
largely of bruised and rounded quartzites and schists, which in the upper part
are embedded in a matrix of glacial clay. This material appears to come
from beneath the glacier that covers the western slope of Asulkan Ridge,

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XL.

Fic. 1.—Left Asulkan moraine shedding its rocky covering and exposing the ice core.

Fic. 2.—Deébris-covered nose of Asulkan Glacier, August, 1904. Glacier had been advancing for some three or four
years,

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. Tor

by which it is delivered to the main stream from the Pass, as previously noted,
upon a level with its surface. This névé-covered glacier sustains the same relation
to the Asulkan that the Collie and Gordon glaciers do to the Yoho. Between
the crest of the moraine and the glacier there intervenes a steep boulder slope,
about 300 feet broad in the lower portion near the nose, but narrowing gradually
for a half-mile, when the moraine and ice meet. Opposite the nose, upon the
eastern side there is an outcrop of a silvery schist, with its strata upon edge,
which has been glaciated and plucked.

The development of the left lateral of the eastern stream from a former medial
has already been described and its very straight course obliquely across the valley
been accounted for. Originally, this moraine may have been largely subglacial,
or englacial, the material being derived from the basal layers of the ice. The
slope of the valley floor is northward, while this lower half-mile of the moraine
bears northeastward. After making the very abrupt bend noted, the moraine
continues for a quarter-mile farther, resting upon the rocky ridge of quartzite
anda greenish schist. This ridge raises the base of the middle stream above
the present surface of this portion of the eastern stream. The moraine consists
very largely of ground moraine, supplied apparently in large part by the middle
ice stream, but instead of clay the filling is a glacial sand. The finer material
may have been removed by currents too gentle to transport this sand. The
inner slope of the moraine is steep, the outer is more gentle down to the drainage
brook from the middle nose. The crest of the moraine rises 125 to 150 feet above
the floor of this valley. About one-quarter mile back from the nose this moraine

begins to shed its cover of rock débris, revealing in a most interesting manner

the real structure of such a moraine. From this point up the valley the moraine
is a typical, sharp-crested structure (plate xxx1x, figure 1), but here the débris has
begun to slip to either side, forming a double ridge with a continuous ice crest
between. Plate x1, figure 1, gives a view of this exposed ice core, looking up the
glacier along the inner side. The highest portion of the ice ridge attains a height
of 25 to 30 feet, which is being rapidly acted upon by the sun in midsummer.
Where it has been longest exposed the ice has melted below the general level of the
glacier, forming a depression with a ridge of rock débris upon either side, the
outermost one being quite prominent. Into the depression the material from
either side has begun to roll and slide, thus protecting the ice at the bottom of the
depression from the sun. Had the thickness of the ice proved sufficient, in
time the rock débris would have gotten back, in large part, into the depression,
allowing the ice to melt upon either side and starting again the formation of a
single-crested, typical moraine. Thus it appears that moraines may, under certain
circumstances, pass through the same series of stages as those described for sur-
face lakelets upon page 57 of this report.

About the eastern nose there has been pushed up a ridge of ground moraine,
from 12 to 15 feet high, into which the ice nose plunges and buries itself. Upon
this ridge minor ridges, but a foot or two in height, also occur, as seen in plate x1,
figure 2. At times the nose is so deeply buried that it is difficult to find it for pur-

Io2 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

poses of measurement. When the middle stream stood at this lower level, the two
built a series of three or four latero-terminal moraines which curve gently down
the valley from the lower ends of the laterals. The inner terminal of the middle
stream has the appearance of age, compared with the others lying just east,and
is being covered slowly with low shrubs and evergreens. From this difference in
age one would infer that the ridges of the series, lying to the east, had been built
by the eastern stream alone. The nose of the middle stream, especially upon its
eastern side, rests largely upon bedrock, more or less strewn with rock fragments.
The rock here, as elsewhere about the glacier, consists of quartzite and schist,
plucked and glaciated. Along its left side, back as far as it has become sepa-
rated from its neighbor, it has built a sharp-crested, lateral moraine, which
in the lower half is double, curving gently down the bouldery slope. The inner
slope is steep, the outer long and more gentle. The boulders.are rounded and
bruised but only occasionally well glaciated. The double western nose is similar
to the middle, in that it is steep, perched high up on the slope, has bedrock

exposed upon its eastern side, while upon the left it has built a short, sharp~

lateral extending up to the névé. In front of the western and middle streams
there has been uncovered a steep slope of bouldery ground moraine so recently
that trees have not been able yet to get anything more than a start.

5. CREVASSES.

Owing to the irregularities in its bed, the steep slopes, and the apparent thin-
ness of the ice, the Asulkan streams are badly crevassed and faulted. The névé
fields of the western and middle streams cannot be traversed with any degree
of safety, while that of the eastern calls for the greatest skill in locating the
snow-covered death-traps (plate xL1). The crevasses in the névé portions are
mainly of the transverse type, caused by the rapid movement of the ice over
an irregular, steep slope. They stand approximately at right angles to the
direction of motion and furnish a clue as to the general movement of various
portions of the névé field. Just below the névé line the eastern stream en-
counters, in its central portion, an obstruction by which the ice 1s shattered in every
direction, but mainly transversely (plate xxxtx, figure 1). The descent is not
rapid enough to constitute a cascade and the blocks, at first angular, become
sharpened by melting into seracs but retain their vertical position until melted
away at the base of the slope. The development of these seracs is well shown in
plate xLu, figure 1. The ice exposed portion, or dissipator, of this stream shows
numerous marginal crevasses along either side of its course, those upon the
eastern side, in the lower portion, extending beyond the central axis. Those
upon the western side are not so strongly developed. It is upon the middle
stream that the dirt band crevasses occur figured in connection with the discus-
sion of this subject (plate xvir, figure 1). The slope, however, is too steep for
their formation and preservation.

ee eee

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER DRE

Fic. 1.—Stratification of Asulkan Glacier. The dates are added upon the sup-
position that the strata represent seasonal deposition. Copyrighted,
1go2, by the Detroit Photographic Co.

Fic, 2.—General view of Asulkan Glacier in 1898. Reproduced through courtesy of the Messrs. Vaux, Compare with plate xxxrx,
figure T,

iv

.

GLACIERS OF THE CANADIAN, ROGKIES AND SELKIRKS. Io3

6. Ice STRUCTURE.

The névé-covered portion of the ice acquires a very perfect stratification as
the result of wind distributed dust and periodic melting over the surface. This
structure is beautifully shown in the crevasse walls and the faces of the numerous
faults in the ice. In the photograph of the Detroit Publishing Co., reproduced
in plate xt, the successive layers, with the minor stratification seams, are clearly

-shown. The correspondence of the strata upon opposite sides of the crevasse

shows that there had been no faulting. From his heel to the crown of his hat
this guide pictured is about six feet in length and, by comparison, we ascertain
that the strata shown range from three to twelve feet in thickness. The picture
was taken during the summer of 1902, and in looking at the uppermost stratum
it is forced upon one’s belief that this represents the compacted snow that accum-
ulated over this spot during the season of 1901-2. Part of this snow was pre-
cipitated directly, part of it may have been drifted by wind action. It may have
lost some by wind action, as well, during the season of accumulation. It has
been compacted by melting, pressure, and occasional rain into a fine granular
ice. If we are right in supposing that this stratum represents the accumulation
during the season of 1g01~2, minus the loss by the combined agencies, then the
stratum upon which it rests must have accumulated during the season of 1900-1.
Passing down the side of the crevasse we may thus assign dates to the successive
strata, finding that they reach back to the season of 1895-6. It is especially
interesting to note that the deposits supposed to have been laid down between
the summer of 1898 and that of 1902 average considerably thicker than those
between the summers of 1895 and 1898, since this dividing date falls very near
the supposed date of the beginning of the phase of increased precipitation in this
region. It is further to be noted that the stratum marked 1898-9 is the thickest
of the series. It is very unfortunate that our precipitation data are not fuller
for the locality. In order to serve the present purpose in establishing a rela-
tionship between the amount of precipitation and the thickness of the strata in
the névé, a combination should be made of the last three months of the year with
the first nine of the following year. This would unite practically all of the snow-
fall and the rain and melting water of the following summer, as it is combined
in the stratum itself. In a paper upon the Canadian Pacific Railway, from
Laggan to Revelstoke,! Mr. William Vaux gives the average snowfall for Glacier
House from 189s to 1898 as 31 feet, based upon records kept by the station agent,
this being but 83 per cent. of the normal. From October, 1898, to May, 1899, in-
clusive, the snowfall alone amounted to 43 feet 8} inches, being 17 per cent. above
the normal. ‘The records are lacking up to 1902, for which year the Meteorological
Service reports 13.88 inches of rain and 347 inches of snow, or a total equivalent
of 40} feet in snow, or 14 per cent. below the normal. By referring to plate x1
it will be seen that the stratum assigned to the year 1898-1899 is the heaviest of
the series while that for 1901-1902 is light. From 1895 to 1898 the strata are

1 Proceedings of the Engineers’ Club of Philadelphia, vol. xv, 1900, p. 73.

1o4 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

thin, corresponding to the average lighter snowfall above reported for these
years.

A still further confirmation of the conclusion reached above, that from the
middle of the year 1898 there has been a marked increase in the snowfall, is
furnished by the notes and photographs of the Messrs. Vaux, to be noted later.
Their photograph of 1898 shows a large amount of rock exposed along the slopes
of Leda, Pollux, and Castor, as well as between the eastern and middle ice streams.
In 1899 they note that the névé line is lower and the hanging glaciers are much more
active, giving rise to frequent avalanches, which were very infrequent in 1898.
When their photograph of 1898, reproduced here as plate x11, figure 2, is com-
pared with that of the Detroit Publishing Co., taken in 1902 (plate xxx1x, figure
1), the increase in the amount of névé is striking, the presence of the bergschrunds,
in areas that were bare rock, indicating that glaciers had formed in the meantime
and that there is not simply a covering of loose snow, such as might fall in a
night. In looking over the broad snow expanse one does not think of there being
hanging glaciers upon the slopes of Castor and Pollux, as they seem to blend with
and be an integral part of the general névé field. In 1898 they were separated
sufficiently so that it was natural to think of them as being detached. In two
weeks of August, camping in plain sight of the region, in 1904 and 1905, the writer
does not remember to have seen or heard a single avalanche from this quarter.
They were frequent in the summer of 1899, and, presumably, continued so un-
til the space between the lower névé and the upper became so filled in as to
prevent further slides. From all the evidence obtainable it seems most prob-
able that the major stratification planes in the Asulkan névé represent the breaks
between the successive year’s snowfall, and that a phase of deficient precipita-
tion closed in this region about the middle of the year 1898, since which time
the average annual precipitation has been in excess of the normal.

The disturbance of the ice noted upon the eastern stream does not destroy
the stratification, since it extends only part way across the stream and is not
intense. The strata are, however, more or less tilted and distorted. The lower
stratum is wedge-shaped, having apparently lost from its basal portion by sub-
glacial melting. The blue bands in this stratum are not parallel with its upper
surface, but cut it at angles of about 13° to 14°, being more nearly parallel with
the valley floor. In the stratum just above, the blue bands and stratification
planes were conformable. In general, the blue bands were found to be regularly
developed, quite in contrast with the stratification. The dirt stripes showed
well over the surface and margins of the eastern stream, some excessively thin
ones being observed and previously noted (page 54). About the nose, upon the
walls of some of the longitudinal crevasses, the blue bands were found to dip
back into the glacier at angles of 11° to 28°. About the eastern side they were
found to dip downwards and inwards, nine observations giving an average of
46°, with a range from 36° to 57°. The granules about the nose are small,
compared with those seen in the larger glaciers, and will average less than a
half-inch in diameter.

om,

a in; ra al clit th te til

ee

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. TO5

7. DRAINAGE.

Owing to the crevassed condition of the ice the surface streams are small,
dropping into the glacier, or to the bottom, before they can develop any size.
Over the névé area the water resulting from melting, or from rains, is at once
absorbed. Over the ice exposed portions, during hours of melting, small rills
and surface brooks come into existence, carrying water with a temperature of
32°. No lakelets were noted upon the glaciers, or about the margins, but upon
the col, lying between Castor and the Dome, Mr. Wheeler found a lakelet of
sapphire blue water. Under ordinary conditions there is practically no marginal
drainage. In 1904, back some 800 feet from the nose of the eastern ice stream,
a small flow was visible for a short distance. From each of the three noses there
issue two to three drainage brooks, those from the eastern uniting with one
another and with the drainage from Asulkan Ridge, after which is received the
central flow from the middle portion of the glacier. That from the western com-
mensal, along with the drainage from the hanging glaciers lying farther to the west,
cascades into the Asulkan Valley, forming the ‘“‘Seven Waterfalls.” The flow
from the eastern nose is the strongest and carries the most sediment, considerably
more than the Illecillewaet. It fluctuates in volume during the day, reaching its
maximum in the late afternoon, or evening, and being lowest in the early morning.
The combined drainage from the middle and eastern portions of the glacier,
along with that received from the Asulkan Ridge to the eastward, has cut a
gorge 30 to 4o feet deep through the soft schist. This has the appearance of
having been done since the withdrawal of the ice, but it may have been started
by a subglacial stream. Under high velocity and charged with sharp, glacial
sediment the cutting power of water must be rapid upon a soft schist. Its action
upon quartzite boulders is well seen in the bed of the brook from Asulkan
Ridge.

During the last week in August in 1904 and 1905, the average of 28 observa-
tions upon the temperature of the water from the eastern nose was 32.42° F., the
range being from 32.0° to 33.0°. Two observations upon the water from the
middle nose, upon leaving the ice, averaged 33.0°, while from the third nose it was
32.8°. Before receiving the middle drainage the temperature of the brook was 36.9°
and after the two had united just above the schist cut the temperature was 37.8°.
Passing down the valley some two miles, and receiving drainage from either slope,
the temperature at the bridge across the Asulkan Creek averaged 42.6°. The
water is here turbid but assuming more or less of a greenish cast. The stream
from the Asulkan Ridge before receiving the flow from the glacier was found to
average 36.5° (20-observations). These observations seemed to indicate that
the maximum temperature was attained between 11:00 A.M., and 1:00 P.M.,
and that as the volume of water increased as the day advanced the temperature
gradually fell. This is brought out in the table here given.

106 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

TEMPERATURES OF STREAM FROM ASULKAN RIDGE.

1904. Ig05.

Aug. 30. 7:10 A.M. Z5nae Aug. 28. 6:00 A.M. aioe
Aug. 26. 7:15 A.M. 35.6° Aug. 28. 7:00 A.M. RO a1°
Aug. 27. 9:10 A.M. 36.9° Aug. 29. 9100 A.M. 37.-0°
Aug. 31. 10:15 A.M. 36/132 Aug. 27. 11:00 A.M. Berra
Aug. 27. 1:10 P.M. hy cts Aug. 27. 2:00 P.M. 37-9°
Aug. 25. 6:05 P.M. 36.5° Aug. 27. 3:10 P.M. ee
Aug. 28. 6:15 P.M. 35.6° Aug. 27. 4:00 P.M. 37 10m
Aug. 25. : 6:50 P.M. 35.2° Aug. 27. 5:00 P.M. 36.7°

Aug. 27. 6:00 P.M. Bonga

Aug. 27. 7:00 P.M. 36.0°

Aug. 27. 8:00 P.M. 36.0°

8. FRONTAL CHANGES.

Points of reference for the study of the frontal behavior of the lower Asulkan
nose were established August 12, 1899, by the Messrs. Vaux and observations
and photographs have been repeatedly made by them since. At that time they
made an unsuccessful search for reference blocks previously marked by H. W.
Topham. One year previously (August 23, 1898) they had visited the glacier
and obtained a photograph from their ‘‘test rock,’ which was published, along
with a brief description of the glacier, in the paper previously referred to.1
A comparison of their test picture of 1898 (plate x11, figure 2) with that of
1899 showed a slight shrinkage in the height and a slight increase in the breadth,
‘“ while the position of the tongue had not changed to an appreciable extent.”
The ice fall appeared to be less and they note that the névé line was lower, the
glaciers upon the slopes of Castor and Pollux more active, giving rise to a
number of avalanches, which seemed very infrequent in 1898. In marking the
position of the tongue at the time of their visit in 1899 three rocks were selected
in a line with the nose, the magnetic bearing of which was N. 85°35’ E. One
rock was located upon the small, left lateral moraine, a second just below and to
the right of the nose, while the third lay upon the inner side of the higher
right lateral. In 1900 the Messrs. Vaux observed a retreat of 24 feet and “a
marked shrinkage in every dimension.”’ From 1900 to 1901 these observers re-
ported an advance of 4 feet and for the two years rgo1 to 1903 an additional
advance of 36 feet. ?

This glacier was first visited by the writer September 17, 1903, at which time
it was found that the nose of the glacier lay 134 feet beyond the Vaux line, —
which was readily located by the two well marked end rocks. This would indicate —
that the nose had retreated 24 feet between the date vf Vaux’s measurement in
1903 and September 17 of the same year. The stone that had been marked
near the nose had been pushed forward some 14 to 15 feet, turned on end and was
about to topple over. Upon August 27, 1904, the nose lay 12} feet beyond the
line, indicating practically no change, when allowance is made for difference in

1‘*Some Observations on the Miccclemaee eat! Asulkan Glaciers of British Columbia,’ Proceedings of -
the Acad. of Sci. of Phil., 1899, p. 121.
2‘* Variations of Glaciers,’’ H. F. Reid, Jour. of Geol., vol. x11, 1905, p. 316.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 107

dates of observation. Exactly one year later (August 27, 1905) the nose had
made a retreat of 34 feet from the position held in 1904, standing now 21} feet
back from the reference line established in 1899. The nose consisted at this time
of a thin slab of ice, sloping to the west and coated with fine débris. A relatively
small amount of melting would cause a further recession of 30 to 35 feet. The
ice in the left lateral moraine was found to extend four feet beyond the reference
line and 254 feet beyond the nose. Owing to the rock cover it could not be
ascertained how much farther the morainic ice core extended. The movements
of this nose may be summarized as follows:

CHANGES IN THE NOSE OF THE ASULKAN GLACIER.
(Eastern Ice Stream.)

1898-1899. ‘‘Practically no change.”
1899-1900. Recession of 24 feet.
1g00o—1g01. Advance of 4 feet.
1901-1903. Average advance of 18 feet.
1903-1904. Retreat of 1 foot.
1904-1905. Retreat of 34 feet.
1905-1906. No change.

9. FoRMER ACTIVITY.

a. Development and decadence. At a much earlier stage, presumably in
Pleistocene time, the combined snows of the Asulkan Valley united into a great
Alpine glacier, the ancient Asulkan, which was a tributary of the ancient Ille-
cillewaet, and this, in turn, a tributary of the great trunk glacier that flowed
southward in the Columbia Valley, to the west of the Selkirks. With the diminu-
tion of snowfall, and possibly also an amelioration of the climate, the glaciers
disappeared from the main valleys and withdrew into the tributary valleys
and alongside the steep, higher slopes. An Alpine glacier occupied the Asulkan
Valley from the Pass to where the valley joins the Illecillewaet, some four
miles in length, which was in part nourished by a hanging, piedmont glacier
extending from the Pass to Mt. Abbott. This glacier sustained, during this
stage, the same relation to the Asulkan lying in the valley, that the hanging
Victoria sustains to the lower ice stream. The effect of these great ice masses
upon the valley floor and sides was similar to that already discussed for the
Victoria and Yoho glaciers (pages 61 and 80).

b. Bear-den moraines. Just before the complete and final separation of the
Asulkan from the Ilecillewaet, the Asulkan became loaded with very coarse,
angular rock fragments, and only a minimum of fine material. This was at the same
time that the Illecillewaet was similarly laden and, conjointly, they deposited
the massive bear-den moraine described upon page 96. The most of its
material was deposited upon its right, showing that it must have been received
from the western side of Glacier Crest and Mt. Lookout. The amount carried
was notably less than that brought down by its neighbor lying to the east of the
Crest. The ridges about the head of the valley and along the western side were
largely under snow and ice and could supply no such débris. After this load

108 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

of rock had been deposited the Asulkan began to retreat, withdrawing a distance
of 3,500 feet up the Asulkan Valley. The front now halted and there was built
a moraine of the ordinary type across the valley, consisting of fine and coarse
material, intermingled with but few coarse blocks. From this we conclude that
the glacier was, at this time, carrying the ordinary kind of load. The retreat was
resumed and in the meantime the glacier became a second time laden with coarse
fragments of the adjoining cliffs. At a distance of about 1,000 feet from the
previously formed moraine these blocks began to be deposited and were dropped
over a distance of some 500 to 600 feet, not so concentrated or imposing as the
outer bear-den moraine. For the next 1,500 feet these blocks were scattered
along the valley, implying that the supply over the surface had not been suf-
cient to bring about a halt. The retreat continued towards the head of the
valley and at a distance of 2,000 feet farther a halt occurred and a moraine of the
ordinary type was again built, with the usual quota of fine and coarse material.
From the time then that the Asulkan was about to separate from the Illecillewaet
it became twice loaded with coarse, angular fragments of quartzite, building a
moraine of the bear-den type. In the interval it carried material of the ordinary
kind found upon and within the ice and built a moraine of the ordinary type.
Subsequently to the formation of the second, straggling, bear-den moraine, it
has been carrying and depositing the usual class of material. It differs from
the Victoria in that the ancient moraine of the ordinary type was deposited
between the two bear-den moraines instead of outside the two, as in the case of
the latter.

c. Rate oj retreat. The only possible data for any estimates upon the rate
of retreat up the valley must be drawn from a study of the forest trees and no
one realizes any more strongly than the writer how unreliable and misleading
such data may be. However, we may obtain an approximate minimum estimate
by this means, which may have some interest, if not real value. Some excep-
tionally large spruces and hemlocks are found near the mouth of the valley and
within the outer bear-den moraine. Based upon the average thickness of the
rings of growth, noted upon page 96, two of the largest seen should be 525 and 598
years of age, respectively. Toward the schist cut, at the head of the valley, the
rings of growth become coarser and the trees smaller, the difference in elevation
amounting to about goo feet. One of the largest firs showed 161 rings of growth
and a still larger hemlock growing near was estimated to have lived about 250
years. Assuming thatit required about the same length of time for the trees to
get started at either end of the valley, it took the Asulkan about 350 years to
retreat the two miles from the mouth of the valley to the schist cut, or at the
average rate of about 30 feet a year. From the schist cut to the present nose, ~
about one-quarter mile, the valley opens and the retreat must have been much
slower, owing to the volume of ice to be melted away. If we assume that it
required 50 years for the hemlock noted to get started, the minimum time in-
volved would be 300 years and the maximum average rate of retreat for this part
of the glacier would have been about 4.4 feet per annum. If the cut in the schist

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE—SHERZER. PLATE XLII.

Fic. 2,—Disrupted quartzite blocks, near head of Paradise Valley, Canadian Rockies, Tlus-
trating plucking power of glaciers.

CLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 10g

has been made entirely since the withdrawal of the ice from that part of the
valley, then the rate of cutting would be over an inch a year, which is probably
too fast for even water at high velocity, charged with glacial sediment and oper-
ating upon rather soft rock. This seems especially true when we consider that
the supply of water is much reduced, or possibly entirely shut off during the
greater part of the year. It is very probable, however, that the narrow gorge may
have been largely formed subglacially, while the glacier extended far down the
valley. Schist layers upon edge do not well record ice action, but even if such
evidence of glaciation was present it may have been destroyed by subsequent
weathering. It must be noted that the time of retreat determined as above
would represent only a minimum value and the rate of movement per annum
for a definite distance would represent a maximum.

CHAPTER VIII.
SUMMARY AND CONCLUSIONS.

In the closing chapter it is desired to give a concise statement of the most
important results secured in the two seasons’ work and the conclusions reached.
The writer desires further to express for the benefit of those who may be interested
his conviction concerning some of the theoretic questions that have arisen in
connection with the study of these Canadian glaciers.

1. PHYSIOGRAPHIC CHANGES IN THE REGION.

a. Mesozoic peneplain. From the close of the Archzean to the end of the
Laramie, conditions were very favorable for the accumulation of sedimentary
deposits in the region now covered by the Canadian Rockies and Selkirks.
Strata belonging to the Paleozoic and Mesozoic eras of the world’s history reached
the extraordinary thickness, according to the work of Dawson and McConnell,
of 50,000 to 60,000 feet. Much of this was brought above sea-level”during the
Mesozoic era and further sedimentation ceased except in certain restricted
regions in the eastern part of the area,where conditions were still favorable for
marine or fresh-water accumulations. During countless ages of exposure
to the manifold atmospheric agencies there was developed a broad Mesozoic
peneplain, extending in a direction to the west of north and sloping east-
ward and westward, determining the general direction of flow of the drainage
streams. It was during this stage, probably, that the mountains suffered
their greatest denudation, rather than since. The great Laramide revolu-
tion of the western United States and Canada completed the formation of
these mountains, the pressure coming from the west in the region under con-
sideration, and producing a series of parallel folds and troughs, with numerous
overthrust faults, all having a north-northwest to south-southeast trend. The
upheaval was slow enough so that many of the original drainage streams were
able to maintain their general direction of flow, cutting their way continuously

Ilo GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

across the gradually rising ridges of the mountains and the lesser folds of the
foot-hills. In many cases, the troughs, lying between parallel ridges, or the
gaping crevasses in the rock strata parallel with them, captured the drainage
and a system of longitudinal river courses was developed, much younger than
the transverse system. Asa result of these great orogenic movements, combined
with the atmospheric and aqueous agencies operating since, we have an uplifted
and dissected peneplain.

b. Pre-pleistocene erosion. With the completion of the mountains at the
close of the Mesozoic, the more or less sluggish streams of the ancient peneplain ac-
quired velocity and renewed their activity, cutting deeply into their former beds.
The newly born longitudinal streams incised still further the channels provided
for them and there were developed two systems of V-shaped valleys more or less
intimately connected. The agencies of weathering broadened the valleys
above and delivered the rock fragments to the stream below, by which tools
the water still further deepened its beds. This action went along slowly from
the beginning of Cenozoic time to the beginning of the Pleistocene, during which
time the roughly angular blocks were carved into jagged peaks and many
of the divides into sharp-crested ridges. The outline of the old peneplain is to
be recognized only when one ascends until his eye is on a level with its uplifted
surface, when peaks and ridges all blend into the common level that cuts the
sky at the limit of vision. See plates n, xv1, and xxxu1.

c. Pleistocene erosion. The opening of the Pleistocene and the advent of
the glaciers introduced a new geological agent into the region. A reduction
in the mean annual temperature, combined with an increase in precipitation,
allowed the snow to accumulate about the higher peaks and ridges more rapidly
than it could melt away during the warmer season. Year after year the snow
banks thickened, sent their avalanches into the valleys faster than they could melt
away, and thus the mountains became enveloped in snow and ice. Glaciers
moved down from the more elevated valleys, joined forces with their neighbors,
grew in volume and power, took possession of the river valleys, and sent massive
tongues of ice far beyond the limits of the mountains. The valleys were filled
to depths of 4,000 feet from their floors, in certain cases, the actual elevation
rarely falling below 7,000 feet above sea-level. These ice streams exercised a
powerful effect upon the rock strata over which they passed; in general, rounding
and smoothing their outlines, cutting down prominences, and truncating moun-
tain spurs. In some cases where plucking was most active the rocks were made
still more jagged and irregular than the ice had found them. The lower half
of the valleys, which had been invaded by the ice, had their floors broadened
and their sides correspondingly steepened, giving this portion of the valley a U-
shaped cross-section. The upper portion, under less pressure of ice, still retains
more or less of its pre-pleistocene V-shaped form, the sides being simply smoothed
and fluted. The extension of these V-slopes until they intersect in the valley
may be assumed to mark the level, approximately, from which the glaciers began
deepening their beds. In the floors of the valleys at certain places rock-basins

= + -%

os oe eee le

SS Ewe

ores

£
%
5
|

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. Tit

were gouged out, either because of the structure or softness of the rock or be-
cause of the more vigorous ice action for a limited distance. At the heads of
the separate valleys, broad semicircular amphitheaters, or cirques, were cut out,
an interesting series of which is shown in plate xxx1x, figure 2, at the right side
of the view.

From observations made upon the plucking power of glaciers in the various
valleys, the writer is quite prepared to admit the sufficiency of glaciers as engines
of erosion, especially where the ice is very thick, concentrated in its action, and op-
erates for long time over stratified, or much jointed formations. In addition
to the plucked mountain peaks observed in the Yoho Valley (page 79), there
is to be seen in Paradise Valley, lying between Lake Louise and Moraine
Lake, a very striking case of plucking, in which very heavily bedded quartzite
has been bodily removed. The upper stratum is to to r2 feet thick, and about
the margin of the stratum, the upper surface of which is very perfectly glaciated,
immense blocks, some of them as large as small houses, have been started a short
distance and then left. Apparently in the waning stages of the glacier the
ice had been unable to get hold of blocks which it had been able to pry loose
from the parent bed. This occurs near the head of the valley and it is difficult
to resist the conclusion that hundreds of feet of similar, or less resistant rock
may have been removed between this ledge and the mouth of the valley. Glaciers
with their basal layers shod with hard rock débris would be able to erode slowly.
The amount of erosion accomplished in this way would depend simply upon the
time, but it has probably always been small, when compared with that due to
plucking. It seems likely that pure ice can have only an insignificant effect
upon ordinary rock, when simply pushed across it. The greater the pressure the
more the melting, and unless disruption of the rocks occurs, the only effect
would be to give the rock a polish.

d. Pleistocene deposition. During the maximum stages of glaciation there
were so few overtowering cliffs above the névé fields and the ice streams them-
selves that very little supra- and englacial material was carried. In consequence
during the stages of halt, that must have succeeded one another in the retreat
from the outermost position attained by the trunk streams, no great, conspicuous
moraines were formed. Not until the glaciers had retreated to near the heads
of the valleys do we find prominent terminal moraines. At this stage the level
of the ice and snow has dropped below the cliffs so that it is possible for the
glaciers to acquire, in many cases, a heavy load of rock débris upon their upper
surfaces. Glaciers like the Yoho have still been unable to build prominent
moraines from materials carried supraglacially. The detritus resulting from
the destruction of rock strata in the valleys and over the rocky slopes was
carried near the bases of the glaciers, or pushed and rolled along between the
ice and its bed. This resulted in the making of much ground moraine, much
of which remained in the valleys in places favorable for its lodgment. During
the maximum stages of glacial development much of this subglacial material was
carried beyond the mountains and deposited as till, or it found its way into the

II2 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

rapid streams, where it was immediately assorted into boulders, cobbles, gravel,
sand, and clay. In these various forms it was built into the terraces, flood plains,
deltas, etc., which characterize the drainage streams. The finer materials made
their way to the Pacific and Hudson Bay. The argument against great glacial
erosion that the material removed cannot be found, seems to the writer to carry
little weight. If one looks far enough and is able to distinguish the Pleistocene
and post-pleistocene deposits from the earlier, it seems probable that enough
would be located to restore the mountains and valleys to the condition in which
the glaciers found them. In the Bow and Cascade valleys, near Banff, Wilcox
discovered two distinct till sheets, indicating that there were, at least, two main
advances of ice through this section of the mountains. Eastward from the moun-
tains McConnell and Dawson found three such sheets, derived either in whole
or in part from the Rockies.

2. PRECIPITATION.

a. Geographic distribution. Owing to the north to south trend of the four
mountain systems that here constitute the great Cordillera, their limited
_ breadth, their nearness to the warm waters of the Pacific, and the relation of the
region to the great cyclonic areas that enter from the Pacific, conditions are
favorable for an abundant precipitation upon the western slopes of the moun-
tain systems. The arrangement of the four systems being such that they
increase in height successively from the Coast Range to the Rockies, enables
all of them to get a fair share of the available precipitation. The prevailing
winds are from the west and laden with moisture. In ascending the windward
slopes much of this moisture is precipitated as rain, or snow, owing to the expan-
sion and consequent cooling of the air. In the condensation of this moisture
its latent heat is liberated and raises the temperature of the air. In being
drawn down the leeward slope by the general cyclonic movement of the atmos-
phere, the air is still further warmed by the compression to which it is subjected,
its capacity for holding moisture is increased, and it reaches the same elevation
upon the leeward slope much dryer and warmer than it was at the corresponding
level upon the windward slope. This gives rise to the well-known “chinook
wind,” the equivalent of the Alpine foehn. The Selkirks, lying to the west of the
Rockies, receive the heaviest precipitation, are more completely forested, expe-
rience more frequent avalanches of snow, and send their névés and glaciers
to lower levels. The shifting of the centers of the cyclonic areas to the south
of this region would give rise to prevailing easterly winds, which in the winter
would be colder and dryer and in the summer warmer than those which now
prevail, and, without doubt, bring about the disappearance of gincicts from this
part of the mountains.

b. Climatic cycles. Precipitation records are too scanty and fragmentary
for safe generalization concerning the occurrence in this region of oscillations
known to occur in the other parts of the world. Still there are several lines of evi-
dence which indicate that a phase of reduced precipitation closed in the Selkirks

oleae

wires y

oh, =f

.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 113

and Rockies about the year 1897 or 1898, and that since then the average
for the series of years is in excess of the normal. These lines of evidence consist

(1) in the records kept by the station agent at Glacier House of the snowfall,

and of the records of the Canadian Meteorological Service for Banff and Calgary.
With the exception of Agassiz, which appears to be one of Brtickner’s ‘‘excep-
tional coast stations,” the other records do not reach far enough back to be of
service. (2) The notes and photographs of the Messrs. Vaux in the Asulkan
Valley, made in 1898 and 1899, when contrasted with those of later date, indi-
cate the close of a series of years with 1897--8, during which the snowfall was
much less than since. (3) The thickness of the strata in the névé of the Asulkan
Glacier, assuming that they represent annual accumulations, indicates at the
point photographed in 1902 that three years of diminished precipitation closed
with 1897-8 and were followed by four years during which the average precipita-
tion was inexcess. (4) In 1883 and 1884 Dawson found over a belt of country
r40 miles long in the western part of the Rocky System, evidence of a recent
high-water stage of the lakes, which resulted in the killing of trees that must have
grown during a prolonged low-water stage that preceded. The condition of the
trees indicated that they had “‘ been killed within a few years.” If we assume that
the wet phase that gave rise to this condition of the lakes closed about the year
1880, then we should expect the inauguration of another wet phase about the
years 1897 or 1898. Finally (5), from the photographs that have been made of
the Illecillewaet Glacier we have proof of the long-time oscillations of the level
of the ice about the névé line, giving rise to

c. Ice waves. When photographs taken from the identical view-point in
1888, 1897, and 1905 (plates xxxvi and xxxvil) are compared, they show a
marked fluctuation in the height of the ice along the sky-line. The ice appears
to have been at a minimum about 1888 and to have been approaching the
same condition in 1905, possibly attaining it during the current season of 1906.
The crest of a wave, or impulse of ice from the névé appears to have reached the °
sky-line about 1897 to 1899. The time from trough to crest would represent a
quarter of the period, that from trough to trough, half of the period of the com-
plete oscillation of the ice wave. In this case our data would indicate a period
of 36 to 40 years, agreeing well with the precipitation cycles to which these ice
waves are to be ascribed. It has been shown by the work of Finsterwalder,
Blimcke, and Hess that the advance of a glacier is due to the progress of an ice
wave along its length, moving more rapidly than the ice itself. The Ilecillewaet
in 1887 was experiencing about its nose the last stages in the effect of the arrival
of such a wave. The trough, then at the sky-line, moved valleyward and _ per-
mitted the retreat of the nose of the glacier, which retreat was probably at its
maximum about the year 1895 or 6, or some 8 or 9 years after the start. The
data for tg02 to 1905 seemed to indicate that an advance was about to be
inaugurated but the very marked retreat of 1905-6 shows that this ad-
vance has been somewhat delayed. When later the year is known at which date
the advance was most rapid we may figure the rate at which the wave travelled the

ri4 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

length of the glacier. It is a matter of much interest to try to connect the
crests and troughs of these ice waves with the corresponding wet and dry phases
of the precipitation cycles noted above. Since the crest of the wave arrived at
the sky-line in 1897-9, just at the close of the dry phase and the beginning of a
damp one, the wave must be referred back to the damp phase of the preceding
climatic cycle, closing in the late 70’s, or early 80’s, so far as we may judge from
the observations of Dawson upon the level of the lakes in the western Rockies.
This would give the “‘reservoir lag’’ of one-quarter of the period, required by
Reid’s calculations, and an additional 16 or 18 years for the impulse to reach the
crest of the rim. The trough resulting from the dry phase closing in 1897 appears
to have moved out from the reservoir more promptly, possibly owing to certain —
local conditions.

3. PrepMONT Type OF GLACIERS.

Three representatives of this unusual type of glacier were found, two of which,
the Wenkchemna and Asulkan, are here described; the third is the Horseshoe Glacier
at the head of Paradise Valley, in the Rockies. This type of glacier is always
compound, being made up of a series of glaciers of the common Alpine type, all
of coordinate importance, which coalesce laterally but retain their individuality
from névé to nose. Since none of them are tributary to any of the others, but
independent in all essential respects, they are here referred to as commensal
streams, in order to indicate this relationship. These separate streams have
temporarily united forces and found strength in the union. In the case of the
Wenkchemna it is very probable that very few, if any, of the commensals could
exist separately. In its earliest stage of development the piedmont glacier begins
as a series of Alpine glaciers, either with or without tributaries, lying in neighbor-
ing valleys. With the increase in precipitation the level of the surface of the sepa-
rate streams rises until they cover the divides between adjacent streams, or the, at
first, separate Alpine glaciers reach out upon the pied-mont and there coalesce
laterally. In its senile condition, a stage to be reached sooner or later, the pied-
mont glacier returns to its condition of youth and disintegrates into its component
streams, as illustrated by the Asulkan of to-day. In the case of the Horseshoe
Glacier some sixteen different commensal streams may be recognized, the most
western four or five of which have almost completely separated from the
others. The glacier has a meager snow-field, the supply for which is ava-
lanched from the slopes of Mts. Ringrose (No. 10), Hungabee, Lefroy, and the
southern side of the Mitre. Observations upon the Wenkchemna showed that each
separate nose may have its own independent behavior and that the movements
of the glacier as a whole cannot be known unless data are collected for each
component stream.

4. Parasitic GLACIER.

From the hanging glacier upon the eastern shoulder of Mt. Lefroy there is
avalanched to the back of the Mitre Glacier quantities of snow and ice, falling

7

Oe oe

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. I15

a vertical distance of some 2,000 feet and accumulating along the base of the
cliff. The ice of the glacier is broken into fragments, some of it disintegrating
into its component granules and much of it ground into ice dust, destroying
completely the stratification and lamination of the hanging glacier. The ava-
lanches occur mainly during the late spring, summer, and early fall, and as a
result of the spreading of the fragments from sliding and rolling there is made ~
each season a stratum of ice similar to those ordinarily found in the névé region.
Regelation is complete and there arises what is known as a reconstructed, or
regenerated glacier, with its strata leading to and dipping towards the region
of accumulation. The weight of the ice here forces the lower strata to move
out at right angles to the cliff face and a forward movement is imparted to the
ice directly across the Mitre Glacier upon which it rests. This regenerated Lefroy
moves about one-half as fast as the underlying Mitre, so that before the latter has
reached the Victoria, the Lefroy has crossed to the opposite side of the valley.
Between the hanging Lefroy Glacier and its bed there is being manufactured a cer-
tain amount of ground-morainic material, which is incorporated into the strata
of the regenerated Lefroy, and moved across the valley as a result of its motion.
While this is taking place, however, the Mitre is carrying the entire Lefroy down
the valley and the actual motion of the débris is the resultant of these two
motions by which there is accumulated at the base of Mt. Aberdeen a great heap
of ground-morainic matter, with a dressing of angular material from the face of
the latter mountain. The ground moraine rests upon the back of the Mitre
and some of it is ridged parallel with its side, in which form it is dealt out to
the Victoria and constitutes the main bulk of its right lateral moraine. This
Lefroy Glacier is distinct from the Mitre, upon which it rests, in that it is a differ-
ent type, is nourished differently, has a different form, a distinct set of strata
unconformable with those of the Mitre, has a different direction of motion, a dif-
ferent rate of motion, and is accomplishing a wholly different geological work.
The glacier is parasitic in the sense that it is carried by its host and is nourished
from snow and ice that might otherwise be available for it. It is not parasitic in
the sense that it draws its sustenance from the Mitre itself.

It is probable that glaciers of this type are now, and have been, more common
than has been generally recognized. It seems likely that at a certain stage the
glacier in a hanging valley would sustain more or less of this relation to the trunk
glacier. By means of such a glacier we may account for the lateral transporta-
tion of materials across a valley and a transportation that would leave no record
upon the bedrock. If two distinct glaciers may occupy the same valley simul-
taneously, it seems probable that two ice sheets of the continental type might
be superposed, flowing in different directions, the upper delivering material to
the lower.

5. BEAR-DEN MORAINES.

The bear-den type of moraine is so exceptional that some special explanation
must be found by which we may account for the accumulation of coarse mountain

[16 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

fragments without the usual filling of fine materials. The size of the blocks
themselves is not so remarkable, knowing what a transporting agent a glacier
is, as the average size of the fragments making up the moraines. In the case of
four of the five glaciers studied, two of these moraines were found and only two.
The absence of them in the case of the Yoho is readily understood when the
lack of high cliffs is noted. Not one of the glaciers at the present time could
form such a moraine, no matter how prolonged the halt. The blocks are angular

and show no more glaciation than they might have received upon one face while |

they were in their original position in the cliff. The blocks were carried upon
the ice and were not pushed or dragged along in front of, or beneath it. The fine
material was not removed by the action of running water, as might have been
done in other cases; but was absent from the first. If we are to account for these

moraines we must load the glaciers with a mass ot exceptionally coarse blocks —

and only a minimum of fine débris. This cannot be done by assuming two periods
of excessive weathering for they would produce as much fine material as coarse,
and very probably a great deal more. The prevalence of the phenomenon pre-
vents our resorting to the ordinary rock slide for our explanation. In the case
of the Victoria it built a moraine of the ordinary type, then the two bear-den
moraines, and then the present modern moraine essentially like the first. The
Asulkan built its outer coarse moraine, then one of the ordinary type, then its
younger coarse moraine and subsequently a series of the common variety. An
examination of the various cliffs, in connection with each of the four glaciers,
from which the material was most certainly derived shows that they all have
a trend from north-northwest to west-northwest. A further suggestive fact
is that in all cases the bulk of the material fell to the eastward.

During the season of 1904 no plausible explanation occurred to the writer,
but upon leaving the field the idea of a double seismic disturbance came up and
was carried back to the mountains in 1905. It seems now to be the only explana-
tion by which to account for the phenomenon. Slipping may have occurred
along some of the numerous fault planes traversing the eastern Rockies in a
north-northwest direction and the region crossed by westerly moving seismic
waves. From cliffs having a general northwest-southeast trend, blocks, already
much weathered, would be detached and thrown eastward, comparatively few
falling from the westerly facing cliffs. Glaciers most favorably situated for
acquiring a load by this means, as the Wenkchemna, have the most massive
deposits of the nature described; those unfavorably related to high cliffs, as the
Yoho, appear to have made none. The great blocks detached by the earth
jars fell into soft névé, or upon the yielding ice, and were not ground into small
fragments as they usually are when they descend to the valley floors. The
protection afforded the ice by the material brought about a halt of the front,
until the blocks were deposited, when the general retreat was resumed.

It was reasoned that if the disturbances assumed reached from the Great Divide
to the Selkirks, then many other glaciers, equally favorably situated for acquiring
a load by this means, should show the same type of moraine, if they were not

> >

eer |

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. Tal 7,

tributary to other ice streams at the time. Further we should expect to find
occasional rock slides of the same age as the moraines and cliff débris that did
not reach the back of a glacier. The latter material not being concentrated
would be inconspicuous. In 1905 there was but little time for a general examin-
tion of the region but visits were made to the Horseshoe and Geikie glaciers.
The former lies between the Wenkchemna and Victoria, with its main extent
of vertical cliff extending to the northeast, but with a considerable portion
extending from Hungabee to the Wastach Pass, with a westerly to northwesterly
trend. Opposite this portion of the glacier there is a deposit of very coarse
blocks, that were dropped upon the crest and outer slope of a still more ancient
moraine, consisting largely of a stony till. The number, however, is very meager
compared with those in the Valley of Ten Peaks, and would call for no exceptional
explanation. A low ridge of coarse blocks occurs just inside, showing best about
the front of the nearly -detached western portion of the glacier, which is correlated
with the inner of the bear-den moraines. In the case of the Geikie Glacier, lying
at the head of Fish Creek Valley and nourished from the southern portion of the
Ilecillewaet névé (plate xx x11), no moraines of the type sought were found within
a distance of one and one-half miles of the nose. Although the cliffs are suffi-
ciently steep to have supplied the material their general trend 1s northeast-south-
west, i. ¢., in the direction of supposed earth movement, and they suffered
relatively little destruction. From the eastern face of Mt. Burgess there has
been dropped a mass of coarse rock, which more strongly suggests the morainic
deposit seen in the valleys than that seen anywhere else outside the reach of
the glaciers. In many of the talus slopes there are many coarse and fine
blocks, which look to be of nearly the same age, instead of showing the
gradation that we might expect. In their work referred to upon page 4,
Collie and Stutfield describe a mass of rock débris in the valley of the
Athabasca (page 126) that may represent one of these ancient coarse
moraines or a modern one in process of forming. In referring to peaks
Woolley and Stutfield they say, ‘““These two last mountains appeared to
have been conducting themselves in a most erratic manner in bygone ages.
A tremendous rock-fali had evidently taken place from their ugly bare lime-
stone cliffs; and the whole valley, nearly half a mile wide, was covered to a
depth of some hundreds of feet with boulders and débris. What had happened,
apparently, was this. The immense amount of rock that had fallen on the glacier
below Peak Stutfield had prevented the ice from melting. Consequently the
glacier, filling up the valley to a depth of at least two hundred feet, had moved
bodily down; and its snout, a couple of hundred feet high, covered with blocks
of stone the size of small houses, was playing havoc with the pine-woods before
it and on either side. In our united experiences, extending over the Alps, the
Caucasus, the Himalaya, and other mountain ranges, we had never seen indica-
tions of a landslide on so colossal a scale.” In a footnote they add, ‘‘ The re-
mains of a similar landslide were afterwards noticed blocking the outlet to
Moraine Lake in Desolation Valley.”

>

118 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

If the seismic theory furnishes the true explanation of these double massive
moraines, then we have a means of correlating the positions of the extremities
of all the glaciers showing them at these two stages in their history, also data
for determining their actual retreat since and their relative rates. Glaciers, that
were not tributary to others at the time, confined between steep cliffs, having
a northwest to southeast trend, may be expected to show such moraines. There
is the possibility that any particular glacier may have advanced since and have
overridden one, or both, as the Victoria and Wenkchemna have partially done
with the inner of their series. Numerous earthquakes must have occurred
during the long Pleistocene period, but the cliffs were so completely blanketed
in snow that we find no such records left in the trunk valleys. Similar moraines
should be found in other sections of the world, but they might originate from
the removal of the finer materials by running water, as well as by earthquakes
and simultaneous rock slides. As to the actual age of these moraines we may
only loosely speculate. The blocks look old and the schists and sandstones have
disintegrated more or less, im situ, but undoubtedly they were badly weathered
before the glaciers got possession of them. The age of the moraines is to be
expressed in centuries rather than thousands of years. Based upon our vegeta-
tion data we may conclude that the inner of the two moraines was completed
about five or six centuries ago and that the earthquake disturbance respon-
sible for it may have occurred two centuries earlier. The outer of the two
moraines seems to be about two centuries older.

6. SURFACE FEATURES.

a. Dirt bands, zones, and stripes. In Chapter III of this report the writer
has described and figured these three glacial features and has suggested that
certain terms, used rather indiscriminately for any one, be restricted to a single
feature. The first two are very often confused, one with the other, but are so
essentially different in their real nature, if not always in their appearance, that
they should be sharply separated and differently named. The dirt zones, or
simply the zones, when the foreign matter is not present to discolor them, are .
the outcropping edges of the strata of which the glacier is composed. They show
to best advantage about the nose and lower margins of the glacier that is suffi-
ciently free from débris, as broad, parallel zones encircling the lower extremity
and passing around to the sides where they disappear. They are usually convex
down-stream, but the form they assume is determined by the configuration of the
glacier’s extremity. In case the stratification in the glacier is absent for any
cause, there can be no zones seen. 5

The dirt bands are entirely superficial and result from the collection of fine
débris in long hollows or troughs that first extend transversely across the glacier,
but which become convex down-stream from the more rapid central motion of the
ice. They occur in series, roughly parallel and regularly spaced, and assume
finally a pointed, or hyperbolic form, which probably suggested to Schlagintweit
the term ‘‘ogiven.” The name “dirt band,” however, was originally assigned

Pe ei

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 11g

to them by Forbes, their discoverer, and is in more general use. The trans-
verse, parallel troughs, in which the dirt bands have their origin, arise from the
incomplete healing of .ransverse crevasses which occur at the crest of a steep
ice slope. The lips of a crevasse, exposed to intense solar action are rounded
more or less and when the crevasse closes there is left a trough which marks the
position of the original crevasse. Into this depression wind-blown dust collects
and is washed from the adjacent slopes. By absorbing heat this dust may em-
phasize the depression slightly and may render the ice somewhat spongy, as
pointed out by Tyndall. If the ice slope is too steep, a cascade results and the
ice is too much shattered to show the bands, or to allow them to form. If the
slope is steep, but regular, with much melting over the surface, the site of
the bands will be destroyed before any complete series can develop. Conditions
are most favorable for their production upon the face of a moderately steep slope,
which is immediately followed by a long stretch of gently inclined ice. They
sustain no necessary relation, whatever, to the dirt zones, being present when
the zones are absent. When both zones and bands are present they may be
comformable for a greater or less distance and may be difficult to distinguish
from one another. In the case of such a glacier as the parasitic Lefroy the zones
and dirt bands may be discordant and intersect at high angles. There is reason
for thinking that the dirt bands are produced annually, only the summer formed
crevasses furnishing the necessary troughs, while the few winter crevasses com-
pletely and perfectly heal in passing down the slope. If this proves to be the
case we have a means of determining the approximate yearly motion of the ice
along the slope and a clue to the extent of the longitudinal compression, or
extension, of the ice subsequently.

Where the edges of the blue bands, embedded in the more porous whiter ice,
outcrop upon the surface, particularly along the margins of the glacier pressing
firmly against the valley wall, there is developed a further miniature banding.
The firmer blue ice melts less rapidly than the more vesicular layers and a series
of parallel ridges and troughs results, the course, distance, and average breadth
of which is determined by the ice structure itself. In the narrow troughs the fine
dirt collects and the ice is marked with a series of delicate parallel dirt streaks.
Tyndall compared them with the marks left in a gravel walk by a garden rake.
Drygalski describes them under the name of Schmutzbander, but this term must
be reserved for the true dirt bands of Forbes. Dirt stripes suggests their appear-
ance and will enable them to be distinguished from allthe other dirt features. In
that they owe their existence to the actual structure of the ice they have some
relationship with the dirt zones, but in that the dirt of which they are composed
is purely superficial, they are more nearly related to the dirt bands. They are
to be seen at only a short distance, while the zones and bands are best brought
out from a distant, elevated view.

b. Differential melting effects. Under favorable circumstances an interesting
series of stages may be passed through by dust wells; dirt, sand, and gravel
cones; boulder mounds; lakelets and morainic ridges. This was first worked out

120 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

by Russell upon the Malaspina for the lakelets and boulder mounds, but it
applies also to the other features as well. Dust wells may persist through a
season, or a series of seasons presumably, the dirt patches to which they owe
their existence being continuously retained in the miniature wells. Although
very shallow at any one time, their total depth might measure many feet. -From
wind action and small trickles of water more dust is being added slowly and in
time there may be enough to protect the bottom, instead of causing its melting.
The dirt now appears at the surface and the ice beneath melts less rapidly than
the unprotected adjacent ice, giving rise to a miniature cone, marking the original
site of the well. Such cones are found of various sizes and covered with dirt,
sand, or gravel. By lateral melting the slopes eventually become so steep that
the veneering slides off, or it may be washed down by heavy rains and distributed
about the base of the ice cone. The bare ice is now attacked by the sun and a
hollow is produced where the cone stood, about the rim of which stands more or
less of the material by which it was covered. This material rolls and slides back —
into the depression as the sides are widened and steepened by melting. When
enough has been concentrated at the bottom and about the sides to prevent
further melting, the adjacent ice which has lost its protective cover, just in
proportion as the depression has gained, now melts away to a level with the
bottom and then still lower, causing the material collected in the basin to again
assume the form of the cone. The miniature examples of this action might pass
through these stages several times in the course of the season, while the boulder
mounds and lakelets would require many seasons for the completion of a single
cycle. In the case of a medial moraine, or a lateral resting upon ice of sufficient
thickness, the same stages may be passed through, except that when the material
is shed it assumes the form of a double ridge, between which the elongated
trough is developed and into which the débris may slide to produce a single
ridge again. In this way the superficial débris of a glacier may be subjected
to much tossing and bruising before it comes to rest in the frontal or ground
moraine. In the case of a débris-covered ice surface all that is necessary to start
the process is to have the material unevenly distributed, a little thinner or a
little thicker patch of foreign matter.

7. IcE STRUCTURE

a. Stratification. From a comparison of the thickness of the strata in the
Asulkan with the available records of snowfall it seems probable that the strata in
this glacier, as well as in the Illecillewaet and Yoho glaciers, represent the
annual accumulation of snow in the region. The fall snows are combined with
those of the following winter and spring, compacted by the summer’s melting
and rainfall into a white, porous stratum of granular ice. At any given place
upon the névé by means of wind action a stratum may, have gained, or lost in
thickness. Owing to the deposition of the snow in successive layers and the
periodic distribution of wind-blown rock débris, each stratum acquires a more
or less distinct lamination; conformable with the stratum itself. During the

——

Pe en eee

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. I21

summer melting the fine dirt is concentrated at the surface, forming a soiled
streak which contrasts strongly with the fresh snowfall of the fall. The water
resulting from the surface melting and rainfall sinks into the stratum and con-
tributes to the growth of the névé granules, forming a crust of different texture
and color, by which the strata may be distinguished when no dust is present. In
the case of a regenerated glacier, such as the Lefroy, the stratification results
from periodic avalanching of snow and ice during the late spring, summer, and
early fall. The strata may vary much in thickness and have no immediate con-
nection with the amount of precipitation. They may become charged through-
out with ground-morainic material and give rise to very distinct zoning. When
a glacier is fed in part by névé snow, and in part by avalanches from hanging
glaciers the stratification may appear very irregular, as in the case of the Victoria.
In passing an ice cascade the stratification and lamination may be completely
destroyed, or the uppermost strata may be destroyed and the lower more or
less perfectly preserved, as pointed out by Reid. It is not supposable that the
stratification could be thus destroyed and the more delicate lamination preserved.
In the case of the regenerated Lefroy the stratification is restored, after having
been lost, but it is not possible to restore the lamination completely, or regularly,
in the case of such a glacier. It should be noted in this connection that under
exceptional conditions shearing planes may be developed in the body of the
glacier which do not coincide with the limiting planes of the depositional strata.
In this way there may be acquired another type of secondary stratification
having no relation whatever to that which originates in the névé.

b. Shearing. Observations upon the oblique front of the Victoria in 1904
indicated that the upper strata were moving bodily over those upon which they
rested. The upper strata projected more and more daily, when there was not
enough additional débris in the lower to account for the phenomenon by differ-
ential melting. A small amount of sand and fine gravel, washed down from
above, collected in the lee of the upper projecting layers. Some days this was
in small enough quantity to accelerate the melting of a narrow strip of ice upon
which it rested, but quite as often melting was retarded by the material. At
one place where the shearing action seemed pronounced three heavy spikes were
driven into the base of the upper stratum and three corresponding ones in the
face of the subjacent layer. These spikes were six inches in length and were
driven horizontally into the ice until their heads were flush with the surface,
about eighteen inches apart. The average surface slope of the ice was 46° and
the vertical height of the ice 50 to 52 feet. The upper stratum had a thickness
of about three feet, the lower two feet, and each contained, apparently, about
the same amount of foreign matter, and this small in amount. At the beginning
of the observations the upper stratum projected 19.7 inches beyond the lower
(July 21), and by August 3, 25.6 inches, showing a gain in the 13 days of about
six inches of the upper beyond the lower. The spikes were visited daily and
reset and showed that while the upper stratum was advancing with reference
to the lower it was also melting back more rapidly, because of tts more exposed

122 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

position. The average daily melting about the spikes in the upper stratum was
0.23 of an inch in excess of that about those in the lower and proved that a differ-
ential movement of the strata was taking place.

c. Bluebands. It seems highly desirable to distinguish the minor stratifica-
tion seams, originating in the névé, from the blue bands, blue veins, or ribbon
structures, that have had an entirely different origin. The first step toward
such distinction is to have a separate term for each of the two types of structure
and the writer suggests that lamine be used exclusively for the minor layers of
which the strata are composed and that blue bands,‘ already in such general use,
be restricted to the structures commonly included under the term, however
they may have been produced. When they are each made the object of com-
parative study it should be possible to distinguish them. We should naturally
expect the laminz to become less and less distinct toward the nose, and to appear
continuous, while we find the blue bands there showing very typically and being
discontinuous. The same stratum might show both structures, either conform-
able, or cutting one another at various angles. In the case of simple glaciers,
Agassiz and Reid have succeeded in tracing the laminz from the névé to the nose.
The structures seen in the Canadian glaciers are blue bands, rather than
laminz, since they are developed in great perfection where the strata have been
completely destroyed, as in the case of the regenerated Lefroy and almost
obliterated as in the Yoho and IIlecillewaet glaciers. In general their position is
at right angles to what may be assumed to be, or to have been, the direction of
maximum pressure. They are seen best along the margins where the glacier is
closely confined between rocky walls, extending parallel with the sides, dipping
downward and inward at a steep angle. Beneath the medial moraine upon the
Victoria they are vertical to fan-shaped. At the foot of ice cascades they may
extend crosswise of the glacier. Having the same origin and being essentially
alike, it does not seem wise to use different terms by which to separate these,
such as marginal structure, longitudinal structure, and transverse structure, as
suggested by Tyndall. Contorted patterns and faultings are to be accounted for
by assuming differential movements in the ice after the formation of the bands.

When followed for a short distance, in either direction, blue bands are found
to thin out to an edge and disappear, showing that they have a very flat,
lenticular shape. Separate bands overlap and are felted together as are the
bands in a schistose or gneissic rock. They strongly suggest schistosity in rocks
and not stratification. The ice of which each band is composed is more compact,
more free from air bubbles, and a deeper blue than the ice in which it is embedded.
That they have been produced by pressure and stand at right angles to it, when in
process of formation, as demonstrated by Tyndall, seems most probable. That

1 The term band alone, or banding, as suggested by the glacial conference in August, 1899, is not fully
satisfactory since it does not distinguish this structure from the dirt bands of Forbes. The following terms
have been applied to this structure by various writers; Bandstruktur, Banderung, Blaubanderung, Blatter-
struktur, Blaublatterung, Blaublatterstruktur, blaue Bander, blaue Streifen, Schieferung, Schichtung,
parallele Struktur, stucture rubanée, structure lamellaire, ribboned structure, blue veins, blue leaves, blue
bands, lamellz, laminz, lamination, and stratification.

ee

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 123

these bands represent portions of the glacier that have been completely liquified
by pressure, allowing the air bubbles to escape, seems, to the writer, very im-
probable, for four reasons. (1.) Blue bands occur in the basal layers, parallel with

_ the valley floor. The thickness of the ice of an ordinary glacier is not sufficient

to induce general melting by its simple weight. (2.) If the granular structure of
the glacier is completely destroyed by melting, it cannot be reproduced by simple
freezing, and still the granules are best developed in the blue band areas. (3.) Oc-
casional granules may be found which extend from the blue bands into the ad-
jacent vesicular ice. (4.) Water freezing in cavities in the body of a glacier should
form a series of prisms, standing with their main axes at right angles to the ice
surfaces bounding the cavity. Such filled cavities are found in the ice but they
do not constitute blue bands.

d. Ice dykes. In connection with the Lefroy Glacier chiefly, there were
noted in the early summer what appeared to be former crevasses, filled with ice and
forming ice dykes in the body of the glacier. Some of these were cut by crevasses,
testifying to their greater relative age and suggesting that they might persist
from one season to another. A few of the dykes contained granular ice, the
granules being moderately coarse, and were assumed to have been formed by the
filling in of crevasses with ice avalanched from the hanging glacier upon Mt.
Lefroy. Most of the dykes, however, were completely filled with a double tier
of ice prisms, having their bases attached to the walls of the crevasse and
extending horizontally out into the cavity, at approximately right angles.
Generally the prisms met at the centre those from the opposite face of the crevasse
and their inner ends interlocked. Sometimes a space was left between the oppo-
site tiers of prisms. Ellipsoidal shaped spaces were also found completely filled
with radially arranged prisms meeting at the centre. The explanation given for
these features is that they were formed by the freezing of water in crevasses,
and other cavities, in the spring, or early summer, while the glacier still retained
a sufficient degree of its winter’s temperature. The water was supplied by the
early melting, or rains, and the freezing surfaces were the walls of the crevasse,
instead of the lower stratum of the atmosphere, as is usually the case. Since in
freezing, water forms a series of parallel prisms, with their axes lying, as a rule,
at right angles to the surface of refrigeration, these prisms have the abnor-
mal horizontal position, instead of the usual vertical one. Although the upper
part of the dyke may have been lost by melting, there was no evidence that a
horizontal stratum of ice had formed across the top from freezing induced directly
by the atmosphere.

Drygalski has argued that it is pressure that determines the direction
that the crystalline plates will assume when water is freezing, and that the
main prismatic axes will lie parallel with this pressure. In the case of
the ice of a lakelet or basin, after it has once been enclosed by the ice
cover, the under side will be subjected to an upward pressure owing to
the gradual expansion of the water as it is brought to the temperature of
freezing. But the orientation of the basal plates, parallel with the upper

[24 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

surface, began before the cover was completely formed and hence before
such pressure could have come into operation. As pointed out by Migge
they also assume this position when an opening through the ice cover is
artificially maintained. Mugge believes that the plates are simply floating
in their position of equilibrium and that the pressure has nothing to do
with the orientation of the plates. That this, however, is not the cause of
the orientation is shown by the position of the columns in the ice dykes
above described, where the plates have formed in a vertical position, while in
the case of the ellipsoidal water-filled cavities they have formed at all angles
between the vertical and the horizontal. In the case of the ice dykes the forma-
tion of an ice cover would have given rise to a lateral pressure, as well as an
upward one, and the position of the plates in the horizontal columns would have
been in harmony with the view of Drygalski. No trace of this cover, however,
was seen, and it seems probable that the columns would still have formed at
right angles to the cold walls of the crevasse, under none other than hydrostatic
pressure.

Some experiments still in progress in the freezing of water in variously shaped
vessels lead the author to believe that the basal plates are placed parallel to the
surface of refrigeration, independently of pressure or position of equilibrium.
The actual congealing temperature enters quiet water at right angles to this
freezing surface, regardless of its position, and each successive plane of mole-
cules in turn feels the effect of the crystallizing force. The result is that sheets
of molecules are successively frozen parallel with the requisite isothermal surface
as it slowly works its way into the body of the water. The orientation of the
plates is facilitated by the fact that in making the ice crystal the molecules ar-
range themselves more readily (because of the superior crystallizing force) in
the plane of the secondary axes than in the direction of the principal axis. This
is Shown by the form of the snowflake which has been produced supposedly under
conditions in which the crystal was free to grow in any direction, so far as the
supply of moisture and suitable temperature are concerned. As is well known
the molecules are arranged mainly about the short main axis in the plane of
the secondary axes. The principle is illustrated further by the frost crystals which
form upon the window-pane, with cold air upon one side and a relatively warm,
moist atmosphere upon the other. At first only a very thin layer of moisture,
parallel with the surface of the glass, can congeal, and in this layer the molecules
at once arrange themselves in the plane of the secondary axes. As the atmos-
phere supplying the moisture becomes cooled for some distance back from the
glass the crystals may grow more or less irregularly. That the cohesive force in
the ice crystal is much more powerful in the direction of the basal planes than
in the direction of the principal axis, is demonstrated in the experiments to be
noted later (p. 130). Pressures in a direction at right angles to the main axis
will cause the basal plates to slide over one another, as in a bunch of tickets,
but no such shearing action can be secured when the direction of pressure is par-
allel with this axis. According to the view of the writer the temperature con-

oe. ': ——=—-s> |. © -P

—_— = Se”) lll es oe

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 125

dition for the crystallization of the water is supplied successively parallel to the
refrigerating surface, whatever may be its position or form, and the molecules
yield to the relatively more powerful forces which are operative in the planes of
the secondary axes.

e. Glacial granules. Glacial ice, which has not been subjected to a melting
temperature, is firm, solid, and, apparently, homogeneous, except for air bubbles
and foreign matter that it may contain. It is brittle, breaks without cleavage,
and in quantity, when pure, has a rich blue color by transmitted light. Sub-
jected slowly to a melting temperature there is developed a system of delicate
capillary tubes, which form a network throughout all the ice affected, and ex-
tend into the body of the glacier a number of feet. These tubes outline the
granules, more or less perfectly, of which the entire glacier is composed. These
granules are irregular polyhedrons, of variable size, with curved faces which
interlock with one another. Ordinarily there are no spaces between them that
can be recognized and there is no cementing material to bind the granules
together. They are observed to increase in size from the névé to, the nose in any
particular glacier and there can be no doubt but that the granules formerly in
the névé are directly related to those seen in the lower part of the glacier. In the
Canadian glaciers studied the largest granules were seen in the basal layers
about the nose; the Asulkan, the smallest of the glaciers, having the smallest
average granules, and the Yoho, the largest glacier, having the largest average
granules. When subjected to considerable melting the capillary tubes become
irregular and very thin spaces open between the faces of adjoining granules,
allowing the granules eventually to fall apart, or to be easily pulled apart.

Each granule is an incomplete ice crystal, incomplete because its development
has been interfered with by the neighboring crystals. Belonging to the hex-
agonal system of minerals, it has a single main axis, which is also its principal
optic axis. In common with all known ice crystals it appears to be made up of a
bundle of very thin plates, placed with their flat faces together, the axis standing
at right angles to these plates. When the granules have melted apart the very
delicate edges of these plates, or more probably sets of these plates, may often
be recognized extending as delicate parallel lines about the granule and thus indi-
cating the positions of the planes of the secondary axes. These lines are known as
“Forel’s stripes.” They are referred to by Miigge as the “‘ Translations Streifung,”’
and were regarded by him as due to the partial shearing of the basal planes over
one another. They are found, however, in newly forming crystals of ordinary
lake or pond ice which have not been subjected to any shearing stress. Within
the body of the granule there are seen, at times, circular disks of excessive
thinness, with their flat faces perfectly parallel and all at right angles to the optic
axis. They are of silvery whiteness and appear like “flattened air bubbles,”
as they were originally described by Agassiz (plate v1 of Atlas, figure ro). These
are ‘‘Tyndall’s melting figures’ and are cavities, “‘vacuous space,” con-
taining nothing more than water vapor, resulting from the internal contrac-
tion of the water, as it changes from dits solid to its denser liquid condition.

126 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

The melting begins at certain points between the crystalline plates and spreads
in a direction parallel to them, instead of across, or through them. The planes
of these melting figures are parallel to Forel’s stripes, and either feature when seen
may be used for orienting the crystal. In addition to the stripes of Forel, there
is to be seen a very conspicuous system of parallel ridges and furrows, covering
the outside of softened granules, which can have no connection whatever with the
crystalline structure. The ridges are either continuous, or consist of a series of
regularly placed points, forming a wavy, irregular pattern about the crystal.
The appearance suggests that seen upon the inside of one’s finger-tips and thumb.
It shows itself when the adjacent faces of the granules begin to separate and is
due to differential melting at the surface, but it is far from clear what could give
rise to such a regularly irregular pattern.

By means of the polariscope it was found that there is a tendency towards the
orientation of the granules about the nose of the Victoria, Yoho, and Illecillewaet
glaciers, the other two not being tested. The Victoria shows distinct stratifica-
tion about the oblique front, the Yoho indistinct, and in the case of the Mlecille-
waet, the stratification about the nose seems to have been completely destroyed.
Vertical sections of the ice were prepared, cut crosswise and lengthwise, and these
were compared with horizontal sections and oblique sections. It was found that
there is a marked tendency to arrange the optic axes of the granules in the basal
layers near the nose in a vertical position, from one-fourth to one-third of them
being estimated to be so oriented. The cause of this orientation is not yet
apparent, but connected, undoubtedly, with the method of growth of the granules
themselves. In order to account for the orientation which he found in the
Greenland glaciers, Drygalski assumed that the granules were separately melted
and refrozen with their axes parallel with the direction of pressure, which he
considers at right angles to the strata. If it is true that the direction of pressure
determines the position that the crystalline plates will assume, and hence the
position of the optic axes, which the writer seriously questions, then the space
occupied by a single crystal, which has been completely melted, should contain
a large number of radially arranged prisms, each standing at approximately right
angles to the portion of the ice surface to which its base is attached. Owing
to the law of transmission of forces by a liquid the pressure is equal in all directions
whether this pressure arises from the weight of the superincumbent ice, or because
of the expansion of the water in the closed cavity just before freezing. Drygalski
is in error in supposing that the pressure experienced by the liquified granule is
vertical only, since, if confined, the water would press outward in all directions.
In case the position of the refrigerating surface, or surfaces, is the cause of the
orientation of the plates, then in the closed cavity occupied by the liquid granule,
there should be formed a mass of radially arranged prisms, similar to those
observed by the writer upon the Lefroy and by Agassiz upon the Aar. In
either case, the cavity should be filled with small radially arranged prisms and
not by a single crystal with its axis in a vertical position. This furnishes rather
conclusive evidence that granules and blue bands never have existed in a com-

ON Oe ae

Oe ae ee

av’?

EOE Foe

*

-

*

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 127

pletely liquified condition. That they may have been melted partly upon
one face and frozen upon another, or the water derived from one by melting
added to another granule, is in harmony with known properties of ice. Pulfrich
found that when an ice crystal was pressed against a wet surface of glass and
allowed to freeze the water between the ice and plate was incorporated into
the crystal, so as to make a homogeneous mass.

Much interest is attached to the methods of granular development, since the
more modern theories of glacial movement are more or less dependent thereon.
It has been shown by Emden, Drygalski, Crammer and others that when névé
granules are made into a water slush, such as might originate from excessive melt-
ing, or heavy rainfall, the granules grow in size and, under favorable conditions,
quite rapidly. In the névé as well as in the body of the glacier, however, there
must be a maximum limit which the granules may attain by this means for the
ice will presently become too compact for more water to enter and no space will
be left for the growth of individual crystals.! That further growth of the granules
does not take place by the simple freezing together of neighboring granules, is
conclusively shown by the homogeneous structure of the mature granule. That
new granules cannot originate by the complete and simultaneous melting of a
number of adjacent smaller ones, is believed to have been just shown in the
preceding paragraph. Of the various theories of granular growth remaining
we may recognize three divisions, based upon evaporation, melting, and “dry
union.”

First.—It has been shown by Chamberlin and Salisbury that in dry granular
snow, kept continually below the freezing temperature, certain granules will
grow in size at the expense of the others, presumably by the giving off and con-
gealing of water vapor.? In the porous snow of the névé it seems probable that
the principle would be operative and that the granules would diminish in numbers
and increase in size, even when not immersed in water. For the body of the
glacier, with the granules in such intimate contact, the authors do not believe
that evaporation and condensation can take place to any appreciable extent.

Second.—Making use of the principle of Thompson that ice may be melted
by pressure, without any change in temperature, many investigators, as Migge,
Drygalski, Chamberlin, Crammer, etc., have accounted for the growth of the gran-
ules in the main body of the glacier by assuming a partial or complete melting
and refreezing. Those granules which owing to their location are subjected to
the greatest pressure, or internal friction, or those portions of granules similarly
affected will melt, thus redistributing the pressure and allowing the free molecules
to attach themselves to the most favorably located granules. The liquefaction of
the granules may be confined to their outer surfaces, or, as Drygalski believes, take
place locally in the bodies of the granules. Chamberlin believes that the granules

1 See Hagenbach-Bischoft’s criticism of Forel’s infiltration theory, ‘‘Weiteres uber Gletschereis,’’ Ver-
handlungen der Naturforschenden Gesellschajt in Basel, v111, 1889, p. 822.

2 Geology, vol. 1, Chamberlin and Salisbury, p. 296 (Chamberlin, Peet, and Perisho). “A Contribu-
tion to the Theory of Glacial Motion,’ Chamberlin, Decennial Publications of the Univ. of Chicago, vol.

IX, p. 194.

128 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

are subjected to more or less of a rotary movement and’a sliding along their
limiting surfaces, by which the internal stresses of the glacier are undergoing
constant readjustment and the ice mass permitted to move under the in-
fluence of gravity. These views of granular growth would call for constant
changes in the form, size, and number of the granules and in their relative
position.

If the principles underlying these views—melting under pressure or friction—
were alone operative in granular growth there should occur a much larger number
of smaller granules mingled with the larger in the basal layers about the nose of
a glacier. Owing to the manner in which the granules are keyed together the strain
upon the smaller would be relieved as they diminished in size and would be
transferred to the faces of the larger neighbors. Lying in between the coarser
granules we should expect a considerable number of these smaller remnants,
but such occur only somewhat sparingly. The remarkably well preserved
blue bands in the basal layers about the nose of the glacier furnish conclusive
evidence it seems to the writer that the granules have not been destroyed since
the bands were produced and that they have not materially shifted their position
with reference to their neighbors. Upon the surface of the lower Asulkan
Glacier these bands were found so thin that thirty were included within the dis-
tance of four inches, several necessarily cutting across adjacent granules. Any
perceptible shifting of the granules as the result of sliding or rotation would give
rise to faulting of these bands, while their destruction by either slow or rapid
melting would cause abrupt gaps in the continuity of these bands. The preser-
vation of the depositional laminz from the névé to the nose would seem impossible
if the granules are being destroyed and reformed or rotated out of their original
position with respect to their neighbors.

Third.—The ‘“‘dry union” of granules described on page 40 of this report
accounts for the reduction in the number and an increase in their size toward
the nose of the glacier. According to this theory the molecules of the yielding
granule give up their own crystalline arrangement and without any apparent
melting are immediately incorporated into the body of the controlling. granule.
Heim’s view was that such a union could occur only when the main axes of the two
granules were placed in approximately parallel positions,! but the experiments
of Hagenbach-Bischoff showed that such union could occur regardless of the posi-
tion of the axes,? and this he regarded as the true cause of granular growth
in the glacier. This view was accepted by Emden in his prize essay, ‘‘ Ueber das
Gletscherkorn,”’ published in 1890. It furnishes the simplest theory of granular
growth, not of glacial motion; accounts for whatever uniformity exists in the
size of the granules, calls for no shifting of the granule relative to its neighbors,
and hence permits the continuity of laminz and blue bands. :

1 Handbuch der Gletscherkunde, 1885, s. 330.
2Verhandlungen der Naturforschenden Gesellschaft in Basel. Bd. vii, 1888, s. 192; Bd. viii., 1889, s. 635
und 821. Archives des Sciences physiques ef naturelles, T. xxiii., 1890, p. 373-

Bd te ee

~~

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 129

8. THEORIES OF GLACIAL MoTION.

The one question that continually arises in the minds of all glacial students,
with tantalizing frequency is—what is the nature of this glacial motion? Are
we any nearer an acceptable hypothesis than we were nearly seventy years
ago when the serious study of glaciers was begun? Possibly! Without attempt-
ing the discussion here of the various theories that have been proposed, the writer
desires to record his convictions after his Pleistocene studies in the Lake Erie
region and his four consecutive seasons about the Canadian glaciers. All are
now agreed that sliding, expansion by freezing or changes in temperature,
general melting under pressure and regelation cannot fully and completely
account for the known facts of glacial motion. Probably all will admit that,
under certain circumstances, every one of these factors may find its application.
In the névé region it is possible that a certain amount of rolling and sliding may
occur amongst the granules, producing some motion, such as we may see in a pile
of beans or peas. Farther down in the glacier, although the granules are inti-
mately interlocked, it is quite probable that they would permit of a certain
amount, possibly considerable, motion between their adjacent faces. If we
introduce the idea of a partial melting of the granules, those portions of them
subjected to especial stress, or friction, will yield under the action of gravity
working from above, or behind, and will permit other granules to yield. Upon
relief of pressure, the water will be frozen to the original granule or distributed
to neighboring granules, as discussed in the preceding paragraph, and thus
the movement of the glacier may arise entirely from the alteration and growth
of its component granules. We must assume that the heat necessary for the
partial liquefaction of the granules is developed within the glacier as the result
of pressure and friction and not that it is derived from the atmosphere, or the bed.1
However, any heat communicated from such a source will make it that much
easier for internal changes to take place. As presented by Chamberlin and
Salisbury, this theory of granular change accounts more satisfactorily for the
glacial phenomena observed than any other, in which no molecular movement of
the solid granules is assumed. ;

The original idea of plasticity ascribed to glaciers by Rendu and developed
by Forbes is based upon the conception that the molecules of firm ice will yield
continuously to a stress, without producing visible rupture. The stress may be
of the nature of a thrust, or of tension. This theory has been rejected by the most
prominent physicists who have turned their attention to the problem of glacial
motion, because of their unwillingness to admit that ice could possess this and cer-
tain other properties apparently inconsistent with plasticity. The difficulty so far
as rigidity alone is concerned is removed by our knowledge that such substances
as lead, tin, and iron may be made to flow by pressure, under ordinary tempera-
tures. As soon as direct experiments were made to test the plasticity of ice it

1 In a recent’ pamphlet entitled, “ThefViscous'vs.‘the Granular Theory of Glacial Motion,” Mr. O. W.
Willcox has endeavored to’show that the heat developed through pressure or impact of the granules would
be cuuducted away as rapidly as it is generated (p. 18).

130 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

was found that bars of ice frozen in a mould, or cut from-a glacier, may be bent,
elongated, compressed, and twisted, without visible rupture, even when kept
continuously below the freezing temperature. These experiments were made by
Main, McConnel, Kock, and Migge and show that solid ice, made up of a collection
of irregular crystals, is decidedly plastic. _McConnel calculated that the amount
of extension required of the ice in the Rhone Glacier, because of the more rapid
central movement when compared with its sides, amounted to 0.0029 millimeter
per hour, for each ro centimeters of length. In his experiments with bars of
glacial ice, but one of the three bars tested showed as small amount as this and
that for only a portion of the experiment. In the case of single crystals they
were found capable of continuous yielding without rupture, providing the pressure
was applied at right angles to the optic axis, the movement appearing to consist
of a sliding be.ween adjacent crystalline plates. When the force of compression,
or tension, was applied parallel with the axis, the result was exceedingly small, or
nil.! The verification of these results by other investigators leads to the conclu-
sion that ice is capable of showing a certain type of plasticity, although different
from that ascribed to it by Rendu and Forbes. An amorphous plastic substance
yields under a suitable stress in any direction without visible rupture. A crystal-
line substance which maintains its definite molecular arrangement will be
limited in the number of directions in which it may yield. If ice crystallized in
cubes it seems likely that it might have yielded without rupture in three direc-
tions. Had it crystallized in square prisms we may now conceive of a movement
in two directions. In the hexagonal system in which it actually crystallizes
the molecular cohesion measured in the direction of the main axis is of a suffi-
ciently different nature from that at right angles to it to permit of a gliding of the
basal plates without rupture and the destruction of their molecular arrangement.
This 1s plasticity in a crystalline substance. The experimental results were
obtained with moderate stresses and much below the freezing temperature. In
the case of a glacier under great stress and a temperature near the freezing
point it seems absolutely necessary chat the glacial granules should manifest
this property to a greater or less extent, regardless of the actual mechanics of
glacial movement.

A number of phenomena, noted upon the Canadian glaciers, has convinced
the writer that a certain amount and kind of plasticity is a fundamental property
of glacial ice. The complicated patterns occasionally shown by the blue bands
are such as might arise from plasticity, but not from melting or rotating of the
granules. When similar effects are seen in ordinary rocks they are commonly
referred back to a plastic condition of the matrix. It was noted that when the
difference in the rate of movement between the centre and margins of the glacier
is sufficiently small, there are no marginal crevasses to be seen. This implies that
the ice permits a certain amount of stretching, without visible rupture. If ice
were absolutely incapable of yielding under tension, any appreciable difference be-

1“ On the Plasticity of Glacier and other Ice,” James C. McConnel, Proc. Roy. Soc. of London, vol. 44,
1888, p. 331; “On the Plasticity of an Ice Crystal,” vol. 49, 1891, p. 323.

ee ee ieee te le tl eh) bt ii el
Sea ME Se CA RA

Rhlgae no" gum £y

ee a ee

‘>

Se Op a re

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 131

tween the rate of movement of center and sides must open marginal crevasses.
In the rock scorings seen near the nose of the Ilecillewaet, the frontal and lateral
grooves about the knobs and trails suggest very strongly a plastic condition
of the ice. In the case of the same glacier, but farther up between the ice and
the left lateral moraine, the ice was seen moving over a knob of bedrock not
fully exposed in 1904. Transverse crevasses formed above while the ice melted
below and there were formed strips of ice, 20 to 25 feet long, supported at either

-end. Moving downward these strips were suspended in the air, and in

the course of ten days in September one bar had sagged so as to be very no-
ticeable to the eye, without forming any crevasse large enough to be noticed, or
to permit of the destruction of the ice bar itself. This phenomenon seemed to
indicate that the ice could, to some extent, yield to a tensional stress.

Now that the question of the plasticity of granular ice has been settled by
experiment, what objections are there that may be urged against its application
to the glaciers? Crevasses and faultings indicate simply that there is a limit to
its plasticity, as indeed there is to the most typically plastic solid. The tendency
to flow must be greater in the basal layers, but it does not necessarily follow that
with the friction of the bed to combat, the velocity here will be greater than or even
equal to that of the upper layers. The hold which glaciers have upon rocks in
their basal layers is probably not a firm one. It seems to the writer that the
chatter-marks, crescentic gouges, the shape and often sudden termination of
the coarse striz, as well as the faceted condition of the boulders in the ground
moraine, all indicate that the glacier was persistent, rather than firm, and that
it very often lost its grip. All of the phenomena of rock scoring, the subglacial
fluting of the ice, the compression of the ice at the base of a slope, the phenomenon
of shearing, and the mounting of reversed slopes prove that portions of the ice
move bodily under the influence of a more or less rigid thrust from behind. This
necessary amount of rigidity in the ice is not inconsistent with the degree of plas-
ticity ascribed to it. When the flow is not sufficiently rapid at any point the
ice must be thrust forward bodily. The most forcible argument against this
modification of the viscous theory of glacial movement is brought forward by
Chamberlin. It would seem that the granules should be distorted noticeably in
the direction of flow. That this distortion is not more apparent may be due to
the complete mechanism of granular growth. It may be disguised by the shear-
ing of the granules in the direction of the basal planes and their later dry union
by the principle of Hagenbach-Bischoff.

g. CoLor oF IcE AND GLACIAL WATER.

The exquisite richness and variety of coloring seen in glaciers and glacial
lakes constantly arouses the wonder and admiration of those privileged to gaze
upon them. The colored photograph fails to reproduce it and it eludes the
brush of even the most skilful artist. An explanation of the cause of this color-
ation can scarcely fail to be of interest. In 1904 a study was made of several

132 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS

of the lakes by means of standard solutions of copper and nickel sulphate and an
instrument devised from a stereoscope. By using measured amounts of the solu-
tions and mixing with pure water it was possible to match the water and to express
its color as an equation, from which the depth and shade of color might be at
any time reproduced. The solutions were prepared by dissolving 30 grams of the
chemically pure salt in 100 cubic centimeters of distilled water and were placed

in a thin glass‘ cell,’ with parallel sides, the inside measure of which was - 4 “

8 millimeters. A reflector was so arranged that the water could be viewed
directly, while the fluid mixture was seen by reflected diffused light. The proper
proportions could then be obtained by experiment. The shade of color changed
somewhat by the condition of the sky, the position of the observer, the time
of day, and the strength of the wind. In the case of Moraine Lake the depth
of blue was too intense in the quiet of the early morning to be matched by
the pure copper sulphate solution, in a cell of the above thickness. A slight
breeze sprang up and lightened the shade sufficiently. The table below gives the
color simply at the time of observation and under the conditions then prevailing.

COLOR OBSERVATIONS UPON ROCKY MOUNTAIN LAKES.

Lake. Date. Time. Proportions in cubic centimeters. Sky.
Lake Louise Aug. 3. 8:45 A.M. 29 c.c. green + 10 c.c. blue + 65 c.c. water. Clear but hazy.
Emerald Lake Aug.16. 8:00 A.M. Io c.c. green + 10 c.c. blue + 50 c.c. water. Fair.
Moraine Lake Aug. I0. 10:00 A.M. 3 ¢.c. green + 30 c.c. blue + 23 c.c. water. Sunny.

As the season advances a marked change occurs in the color of the water of -
Lake Louise, there being much more blue in the lake in the early part of the sum-
mer and more green towards the fall. Mr. Robert Campbell informed me that a de-
cided change has occurred in the color of the water of Emerald Lake in the last
few years, it being more of an emerald when he first saw it and now he considers
it more of a turquoise. In discussing this change with Mr. Bell-Smith, the
Canadian artist, I learned that he had observed the same change since 1888, there
being a very noticeable increase in the amount of blue.

Some simple observations and experiments in the field made clear to the
writer the cause of the differences in color of the water of different lakes and the
changes that occur in the same lake. It had been remarked before, but it was left
for Bunsen to demonstrate that absolutely pure water is blue, as seen by trans-
mitted light... The colors with the longer wave-lengths, as yellow, orange,

1‘ Ueber den innern Tasaieneniaan der ‘peoudooumnnrecnes Erscheinungen Islands,” Annalen der
Chemie und Pharmacie, Bd. Lx11, 1847, p. 1. Previous to the time of Bunsen the illustrious scientists, Newton
Humboldt, Davy, Arago, and Forbes had speculated upon the problem but with only meager results. Later
Beetz, Tyndall, Bezold, Boas, and Aitkin had more or less completely grasped the idea of selective absorption
and gave us a satisfactory theory of the various modifications of the color of water in nature.

“Ueber die Farbe des Wassers,’’ von W. Beetz, Annalen der Physik und Chemie, Bd. cxv, 1862, Ss.
137 ZU 147.

The Glaciers of the Alps, chapter 6, Pt. 11, Color of Water and Ice, 1860, Tyndall.

Lectures on Light, Tyndall, 1877, p. 35.

Theory of Color, Wilhelm Von Bezold. Translated by Koehler, 1876, pp. 41 and 67.

“On the Color of the Mediterranean and Other Waters,”’ Aitkin, Proc. Roy. Soc. of Edinburgh, x1, 1882
PP- 473 to 453.

af

ae

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 13

Ke

and red, are absorbed if passed through water of sufficient thickness. While of
the colors at the other end of the spectrum, with the short wave-lengths, blue is
the one which water is chiefly able to transmit, violet and green being also trans-
mitted, but less perfectly. Bodies of pure water of a volume sufficient to absorb
the longer waves of light reflected from the bottom, but not so deep as to absorb
it all, will appear blue. This blue is not reflected from the sky, although the con-
dition of the sky will affect the tint. Lakelets in the névé, such as the one dis-
covered upon Sapphire Col by Mr. Wheeler, are a rich blue; those upon the ice
may be blue, or not, depending upon their freedom from sediment, being liable
to change rapidly. Moraine Lake (plate xxv) owes its exquisite blue color to its
purity and depth. Water in the form of ice possesses still the same power
to transmit the colors with the shorter wave-lengths, violet, indigo, blue and
green, with the preference for blue. If a mixture of these four colors, or of all
the others which compound white light, be passed through a block of pure ice,
of sufficient thickness, none but the blue will emerge. If no light whatever is
being transmitted through either ice, or water, it will look black, or will show
whatever color of light is being reflected from its surface.

From this blue as the fundamental and natural color of ice and water by
transmitted light we meet with many modifications in nature. Finely divided
ice, as snow and névé, presents innumerable reflecting surfaces from which light
of any and all colors is sent to the eye. The same is true of water lashed into
foam, or in any finely divided state, as fog, cloud. or condensed steam. In
ordinary light these forms of water and ice appear white, but in the gorgeous
colors of the sunrise and sunsets they transmit to the eye by reflection the greatest
variety of color. Névé begins to show a bluish tinge as soon as the transmitted
light begins to predominate over that which is being reflected. The water which
issues from the glacier is generally charged with sediment, and if this is much
in amount, its color will determine the color of the water of the drainage brook.
It generally appears a milky, or creamy, white but may be a dirty gray. With
the deposition of the coarser yellowish sediment and the retention of the very
finest, if the volume of water is considerable, the water assumes a greenish tinge,
as seen in the Asulkan and Illecillewaet streams. With the loss of this sediment
the stream acquires more and more of its natural blue. When this glacial sedi-
ment is introduced into a lake in sufficient, but not too great, quantity the water
becomes charged with finely divided rock particles in suspension. These
particles are able to reflect the longer waves of the spectrum, particularly yellow,
but also the closely related green and orange, while they very effectually cut out
the shorter wave-lengths giving rise to violet, indigo,and blue. The result is that
there are introduced into the water innumerable reflecting faces which are capa-
ble of sending to the eye only those colors that lie at the centre of the spectrum
and towards the red end. But the only light that is available for reflection is
that which has already passed through the water once and had its yellow, orange,
and red to a greater or less extent filtered out. That which remains to be
reflected by the foreign particles will pass again through the water and will suffer

134 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

still further absorption. Of the blue and green rays, which make their way —
readily through the water, only the green rays are reflected by the particles and
these alone reach the eye. The quantity of light is thus much reduced in amount,
but with the sun shining upon the lake the green becomes quite vivid if the water
carries the requisite amount and kind of foreign particles. If the foreign particles
were pure white they would be capable of reflecting all colors equally well, and
the rich blue of the water would be brought out in perfection.

In the case of Lake Louise, we have a variable amount of sediment entering the
lake during the year, and consequently a seasonal variation in the color of its water.
With the very slack drainage during the winter, the sediment, in considerable part,
settles and the water in the spring shows more of its own bluecolor. With the in-
creased activity and melting of the Victoria Glacier the supply of sediment deliv-
ered to the glacier increases as the summer advances and the water becomes a richer
and richer green. About the delta at the head of the lake the sediment is so abun-

-dant and lies so near the surface that the water is unable to absorb the yellow, and

the color of the sediment itself is seen with little or no modification. In the case
of the change in color noted for Emerald Lake we may infer that the drainage
stream at its head has been carrying less sediment than formerly. It is not
improbable that the diminished activity of the inlet may be connected with the
stage of diminished precipitation recently closed and that the rich emerald
green of the lake in 1888 and earlier was connected with the stage of increased
precipitation, which is supposed to have closed in the early 80’s. With an
increase now in the average annual precipitation it will be interesting to see
whether the lake returns to its former shade of color.

The introduction of green or yellow, organic solid matter, animal or vegetable,
into the body of water would have the same effect. If the lake is sufficiently
shallow and the bottom covered with green vegetation, or yellow sediment,
the water will not be able in the short transmission to cut out the green and this
color may appear in lakes of water free from sediment. About the margin of
Moraine Lake the water has a greenish cast for this reason. Organic matters
in solution quite generally give water a yellowish to brownish color, as seen
naturally in bogs and artificially in tea, coffee, cider, beer, etc. With the above
principles in mind one may infer from a glimpse of a distant lake the condition
of the water and the state of activity of glaciers whose drainage streams empty
into it. If wpon rounding the shoulder of Mt. Temple, in entering the Valley of
Ten Peaks, the first glimpse of Moraine Lake showed that a rich green had been
substituted for its superb blue, one might safely infer that the Wenkchemna
Glacier had begun to erode its bed although a neighboring rock slide might
temporarily give such a result. Even the names of lakes in a glacial region
are suggestive of the amount of glacial activity in the valley, such as Sap-
phire Lake, Turquoise Lake, and Emerald Lake.

The same principles of coloring apply to ice as well as to water. When highly
charged with sediment of any particular color, its own natural color is obscured
and the color of the sediment is sent to the eye with little or no modification.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. TS
If, however, yellowish sediment is distributed through the ice in proper propor-
tions the case is identical with that discussed for water, and the ice assumes a
greenish cast. In the case of the solid ice the sediment can not be assorted and
evenly distributed as in the case of water and hence only greenish patches and
streaks occur, just where conditions are favorable. It has not seemed appropriate
to any one to apply the names “emerald,” or ‘“‘turquoise,” to a glacier while
“sapphire” would not be very distinctive. By applying the principles here
set forth we may account for the coloration of glaciers, the lakes in their neigh-
borhood; the gorgeous pools of blue and green water in the Yellowstone Park
and similar regions, the blue color of the ocean and the seas, the green and final
yellow strip as we approach the shore and such phenomena as the blue and green
grottoes of the island of Capri. No one has yet satisfactorily explained how a
considerable body of pure water may appear limpid.

THE END.

7
*
+
°
\
oF 7. a
or
up ~., — ut; ne 7

te

* os Cie ~t) ~
. ce hee eA
7

SMITHSONIAN INSTITUTION LIBRAR

iy

3 9088 01421 0421