/^7

^ H :^

m m-

THE

JOURNAL

OF THE

COLLEGE OE SCIENCE,

IMPERIAL UNIVERSITY,

J AI>A]Sr.

VOL. VIL

^ m i^ ^ ^\i n

PUBLISHED BY THE ÜN1VEKÖ1TY.

TOKYO, JAPAN.

18 9 5.

MEUT KXVIIl.

Publishing Committee.

Prof. K. Yamagawa, Ph. B., Ri^akuhakushi, Dh-ectoi- of theCoWegerexoßdoJ.

Prof. E. Divers, WI. D., F. R. S., etc.

Prof. D. Kikuchi, M. A., Rigakuhakushi.

Prof. K. Mitsukuri, Ph. D., Rièakuhakushi.

^1^1

AU communications relating to this Journal should be addressed to the
Director of the College of Science.

CONTENTS.

Vol. VII., Pt. 1.

The Manufacture of Calomel in Japan. By E. Divers, M.D., F.E.S.,

Professor, Iraperial University. (Witfi Plates I-III.J 1

Oximidosulphonates or Sulphazotates. By E. Divers, M.D., F.E.S., and

T.B.k(ik,"e.C^., Rigakuhakmhi 15

Constitution of Glycocoll and its Derivatives. (Appendix: General theory
and Nomenclature of Amido-acids.) By J. Sakurai. F.C.S., Rigakuhakvshi,
Professor of Chemistry, Imperial University 87

Vol. VII., Pt. 2.
On the After-shocks of Earthquakes. By F. Ômori, BigakusU. (With

Plates IV-XIX J HI

Vol. VII., Pt. 3.
Mesozoic Plants from Kôzuke, Kii, Awa, aiid Tosa. By Matajirö

YoKOYAJiA, liigakushi, Rigakuhakiisld, Professor of Palaeontology, Imperial
University. ( 11 'ith Plates XX-XXVIU.) 201

On some Organic Remains from the Tertiary Limestone near Sagara,

TötÖmi. By Kyu, NisHiWADA. (With Plates XXl^. J 283

Vol. VII., Pt. 4.
Mercury and Bismuth Hypophosphites. By Seihachi Hada, Higaknshi,

College of Science, Imperial University 245

The Acid Sulphate of Hydi^oxylamine. By Edward Divers, M.D., F.E.S.,

Professor, Imperial University 249

Decomposition of Sulphates by Ammonium Chloride in Analysis

according to Fresenius. P'V Masxmi Chikashige, Higaknshi, College

of Science, Imperial University 251

Ewart Johnstone's Way to prepare Nitric Oxide. By Masumi Chikashige,

/f/r/rz/ri/s/i/. College of Science, Imperial University 253

CONTENTS.

The Acidimetry of Hydrogen Fluoride. By Tamemasa Haga, F. C. S.,

Rigakuha/iiishi, Assistant Professor, and Ynkiclii Osaka, Rigakushi,
Imperial University 255

On the Poisonous Action of Alcohols upon Different Organisms, By

M. Tsukamoto, NOgakiishi 269

Formulas for sn 9u. By 0. Sudö 288

Formulae for sn 10 u, en 10 u, dn 10 u in terms of sn u. By E. Sakai,

Student, College of Science, Imperial University 285

The Diagram of the semi-destructive Earthquake of June 20th, 1894

(Tokyo). By S. Sekiya, Ri(/aki(/t(i/cit.'ihi, Professor of Seismology, and

F. Omori, Ri(j(i/ii(shi, Imperial University 289

Vol. VII., Pt. 5.
Beiträge zur Theorie der Bewegung der Erdatmosphäre und Wir-
belstürme. Von DiRo KiTAO, 7^/. rhil., Professor für Physik und
Mathematik an der laiidwirthschaftlichen Facultät der Kaiserlichen
Universität zu Tokyo 293

The Manufacture of Calomel in Japan.

By

Edward Divers, M. D., F. R. S., Prof.

Imperial University.

(With Plates T-JÎT)

Introductory. — Calomel, in the fonu common in England and all
countries under Western civilisation, is now extensively used and is
even manufactured in Japan, under the name of Imnhö. But mer-
curous chloride is also iarü-ely used here, under tlie name of ' light
powder,' Irifun (Chinese, h'ni(jfuii), in another and very much older
form, which is of signal qmrity, and made hy a simple process as yet
quite unknown in Europe. I witnessed this interesting process from
beginning to end some years ago, and now make this publication of
it with full permission of the proprietor of the works I visited, Mr. H.
Kokubu, who has aided me in every way he could, and notably with
drawings, some of which illustrate this paper.

Historieal. — According to Terajima Hökyö and Ono Ranzan,
writers who lived in the last century, the Hrst-named perhaps a little
earlier, calomel was known in Japan as far back, at least, as the
beginning of the eighth centiu-y, having then been presented to the
Empress Gen-miyô ; but their authority is the Zohi Niliongi, reference
to which, Professor Haga, F.C.S., informs me, makes it clear that
mercnry itself, not its chloride, was the thing presented. In the time
of the writers above named, mercurous chloride was well-known and

9 EDWAKD DIVERS.

was manufactured in Japan, not only tit Isawa, a village in !«(', wliere
it is still made, ])ut also in the city of Osaka and in a town near it,
called Sakai. Mr. H. Kokului, manufacturer, tells me that records
exist at Is:iwa-nun-a of his family having carried on the manufacture
of keifvn there for the last three hundred years.

Far earlier, namely, in the tenth century Minamoto-no-Shitagö,
in his work entitled U uiiiiio-lhiijmho, niakes mention of a mercurial
preparation named köj'an or ' |)o\vdci' (jf mercury.' It is, however,
questionable whether this was mercui'ous chloride or mercuric oxide,
and therefore whether calomel was known or not at this time. But
since calomel, under the name of l-eifun, is mentioned l)y Chinese
writers even earlier than this it may be safely accepted that Japanese
knowledge of this Ixjdy is older than that in Europe. The Western
knowledoj-e of chloride of mercury dates from the first half of the six-
teenth century, liut tlie distinction between calomel and corrosive
sublimate was not recognised till near the end ofthat century.

Literanj. — The literature on Japanese calomel is meagre. Japan-
ese writers of the old school have contented themselves for the most
part with translating Chinese writings. Ono Ranzan mentions that
the Japjinese method differs from the Chinese in making use of water
in place of alum and other chemicals, in which he came near the truth.
The late Dr. Geertz, who in the Government service did much in
establishing Western pharmacy in Japan, treated of keij'nn in some
metal lurofi cal contributions he made to the Transactions of the Asiatic
Societij of JajHOt. AVhat he wnjte is contained in vol. iv (1875), and
consists of information almost exclusively about Chinese calomel, and
gained more from Chinese and Japanese writings than from any ex-
perience of his own. Concerning Chinese calomel English readers
have the Notes on Chinese Materia Medim, among the Science Papers
by the late Daniel Hanbury, F.R.S., edited by J. Ince. Hanbury

THE MANUFACTURE OF CALOMEL IX JArAX. 3

mentions, as the result of his own observation, the characters of kiiig/un
and its great purity but for the presence of minute, transparent,
acicular crystals of calcium sulphate. He refers to Porter Smith's
Contributions towanU the Materia Medica and Natural Histonj of China
for an account of the manufacture. Smith, however, takes his in-
formation solely from Pearson's account on ]). 59, vol. iii of Sir J.
Davis's work on the Chinese. I have not seen this book, but it is clear
from P<3rter Smith that Pearson, again, has only derived his informa-
tion from the Chinese Materia Medica, Pun-tsaou-kang-muh. aud not
from his oww observation, and it amounts to tliis : — Common salt and
mercury, (jf each one oz.; alum, 2 ozs.; or, salt, mercury, copperas, and
saltpetre, in some such prijportions ; are rubbed together and put into
an iron bowl which is then covered with a roomy earthen dish well
luted down. This is exposed to the heat of a strong charcoîd fire for
four or five hours, when water is thrown on the cover and the cover
taken otf. On its inner surface the calomel is found adhering in the
form of a beautiful, feathery, white sublimate. Ten parts of mercury
are said to yield about eight parts of calomel. Dr. Geertz's paper,
alreadv referred to, contains essentially the same account translated
from the Japanese version of the Chinese work.

Tastly, there is a paper, in the Japanese language, on the
manufacture of keij'un at Isc, which is the forerunner of the present
one. That pajjer appeared in 1887 in the Journal of the Tükijö
(lu'iiiical Socielii, l:)y Mr. T. Shiniidzu, ^l.E., F.C.S., my former
pu})il and colleague, ai-d it was his description to me of what he had
seen that led to my (jwn visit to Ise in company Avith Professor Haga
in the following year. In one or two points I have availed myself
of this paper to make my own account more com[)lete.

(Jf the sjiecinc properties of keifun. — Keifun is in very thin
minute scales, lustrous, transparent, and white (jr faintly cream-

4 EDWARD DIVERS.

coJüiired. It uiight be described as micaceous c;ilomeI, To the touch it
is soft and snujoth. Measured in buik, dry, it is four times as
voluminous, more or less, as the t'l'ound calomel prepared by the
European process, and can be readily scattered by a putf of the breath.
Kubbed hard in a porcelain mortar it gives the brown resinous streak
characteristic of calomel and the evidence therefore, accordiner to
pluirmaceutical authorities, of its freedom from corrosive sublimate.
Exposed to bright sun-light it gradually assumes a light brown
colour, a. colour, that is, having no ntHnity to grey or black. Moist-
ure does not seem to 'favoiu- this chnnge which is certainly not
owing to any reduction to metal. European cnlomel suffers a similar
change. Keifun. is free from corrosive sidjlijiiate, and from metallic
mercury.

Hanbury found selenite in Chinese caltjmel, and Geertz found
calomel of this form generally adulterated with selenite and mica, but
whether what he examined was ever Ja])anese and not always import-
ed Chinese calomel he does not show. I have found kcifini, as it came
direct from Ise, (piite free from adulteration, and ha\e not met with
any adulterated.

Of the material used in maldiuj calomel in Ise, Japan. — The materials
I'or making Japanese calomel are — mercury, an arenace(jus red clayey
earth, bay-salt, bittern or salt-mothers, :iiid air. The mercury is
im]X)rted from Europe, but in old limes is^said to have been found in
the neighbourhood of l.^e as cinnabar.

The eai'th, called mitsuchi (' seed -earth '), is all taken ïvom a
neighbouring hill, Shunakayama, and according to ^li-. Ivokubu,
many other clays have been tried in place ol it, always \vith Ijad
results. It is of a. rather lii»ht bricht red colour, which changes to a
duller and somewhat l)rown red on drying and gently heating the
earth, and to a light ordinarv bi-ick red bv a strong heat. As mined,

THE MANUFACTURE OE CALOMEL IN JAPAN. 5

the earth is seen to consist largely of colourless quartz grains. Besides
the quartz a very little biotite is seen sparkling through it. The fresh
damp earth does not form a compact mass, but u slightly cohering
aggregate of damp crumbs. This texture aj)pears to be due to the
earth being a mass of quartz in small grains from the size of a hemp
seed down to that of impalpable particles, held together by plastic
clay. For use that which does not contain coarse quartz grains too
abundantly is selected, and is made into briquettes and moderately
baked on the hearth of tlie tire-place under the calomel pots. These
briquettes are then as light and porous as the ])repared porous clay
used in Fletcher's gas-furnaces. The raw earth air-dried is readily
rubbed into its constituents by the fingers ; and the baked briquettes
very easily and ra])idly reduce to a soft powder, quartz grains and all,
in the ai^ate mortar. I'lie larger grains of quartz in the raw earth are
also verv brittle. I ha\e treated thus fullv of the mechanical charac-
ters of tlie earth, because probably nuidi of its efficiency is due to
them. l:>ut its chemical character also calls for notice. As baked
ready for use it contains in the thoroughly air-dry condition still ô
per cent, and mui'e of water. Before ignition it is almost entirely
decomposed l)y sulphuric acid, either in some days in the cold or
quickly by heat. It is also largely acted u})on by hot hydrochloric
acid, and heated in sealed tubes with this acid to 11^0-150,° it is almost
as fully de('om])oscd as Ijy sul|)liuric acid. It contains practically no
silica soluble in hot sodium-carbonate solution, ])ut after acid
treatment yields, of course, uiuch silica to this reagent. The com-
position of the eai-th, as found in use at the works, but rendered
anhydrous is as follows : —

g EDWARD DIVERS.

Quartz 38.4

Combined silica '24.2

Alumina 2(5.3

Ferrie oxide 10.5

Magnesia 0.2

The magnesia is only got by fusion of the finely ground earth
with alkaJi carbonate, and belongs to the particles of mica scattered,
through it. Only doubtful traces of plios])horic acid could be found
and, what is specially to be noted, no hnie wli;itever. The earth is
thus nothing but (|uartz, kaolin, ferric hydroxide, and a ver}'^ little
biotite, and is probably valuable to the calomel maker not only for its
highly porous texture, l)ut also for its negative chemical qualities.

The bittern and even the rough bay-salt ccmtain magnesium
chloride, and this rather than sodium chloride must be the source of
hydrochloric acid in the process. That air finds a graduated entrance
to the other materials by diffusion during the jjrocess, will become
evident from a consideration of the set-up of the a])paratus.

(.'/" the plant. — The apparatus for making 'bipanese calomel con-
sists of a tal)]e-fiu-nace supporting sixty cast-iron ]jots lined and sur-
mounted with the Shunakayama e;irth, on which rest, as covers, and
condensers and receivers of the calomel, unglazed chiy cups bottom
upwards. In PI. I. the furnace as it is when in action is seen from
the working side ; ten ])ots on the left side are shown still to be
charged and co\ered. The wooden step in front is to enable the
workman tu reach over the table easily when charging the pots or
em|)tying them. In PI. II, fig. 1 the furnace is seen from the back, or
firing side, and with the walls broken away to show its interior and the
method of firing. In 1^1. II, tig. 2 the mounting of the pots is shoAvn
in three stau'es bv sectional i)lans (jf the table.

THE .MANUFACTURE OF OALOMEL TX JAPA^'. 7

On :i smoothed dav hearth tlie walls of the furnace are raised in
cl:iy, building", in the three stones which frame tlie stoke hole (Fig. -).
The walls are 2.() ft. high and the enclosure is 7.6 ft. by 4.7 ft.
measured outside. The stoke hole is 1 ft. by 1 ft., but a little wider
than tliis at the base, and is without dooi-, Tlie table of pots and roof
of the furnace is construeted (PI. I and PI. II, fig. i) by laying a square
iron rod on each of the hing walls, and on these eleven cross rods also
square on which are to rest the flanges of the pots. The j)ots are
then put in position as close together as possible, hanging by their
flanges, in ten rows of six each, and plastic clay pressed int(j the o]ien-
ino-s left between the flano-es and the rods, and the rods and flanijes
covered in so that only the mouths of the pots remain visible, as
shown by the middle rows in Fig. '2. The furnaee clay being
thoroughly dry, it is deeply laid over with the red earth mixed with
a little bay salt and moistened with bittern in small quantity. The
pots are also tilled with the same moist red earth, exc-ept a central
cylindrical shaft (see the left side of the furnace-table in PI. I or the
right side in PI. II, fig. 2) reaching to tlie bottom of the pot, which is
left bare. The pot is 0.5 ft. dee]:> inside, and across its mouth, inside, is
0.45 ft. It is shown in PI. Ill, fig. 4. The shaft or cavity left in the
filling is 0. IS ft. in diameter, and is shaped by resting a wooden core
on the bottom of the empty pot, and then pressing-in the moist earth
round it, smootliing-oif the top, and dropping a perforated lioai-d over
tlie projecting core to hold down the earth Avhile withdrawing tlie core
which is then removed by its handle. The furnace is now ready for
work. It should have been mentioned that after the clay walls of the
furnace are built rhey nve framed-in witli wood to increase their
stability and to give support to a wooden back and to shelving above
the table, as seen in PI. I.

O/' tlie firing.- — The firing the pots is kept far below what are

8

EDWARD DIVERS.

iisnnlly niu-ardcd as fiiniacc heats. Tlu! fuel employed is wood, and
the Japanese are certainly clever in the use of this, in tlie old style of
fnrnaces, economically and effectively in firino- pots for hoilinii',
evaporatinu', distillini;-, or suhliming. Tlie method of hcatini;- is
seen in PL II, fi^'. 1. Five leni>-ths of tire -wood are ranged along the
hack and front w;ills on the hearth, generally raised at one end
b\' restinf 'on a lump of clav- In the ordinary working^ of the
furnace, as T saw it, the lieat from pi'evious work is sufficient to
kindle the fresh wood. The flame's rise up the sides and run over
the hottoms of rlie ])ots, IcaAing the central space in t,he cliamher free
from flame, 'flie air enters by the lower ])art of tlie stoke hole and
the ])roducts of comhustion escape, invisible, l)y its upper part, so
perfect is the combustion. At the time of first lig-htiiig the fire and
of irreg-ularities in stoking, some smoke is inia voidable, and to keep
the top and table free from this smoke and from aslies, a wooden back
is ])ut in above the table as shown in the figures. There is also a
wooden hood and fine above the stoke hole, to carry off any smoke ;
this is iKjt sliown in the figures, but is similar to those put up in
England and elsewhere, over the working doors of fui-naces to protect
the workmc^n from arsenic, aulpliur, or other noxious fumes. The
heating is so well effected that the pots two feet above the 1)urning'
logs are made siifficiently hot, barely red-hot at the l)ottom, and yet
the wooden frame on the outside of the furnace is not charred, and the
work-room is not unpleasantly wai'med. About three bundles or 40
lbs. of wood serve for one firing, and it is remarkable to see so little
fuel working so many pots.

Of the u'orldiui. — A compost of liurnt niitsvchi with about a fourth
of its weight of bay salt is made u]) with bittern into lumps the size of
larfjfe chestnuts. The furnace being hot enough, the initsiii-lii surface of
the tabic! is once lor all freely wetted by a watering-])ot, [)erhaps half a

THK MAXlTFAt'TURE OF OALOMF]!, TX JAPAN. 9

o'.'illon of wiitcr beinfr iisofl, nil of wliifli is nljsorlicd ; a lump nv two
of conijiost is (]rop])ed l^y tono-s (^r l)y liaiid into each pot in ra]»i(l
succession ; a very small s])oonfiil of men-nrv ponred into each pot,
the total <-liarire for the >jixty pots hoinu" somewhat less than one
pound avoirdupois, (more exactly f of a Ih.). and a clav cnp. hottom
up, i^laced over each ]^ot and adjusted hv u'entlv pressing', and tni'iiiiiLi-
it round slio-htly. The cup is thus made to fit neatlv <m the earthen
top without adhei'ino- to it in tlie least. The eups are thick and
nnglazed. hut hecome verv smootli inside hv use. They have an
inside diameter at the mouth of 0.5 ft. and a depth of O.l^') ft.

Thus arranii'cd. the ])Ots and eups arc left for three hours, and
during the latter part of this time the furnace is coolinii-. Wlien cool
enough, eacli eu]) is lifted in turn. and. witli two turns of a teather,
the hnt'nii or calomel, which fills it in the form of a sparkling network
of delicate crystalline scales, is transferred to a sheet of paper held
under it, and the cup, mouth downwards, placed on a shelf of the
lurnace ready for use in thenext operation.

T'he calomel, it will tlius he seen, forms 710 adherent cake in tlie
eii[). hut particles only loosely hanging together. So obtained it is
ready for the market, needing only to 1)e packed in small wooden boxes
for sale.

The spent lumps of earth and salt are liite'd out of tlie pots l)v
tlie tool shown in 1*1. HI, fig. ô, and when the furnace has become still
(^ooler. the fire is made up again, the furnace top freely wetted, and a
new o])eration set going as before. The furnace is Avorked twice
each day. Nothing could well be carried out with greater siinplicity
and less expenditure of labour, time, and fuel.

Ol' tlie ijiehl. — Tdie loss T am told is about sixteen per cent, of the
theoretical amount of calomel which is. I believe, about twice as much
as is lost in the ^Vestern ])rocess.

-j^Q EDWARD DIVKKS.

Expérimenta}. — If one of the cups is removed ii few minutes after
the operation has been started, much of the mercury is found m it as
a sublimate of tine i^loljules. mixed with only a Httle almost amor-
phous calomel, from which it would appear that the formation of the
calomel resuUs from reaction in the cup between th.' mercm-y in
va|)our and the active gases.

On dipping- into the pot, uncovered during the ])rocess, a glass
rod with a drop of water hanging to it and then withdrawing it and
testing the remainder of the water with potassium iodide and starch
no chlorine can thus be detected. The same is true when a drop of
solution of potassium hydroxide is used. Xor can the slightest
odour of chlorine be detected in the N^apours issuing from the un-
covered pot.

On passing air containing a little hydrochloric acid gas through
a tul)e in which mercury is freely boiling, sparkling <-aJomel is formed
close to and mixed up with the mercury.

Red earth which has been used in tlie process turns moist red
litmus-paper blue, while fresh red earth is neuti-al.

TheoreticaL — The nature of the materinls used and tlie observa-
tions gained bv the preceding experiments arc sufdcient to cstablisli
that the calomel is formed by a reaction Ijetween mercury vapour,
oxygen, and hydrochloric acid gas, in which along with mercurous
chloride, water is formed : —

4 Kg + 4 HCl + 0, - 4 HgCl + 2 \{,0

and that the formation takes place at a temperature near — above or
below — the boiling ])oint of mercury and nmch below that at which
«•alomel freely volatilises. The source of the hydrochloric acid is
cei-tainlv the magnesium chloride of the bittei'n and bay salt, which
heated in a moist atmosphere, even i]j the presence of sodium chloride,

THE .MA\UFACTUJ;E of CALUMEL IX JAPAN. l\

is, as i« well known, p:irtly (.'oii verted into magnesia and liydrocliloric
acid. Hence the alkalinity of used iiiilsuchi.

Tlie porosity of the walls of tlu' ai)[)aratus particnlarlv of the
layer of initsuclii on which the clay cup rests, nuist he more than
sufficient to allow enough air to enter during the working;-. I had
supposed that the hot hydrochloi'ic acid and air might, in contact with
the earth, have yielded a little chlorine, l)ut the temperature reached
in the process, normally worked, seems never to reach that required for
the liheration of chlorine.

The Chinese process, if cürrectlij da^crihctU ditters from that used
in Jajian in several material [)oints, one of which is that the mercury
is intimately ground up with the other materials, and one fails io see
what reaction can take place 1)et\veen it and the two others, namelv,
salt and alum. Heated, the mixture must gi\e ott* mercurv and
hydrochloric acid, and then these as in the Japanese process will with
air give the cahmiel ; but this is independent of the previous intimate
mixture of the mercury with the salt and the alum. An(jther p(jint
is that in place of the magnesium chloride of sea-water as the source
oi' hydrochloric acid, the Chinese are said to use alum, or coi)peras.
which with the salt will react to give hvdrochloric acid. A third
point is that the cover is said to he closely liUcd to the \vo\\. poi , which
must nearly exclude the air, «without \\ hich it is impossil)le to explain
the formation of the calomel. Perhaps this is the reason whv the
C'hinese process is said to take four or ti\e hours tiring, since this mav
give time enough for the needed oxygen U) diffuse through the cup
and luting : ir will also account for the fact, if it is one. that the vield
of calomel is markedly less in weight than the nierciu-v used. Ai;-ain
there are to Ije noticed the statements that the ii'on [^ot is exposed for
hours to the strong heat of a charcoal fire, and that the hot clav cover
is cooled by throwing cold water <jn it, statements which must he

12

l]r>\VAIir) DTVETÎS.

reuarded ;is frreativ exagü'eraied, if not eiToneou.s. Lastly, the
ca]()nu'l-"i\ ill"- \ai)i)iir.s are allowed to remain in eontaet witli tlie liot
iron of the ])ot instead of l)eiii,u' kei)t from it l)y the tliick lining of
earth provided in the Japane.se process, a contact which such va|)Ours
could not stand Avithout destruction. I think, tlierefore, that ^ve may
be foirlv doubtful whether any reliable description has yet been given
concerning the Chinese process, ^vhich Ave may expect to find to differ
little, il' at all, from the Japanese |)rocess, exce]»t in sali and alum
being used in ])]ace of the mother-li(pior of sea-salt, (jr ' water ' as Uno
Ranzan supposed it to be.

Oiic thing, lo wliich atfeiition mav be called, isthat tlie Chinese
are! stated to add s(*me nitre to a simijai" niixtnre wlien empIoye<l to
«ive con<)si\e sublimate. For that luMUg tlie case, it is seen that tree
chlorine, whicli woidd here be develo])ed from the salt, nitre, and
alum, is necessary ibr the j-rcdiK'tion of the highei" chloride, and that
air and hydr<-chl('ric acid can only yield the lower cld<jride. a difference
of much theoretical interest, and indeed of ])racti<'al moment also.

It is kiiown tliat re-snblimation of calomel generates some corro-
sive sulilimate. aiid, although aiithoi'ities are not (piite agreed as to
whether reaction occurs l)etween gold leaf and calomel va}»our, it is
liardlv to be doubted that such reaction does occur. Now I have
ibund tliat if in the Japanese apj'aratus the temperature of the cover
is raised sufficiently to volatilise much of th(; calomel, the remaining
calomel is no longer free from corrosixe sublimate. It must therefore
be borne in mind that the calomel formed in the Ja]'anese jti'ocess is
not, and cannot l)e, the result of true sublimation, Ijut of precipitatioii
as fist as formed from the tlu-ee gaseous bodies which give rise to it.
At the temperature at which mercury boils, calomel is either (piite
fixed or at most has a \apour of ex<'eedingly small tension. In the
two facts, that the three gases do not i-eact to vield corrosive sul)limate

THE IMANUFACTÜliE OF CALOMEL IN JAPAN. 13

and thnt the calimiel is not heated to its volatilisinfi- point, lie the
explanation and. at the same time, the assnran<"e. tliat .Japanese (and
Chinese) calomel contains no corrosive snhlimate.

Suniniarij. — The Japanese pre])are ealomel pure, ahove all thinus
free ti-om corrosive sublimate. They heat l)alls of porons earth and
salt, soaked in bittern, alono- with mercnrv, in iron pots lined with
earth. The heat forms hydrt^chloric a<tid from the masrnesinm
chloride in the bittern, and tlie mercury sublimes into the closely-
fitting but unattached clay covers of the pots. Air enters by diffusion,
and oxygen and hvdrochloric acid ,o-as act together in the hollow cover
on the vapour given off from the sublimate of mei'cnrv there formed.
The cover thus becomes filled, with a network of micaceous particles of
calomel, precipitated, at a temperatui-e below its subliming point, at
the moment of its formation. "' "

Tliis investigation of an intei"esting product of Japanese industy
has been carried out luider the authority of the Imperial University of
Japan. I cannot acknowledge fidly en(nigh the indispensable assist-
ance I have received from my colleague, Assistant- l*rofessor Haga.
F.C.S.

14

EDWARD DIVERS.

Description of PI. III.

Flo;. 1. Tongs ïov liftiiig and droppinp; the Innip« of wet compost into tlie pots.

„ 2. Bras.s mercury spoon.

,, 8. Clay cnp condenser.

,, 4. Iron fin-nace-pot.

,, Ü. Peel for lifting compost l>all ont of pot.

„ H. <}nill feather for emptying cnps.

,, 7. Wooden watev-pot, for wetting fnrnace top.

,, 8. Wooden coi-e and board for lining pots.

,, 9. Rectangular iron bars for pots.

Oximidosulphonates or Sulphazotates,

By
Edward Divers, MD., FR. S.

and

Tamemasa Haga, F.C.S.

Hy treatioir a solution of potassium nitrite and hyciroxide with
-ulplmr dioxide, Fremy, in 1845, discovered a series of salts of com-
plex composition, of which, without succeedinçf in getting" any definite
insight into their chemical constitution, he was yet ahle to indicate to
some extent a classification into tlu'ce grou])s : the sulphazatcH, the
sulphazotates, and the snlphammonate. He attached, and justly so,
special importance to the sulphazotates, the salts which are the subject
of the present pa})er, and from which by hydrolysis are derived the
oxijamidosulphotiatcs treated of in an earlier communication by us
to this Journal (3, 211, where will be found references to the
papers of Fremy, Clans, and Haschig, which will serve for the present
paper).

A slight diiferciice in procediu-e gave Fremy one or other of
two salts, which he named respectively hasie snlphazotate and nciiind
sulphazotatc, because fi"r)m tiieir chemical composition and their ready
passage into encli other he held them to be salts of the same acid. He
also described still more basic sulphazotates of potassium with barium
and with lead.

]ß E. DIVERS AND T. HAGA.

Clans in 1871 worked upon the two sulphazotates, and pive an
account very different from Fremy's of their composition, properties,
and relations to each other. Finding them inconvertible into each
other, he referred (hem to different acids and renamed them. He
brought to light the sulphonic constitution of these salts (which lind
been, however, foreshadowed by Fremy), and consequently named the
neutral sulphazotate of Fremy disnlphydroxyazate, while f )r the Intter's
basic sulphazotate he retained the name sulphazotate simply.

The sulphazotates were again examined in 1887 by Raschig, who
established their constitution as derivatives of hydroxylamine* and
made discovery of a potassium salt still more alkahne than Fremy's
basic salt. He, in his turn, differed greatly from Claus and found
Fremy's account of the salts in some respects more in accordance with
the ficts, but, on other grounds than those Claus had taken, retained
the distinction made by this chemist between the neutral sulphazotate,
which he renamed luidroxijlamive-distdpliotiate, and the basic sul])hazo-
tate, which he agreed with Claus in calling simply sulpha :otate. To
his own discovered third potassium salt he gave the name of basic
sulphazotate.

In the present contribution to tVie subject the existence of sodium
oximidosulphonates is established ; these and salts of ammonium,
calcium, strontium, bnrium, and lead ;u-e described ; much-needed
methods, definite and productive, for preparing both sodium and
potassium oximidosulphonates :ire given ; the reversion of these salts
to sulphite and nitrite made known ; and the interrelations of the salts

* Claus liad pointed out the hydroxylamiue derivation of oxyamidosulphonates but had
decided :i gainst such a derivation for the (neutral) sulphazotate. In a foot-note to our preliminary
paper on the Reaction hettoeen ttilphites and nitrites, J. Ch. S. 51, 659, we erroneously represented
Raschig to be not quite accurate in stating tliat Claus had so decided. We regret our error.
The facts are that while in one place, overlooked by us, in his several long papers Claus ex-
prossly makes this decision, hf in another place, indicated in our note, gives thf- foruiulœ—
< >NII(SOgK)2 and HON(S03K)2— as alternative, without deciding between them.

OXIMIDOSULPHOXATES OR SULPHAZOTATES. ;[ 7

classed apart by Claus and by Raschig shown to be such as tn demon-
strate the unity of their coiistifutioii as oximidosulphonates, or
Fremy's sulphazotates.

Preparation of sodium and potassium oximidosulphonates.

In practice two courses are open by which to proceed in prepar-
ing oximidosulphonates from nitrites : * to mix together solutions of
the nitrite and sulphite ; and to treat a solution of the nitrite and
hydroxide or carbonate with sulphur dioxide. The question, already
discussed by Claus, whether these methods are in [)rinciple identical,
need not here be considered, and will l^e taken up in a future paper.

Fremy succeeded only by the second of these methods in getting
oximidosulphonates. Claus found both successful but the second to
be much more productive. Raschig recommended ÇAnnalen, 241)
the first method as being the more convenient and productive when
sodium salts are worked with, neutral potassium oximidosulphonate
and also an alkaline potassium-sodium oximidosulphonate being tlien
got by double decomposition wäth potassium chloride. Only in-
cidentally, in discussing Claus's views, did he mention that the second
method of getting tlie potassium oximidosulphonates is occasionally
successful. He included, however, this method in the specification of
his patent (1887)** for the manufacture of hydroxylamine. Both
methods have been investigated by us, and in what follows it will be
seen that we have found the second method to be much the better one,
whether for the potassium or for the sodium salt, but that the first
can be made far more successful than it has hitherto proved to be.

* It has been shown by Raschig that oximidosulphonates are also obtainable from nitric
oxide, namely, by first converting it into nitrososulphonate, Pélouze's salt, and then letting this
decompose in alkaline solution.

** Our references concerning the patent are abstracts in the Berichte, J. Cli. S., and
J. S. Ch. Ind.

E. DIVEKS AND T. HAGA

Oximidosulphcntaics prepared by the direct use
of sulphite.

Introduction. — Claus supposed himself to use the norma] or
dipotassmm sulphite but, it is clearly evident, used in reality mainly
the metasulphite. The 'neutral ' (or meta) sulphite therefore is to be
taken in the proportion of ' Jess than four' molecules, or "2 KoSaU,,, to
one of the nitrite, KNO. ; Avith more than tour molecules only nitrilo-
sulphonate — N(S03K)3 — separates, and even with less than four much
of this salt is produced ; both salts are almost insoluble, but the l<jose
tine needles of the nitrile can l)e mechanically washed away from the
large crystals of the oximide ; the solutions used are not to be too
dilute; the yield of oximidosulphonate is very small. We can confirm
Claus's account of this very unproductive process.

To prepare the neutral salt Kaschig runs into a mixtm-e of the
solution of one molecular proportion ofsodiimi nitrite with ice, a solu-
tion of two molecular proportions of sodium disulphite (/. t'., one mol.
of metasulphite), adds two molecules of potassium chloride in (^old
saturated solution, and leaves the whole to crystallise. Nitrile forms,
as well as oximide, and the two salts are to be separated by elutriation,
as directed by Claus. The yield of j)otassium salt is stated to be about
half the calculated quantity. No mention is made in the memoir in
the Anualen, or in the papers in the Berichte, of a similar proceeding
with potassium nitrite and sulphite to get the potassium oximidosul-
phonate direct, the above indirect process obviously being resorted to
as superior to it. But in the specification for the patent, which of
course preceded the publication of the memoir, Raschig did give the
direct method. His process has proved successful in our hands but
less so than in his.

OXIMIDOSULPHONATES OR S ULPH AZOTATES.

19

The saine investigator records tlie preparation, from sulphite, of
a potassiimi-sodiuin oxiniidosulphonate haAing- tlie basicity of Fremy's
basic potassium siilphazotate. Having on one <3ccasion poured com-
mercial soluti(jn of sodium bisulphite upon cold sodium nitrite he
found the mixture become boiling hot, and by addition of nmch potas-
sium chloride to yield nmch potassium -sodium nitrilosulphonate.
Then, during the twenty four hours following, the mother-liquor
deposited the new salt togetlier with some i)otassium nitrilosulphonate.
We have not been able to confirm his ex[)erience. The potassium-
sodium salts we can get, but according tu om- own observation, the
addition of alkali hydroxide, of which he says nothing, is requisite to
iorm an alkaline oxiniidosulphonate. And from Raschig's own sound
criticism (op. cit., p. :i'J'J) of Claus's method of getting the alkaline
potassium salt, it might be supposed that he too believed in this neces-
sity of adding hydroxide in order to form any basic oximide.

Préparation of the neutral potassium oximidosulphouatc j'non metasul-
phite. — To prepare the potassium oxiniidosulphonate by using meta-
sulphite we ourselves work from the first with potassium salts. The
solutions of potassium nitrite and metasulphite are used w4th ice float-
ing in them, the mixture is kept in ice for a day or more, and the ice-
cold mother-liquor is then drained off from the crystals which will
have formed. To allow, as Raschig would, as we understand from
abstracts of the specification of his patent, the temperature to rise even
up to 40°, greatly interferes, and gives such a result as Claus obiain-
ed, — very much nitrile and very little oximide. Some nitrile forms
even in the ice-cold mixture and has to be washed away fn^ni the
oximide crystals ; and wlien the mother-liquor is afterwards allowed lo
acquire the common temperature it produces and deposits more nitrile.
The ]jroportion of tlie quantities of the salts to be taken is some-
what more than that of three molecules of metasulphite to two

20

E. DIVERS AND T. H AG A.

molecules of nitrite, instead of one to one as when Rîischig's direc-
tions are followed. In the niother-liqnor remains much sul])hite but
only insignificant ([uantities, at most, of nitrite. In order to measure
with some degree of accuracy the qujintities of nitrite and metasulphite
taken, it is best to use good commercial nitrite, the real strength of
which in this salt is known from a previous assay ; and to prepare the
metasulphite, shortly before it is wanted, by passing-in sulphur dioxide
just to the point when the solution beromes neutral to methyl-orange or
to lacmoid paper. Some incimvenience in preparing the ice-cold
soUitions is caused by the sparing solubility of the metasulphite,
which entails the cooling of large volumes of solution. The yield of
washed crystals of the oximide is about 60 (63) per cent, of the
calculated quantity, and such therefore as seems not to have been even
remotely approached by previous vv(3rkers, although greatly surpassed
by that of the sulphur-dioxide method described on pj). '27 el seii.
By working at temperatures some degrees below 0° the pi'odnction of
nitriie seems to be further lessened and that of oximide somewhat
increased.

It is besides almost certain that the formation of t lie nitriie in-
creases relatively to that of the oximide as the nitrite remaining
grows less, the sulphite having more and more already -formed oximide
to work u]3ou along with less and less nitrite. This view of the
matter is supp(jrted by experience. For when the metasulphite
is added to an excess of nitrite, the crystals of oximide produced
amount, wlien washed frtnn the little nitriie accompanying them, to
80 per cent, of the quantity theoretic^ally possible from the quantity of
metasulphite taken.

Formation of tJie neutral sodiuni nxinudosulphonate from metasuU
fjhitc.-'Viy treating one molecule of sodium nitrite with two, or a
little less than two, of sodium metasulphite, the nitrite may be wholly

OXIMIDOSULPIIO.N'ATES OU SÜLPH.VZOTATES. 21

sulphonated, principnlly into oxiuiide })ut pnitiy into nitrile. To
prevent heating u]) the solutions are best mixed ice-cold, but the
mixture may then be exposed to the ordinary temperature and left
so for a day. This process though successful is of little practical
value because the very soluble sodium oximidosulphonate cannot be
separated from the sulphite and nitrilosulphonate which accompany it.
rhe actual success of the process in forming much oximidosulphonate
has been ascertained therefore only by analysis : the excess of sulphite
having been precipitated as barium salt, the solution has been hvdro-
lysed, and the hydroxylamine estimated. In this way evidence has
been got that there is formed about 83 per cent, of the oximidosul-
phonate equivalent to the nitrite.

The solution of rhe sodium salt thus obtained can be used to get
the potassium salt by double decomposition, like that prepared after
Raschig's directions, than which it is markedlv more productive and
gives a purer product. Raschig used only half as much inetasulphite
as we use and therefore, by our finding, left <me-third of the nitrite
undecomposed in the solution. He also added the potassium chloride
just after mixing. the sodium salts, and did not, it would seem, pre-
serve the coldness of the solutions necessary in presence of potassium
salts. By his method, consequently, much more nitrile is got than by
ours, in which the solution of the sodium salts is only treated with the
potassium chloride after it h;is stood a day, so as nearly to complete
the reaction, and then been cooled again in ice, to prevent as much as
possible the formation of nitrile which otherwise goes <jn, especially in
the potassium salt. Following this method the oximide is obtained
with so little nitrile accompanying it ;is onlv to become visible during
recrystallisation of the product. But these points are now^ of no im-
portance as regards the preparation of the potassium salt, since there
are according to our experience, the Tnu<;h better direct processes for

9 0 E. "DIVERS AND T. H Ar, A.

its preparation, one just iJ-iven (p. lî') and the other now to follow
(p. 27).

Oximidosiilphonates prepared by the iifie of siilplnir-dioxide.

Jiitrodvctiiuu — F'remy's process for the nevtnti jjotassium salt is to
pass siilphnr dioxide into a snitably concentrated solution of potassium
nitrite and hydroxide, until the s:dt separates. Should the solution,
however, he sufficiently concentrated, the basic oximidosulphonate will
make its appearance instead, he found ; and with higVi concentratioji
the Aulphazafe, and occasionally tlie wetnsiilpliazate will first form.
Then water must be added, enough to dissolve up any of these salts,
and the passage of sulphur dioxide be continued until the neutral salt
begins to crystallise.

Claus found Fremy's juvsci-iption too indefinite and uncertairj,
and therefore modified it, mainly so far as to employ equivalent
quantities of nitrite and hydroxide. He could <ret no ^uoh simple
succession of salts as Fremy had obtained, and he had ;d ways much
nitrile to de:ii with, a salt which giive no trouble to F'reniy. Keeping
the mixture cool during the entry of the sulphur dioxide, he got a
solution which, not immediately, as Fremy had found, but oidy on
standing, gave crystals of the neutral oximidosulphonate. Letting
the mixture grow hot by the action of the sulphur dioxide, crystals
wseparated, very much like Fremy's sulphazate, and these treated with
enough water left much insoluble nitrile and gave a solution wliich
slowly deposited crystals of the neutral oximide along with those of
the basic oximide.

Raschig, in the specification of his patent, appends to the account
of the process he gives for the preparation of the neutral oximidosul-
phonate a statement to the effect tint this may also be obtained by
passing sulphur dioxide in excess into a solution of alkali nitrite and

(>XlMll)(<SL'[-lMloXA'rKS (»1! S L'LPH AZOTATES. -^ß

its equivalent of either hydroxide or carbonate. l^>iit in his memoir in
the Aininh'ii, [xiblished later, he makes no mention of this, and con-
demns Fremy's and Claus's similar processes as iDeing frequently un-
productive, althouoh Claus had also used equivalent quantities of
nitrite and hydroxide. Evidently he thought little of this method
because of its uncertainty ; yet, as we shall show, it is. when pro[)erly
modified, regularly very productive in the case of both sodium and
potassium salts, and indeed is the only one by wliich sodium oximido-
sulphonate can be isolated.

Fremy prepared his ha sic salt nearly in the same way as his
neutral salt, varyino- this onlv to the extent of workino- with more
concentrated solutions, and so regulating the passage of sulphur
dioxide (assisted, it would seem, by addition of more potassium hydrox-
ide), that separation of crystals should take place with the solution
still strongly alkaline. Some care was required as to the degree of
concentration, which he adiusted bv beo-innino- witli solutions strono-
enough to deposit sulphaztite, and then adding just sufficient water
to redissolve this before passing in more sulphur dioxide. The
yield of salt was large. A secondary method of his was to make
sulphazate and salts allied to it and then treat these with water,
when the alkaline oximidosulphonate slowly crystallised from their
solution.

Clans in following Fremy's main process could get the alkaline
salt only in admixture with the neutral salt and other compounds.
By fractional crystallisation it could indeed be separated from these,
but he preferred a modification of Fremy's secondary process, in which
salts nearly the same as Fremy's sulphazate, with their adhering alka-
line mother-li([uor, are boiled witli two or three times their volume of
water, the sc^lution filtered hot. and when barely cold decanted from
the crystals tliat have formed of the basic salt, Tliis Fremv-Claus

24 E. DIVERS AND T. HAdA.

process is no longer, we consider, of any value, the two alkaline salts
being- readily obtainable from tlie neutral salt.

Prepanition of the neutral sodiuiii (Miniidosulphonate hy the, sidplnir-
dioxide melliod. — Tlie stdts are to be in the proportion of two molecules
of sodium nitrite to one of sodium carbonate or to two of sodium
hydroxide, but with about a tenth extra of the carbonate or hydrox-
ide, (called for by the unavoidable* conversion of about a tenth of the
nitrite into nitrile). In our experience this proportion gives about
the highest yield of oximide along with decomposition of all the
nitrite. Suppose the nitrite used to be 9() per cent, pure, and the
quantity ttiken 50 grams. Then this is put with 110 grams of sodium
carbonate crystals, or 80.8 grams of real sodium hydroxide, a piece
of lacmoïd ])a]ieï-. and 150 cub. cents, of water when carbonate is
taken, or 200 cub. cents, when hydroxide is taken, into a 500 cub.
cents, tiiisk. firted with cork carrying inlet and exit tubes and, if con-
venient, a thermometci-. Tiie inlet tube dipping into the solution is
connected with the sulphur-dioxide apparatus, and the exit tube with
a washing bottle of water, by (^aoutchouc tubes of adequate length
to allow of free movement of the flask. The stream of sulphur dioxide
may be r.apid (in the case of the carbonate being used, very rapid), but
not so as to cause any fumes either wliite oi* red in the flask ke])t in
active motion and immersed in water with ice floating in it. Hardly
any sulj)hur dioxide at all will escape consumption. In about 70
minutes the sohition will have become acid or nearly so, if the sulphur
dioxide have been passed in at a, good rate. Notice of the remote
approach (^f neutralisation of the sohition is gi^^en, when carbonate is
being used, by the disa])pea ranee <^f the last of the sodium hydrogen
carbonate which has heevi ])reci{)itated previously by the sul])hur
dioxide and, when sodium hydroxide has been used, by the disa)»-

* No lon^-er nuavoidable -, March, 1894.

OXlMIDOÖULi'UUXATEc; OK .SULi'HAZuTA'J'ES. 95

peamncc (if the precipitated siilpliite. lîy efïeeliui^- iieutralisation very
slowly at the last until acidity to lacinoïd paper i« jii.st readied, the
finish i){ the process may be so hit nÛ' that tlie s(jlution is left free not
only from nitrite bnt also from more tlian traces of sulphite. In
practice we tind it better to enter the i^'as quickly up to distinct acidity
and then at once to stop its flow, for although the solution is then
found in contain a not inconsiderable (luantitv of sulphite and to o-ive
off a small quantity of nitrous fumes, this hardly affects the vield of
oximide, and there is great saving of time and attenticjn. When the
method of slow finishing has been followed it is usuallv necessary to
add a drop of dilute sulphuric acid at last, to destroy nitrile.

In the solution, either way prepared, the nitrile present now
suffers hydrolysis into imidosulphonate and acid sulphate. To make
sure of the disappearance of every trace of this very unstable and
therefore injurious salt, the solution is well cooled after the rapidly
occurring hydrolysis, and left in its' acid condition for ten or fifteen
minutes, while in the case of its containing sulphur dioxide a free
stream of air is blown through it to carry this awa3\ For fifteen
minutes or so in such a solution, the oximidosulphonate proves able
to resist the hych-olysing action of a little sulphuric acid. To the
solution deprived of ail nitrile and almost all sulphur dioxide, con-
centrated solution of sodium carbonate is added until alkalinity is
reached. For this ])urpose about 10 grams of the cai-bonate will be
wanted Ijut moiv. sliould be at hand to at once arrest liydrolvsis sliould
it happen to have set up in the oximide.

The liquor is now a solution (jf about ioO grams oximidosul-
phonate in about 220 grams of water, along with 15 grams of diso
dium imidosulphonate, and 22 of hydrated sodium sulphate. After
filtration from any impurities derived from the nitrite it has to be
evaporated and this cannot be safelv effected at a steam-lieat. It has

26 E- DIVERS AND 1'. HAöA.

therefore to be done either at u gentle heat in the air. or in a vacuum
over sulphuric acid. When the solution has been somewhat con-
centrated, say to a weight of oOO grams, it is cooled in ice and after
some hours strained, still in the refrigerator, from a transparent
magma of the crystals which have separated of sodium sulphate.
Sliould the oximidosulplionate be wanted for the })reparation of
oxyamidosulphonate or. through this, of hyponitrite (77//.s- JonrnaL
3, '211). the solution at this stage is serviceable without further
preparation.

The solution, deprived of much of its sulphate by cooling, soon
yields when restored to the vacuum-evaporator much of the sodium
oximidosulplionate. 90 or 100 grams of tliis having crystallised out,
its mother-liquor may be deprived of more sulphate by cooling and
then yields a further crop of oximide crystals, nearly pure, by
evaporation.

Even as first obtained, disodium oximidosulplionate is nearly
pure, l^eing an anhydrous salt in hard and dense small thick prisms
deposited as a thick crust on the bottom of the crystallising pan. It
may be recrystaliised from its solution, made with a little hot water
rendered slightly alkaline by ammonia, by evaporation.

Titration of some of the original liquor after hydrolysis of the
salt into hydroxylamine, shows the yield to be 90 per cent., more
or less, of the quantity calculated from the nitrite, but the crystals
obtained do not amount to much moi-e tlian 70 per cent., because
the magma of sulphate retains mucli of the solution, which is very
concentrated.

The theorv of the formation of tlie oximidosul[)honatcs is Qon-
tained in the theory of the reaction between nitrites and sulphites, and
this we hope to discuss in a future paper. It will be enough to point
out here that one molecide of 'nitrite, one of hvdroxide. and two of

OX1M]I)OSULPHOXATE8 OK SULPHAZOTATES. £7

su]])hur dioxide are by calculation convertible into oximidosulphonate
and nothing el.se ; —

XaXOo + XaOH + l>SO., = HUX(>S03Xa)2.

It will therefore be seen that, with tliese proportions, were sulphite
used instead of hydroxide or carbonate to begin with, the use of
sulphur dioxide could still not be dispensed with. Cooling during
the preparation of the solution guards against premature hydrolvsis
and lessens the production of ni true. The quantity of water is limited
because it has to be evaporated afterwards in the cold. Much less
than the quantity jjrescribed will not be enough because then so much
acid-carbon.ate or normal sulphite, as the case may be, mav separate
ont in the middle of t lie process as to thicken the solution so that it
cannot be sufficiently rapidly agitated with the sulphur dioxide to
prevent the injiu'ious action of local excess of the latter.

It is generally more convenient to work with sodiimi carbcjnate
than with the hydroxide ; and on a large scale especially or on a
moderately large scale, much more rapid working is possible with the
carbonate than with the hydroxide, because of the much greater heat
to be dealt with when the latter is used.

Preparation of tlw neutral poiassuDn oxinudmiilplionate hij using
sulj>Jiur dioxide. — The ]^rocess just described foi' getting sodium
oximidosulphonate, but modified as to temperature, is eminently suc-
cessful, and even simpler than when used for thtit salt, when employed
for getting potassium oximidosulphonate from potassium nitrite and
either carbonate or hydroxide. Other j)rocesses, already published,
cannot approach it in certainty and in purity of the salt yielded. 0\er
the process described in this pa])er (p. 19), in which are mixed some-
what more than three molecules of potassium metasulphite with two
of nitrite, it lias the same advautau'e. althcjuiiii in a less deo:ree, and in

23 È. mvÈRS Axb t. ha<ia.

addition that of not requiri ug voluminous solutions to be worked with.
For it is still more certain than that process, and gives a yield in
crystals of nearly 97 ])er cent, of the calculated quantity of the salt
when carbonate has l)een used, and nearly 95 per cent, when hydrox-
ide has l)een used, and with not enough nitrile with it to be visible,
even on recrystallising. It is tlie small solubility of the potassium
salt -which makes its preparation simpler than that of the sodium salt,
and also which, by throwing it our of solution removes most of it from
the influence of the sulphite and sidphurous acid, and thus lessens the
fbi-mation of nitrile. Our success by this method has been so much
greater tban that reached by either Claus or llaschig mainlv Ijecause
we have taken the nitrite and the hvdi'oxide in the right proportion,
besides "'uardino' ao'ainst local excess of sulphur dioxide, and beiniz'
careful to preserve a low temperature.

Potassium nitrite, assayed for real salt, and either powdered
potassiiun acid-carbonate or concentrated solution of potassium hvdrox-
ide in equal molecular proportions are put, witli a weiglit of water
alxnit eight times that of the real nitrite, in a roomy flask ke[)t at 0°,
or even better at Ü° or so below zero. l)v means of a brine bath with
ice lioating in it. To preser\e this tem])erature during the jiassage of
the sulphur dioxide tlie flask must be kept in active motion, witli a
thermometer in it foi* observation. Bv a wide inlet tube terminating
high al)0\e the surface of tlie solution, in ordei" to guard against its
being choked, the sul[)hur di<jxide is i-apidlv passed in until the car-
bonate has ull dissohcd and etter\escence has almost ceased, or until
the hydroxide, if tliat has been used, is nearlv neutralised, and then
passed slowly until the solution has bectmie neutral to lacmoïd papei".
The oximidosulphonate separates during the latter part of the process
as ;i ci-ystalline ])Owder. T'he mother-liquoi- retains a little of it and
contains, 1)esides, some niti'ile along with small ([uantities of both

(^XlMIDüSULPHOXATES Ol! SÜLPHAZOTATES. 29

nitrite and sulphite. The precipitnterl oximidosiilplionnte can he
dissolved in hot water containing for safety a little ammonia, and from
the solntioM he got in sfood and characteristic crystals of uTeat ijuritv
and comparative stahility. The limit to the (|uantity of water to be
used is given hy the conditicjii that there must be enough not to lie
thickened by the separation of the insoluble oximidosul])honate. When
passing the sulphur dioxide into tlie flask closed with a cork holding-
inlet and outlet tubes and the thermometer, and with the inlet tube
endinii' seven or eio-lit centimeters above the surface of the solution,
and even when the alkali is used as carbonate so that carbon dioxide
is freely escaping, it is remarkable to observe the almost [M-rfect ab-
sorption of the sul])hui- dioxide by the well-agitated s<ilutioii.

l^rrjxiratidii nf ((ll:nli)ii- n.riiniih)sii1ii]ioii(it<s from tlw
Ht'iitnil soils.

Alkaline sodium o.i'iniiilo.suli)]ionates. — Tlie alkaline s<^dium oximido-
sulphonate, Na.NSoO;. corresponding to Raschig's haHic potassiuin
siilpliazotate, can be got by dissolving tlie neutral salt in water, adding
just the calculated quantity of sodium hydroxide, and evaporating to
the crystallising ])oint in a vacuum desiccator. The possibility of
])reparing the salt in this way is of theoretical importance, but the salt,
iDeing less solid^le in presence of sodium hydroxide, can also be pre-
pared at once as a crystalline precipitate by adding the sodium hydrox-
ide in excetss to a «concentrated solution of neutral sodium oximido-
sulphonate. The ci-ystals are drained on a tile and recrystallised once
or more by dissolving and evaporating.

A Hess alkaline sodium salt,, NasH(iSrS., (,)-):.,. is ol)tained when a
solution of the neutral oximidosulphonat«^ is evaporated with some-
what less sodium hvdroxide than is needed to form the above salt,

30

E. DIVERS AND T. HAGA.

such as half or more oftliJit (|iiMiitify. It can l)e purified l)y recrystal-
lisatioM from water. It does not correspoFid in composition to
Fremv's basic aidjtliazotalf of potassium, heing more alkaline than that

salt.

Alkaline polüsahnii nxlinidosulphonatt's. — The alkaline potassium
salt, K^NSoO;. discovered Iw Raschig, was obtained by him by dis-
solving the neuti-al salt in hot alkalised water, cooling rapidly, adding
to the supersaturated solution cold concentrated potassimn-hydroxide
solution until it caused a troubling, and finally leaving the mixture
to de})Osit the new salt in crystals. lie also ])repared it by adding an
excess of very strong potassiuin-liydroxide solution to Fremy'x basic
salt alreadv dissoUed in a little warm water, and precipitating by
alcohol.

Xot ordv is excess of alkali useful here, as when pre])aring the
sodivun salt, by diminishing the solubility of the salt in its aqueous
mother-liquor, it is also necessary, whicli is not the case with tlie
sodium salt, to preserve the salt from decomposition by water. But
when alcohol is to be used as the precipitating agent, tlie excess of alkali
used by Kaschig is uncalled for. equally as in the case of the sodiiun
salt, and only the calculated quantity of potassium hydroxide having
been added, a sufficiency of alci^hol precipitates the salt. Indeed the
best way to purify the salt from adhering alkali is to dissohe it
in water and precipitate by alcohol, repeating the process ont-e or
twice.

The less alkaline polassiuni salt, KJi(NS20^).2, — Fremy found that the
neutral salt in the solid state combines at once with potassium hydrox-
ide when treated with an excess of it in solution, and that the product
dissolved u]) in hot water deposits his hasic sulpha :otate in crystals on
cooling. We have fully confirmed Freiny's ex])erience although from
the fact of both salts being very sparingly soluble, the change of the

OXlMlD0SULl>li(_)XATES Ol.' SULPIIAZOTATES. p,]

powdered nentral salt int«» the alkaline one is not vei'v apparent to the
eye. Yet the potas^inm liych-oxide is withrh-awn by the powdered
salt rapidly from the solution, which consequentlv loses all its caus-
ticity to the toniiMie. When used concentrated the alkali causes, as
Fremv noted, a heatinu' up. wliile when it is dilute it causes, we find,
a fall of temperature amountinu' to about three degcees, due no doubt
not to dissolution of s;iJt so much as to liquefaction of three-fourths of
the water of crystallisation of the neutral salt transformed.

Claus denied altog-ether that the neutral salt could be converted
into the alkaline salt, and Raschig found it necessary for success to
modify Fremy's process. The latter uses twice the calculated quantity
of potassium hydroxide and in concentrated solution, heats to boiling,
and crystallises the salt by cooling. The salt is purified by recrystal-
lisation from water. Weak alkaline solutions of the neutral salt
did indeed yield him crystals of the alkaline salt, but in order to
do so had to be left to stand for some weeks. According to our
experience, however, any excess of alkali is unnecessary and best
avoided, as then the alkaline salt is at once obtained pure. Neither
need the solution of the alkali be concentrated ; nor the mixture be
heated to boiling ; nc^r time be given, more than is usual for crystal-
lising out salts. When crystallisation happens to be slow in setting
in, a particle of the solid salt, previously obtained, at once determines
it when dropped into the solution.

Alkaline jiotnssium sodiwn oxùnidostdph ouates.- — There is some dif-
ficulty in preparing potassium-sodium salts having a basicity as great
as that of the more alkaline sodium or potassium salt, but several less
alkaline can be obtained liy mixing in solution the neutral potassium
salt with sodium hydroxide, (^r the neutral sodium salt with potassium
hydroxide, and either evaporating or adding alcoliol.

AJhtlinr hariinu, ><trn)ititnii, and lead mils. — Addition of barium

32 K. DIVEES AND T. HAGA.

hydroxide or of oxy-lead acetate to a solution of neutral potassium,
sodium, or ammonium oximidosulphonate causes precipitation of an
alkaline or ' basic ' salt. Strontium hydroxide gives a precipitate of
alkaline salt with potassium oximidosulphonate almost immediately ;
it gives no precipitate with sodium or ammonium oximidosulphonate,
but causes an alkaline salt to crystallise out in a few hours.

Decomposition of chlorides and nitrates by the neutral oximidosnl-
plionates. — The neutral potassium salt in solution with potassium
chloride, potnssium nitrate, or sodium chloride, yields when treated
with strong ammonia-water, alkaline oximidosulphonates either at
once or after evaporation, ammonium «^hloride or nitrate being formed.

Decomposition of carbonates and of acetates by the neutral oximidosvl-
phonates. — AVhen a warm solution of sodium or potassium carbonate
is saturated with the neutral potassium oximidosulphonate, the less
alkaline potassium-s;dt or an alkaline potnssium-sodium salt crystal-
lises out on cooling. Neutral sodium or potassium oximidosulphonate
evaporated with either sodium or potassium acetate gives off acetic-
acid vapours and yields a highly alkaline solution (of alkaline
oximidosul phonate) .

Eeconversion of alkaline into neutral and less alkaline
oximidosulphonates.

The conversion of the more alkaline back into the alkaline, and of
these aoain into the neutral oximidosulphonates, described in this
section, and the conversion of neutral into alkaline oximidosul-
phonates, described in the preceding section, demand particular atten-
tion as being facts opposed to the view maintained by Claus and by
Raschig that Fremy's sulphazotates include the salts of two distinct
acids.

Solutions of the inore alkaline and the neutral sodium oximidosul-

0XIMID0SULPHOXATE8 OR 8ULPHAZ0TATES. 33

phonates evaporated togetlier. best in a vacuum desiccator, are con-
verted int<3 the less alkaline salt in crystals. Two equivalents of the
former combine with one of the latter, and this happens even when
only one equivalent of the former salt is taken, excess of the latter
then remaining in the mother-Hquor. We find also that, when a
concentrated solution of the more alkaline potassium salt is mixed
with a cold supersaturated solution of the neutral potassium salt
in equivalent quantity, crystallisation of the less alkaline salt soon
sets in, or else can be at once determined by adding a crystal of
this salt.

Conversely, in a manner, as Fremy found, the less alkaline salt
is resolved, when treated with the solution of a lead or a barium salt,
into the neutral salt and the more alkaline salt, the latter becoming
bv the reaction an insoluble salt of lead or barium with potassium.
Indeed, this conversion is the basis of one of Fremy 's methods
of getting the neutral potassium salt. Fremy's observations were
emphatically discredited by Claus and ignored by Raschig.

Weak acids— including sulphurous acid (cf. Fremy) — and even
strong acids diluted and cautiously used, convert the alkaline salts
into the neutral, when added not in excess so as to hydrolyse the
oximidosulphonate. Claus strongly denied this to be the case, but
Fremy was right. Metasulphites also, added not in excess, remove
alkali from the alkaline salts, they themselves becoming thereby
normal sulphites, as shown by tlie mixed solutions being neutral to
rosolic acid.

Several metallic salts such as those of zinc and manganese Ijehave
like acids with the alkaline salts, converting them into the neutral
salts, and depositing their own metal as hydroxide, as was observed
by Fremy. Water alone when ])resent in large quantity suffices to
convert the more alkaline potassium salt into the less alkaline salt

34 ^- DIVERS AND T. HAGA.

along with potassium hydroxide (cf. Kaschig), but has little if any
action upon the sodium salt.

Constitution. — Unity of oximidosidphonates and
sidphazotates.

The constitution of the oximidosulphonates as sulphonic deriva-
tives of hydroxylamine follows from the fact of their formation by the
union of nitrite with sulphite, and from that of their hydrolysing in
stages into hydroxylamine and sulphate. From the account we have
given — and Fremy before us — of the conversion of the neutral into the
alkaline, and of the alkaline into the neutral sii Its, it might also be
accepted without hesitation that they are aJl the salts of one acid,
were it not for the dissent of Claus and of Raschig from this
conclusion.

All the compounds described in this paper seem to us to be as
clearly salts of one acid as are the acid, neutral, and alkaline ortho-
phosphates all salts of one phosphoric acid. This acid, Fremy 's
Bulphazotic acid, is oximidosulphonic acid, a tribasic acid with the
constitutional formula : —

HO-SOo

i

I
HO-SO.

and unknown as the wholly hydrogen salt. Of tliis acid the salts
with neutral or slightly acid reaction are monohydrogen salts, and
those with three monad atoms of metal (or their equivalent) which we
have hitherto designated the more alkaline salts are normal salts. For
example : —

/ON(803).,Ba
HONrSOaNa)., ; A'aON(SaNa)., ; r.a<

OXIMIDOSüLPHOXATES OR SüLPHAZOTa.TES. 35

The le.'is alkaline salts, called in the case of potassium, basic .ml-
phazotate by Fremy and sulphazotate by Claus and by Raschig, are
double salts of the normal and the acid or hydrogen salts, thus : —

HOXC^OsK),, KONCSOsK)^ and H0N(.S03Na)„ :i NaON(S03Na)-,,

pentapotassiiiiii hi-oximidosidphonate and octasodiuui fer-oximidosulphonate.

ISTothing is wanting in the evidence to the truth of the views here
set down. Any acid serves to replace in the normal salts the third
atom of metal by hydrogen ; many salts, such as zinc sulphate, do the
same (p. 3o) ; potassium or sodium hydroxide directly reacts with
the hydrogen salts to replace some or all of the hydrogen by metal,
and just as readily as it does with phosphoric acid, which, as is well
known, does not excliange all its hydrogen for metal unless the alkali
is used in good excess.

That the pentapotassium and octasodiuui comj)ounds are double
salts is a conclusion in accordance with all that is known of them.
They are formed by the simple union of the hydrogen salt A\ith
the normal salt in solution, even when the quantities of these
salts deviate to a not inconsiderable extent ïvoin their proportions
in the compound salts, and they have an action on litmus, Phenol-
phthalein, etc., the same as that of the normal salts alone. In reac-
tion they decompose into their component salts, one only being active.
Fremy observed that lead and barium salts added to a solution of the
pentapotassium salt gave precipitates more basic than it and left in
solution the neutral potassium salt ; but as he did not know of the
existence of the normal potassium salt, and got complex basic pre-
cipitates, he could not represent the change quite as we do. E\en
now it still remains open to say with him that a basic salt precipitat-
ing, the acid of the precipitant converts anc^ther portion nï the
alkaline potassium salt into the neutral one, though tu do so is to

gg É. DIVERS AXT) T. HAGA.

pay no consideration to the fact that normal salts are known and that
the barium salt is one of them and the lead salt another, though in
this the lead is half as hydr<3xide. It seemed of interest to examine
the precipitation by barium chloride quantitatively. Adding this salt
in moderate excess to a solution of the pentapotassium salt, the propor-
tion of oximidosulphonic radical precipitated was found to be eight-
nineteenths of the whole, a result which, after allowing for the solu-
bility of bai'ium oximidosulphonate, may be accepted as jjroof that the
pentapotassium salt had acted as tripotassium salt united with in-
different dipotassium salt.

One thing more is to be pointed out in sup))ort of the double-salt
character of the pentapotassium oximidosulphonate. Although this salt
so readily and frequently crystallises from solution as to have become
the best known of the three potassium salts, matters are quite otherwise
with the sodium salts among which the pentasodium salt cannot even
be prepared ;it all ; the salt intermediate to the normal sodium and the
disodiinu salts which can be formed is troublesome to get ; and the
normal salt itself is that which readily crystallises nut from a suffi-
ciently alkaline solution.

There is a strikingly characteristic test for these salts in common,
from which it may be assumed tbat they have a common constitution,
lîoth neutral and alkaline oximidosulphonates, and no other salts,
yield in solution, at once and in the cold, an intensely coloured salt
when gently oxidised in the absence of acids, the reaction being one of
the most remarkal)le in the domain of inorganic chemistry (p. 4(5).
This reaction was described l)y Fremy and his accuracy fully confirm-
ed by Raschig.

Failing entirely to convert Fremy's two potassium sulphazotates
into each other, or believing that he had failed, Claus held them to be
salts of distinct acids and gave them the distinctive formula —

OXIMJDOSULPHONATES OR 8ULPHAZOTATES. 37

NHO(SOJv), and XH(S03K),-N0(0K)(S0,K)^-M-itl, tl.e nitroo-en
qninquevalent. He believed that he had evidence of the latter salt
beino- always formed from the nitriJe. ^(SOgK).,, which for him way
trisnlphammonate, NH2(S()J\)3. Claiis's observations and conclusions
were examined at lenirth by Raschig and satisfactorilv refuted bv him:
they need not, therefore, be discussed by us.

Having rejected Claiis's views, liaschio- gave the hvdroxvlamine
constitution to the neutral salts, but proposed another constiriition
for the two alkaline salts, in which hke Claus he made the nitroo-en
quinquevalent :

(S03K),NK<^ >NH(SOJv), and (SO,K),N'K<^ )>A^Fv(SO,K),.

The character of this constitution need iiof be noticed here : we shall
probably retui-n to it in our paper on the nalphazilatea.

One argument given by Raschig for regarding as distinct the
acids of the neutral and alkaline salts was, that althougli acids at once
convert the latter salts into the former, alkalis effect the opposite
change only with difficulty. Weak alkali, he found, requires weeks
and to be in excess in order to effect the conversion ; while a con-
centrated solution of alkali must be, in order to act quicklv, boilino-
hot and also in excess. Our experience, recorded on page 31, lends
no support to this finding. The delay he observed in the appearance
of crystals of the salt, a fact to which he attached weight, must have
been due to the salt remaining in supersaturated solution, although he
himself dismissed this explanation of the matter as inapplicable, aver-
ring that this salt when obtained does not show the phenomenon of
supersaturation, without however testing the matter bv adding- ;i bit
of the salt to excite crystallisation. Tlie salt does not indeed super-
saturate its solutions to the degree shown bv the neutral potassium

38

E. DIVERS AND T. HAÔA.

salt, but nevertheless its hot strong- solution in simple writer can be
rapifHy cooled and then kept for an hour or more without crystallisino;,
and nothino; can l»e objected to allowing that it may show still greater
supersatnration in presence of the other salts of its mother-liquoi".
Mixtures made to prepare it. yield it at once if cold when a crystal of
it is dropped iti or when alcobol is added, free alkali remaining in
solution in the latter case, oidy when in excess ofthat required to form
the normal salt, and this salt occurring in the alcoholic pi'ecipitate
when the alkali is in excess ofthat îiecessarv to tVie composition <jf the
pentapotassium salt. But apart entirely from this matter of su])er-
saturation. there is in proof that cold dilute alkali converts the neutral
salt at once into the alkaline salt, Fremy's experiment, repeated
l)y us, of acting u])on the solid neutral salt with potassium hydroxide
in aqueous solution (p. 80). The neutral sodium salt also passes
readily and at once into the ncjrmal salt on adding sodium hydroxide
to it.

Rasch ig gave as a chemical property distinguishing between the
neutral and alkaline salts, the activity of alkali sulphite upon the
former and inactivity upon the latter. We have not succeeded our-
selves in finding anything in Claus's writings upon this subject, but
Raschig adduced Glaus in evidence of his statement. Fremy's wi-it-
ino-s, however, are sufficient to show that the above distinction, so far
as it holds good, serves only to mark the alkalinity of the alkaline
salt, for he points out that sulphurous acid first converts the basic
sulphazotate to the neutral sulphazotate and then acts upon this salt.
To this we may add that metasulphite and alkaline oximidosul]ihonjite
change together into normal sulphite and neutral oximidosulphonate,
and that when enough metasulphite is added all oximidosulphonate is
quickly destroyed.

There yet remains for consideration one other ground taken by

OXIMIÜOSULPHOXATES OU SULPHAZoTATES. 39

Raschio- for separatiiio; the alkaline siilphazotates from the neutral
salts. According to liim, there is a salt isomeric with Fremy's basic
siilphazotate which, unlike tliis, cannot be recrystallised from water, its
alkaline solution de])ositinü- neutral (^ximidosuljihonat«. It, tlierefore,
rather than Fremv's basic snlphazot;ite, must be held to be li:df-basic
oximidosulplionate. The Fremy-Claiis [)rooess for preparing neutral
oximidosulphonate had yielded him the new salt but this was only
once, and many repetitions of the process failed to reproduce it.

The difficulties in the way of accepting the existence of this salt
as proved are very great. As just stated it could not be got agaiu by
the process which had apparently yielded it. Moreover it cannot begot
by treating the neutral salt with potassium hydroxide, and it is hard
to see why, if it is a basic derivative of that salt decomposable by
water. Xeutral oximidosulphonate recrystallised from water holding
potassium hydroxide is strongly alkaline, so that Claus found the
percentage of ])()tassium to ])ecome some units higher than in the pure
salt, and such a prepai-ation behaves with water like Raschig's salt.
But. according to Fremy and us, there is here only the production of
some basic sulphazotate, and recrystallisation from water separates the
two salts. Now, the process which gave Raschig his salt may, accord-
ing to Claus, give not only the neutral salt but also some basic
salt, and therefore a combined product behaving with water like
Raschig's preparation.

But if we accept the re'sults of the analysis of the salt, it could
not have been other than a new salt, for Fremy's basic sulphazotate does
not decompose with water into the neutral salt. We are strongly of
opinion, and venture here to suggest that some mistake such as may
happen to any one at times, crept into Raschig's work or the record
of his work. He made but one analysis, and in that determined only
the p(^tassium and the sulphur. Of these two the sulphur content

40 E- DIVERS AND T. HAGA.

happens to be a factor of no account in the present case, f(^r the per-
centage of sulphur is nearly the same in the neutral and the basic
sulphazotate (or its isomer), namely, 20*1)7 and 21'.53, and Kaschig's
preparation could not from the circumstances claim any high degree
of purity. There is, however, a wide degree of difference between the
percentage of potassium in the neutral and thnt in the basic salt, and
that got by Raschig agrees well with that for the l:)asic sulphazotate
with its one molecule of water of crvstallisation. But here liaschiof
himself comes, as it were, to our assistance, and this emboldens us to
ask for some distrust of his potassium number, f^r he actually
discusses (though only to reject on experiment!) I grounds) the pos-
sibility of his salt being Fremy's sulphazate, wliich does not differ • so
very much ' from it, he says, in the Mmount of snl])]mr, — although
that difference is really S'/. i)er cent. Now, let such a l(3wering as this
be made in the potassium percentage and the sulphur be allowed to
stand, and we should have a sul])hazotate mixture or compound
resolving itself, as Raschig's preparation did when dissolved in water,
into the two salts. In concluding this criticism we may add that
Raschig ends by affixing a note of interrogation to the formula of his
salt.

Berenian of oximidosnlplwnates to nitrite and sulphite.

Sodium or potassium oximidosulphonate digested in the cold for
some hours with highly concentrated solution of sodium or potassium
hydroxide, or boiled with it for a few minutes, is largely converted into
nitrite and sulphite. Some of the latter salt crystallises out, and the
solution evolves sulphur dioxide ;md oxides of nitrogen when acidified,
or when neutralised and mixed with barium chloride gives a pre-
cipitate of barium sulphite, leaving nitrite in plenty in solution, easy
then to detect. The following equation ex[)resses this change : —

UXiMlDUöULl'UUXATEt; OK SULPHAZUTATES. 4^

KON(S03lv), + iMvOH = H.Ü + KNO,+ 2K2S03;

but the reversäion is attended Avith some other changes resuhiiiü; in the
escape of nitrous oxide and perhaps nitr()<ien, a consideration of which
will be found following the paragraphs on hydrolysis (}>. 42).

The normal salts in solution, or as solids not quite free from
moisture, soon suffer s(jnie reversion in the cold, and very much more
when the solutions are boiled, without any alkali being added : —

The tive-sixths normal potassium salt, althougli when well
prepared and preserved in a dry atmosphere it can be kept for months
without suffering noticealjle change, is when damp unstable, as found
by Claus and Raschig. But even the best preparations of it prove at
last unstable, and we have found that this is due to the reversion
which slowly goes on in it : —

K5H(X8,0,), = K,HXS,U7 + KN(J, + K,.S,0,

Such a mixture of salts must be very unstable, liable to hydrolysis and
other changes, sufficient to account for that entire break-up of the salt
which finally hapjjens. Even l^efore the cViange has proceeded far
encjugh to aff'ect the appearance of the cryst;ils, the reversion can be
detected by the presence of sulphite. For this detection direct acidi-
fication of the salt is hardly effective, because of sec(mdary relictions
which then consume the sulphurous acid. But by dissolving the
salt, thus slightly decomposed, in water, adding barium chloride, and
acidifying the washed precipitate, the sulphite can be readily found.

The disodium and di potassium salts should, as neutral or nearly
neutral salts, be liable to neither reversion nor hydrcjlysis, but their
equilibrium is so unstable that the fact is that they are especially
prone to rapid change. It is probable that they do really suffer no

42 E. DIVERS AND T. HAGA.

reversion, but in practice, in consequence doubtless of the traces of
alkaline salt purposely left in them as a guard against rapid hydro-
lysis, they do when kept dry develope a little sulphite before they
hydrolyse. The disodium salt, which is an anhydrous salt, may
however be kept for years in a desiccator \\ithout suliering sensible
reversion. In comparison with the others, the live-sixths normal
potassium salt and the disodium salt are the salts of their respective
metals best fitted to keep in stock for any considerable time.

The dry lead salt reverts wdien heated. This salt is a basic or
hydroxy-salt, and when moderately heated is decomposed, thus : —

(HO?b)3NS207 = HOPbNO, + 2PbSO:3 + HoO

The residue if further heated evolves red fumes, and if moistened with
sulphuric or hydrochloric acid evolves much sulphur dioxide.

Other clieritical projicrties of oxiniidü6ülj>ho)iates.

Hydrolysis. — Tlie hydrolytic decomposition of oximidosulphonates
has been repeatedly referred to in this paper, and is besides a fact long-
known. Its almost inevitable occurrence in every alkali oximidosul-
phonate is all that needs further notice. When by circumstances or
by intention the salt is rendered acid, h3'drolysis always quickly
ensues. An alkaline salt when there is any water available, always
becomes acid in time by suffering reversion, and thus its hydrolysis is
brought about.

In dilute solutions of oximidosulphonates acidified, hydrolysis
proceeds without complication, but in strong solutions, and when the
change is carried through, l)y heating, to hydroxy lamine sulphate,
some gas is produced, — nitrous oxide and nitrogen. In hydrolysing
for analytical purposes, this is manifested by pressure in the sealed tube,
and by a deficiency in the quantity of hydroxy lamine (cf. Raschig).

OXIMIDÜSULPHONATES OK SULP U AZOTATES. 43

When iiioisteiied dipotassium oximidosulphonate hydrolyses
spontaneously, it ;iJso ,i;ives out at the same time much nitrous oxide
and a little nitrogen, while strons* hot solutions of it or of the
disodium salt effervesce a little when acidified.

Dipotassium salt which has been preserved as far as possible in a
desiccat(3r, also effervesces when dis.solved in water, even iu ])resence of
an alkali. In these chanofes a little ammonia is also a'enerated.

It thus appears that oximidosulphonates decompose in several
ways combined ; by reversion, by hydrolysis, and by secondary
reactions with tlie sulphurous acid generated by the other changes.
As products we therefore get in the first line, nitrite, sulphite,
oxyamidosulphonate, and sulphate. From the sulphite and oxyamido-
sulphonate we may get imidosulphonate liydrolysing to amidosul-
phonate. and from the sulphite and oximidosulphonate, nitrilosul-
phonate. By the hydrolysis oï the oxyamidosulphonate there will })e
got hydroxylamine, and by its reversion, hyponitrite. From the
hydroxylamine (jr from the amidosulphonate (slowly) may come
ammonia, and from the former in alkaline solution also nitrous oxide
and nitrogen. Xitrous oxide may also come from hyponitrite, and
from hvdroxvlamine and nitrous acid, and nitrogen from ammonia
and nitrous acid. Hut with all these possibilities before us, there are
points which remain obscure, particularly that in the acid-decom])Osi-
tion of oximidosulphonates nitrous oxide and nitrogen should be
evolved.

Behaviour iclien heated. — The dipotassium salt hydrolyses with its
water »jf crystallisation quite suddenly when heated to 90° in a limited
s])ace, but by very gradual heating in a roomy air-bath it can be rendered
aidiydrous ;ind raised to 100° and a little above without change (Claus).
The disodium salt, which is anhydrous, can be safely heated to the same
extent. In a current of diied air Ijoth salts can be lieated much abo^^e

44 Ë. DIVERS AXlJ T. HAGÀ.

100" without iDeing changed. The alkaline salts, though they contain
water, may be heated also without change to 100" ov a little above.

About 105° for the two-thirds normal salts and 110° tor the
alkaline salts, (jximidosulphonates slowly take moisture from the air,
increase cc^nsequently in weight, and become hydrolysed and acid.
In this behaviour they are like imidosulphonates (This Jouni..^ 6, (M).
Claus observed that the dipotassium salt gained thus in weight, but
set this down t(3 absorption of oxygen, lîut :iny oxidation would be
attended with a loss of nitrogen or oxides of nitn^gen much more in
weight than the oxygen absorbed. P)esides, we tind that, heated for
hours at 120° in a current of well-dried air, oximidosulphonates, even
the two-thirds normal potassium salt, gain nothing in weight.

The effect of higher temperatures was investigated by Claus, who
found that the anhydrous dipotassium salt gives olf some acid-reacting
vapoiu' not sulphur dioxide, that (3n further heating a little am-
monium sulphate sublimes ( ! ), and that the residue then consists
a])parently of potassium pyrosulphate. Mis experience on heating
the tive-sixths normnl [lotassium salt was. that abo\e '20(f the
crystals swell up and s])ring asunder, gi^e first nitric oxide and then
ammonium sulphate, and leave a residue which does not fuse and is
normal potassium sulphate. Freniy, Claus, and Raschig all state that
this salt decom])Oses suddenly at al)out :^00° (150° Raschig) with
develojmient of red vapours.

We have experimented more particularly with the sodium
oximidosulphonates, the disodium salt being better fitted than the
dipotassium salt for testing the eifects of dry heat upon oximidosul-
phonates because it is anhydrous. With the object of c(jllecting the
resulting' cruses the heatino^ has been conducted in a vaciunn as well as
in open vessels.

Tlie dried normal sodium salt remains unchanged until the tem-

OXTMIDOSULPHONATES OR SULPHAZ( JTATES. 45

perature has risen In l.Si^-3° when it sudden! v decomposes into a loose
unfiised residue, vapours depositing- a very small sublimate, and gases.
The residue is neutral or slightly alkaline and consists of sodium
sulphate witli a very little thiosulphate. The very small sublimate
consists ofsulphiu- and ammoniacal salt which when acidified yields a
solution milky from sulphur and smelling fain tlv of sulphur dioxide.
The gases are sulphur dioxide and nitrogen in nearly eipial volumes.

Neglecting the very small <|uantities of sulphur, thiosulphatt;.
and ammonia, all due to water retained by the salt, the decomposition
may be written : —

2Xa,NS,0, = 3Na,S0,+ SO.+K,.

The anhydrous disodium salt liegins to give off gas in a ^'^^cuum,
but exceedingly slowly, at about I4<P, and continues to do so nearly
as slowly while the temperature is rising to about 170°, or while it is
maintained steadily at any intermediate j^oint. When the tem}3erature
is allowed to rise continuously though slowly the full decomposition
of the salt occurs suddenly at 171°, whilst when it is kept steady for
an hour or more at about 165° and then slightly raised, full decom-
position occurs suddenly at 1(37-8°. At the moment of active decom-
position the salt fuses and effervesce« and, like the trisodium salt,
yields a small sublimate as well as gases. The i-esidue is exceedingly
acid and consists of sodium acid-sul])hate with a not inconsiderable
quantity of ammonium acid-sulphate or more probably the correspond-
ing pyrosulphates. The sublimate is a compound of ammonia and
sulphur dioxide while the gases consist of nitrogen and sulphur
dioxide, the former in somewhat greater volume than the latter.
Nitrous oxide could not be detected by us ; liut in one experiment the
sulphur dioxide constituted more thiin half the total gas, a result only
to be explained as due to some undetected error ov to the presence of

46

E. DIVERS AND T. HAGA.

a little nitrons oxide alonir with the nitrogen. The nitrogen obtained
was about four-fifths of the total nitrogen of the salt, the rest remain-
ing in the residue as ammonia. So far as was asrertained — and
observation is difficult because of the explosive character of the salt —
the first jiortions of gas are richer in sul])hur dioxide than the last.
If this be reallv so, it ]ioints to the not im])robab]e intermediate
formation of some imidosnlphonate.

From the facts observed, the decomposition of the disodiuni salt
seems to ^ arv between that expressed by —

l^XaJlXS,(K-l^Xî«TISO,+ Xa,S(),+ 8(1,+ X,

and that exjiressed by —

SNaJIXS20~3Na,SO,+ r)Na,S,0;+ (NH,)oS,0,+ 8X,-l- SO,.

ft will be seen tliat our ol)servations do not agree with tho.se
made l)y Claus as to the nature of the products of the decomposition
of oxiinidosulphonates liy heat. Fremy, Clans, and Raschig all state
that the five-sixths normal j)otassium salt gives red fumes when
lieated, and this also we cannot confirm ; whether in ci-ucible or in tube,
heated slowlv or quickly, it has never given us such fumes. Only
where reversion to nitrite and sulphite has taken place, as with lead
oximidosulphonate already noticed, does heat generate red fumes.

n.r/(h's((]iil/fii. — Characteristic of the («imidosul])honates are, firstly,
tliat thcv do not reduce cop])er salts in presence of alkali, unless they
have been previously hydrolysed with acid ; secondly, that they yield
bv gentle oxidation Fremv's sulpliazilates, unstable salts, blue-violet in
solution, vellow and crystalline in the solid state. Traces of almcxst
anv oxidising" jigent capable of actin«'- in alkaline solution effect this
change in the alkaline salts, but oxidising agents of a basic character
act also on the neutral salts. Silver oxide and lead peroxide are the
best reagents lor detecting any oximidosul])hoiiate. These oxides

OXIMIDOSULPHONATES OK SULPHAZOTA'l'KS. 47

were used bv Freiny. In hot acid .solution oximidosulphoimtes are
converted into sulphates and oxides of nitrogen by oxidising agents,
such as nitric acid, bromine, and potassium chlorate, but the decom-
position is seldom so complete as to be available for determiniiig sul-
phur quantitatively.

Action of stronfi snlfilmnc acid. — That strong sulphuric acid dis-
engages nitric oxide, as stated bv both Freniy and Claus, v^'e cannot
admit. Heated with this acid, a sodium or potassium salt dissolves and
evolves no gas luitil sulphuric-acid vapours are freely escaping, when
nitrous oxide and sulphur dioxide are slowly generated. Indeed,
addition of sulphuric acid should ])e made before igniting oximido-
sulphonates for analytical purposes, in order to avoid loss through
explosive decomposition.

Behaviour iritli other salts in sohtion. — Disodium, dipotassium, or
diammonium oximid(i><ulj)honate does not precipitate other salts and
appears often not even to react with them, since on evaporation the
salts mixed together crystallise out unchanged. This interesting fact
was observed by Fremy in the case of the dipotassium salt mixed with
those of zinc, manganese, copper, and silver. Claus denied its truth.
We have fully verified it in the case of the disodium salt and copper
sulphate. These salts in about equivalent quantities were dissolved in
water and the solution made slightly turbid with copper hydroxide by
adding a minute quantity of sodium hydroxide, in order to guard
against hydrolysis supervening. By spontaneous evaporation copper
sulphate first crystallised out and then the sodium oximidosulphonate,
that is, in the order of their solubilities. This behaviour of the alkali
oximidosulphonates serves to show that as sulphonates they have the
stability of sulphates themselves.

A normal alkali oximidosulphonate reacts generally as alkali
hydroxide towards salts the hydroxides of whose bases are insoluble ;

48

E. DIVERS AND T. HAGA.

disodium, potassium, or ammoiniim oximidosulphonate remaining in
solution. Silver oxide thus formed acts slowly us an oxidising agent,
ïriammonium oximidosulphonate added in excess to copper sulphate
gives a chromium-green solution, which is turned bkie hy ammonia
and which with excess of copper sulphate behaves as usual, yielding
a precipitate of copper hydroxide and leaving diammonium oximido-
sulphonate in solution.

However, both the two-thirds normal and the normal salts of the
alkalis enter into some reactions of complete double decomposition.
Thus, either a two-thirds normal or a normal oximidosulphonate
precipitates with hydroxy-lead acetate in excess or with barium
hydroxide ; a noruial salt precipitates also with sufficient normal lead
acetate or with barium chloride ; but to seciu-e precipitation of a lead
salt the solutions must be dilute. The barium precipitates, too, are
very slightly soluble in solutions of alkali oximidosulphonates.
Calcium compounds yield no ]jreci[)itates, but its hydroxide forms
salts by decomposing the ammonium oximidosulphonates. Strontium
compounds give with potassium oximidosulphonates clear solutions
for a moment which then become tilled suddenly with a voluminous
silky precipitate. Fremy could get no precipitate with strontium
salts, n(jr Claus either, strange to say. With sodium or ammonium
oximidosulphonates, strontium salts do not precipitate, but in a few
hours the solution deposits silky crystals in large hemispherical
groups on the sides of the vessel. ,

The precipitates of barium, strontium, and lead salts are soluble in
acids and in ammonium salts including ammonium oximidosulpho-
nates. The lead precipitates are also soluble in sodium or potassium
hydroxide.

Barium sulphate is distinctly soluble to a small extent in solutions
of the nc^rmal alkali oximidosulphonates. Thus, barium chloride may

OXlMlDOîsULl'HONA'J'E« UK .SL'LPHAZUTaTES. 49

l)e addrd in very «nuill (juaiitity fo a .solution, not too dilute, of a
normal alkali oximidosulphonate containing- sulphate without oceasion-
ing any permanent precipitate even on standing. But when the
alkalim'ty of the solution is ahout neutralised with hydrochloric acid,
the solution becomes turbid from the separation of barium sulphate.

The purified barium or lead oximidosulphonates can be decom-
posed by sulphates and very completely l)y carbonates, and thus be used
to furnish other oximidosulphonates. The ammonium salts which can
be formed in this way serve in tm-n for the [)reparation of salts of
metals which do not unite with ammonia by evaporating their
solution with the oxide or hydroxide of the metal. However, the
oximidosulphonates obtained in this way are the more unstable ot
these salts and we have not pursued their examination.

Oximidosulphonates comlnne readily with each other and with
other salts, a fact which will be made fully evident in the brief descrip-
tions which follow of the salts we have prepared.

Sodium oximiclosul'phonates.

Xornial sudiuiii salt, Xa3NS207,3H20. — This salt forms rhombic
prisms often a centimetre in length. It is strongly alkaline, soluble
in 1'3 parts of water at 20°, and tends to form supersaturated solutions.
It is only partly precipitated b}' alcohol but without decomposition,
and is less soluble in sodium-hydroxide solution than in water. Its
analysis gave —

Calc. Found.

Sodium 22-05 21'94

Sulphur 20-45 20-38

Oximide residue 9*58 9-32

Water 17-28 17-86

This and other oximidosulphonates were decomposed for the es-

50

E. DIVERS AND T. HAGA.

timation of hydroxylamine and sulphuric acid by hydrolysing in
sealed tubes with dilute hydrochloric acid kept for some time
at 100° and only then heated t(j 180°. Some of the attempts to
estimate hydroxylamine wei-e however defective for the reason already
given (p. 4:2). The sodium was directly estimated by cautious
io-nition with sulphuric acid, the presence of which prevents explosive
decomposition of the salt by heat.

Disodiuiii .^alt. Na.HNSoOy. — This salt is anhydrous and is soluble
in somewhat more than its own weight of water nt 14°. It just red-
dens litmus. Its crystals are usually small, dense, brilliant prisms,
united into hard masses and tii-mly adhering to the glass. On
analysis it has given : —

Calc. Found.

Sodimn 19-41 19-21 19-5Ö

Sulphur 27-00 1^7-06 26*98

Fjight-ninths noniuil .sudiiuii sali, Na.HNSoO;, 2Na3NS207, 3H..0. —
The solutions inlcruiediate in composition to those of the trisodium
and disodium salts which yield this when evaporated may be pre-
pared by adding either sodium hydroxide or ti'isodium sab to a
solution of disodium salt in approximately the calculated quantit}^
Twice, when the solutions were left in the desiccator to evaporate and
were of the composition of the unknown five-sixths normal salt, good
sized prisms and plates of some salt filled the solution which when
the attempt to remove them was made, i'a[)idly melted away very
remarkably, to give place to a precipitate of the eight-ninths sodium
salt. Possibly, they were crystals of tlie five-sixths salt. The eight-
ninths salt, redissolved in water or in its diluted mother-liquor, gives
on evaporation small nodules of the salt. Only under the microscope
can this precipitate and these nodules be made out to be composed of
prismatic crystals. The crystals are eftlorescent. Like the other

OXIMIDOSULPHOXATES OR SULPHAZOT.VTES. 5]^

sodium salts this one is exceedingly soluble in water, requiring less
than 1-5 parts of water at 14° to dissolve it, aud can be repeatedly
recrystaUised unchanged. It tends greatly to foi-in supersaturated solu-
tions. We have made the following analyses of it prepared under
different conditions : —

Calc. Found.

(a) (6) (c) (rf)

Sodium i>2-77 22-51 22*64 22-63 22-77

Sulphur 2:^-73 23-76 24-04 23-73 23-49

Oxinnde residue 1M3 1M4 — 11-13 —

Water 6-68 3-59 3-53 — —

Alkalinity, sodium. .. 5-69 — 5*45 — 5*83

The alkalinity of the preparations was determined with decinoi-mal
acid aud methyl-orange. The water was estimated by drying at IK)**,
and it will be noticed tliat h;df of it was retained at this temperature.
Similar retention of water will be seen in the case of the tive-sixths
potassium salt, of one of the sodium-potassium salts, and in that of
the barium salts, and is not due to the fixation of the water by
hydrolysis. It is remarkable that there should be such a difference in
this respect between this and the normal salt.

Preparation («) was deposited by eva[)oration in the cold of a
soluticjn of two m«jle<'nles of the disodium salt to one of sodium
hydroxide ; preparation (/>) was the same salt recrystaUised troni water
and separated as a preci[)itate from the supersatui'ated S(jlutii)n by
stirring J (t-) was obtained by evaporating the mother-liquor of (/>) ;
aud ((/) was prepared by evaporating a solution of the calculated
quantities of the disodium and trisodium oximidosulpJKjnates.

Futassiti til oxirnidosulphonafes,

Nonwil polassitnii sail. KgNS-.O;, HjO. Also KaNS.O^, 2H„U. —
We have several times prepared and analysed this salt with results

52

E. DIVERS AND T. UAUA.

verv nearly ugreeing with those got by Kaschig. lîut while he has
given t«) the salt only one molecule of water to KeNsS^Oj^, we find
that there is one molecule to K3NS2O7. Neither he nor we hajj})en to
have succeeded in getting, the salt quite free from excess of alkali :
we say ' happen,' because we are sure that we could prepare it so.
Tabulatin o- his and our analyses wdiicli siiow least excess of alkali with
calculations for his formula, and for ours —

A Hit schiff. Divers and Haga.

Calc. li'ouud. Calc. Found.

Potassium o7-09 'Mriu aß-OO 36-29

Sulphur 2()-2o U)-33 19-67 19-42

Wuter 2-85 — 5-53 5-61

^ve see Hiat liis analysis is less incompa(;i))le witli our formula than

with his, for whir'h his sulphin- is much, and inexplicably, tcjo low,
while agreeino- with his otlier and all our determinations of sulphur.
We estimated the water in our preparation, and it will be seen that
potassium, sidphur, and water all agree well with our formula. Our
analysis w^as made on a precipitate got by adding alcohol to a, solution
of dipotassium salt and potassium hydroxide in about the calculated
proportions.

We now tabulate tlie other analyses by Raschig and by us along
with a calculation for Iv3NS„U7,7hKOH,7,H,() wMiich matches Kasdiig's
preparation : —

Ü llasclriff. Divert and Harja.

Calc. («) 0>)

Potassium o7-5o 37-55 37-4(S 37-o7

Sulphur 19-66 19-56 19-23 19-54

Water 4-15 — 3-86 —

Our (d) preparation was got by using excess of concentrated

OXIMIDOSULPHONATES OR SULP H AZOTATES. 5p,

alkali, was drained on the tile, and then washed with alcohol ; (h) was
got by recrystallising from w^ater an unwashed salt like the last. This
was the only time we succeeded in recrystallising the salt from w:iter
without addin«: fresh alkali. It will be seen that botli Rascliicr's
results and ours agree much better in composition with the foi-itinl;i
of an impure salt than with tliat for !i salt with linlf a moJecide of
water t<^ one molecule of oximidosulphonate, (A) and from such results
no deduction can be safely made as to the decree of hvdr;iiinii of tlie
pure salt.

\Ye once ol)tîiined the normal potassium salt crystallised with
two molecules of water. The salt ])recipitated by adding excess of
potassium hydroxide, and drained on the tile, was redissolved in water
and precipitated by alcohol. In other trials in this way we got the
salt with only one molecule of water as above. Our analysis included
the estimation of the oximide radical as hydroxylamine : —

Calc. Found.

Potassium 34-17 34-10

Sulphur 18-64 18-81

Oximide residue 8*74 8-44

Dipotassium salt, K2HNS207,2H20. — Besides the tw^o excellent
methods of preparing this salt from potassium nitrite it can idso be
obtained, sometimes conveniently, from the disodium salt by pre-
cipitating this with potassium chloride, and is then more certainly
free from nitrilosulphonate. The yield in this way can be made as
much as 80 per cent, of the equivalent of the quantity of disodium salt
taken. Fremy's method of preparing it pure from the alkaline live-
sixths potassium salt by precipitating this with zinc or lead or barium
salt thus leaving the dipotassium salt in solution, and which did him
o-ood service, is now onlv of interest as establishing- the two i)otassium
salts as salts of one and the same acid.

r^ I E. DIVEKR ANT) T. HAGA.

The dipota.ssium snlt toi-ni.s detached hard crystals remarkable in
«hape, being like somewhat flattened, Aery acute octahedrons. The
crystals belong to the obli(|ue rhombic system (Raschig and Fock),
The analyses of Claus and Raschig established the composition of this
salt. Fremy's recorded results are improbable and insufficiently con-
<;ordant ; l)ut the identity of his sulplKCotdh' with dipotassium ox-
imidosul))honate is beyond doubt.

According to Claus, the dipotassium salt is ])ractically insoluble
in cold water, and is very difficultly soluble according to Raschig.
Its crystals cnn indeed be washed without niiu-h loss, but the salt in
fine powdei- proves to be far from insoluble. It dissolves to the extent
of about one part in thirtv of water. When quite fi'ee from alkaline
salt, it is faintly acid to litnins. Ft readily forms supersaturated solu-
tions, as observed bv Rasdiig, and in the preparation of the salt the
inothef-liquors retain much more of it tlian could sim])lc water. We
find that of a saturated solution of sodium cliloride about 4 cub. cent,
are al)le to dissolve 1 gram of the dipotassium salt, but the solution
quickly deposits a compoiuid of the two salts. Other salts idso
increase its solubility. Potassium or sodiiun hvdroxide or normal
oximidosidphconate acts u|)oii it and thus affects its solubility.

Fire-sixths normal potassiinii salt, IVoHNSoO;, KgNSA);, IL,f). —
There is really only one way of preparing this salt, namely, by dis-
solving up the dipotassium salt in hot solution of enough potassium
hydroxide (or salts equivalent to it) and crystallising. The dipotassium
salt should not be put into cold water or even potassium hydroxide,
and then the vessel heated to dissolve it, unless the mixture is con-
tinuously stirred ; foi' undissolved salt lying on the hot bottom of the
vessel is liable to hydrolyse. Observing this precaution, no simpler way
could be conceived to prepare the salt, all dilliculties being imaginary,
riie belief that the salt can be formed from the nitrite direct is un-

OXIMIDOÖULPHONATES OK «ULPHAZOTATES. 55

founded in fact, the dipotassiiim salt always preceding it and then
only yielding it by alkaline treatment.

The five-sixths potassium salt forms rhombic plates, crystallising
in masses on the walls of the vessel. It is only sparingly soluble in
water. It was analysed by Fremy and l)y Claus with results identi-
cal ns to the sulphur and not greatly differing from each other as
to the potassium. Raschig got results for potassium exactly
intermediate to theirs, but for sulphur one per cent, less than they got;
and he rightly gave to the composition of the salt a molecule of water,
although Claus had emphatically stated it to be anhydrous, having
found that it can be heated to 120" without any change or loss of
Aveight. Raschig, however, so far from establishing the presence of
this molecule of crystallisation-water, wrote of the salt that it ' remains
w^holly unchanged at 120°,' a statement almost the same as that
which Claus had made in proof that it contained no water. The
facts, we firid, are that it does lose in weight when heated, in fine
powder, to 120°, and therefore contains water, but, as is usual in our
experience with sulphazotised salts, it gives up this water very slowly.
We further find that at 120° it slowly increases in weight again by
re -taking water from the atmosphere and in doing so becomes hydro-
lysed and acid. When the hydrated salt is rapidly raised to, and
maintained at 120°, such hydrolysis also occurs and thus fixes the
water, causing the salt to suffer scarcely any change in weight.

The results of our determinations of sulphur and potassium agree
with tliose got by Raschig, but with the potassium a little lower than
the calculated number. As our methods of getting our preparations
were somewhat exceptional, we submit the results of our analyses
along vvith those got by the other investigators, also interesting from
the peculiarities noticed of their agreements and variations : —

^Q E. DIVERS AND T. HAGA.

Fremij. Claus. Basdiir/. Divers and Hacja.

Calc. Mean. Mean. (a) {h) (c)

Potassium ... 32-89 32-0 33-(J 32-8 32-38 32-41 32-04

Sulphur 21-53 22-2 22-4 21-4 21-43 21-23 21-19

Nitrogen 4-71 4-4 4-9 — — — —

Oximideres. 10-27 _ — — 9-05 — 10-23

Water 2-97 _ — _ 3-00 3-29 —

Sample ((/) was prepared by adding to a cold supersaturated solu-
tion of dipotassium oximidosulphonate the calculated quantity of
potassium hydroxide in solution, and leaving to crystallise. Sample
(/)) was [»repared by dissolving the tripotassium salt in warm water
and cryst'dlising out. S-imple (f) was obtained on adding potassium
chloride to dipotassium-oximidosniphonate solution and then excess
of ammonia.

Dipotasshtm oximidosulphonate and potassium nitrate.

Dipotassium oximidosulphonate unites with potassium nitrate,
but not with sodium nitrate or with sulphates. Again, it unites with
sodium chloride but not with potassium chloride.

The potassium-nitrate compound is obtained when a cold
saturated solution of potassium nitrate is mixed with a warm con-
centrated solution of dipotassium oximidosulphonate. The compound
begins almost at once to crystallise out in long silky needles, which
by their abundance and by interlacing often make the mixed solutions
set. Drained on a tile, the compound proved to be the simplest
double salt of the two radicals, K.dlNS^O;, KNO3, H2O. Two prepara-
tions were analysed : —

Calc. (a) (b)

Potassium 30*15 30-86 —

Sulphur 41-24 41-52 41-83

Nitrown 7-22 6-83 —

ÜXIMIDOSULPHüNATEö OR SULP HAZOTA TES.

57

Heated, it explodes giving off much red fume. Water decom-
poses it into its component salt-s. Its reaction to litmus is neutral.

Dipotassium oximidosulphonate and sodium chloride.

The compound of the dipotassium salt with sodium chloride is
obtained by dissolving the finely j)o\vdered potassium salt in cold
saturated solution of sodium chloride. Operating sufficiently quickly,
35-40 CCS. of the salt solution may be made to dissolve nine o-rams of
the potassium salt before the compound begins to crystallise out.
The compound salt forms small striated t^ood crystals belono-ino- to
the orthorh(jmbic system, and is neutral to litmus. Two preparations
were analysed and gave results agreeing with the formula— 5K2HNS2O-,
8NaCl, 3OH0, but as water decomposes the compound into indefinite
potassium-sodium oximidosulphonates and chlorides of both metals,
this formula can only claim to be the simplest expression of the com-
position (see further p. 60).

Calc. (/;)

Potassium 20-89 20-83 —

Sodium 9-86 9-73 —

Sulphur 17-14 17-32 —

Chlorine 15-21 14*96 14-90

Mixed sulphates ... 76*87 76*50 76*60

Potassium sodium oximidosulphonates.

It is perhaps not impossible, but it is certainly difficult to obtain
potassium-sodium oximidosulphonates of very simple composition.
Nor is it generally easy to get a salt of the same composition again
and again. The salts now to be described must be regarded as
examples only of an apparently indefinite number possible to prepare.

58

E. DIVERS AND T. HAGA.

Most of them occur only in very small crystals, sometimes microscopic,
generally aggregated in hard crusts and nodules. They were all
however obtained in trans[jarent though minute prismatic crystals,
and under the microscope appeared to be liomogeneous.

Normal sodium potassium salt. — Our attempts to produce normal
mixed salts by adding the hydroxide of one metal to a hydrogen
oximidosulphonate of the other metal have been unsuccessful. With
any great excess of alkali, evaporation leads to the destraction of the
oximide and crystallisation of potassium sulphite. With moderate
excess of alkali and evaporation in vacuo, uncrystallisable or nearly
uncrystallisable solutions are obtained. Such mixtures also fail for
the most part to give a solid precipitate with alcohol ; at most, liquid
droplets form which then sometimes slowly solidity to granules of
microscopic crystals. Besides this, sodium oximidosulphonate in
presence of alkali is largely soluble in strong spirit. When the alkali
is used in only slight excess, salts are formed on evaporation, but these
generally fall short of normal salts in composition. However, from
the two sinffle normal salts we did succeed in <'ettin(i- a mixed norm;d
salt nearly pure, which we now describe.

This mixed normal salt, SNasNSoO^, SKgNS.O;, 2E.0, but with
one-fourth of the water replaced by its equivalent of potassium
hydroxide, was obtained by evaporation in the desiccator of a solutic^n
of normal sodium and normal potassium oximidosulphonates in
molecular proportions together with a little sodium and potassium
hydroxides alsc^ in molecular proportions to each other. The salt
appeared as a powdery deposit of microscopic crystals, and was drained
dry on a tile, out of free contact with air. Reference to the descrip-
tion of the normal potassium salt, as obtained both by Raschig and
by ourselves, will show that the small excess of potassium in the
mixed salt is not peculiar to this preparation. Potassium tind sodium

OXIMIDOSULPHOXATES OR SULPHAZOTATES. 59

were estimated by the Fiiikener-Dittmar method. The calcaJatioii A
is for the true normal salt ; B for the salt with the small excess of
potash mentioned above : —

Calculated. Found.

A B

Potassium 23-85 24-85 24*89

Sodium 9-35 9*23 9-40

Sulphur ^21-68 21-40 21-35

A seveii-eigliths-uormal 'potassiim sodiviu salt, Ki8Na3H3(]!^SoO-)8,
2OH2O, in nodular masses of minute transparent prisms was obtained
on evaporating in a vacuum-desiccator a solution of the five-sixths-
normal potassium salt with sodium hydroxide in quantity calculated
to form normal salt. The composition of this salt is not very far
removed from that of a much simpler mixed eight-ninths-normal snlt,
K-iS[aH(XS207)3, 8H2O, corresponding to the eight-ninths sodium salt,
but the deviation is a little beyond the probable errors of analysis.
In the table, the calculated numbers marked A are for the 'eight-
ninths ' formida, and those marked B for the ' .seven-eighths ' formula.

Calculated. Found

A B

Potassium 27-05 26-51 26-58

Sodium 2-27 2-59 2-49

Sulphur 18-98 19-28 19*22

Mtroxy-radical, NO. 8-90 9-03 9*01

A six-sevenths-normal potassium sodium salt, K2Nai6^z(J^^2^7)7i ^HgO,
was obtained in some quantity by the evaporation of a solution of
disodium oximidosulphonate to which a small proportion of potassium
hydroxide and, unintentionally, some sodium hydroxide also had been
added. It formed as a hard crust of small prisms on the bottom of
the vessel. Attempts to prepare this salt again were unsuccessful ;

60

E. DIVERS AND T. HAGA.

the products then obtained will be noticed a few paragraphs later on.
Preserved for many months, the salt was again examined, to confirm
the results first o])tained.

This salt scarcely loses weight in the vacuum-desiccator, although
it contains water ; is not so freely soluble in water as are other
potassium-sodium oximidosulphonates rich in sodium ; and may be
recrystallised from solutif)n almost unaltered. The analysis marked
(a) is of the original pre[)aration, and that marked (/>) of the recrys-
tallised salt : —

Calc. {a) (b)

Potassium 4-17 4-10 -l-^T

Sodium 19-69 19-43 19-79

Sulphur r.2^-97 23-75 23-84

Nitroxy-radical, NO. 11.24 — 10-81

A fivc-sixths-noniial jiolassiuni sodiwu salt, K^J:^i\-Jii(Ny^.2^j\, 9H2O,
or not verv far from the five-sixths normal monosodium salt,
K4NaH(NS207)2, 2H2O, may from its mode of preparation possibly
not be a single salt, yet is interesting on account of the way in which
it was obtained, and for its relation in composition to the salt next
described.

When the dipotassium salt was mixed with the calculated
quantity of sodium hydr(3xide and the solution evaporated to a small
bulk over sulphuric acid, no salt separated, and when alcohol was added,
there occurred only a formation of droplets on the bottom of the vessel.
But gradually these solidified to masses of minute crystalline particles
having the composition formulated above instead of that of a normal
salt or mixture of normal salts as it would have had if compounds of
either metal had alone been present.

The calculation A is for the simpler and B for tlie more complex
formula, given above : —

OXIMIDOSULPHONATES OR SULPHAZOTATES.

61

Calculated. Found.

A B

Potassium 26'23 24-57 24*66

Sodium 3-86 4*82 4-93

Sulphur 21-46 21-44 21*37

The ßi'f-sixths-noniial monopotassium sodium salt, KNü^H(NS.,0-%,
E,0, is nearly represented by two salts obtained in different ways.
Both salts were however slightly more basic than calculated, while in
the first to be noticed the potassium was slightly replaced by sodium,
and in the second salt the sodium was slightly replaced by potassium,
as will be seen fn^m the numbers given. The first salt (a) was in
suuill well-formed prismatic crystals, and was got accidentally by the
spontaneous evaporation of an ammoniacal solution of disodium
oximidosulphonate containing some potassium chloride. It was
readily soluble in water but could not be recrystallised unchanged.
It was free from chlorine. The second salt (/>) was obtained in one of
the attempts to get again the six-sevenths mixed salt already described.
To seven molecules of disodium oximidosulphonate in solution were
added two of sodium hydroxide and two of potassium hydroxide.
The solution was then evaporated in vacuo until crystallisation
occurred : —

Calc. Found.

(a) (b)

Potassium 7*36 6*89 7*61

Sodium 17-86 17*69 17-25

Sulphur 24-15 23-93 23-96

A ßve-sixths-normal 'potassium sodium salt, K^.4Niiz■QH(NS20^\, HoO,
was obtained as a crust of microscopic orthorhombic prisms when a
mixture in unknown proportions of disodium oximidosulphonate in
solution and potassium hydroxide was evaporated in the desiccator.

A salt differing (3nly a little in composition from this was

62

E. DIVERS AND T. HiGA.

obtained in another attempt to form the ' six-sevenths ' mixed salt.
In this case one molecule of dipotassium, two of disodium, and four of
trisodium oximidosulphonates were brought together in solution and
evaporated in the desiccator. The solution became somewhat viscid
and only yielded a crystalline deposit when stirred with a glass rod.
The deposit or precipitate was thoroughly drained on a porous tile,
dissolved in a little water, and the solution evaporated to crystallisa-
tion. A salt now formed as a crust on the bottom of tlie vessel and
resembled the salt sought for, from which however it differed
materially in composition. The results of analysis of this preparation
are given under (b), and those of the other preparation under (a).
This is closely five-sixths normtil, while (h) is a little less basic, and in
other points deviates a little from the calculation : —

Calc. Found.

(a) (b)

Potassium lO'^O 1016 9-71

Sodium 15-43 15-46 15-83

Sulphur 23-86 23-82 24-27

Alkalinity as sodium 4-29 4-29 4-30

Raschig's potassium sodium sidphazotate is a salt a little less basic
than that just described as (/>). He represents it as being — in the
nomenclature of this paper, that is, — the five-sixths normal monosodium
oximidositlplioiiate, \iJ^aHQ^SoOT)2, 2H2O, and was justified in doing
so by his analysis. But from his results, the salt would nevertheless
appear to have had a composition approaching more nearly that of a
four-fifths normal salt than of a five-sixths normal one, and hence in the
order of basicity of these potassium sodium salts stands between that
last described and that to follow.

Raschig describes the salt as forming opaque spherical masses of the
size of millet seeds, and therefore resembling some of our preparations.

OXIMIDOSULPHONATES OK SULPHAZOTATES. 63

A four-fiftlis normal putaaslnm .sodium salt, Ko-iXao-sH^-eXSaO;,
O'Tl^HoO, was obtained by clissolviDg dipotassiiuin oximidosniphonute
in sodium-chloride solution and then addinii' :immonia in excess
(p. 35). It separated out as a precipitate of exceedingly minute
prisms sHghtly opaque when \'iewed under the microscope. It was
free from chloride and ammonia. The atomic ratio of the metals, it
will be seen, is Xa : K;, and of the metals to the hydrogen as
(KXa), : H.

Oalo. Found.

Potassium iVS'OC i^S-08

Sodium 2'od 1^*40

Sulphur il-87 il-86

A j'our-njths normal potassium sodium salt, but a little less basic, so
as to be nineteen-tn-entij-fourths normal, Ksjo^'^i^vioiß^^^:)-! l*8HoO,
was obtained almost in the same way as the last. Di potassium
oximidosulphonate in fine powder was dissolved in such quantity in
a saturated solution of sodium chloride that crystallisation of an
oximidosulphonate-chloride (p. 57) took place. This salt Avas
redissolved by warming in its mother-liquor to which some ammonia
water had been added, and the solution set aside. On cooling, the
above salt crystallised out in minute prisms, which in a good vacuum
over sulphuric acid lost nothing, although they contained water.
The salt was free from chlorine and anmionia.

Calc. I'ouud.

Potassium :^5"lo 'Jd'1\

Sodium o'94 3-Ül'

Sulphur 21-94 iU-95

A secen-nintlts normal potassium sodium salt. KuXaHo(XS.07)3, 2\rl.,0.
forms hard crusts of small rhombic prisms, only moderately soluble-
in water, and recrystallisable unchanged. It is oljtained from the

g4 E. DIVERS AND T. HAGA.

dipotassium oximidosulphoiiate, which exchanges one-third of its
hydrogen for sodium on treatnient with either sodium hydroxide, or
sodium carbonate, or sodium chloride witli ammonia.

In working with sodium hydroxide some small excess of this
may be used without affecting the composition of the crystals.
Analysis (j() was made upon crystals prepared by adding NaOH to
2K2HNS0O7, that is, in the proportion calculated for the five-sixths
normal salt.

The dipotassium salt dissolved in an excess of warm solution of
sodium carbonate, gave on cooling the crystals of Avhich (J)) is the
analysis.

Analysis (c) is of crystals obtained by dissolving together in
warm water, XaCl + l^KjHXSoO^, and then adding concentrated am-
monia-water in moderate (piantity, and leaving to cool. The sodium
chloride was, it will be seen, taken in small excess, namely, in the
quantity calculated to produce the five-sixths normal salt. When the
effect of a great excess of sodium chloride Avas tried, the result was
less satisfactory, the crystals being then somewhat opaque jind less
definite in composition. As best prepared, this stilt made by the use
of sodium chloride, is somewhat less basic than a seven-ninths normal
salt, being half-way, in composition, between this and a three-fourth a
normal salt,

Calc. Found.

(a) {b) (c)

Potassium 27-10 27-01 27.22 26*73

Sodium 2-66 2-53 2-50 2'62

Sulphur 22-18 22-00 22-08 22-27

Twj-thinh normal potassium- sodium salt, KXallNSoO;, SHjO. —
This salt, but slightly more basic from the presence of about -zV! of an
atom of potassium in excess of that in the formula and being therefore

OXIMIDOSULPHONATES OR SULPHAZOTATES. 65

a 14 normal instead of t normal salt, was obtained in long rhombic
prisms of small size by dissolving up 5*2 grams of dipotassium salt
(slightly alkaline) in oOccs. of a satm-ated solution of sodium chloride
and 5ccs. of water, by the aid of a gentle heat, and ]en\'ing the
solution to cool : —

Calc. Found.

Potassium 12-70 14-62

Sodium 7*-19 7-60

Sulphur 20-85 20-87

Reaction between the dipotassium salt and sodium chloride gives
also, as alreadv described, p. 57, a double oximidosulphonate-chloride,
besides other compounds allied to this and the salt just described. It
is a reaction which requires fuller investigation than we have fjund
time to make of it, and what follows is all we can add concerning it.

While from sodium oximidosulphonate and potassium chloride,
potassium oximidosulphonate crystallises out nearly free from sodium,
tliere is obtained from potassium oximidosulphonate dissolved in
sufficient quantity in a hot concentrated but not saturated solution
of sodium chhn'ide, a crystallisation of oximidosulphonate, half
potassium, half sodium. If the sodium-chloride solution is saturated
arjd the potassium oximid()su]['honate is dissolved in it in the cold,
what appears to be a compound of these salts, already described,
quicklv separates. Ihit if l^v the aid of heat more potassium oximido-
sulphonate is dissolved, the salt which crystallises out is what may be
represented as sodium chloride in combination with potassium-sodium
oximidosulphonate. And if in the mother-liquor of this salt, now
containing sodium oximidosulphonate in place of some of its sodium
chloride, potassium oximidosulphonate is again dissolved l)y heat, the
crystals wdiich form on cooling consist of an oximidosulphonate. more
of sodium than of potassium, and with only very little sodium

Q(\ E. DIVERS AND T. HAGA.

chloride alono- with it. We give an example of the composition of
snch a salt, wliich was obtained in good transparent small rhombic
prisms : —

Potassinm 7*45

Sodium 1^-98

Sulphur 23-65*

Chlorine 0-03

The potassium is to the sodium as K : Xa., pins a very little for the
chlorine.

By dissolving dipotassium oximidosulj^honate in a warm con-
centrated solution of disodium oximidosulphonate. it is practically
certain from what precedes that a potassium-sodium sfdt would
crystallise out. We regret we have no experience to record on tliis
point.

Ammonium oximidosiilplioiiales.

Hydroxy -lead oximidosulphonate (p. (SO) yields at once, when
shaken with enough solution of ammonium acid-carbonate to convert
its lead to carbonate, a solution of normal ammonium oximidosul-
phonate almost pure. Barium oximidosulphonate may be used in
place of the lead salt, but not being basic it requires the normal instead
of the acid ammonium carbonate to be used with it. The solution
smells mildly of ammonia, but in a closed vessel may be preserved
unchanged. Heated with exposure to air, it becomes a solution of
the diammonium salt, l)y loss of ammonia. Guarding against hydro-
lysis, to whicli this salt is very liable, bv adding a drop of strong
solution of ammonia occasionally so as to maintain its alkalinity, the
solution of the diammonium salt may be evaporated on the water-
bath to an exceedingly small volume without decomposing. If. now,

* This number is calculated, as the sulphur was lost.

OXIMIDOSULPHONATES OR SULPHAZOTATES. QJ

the very concentrated solution be quickly evaporated cold over sul-
])])nric ncirl, it is possible to o-et the (lianntinuiinii salt in good prismatic
crystals, very soluble in water. If instead, the solution is mixed with
concentrated ammonin-watei* and evaporated in an ammoniacal atmos-
phere over solid potassium hydroxide, crystals of another very sohible
salt are obtained, having- an unconquerable tendency to climb the
sides of the vessel and there exfoliate. These crvstals slowly effloresce
in the desiccator, and (^crui- in masses of overlapping flat prisms.
They are believed to be the ßre-sixfh s normal ammonium oximido.wîpho-
natr. corresponding to Fremy's basic potassium sulphazotate.

We have no analyses to bring forward. The dinmmonium salt
has hydrolysed before it could he freed îvo\\\ its motlier-liquor. The
alkaline salt has also not been obtained in (piantity in such definite
form as to promise results of value from its c|uantitative analysis.
Heated, this salt decomposes like other oximidosulphonates suddenly,
and le;ives a residue of ammonium acid sulphate.

On evaporating a solution of normal ammonium oximidosul-
phonate with ammonium acetate over a water- bath, ammonia escaped
as usual and the concentrated solution on cooling yielded a magma
of lustrous needles. This magma slowly drained itself dry on
the tile and for àr^y^ evolved strong acetic-acid vajiour without the
oximidosulphonate hydrolysing or the compact dry mass losing its
silky lustre. In the course of weeks, however, hydrolysis occurred
and the mass became loose opaque crystals of ammonium hydrogen
sulphate. It thus seems that diammonium oximidosulphonate like
the potassium salt combines with other salts such as ammonium acetate.

Barium oximidosulphonates.

Barium chloride does not precipitate a solution of dipotassium or
disodium oximidosuljjhonate, but the mixed solutions are very

68

E. DIVEES AXD T. HAGA.

unstalDle, soon hydrolysing into sulphate and ox3^amidosnlphonate {cf.
Clans). From the normal potassium or sodium salt it precipitates
liarium oximidosulphonate, in combination always with some potas-
sium or sodium oximidosulphonate. and with very much when the
alkali salt is in concentrated solution and kept in excess. The alkali
salts intermediate to the normal and two-thirds normal salts behave
towards barium chloride as mixtures, the two-thirds normal alkali
salt remaining unprecipitated except in concentrated solutions when
much of it, if the potassium salt, becomes insoluble in combination
with the barium salt. The precipitates dissolve to some extent in
warm concentrated solutions of the normal and two-thirds normal
alkali oximidosulphonates but. for the most ]»art. separate on cooling,
sometimes richer in alkali salt.

Barium hydroxide precipitates solutions of the two-thirds normal
as w^ell as normal salts, and all the precipitates are combinations of
barium with potassium or sodium salts. \Yhen the hydroxide is
used in excess, and when barium chloride is used in excess with the
normal or the intermediate salts, the precipitates are generally basic,
that is, have some of the barium as hydroxide.

Annnonium oximidosulphonates behave much like the potassium
and sodium salts, but have a much larger solvent action upon the
barium precipitates. We have made no analyses of the ammonium
barium salts.

There are apparently three barium salts Ba3(XSo07)o, BaHXSoO;,
and (HÜBa)oHNSoO; or (HOBa)Ba¥S,0, ; of which, however, the
second is so unstable as to be known only in solution, while the third
is known only in combination.

Normal barium oximidomJplionaic, Ba3(NSo07)2, 4]rL,0 and 8H0O. —
The precipitate obtained by using barium chloride or hydroxide in
excess with a potassium or sodium oximidosulj)honate is washed and

OXIMIDOSULPHOXATES OR öULPHAZOTATES. QÇ)

then nearly all dissolved by adding^ dilute hydrochloric acid with
continual stirring- until a solution is obtained neutral or only slightly
alkaline to litmus. The turbid solution is as quickly as possible
filtered, by vacuum-pump, into excess of warm baryta-water. The
precipitate thus obtained when washed with boiled-out water, hot or
cold, is the norm;d barium salt free from alkali and from carbonate.
It is a voluminous curdy precipitate at first, but generally changes to
a denser powder, crystalline under the microscope but to the naked
eye chalk-like when dry. It is practically insoluble in water, but
soluble in ammonium chloride. It is fully decomposed in the cold by
ammonium or sodium carbonate solution. Heated dry, it suddenly
decomposes into barium sulphate and gases. It loses water at com-
mon temperatures in dry air, and nearly all at 110°. Its composition,
calculated and found, is shown by the following table, in which A
refers to one pre[)aration with 4 Aq., and 1] to another with 8 Aq. : —

A

B

Calc.

Found.

Calc.

Found.

47-62

47-50

43-96

43-74

l>arium

.Sulphur 14-83 14-80 13-69 13-46

Tivo-thirds normal barium oximidosidphonalcs, BaHNSoO^, can be
obtained in solution by adding just enough acid to the normal barium
salt. When the normal salt is free from alkali and the acid is
sulphuric the solution filtered is pure. When the normal barium salt
is combined with any potassium or sodium salt the solution obtained
by means of an acid is treated as described fnv getting the pure normal
salt, which is then aijain converted to the two-thirds normal salt bv
sulphuric acid. The solution of this salt is distinctly acid to litmus
and hydrolyses too readily to admit apparently of crystallising out the
salt from it.

70 ^' DIVERS AXD T. HA(iA.

Bariiuii sodiiuii oximidosidpJionates. — A normal one-fifth «odium
salt, Ba6Na3(]srS207)5, 7HoO, is obtained when to normal sodium
oximidosulphonate in strong solution, one mol., there is added barium
chloride, one mol. It was also obtained by adding baryta-water to a
strong solution of the normal sodium salt kept in excess. The pre-
cipitate unwashed or slightly washed is dried on a porous tile. It
loses only one-third of its water at 120°. ((<) was obtained from barium
chloride and (h) from barium hydroxide :—

Calc. found.

(a) (h)

ßarium 11-79 42-02 42-24

Sodium 3-51 3-60 3-44

Sulphur lG-27 15-69 15-97

Another normal, only one-seventh sodium, salt, Ba9Na3(NS207)7,
7H2O, can be obtained by adding baryta-water to a mixed soluti(3n of
one mol. of disodium salt and one mol. of barium chloride, and in
other w;)vs. (u) and (A) were se}>arate ])reparations : —

Calc. Found.

(a) (h)

Barium 44-71 44-70 44-18

Sodium 2--")(i 2-40 2-47

Sulphur 16-24 15-.S3 —

A normal salt richer in sodium wa.^ got from 0 mois, of triso-
dium and o ]jiols. of disodium salt ground together with hardly
enouiih water to dissolve all, by addiny' to the mixture very liraduallv
and with tritiu'ation a dilute solution of <jne mol. of barium chloride,
half of which was added befjre it caused any precipitation in tlie now
dissolved salts. The mixture quickly thickened to a nearly opaque
firm jelly which after a time broke up into a thoroughly li(juid

OXIMIDOSULPHOXATES OR «ÜLPHAZOTATES. 71

mixture of a powdery precipitate and an abundant mother-liquor.
The precipitate was drained on a tile, and then had changed from a
powder to a horny hard mass.* This was ground (ine in a mortar
and dried in a good desiccator, the grinding being once or twice
repeated. Analysed it proved to be BaNaKSoO;, ILO with a little
Ba3(iVS,0;),. 9H,0 :—

Bai8^''ii5(^'^L'07)i7. 21IL,0. Found.

Barium aS'lO 38-36 38-01

Sodium o'oo — o'25

Sulphur 1(J-81 — 16-72

Baruini potassium o.riniidosulpliojiates. — The normal potassium salt
decomposes so (piicklv in aqueous solution into tlie iive-sixths normal
salt and potassium hydroxide that constant and simple results would
probably not be obtained with it as a precipitating agent. AVe have
used, as Fremy did, the K5 ><-\lt, or with baryta the Ko salt.

The secen-fiinths nonual salt, Vjn-Js.Jrl^(NS.fl^)s, 9HoO, is obtained
as a crystalline powdery precipitate when a concentrated solution of
barium chloride, two mois., is added to three or more mois, of the
five-sixths normal potassium salt in warm concentrated solution, this
salt, it should be remembered, beiug very little soluble in cold water.
The precipitate first formed redissolves in its mother-liquor up to the
point when only half the barium chloride has been added, but is all
again precipitated on adding the rest of the barium chloride and
letting the mixture cool. The mother-liquor must be decanted soon
and the precipitate drained ou a tile, because the former in about an
hour after cooling begins to deposit potassium salt. The results of
the an.alysis of the precipitate were —

* The behaviour of the mixture in g-elatinising and in breaking up then into a thin liquor
and a precipitate which became horny on drying, is much like that of one of Fromy's potassium
salts, his metasnlpliazate.

'J 2 E. DIVERS AND T. HAGA.

Calc. Found.

]^,arium 20-25 20-40

rotnssium 15-41 lô'GG

Sulplmr 18-92 18-25

Tlie snlphin-, it will be seen, is ii little low for the formula, a deviation
attributable perhaps to the presence of a little K5 salt from the mother-
liquor.

In this salt the oximidosulphonate exists one-third as normal
barium salt and two-thirds as dipotassium salt. In its formation the
Kg salt behaves, as usual, as mixed normal and dipotassium salts, of
Avhich one part of the former suffers decomposition with the barium
chloride and the other remains in solution, while the dipotassium salt
precipitates in combination with the barium salt : —

4KJI(NS,0,)o + 8?MiCL = ?>a3(NSA).,^T^%HNSA + 2K3XS20, + 6K^^

On addinp' to a warm concentrated solution of three mois, of the
K5 salt, one mol. of ])ariiun chloride, agitating to redissolve the
precipitate and then letting the solution cool, a relatively very large
cpiantitv of crystalline powdery precipitate is obtained, which from its
quantitv must consist largely of potassium salt. I'eing uncertain
concerning its freedom from sej^arate potassium salt we have not
analysed it. It can be redissolved in its warmed mother-liquor and
recovered bv coolino-, ao-ajn and ao-ain. Ihit if the mother-liquor
decanted is alMwed to stand some hours to deposit most of its
potassium salt and then re-decanted on to tlie barium precijntate it
fails to dissolve this Avhen warmed with it, though it effects great
change in its c(-)mposition, dissolving out ])otassiLun salt. Thu.s
treated, the pi-ecipitate was found to diffei- only a little from a salt
having the formula BaIvNS,,();, 1 LU and could be regarded as this
salt retaining unchanu'ed a little of the salt from which it liad been

OXIMIDOSULPHONATES OR SULPHAZOTATES. 73

prepared. It liad the composition expressed Ijy the formula,
Ba,,K,,H(NS,0,),3, lOH.O : —

Calc. Fouud.

Barium o3-95 :53-l)3

Potassium ll'o5 11*57

Sulphur 17-18 17-03

Tlie attempt to get tlie salt BalvNS.O- by direct reaction thus: —

K3(XS,U;;, + lkCL,== IkKXS.O; + K,HNS,.U, + iMvCl.

was onlv partly successful. The precipitate obtained on nddiug to a
warm concentrated solution of one mol. of the K., salt one of barium
chloride had the composition shown by the formula (IIO).jB;!6K4ÏI
(NS,0,),3, H,0, or quite probably. H0Ba,KXN^S,0;)5, 1^1,0. This
composition would agree with that of a barium sodium salt already
described, BaeXa3(N'SoO;),5, if an atom of barium were not half dis-
placed by one of potassium.

Calc. Found.

Bariuui 41-49 41-68

Potassium 7-89 8-22

Sulphur 16T5 IG-OG

This precipitate may Ije represented as a mixture or compound of
4BaKNS,0, with (HOBaX,HXS,0„ FLO (see the lead salts 3, p. 81, and 7,
p. 85) and it is probable that the precipitate first forined consists largely
<jf the latter salt, and that when, by the accumulation of dipotassium
salt the mother-liquor lessens in basicity, the BaK salt alone precipi-
tates. The reaction representing the formation of the hydroxy-
barium salt —

2BaCL+2K3XS,,0, + 2H,0 = (HOBa),HXSA + K,HXSA + 4KCl

— ^in which (HOBa)2HNS207 when doubled is equivalent to Ba(0H)2 +

74 E. DIVEES AND T. HAGA.

Ba3(NS207)o + H20, shows the normal sodium salt behaving with a
salt of barium somewhat as it does with zinc, manganese, copper, and
other salts, that is, as dipotassiam salt and potassium hydroxide
(V- 47).

Fremy, by treating the Kg salt with barium chloride, got a
preci|)itate, gelp.tinous at first, afterwards crystalhne, the composition
of which, as given by him, may be expressed (1) pretty closely by
(HO)3]]:!,Ks(NS20,)„ or (2) l)y BaKNS.O; as an approximation suf-
iiciently near to Fremy 's analytical results, considering the lialjility to
impurity of his preparations, and his imperfect methods of analysis : —

Calculated. Found.

(1) (2) (Fremy)

Barium 39-29 37-43 39-90

l^tassium 11-19 10-65 11-55

Sulphur lG-07 17-49 15-80

He must have kept the potassium salt in excess when precipit-
ating. He found that baryta-water gave the same salt. Although
much reliance cannot, we admit, be placed on Fremy's analytical
results as closely accurate, we are not inclined to adopt the simpler
formula rather than the other of the two we have calculated ; fjr oiu-
own results, here and in the case of strontium and lead salts, where
we have been sure of the comparative purity of our preparations and
accuracy of our methods of analysis, have taught us that compound
oximidosulphonates are generally of complex composition. Claus,
after experimenting for himself, pronounced Fremy's results to be
worthless. He found that after ten minutes or so, the barium pre-
cipitates always began to decompose and to contain sulphate. We
have not met with this difficulty, the barium salts (except the tAvo-
thirds normal salt, as already noticed) proving to be stable, and free
from sulphate when analysed. As the barium salts when acidified are

OXIMIDOSULPHOXATES OR öULPHAZOT.ATES. 75

extremely rapidly hydrolysed, the best way to test their freedom from
sulphate is to digest them with sodium carbonate, lilter off and wash
the barium carbonate, and observe whether this dissolves i)erfecrly in
nitric or hydrochloric acid.

Claus's experience may be accounted for by the presence of much
dipotassium oximidosulphonate in tlie mother-liquor, Avhich with
bari nm chloride very rapidly yields sulphate by hydrolysis. He refused
credence to Fremy's statement as to the formation of the dipotassium
salt when the K5 salt is precipitated by barium chloride and other
sdts, and consequently took no steps to protect himself against this
accident.

On adding to a supersaturated solution of the dipotassium salt
warm baryta-water not in excess, a precipitate was obtained by us
somewhat poor in potassium ; but then neither a supersaturated
solution of the dipotassium salt nor a warm one of barium hydroxide
is a solution of much concentration, which is the condition for the
precipitate to contain much potassium. The composition of the
precipitate agrees with that for the formula Ba,K3(XSoO;);, 14HoÜ,
matching which, except as to water, there is a barium sodium salt
already described on a previous page : —

Call'. Found.

Ijarium 42'05 -liM4

Potassium 4-01 4-19

Sulphur 15-28 l'ro2

Two mois, of dipotassium salt react with one mol. of barium
hydroxide to yield IjaKNSoO-, wliicli is resolved by water into
normal pcjtassium salt and normal barium salt retaining a twelfth of
the former, thus :■ —

lSKJ~lXS,O,4-01k(OII),=FxiJv3(N8,O0r + llK3NSA + 18H.,ü.
To sh(3\v the effect of adding the barium hydroxide in greater

76

E. DIVERS AXD T. H AG A.

(iLuiiitity lipon the normal potassiuni .salt otherwise found in the
mother-liquor, we may give the result of adding it in some excess to
a warm solution of the K^ salt. The precipitate obtained had a com-
position Avhich may Ije regarded as resulting from the addition of
harium hydroxide to the preceding salt, ov as that of a hydroxy-
derivative ofthat salt, in which half the Ijarium is as half-hydroxide.
The composition of the salt was (liO)3Ba8K.(NS207)5, 14H,0, Avhich
is nearly equivalent to (HO15a);i3a„K3(NS20,)r, lö^HoO, or aJ3a(0H),
4- Ba9K3(NlS2U7);. It is also related to Fremy's salt, Ijeing that salt
with Kg replaced ))y (H(Jl>:i);jBa : —

Oalc. Found.

Barium 45-r7 45-2i)

Potassium .S-lU 3'21

Sulphur 13'BS 13-11

The reaction has been one in which potassium hydroxide is

largely formed : —

5KJl(XS,ü0.+ l(^Ba(Ull), = 2(llO)3lHK,(xVSA)o + 2lKOH + 5H,O

The mother-li(|Uor was tested and found to contain much potassium
hydroxide and some barium hydroxide and nothing else.

AVhen barium chloride in slight excess and in moderately dilute
solution is added to the K^ salt, it yields a similar compound lo that
obtained by excess of baryta- water, but with some chloride in place
of liydroxide of liarium. The precipitate had the composition
Cl(HO),Ba,,K3(XSoO,X, 7II3U or BaCL, + lM5a(011), + Ba3(XS,ü,X. +
2BaJs:3(NS,0-);-l- 14tL,0 (see lead salt 4, p. 82).

Calc. Eoiiud.

Barium 47-29 47-78

Potassium o-87 o-24

Sulphur 14-73 14-07

Chlorine 1-02 0-96

OXIMIDOSULrHOXATES OR SÜLPHAZOTATES. 77

Strontium oximidosjiJpliofiatefi.

We have not made a a'eneral examination of the istrontium salts.
What we have observed <[iialitativelv and at variance with Fremy we
liave recorded on p. 4S. On adding warm concentrated strontia-water
to the five-sixths normal potassium salt the mixed solutions remain
clear for a few moments, but then become suddenly filled with fine
needles of very silky lustre. Yet this precipitate, which retains its
lustre when drv, has a com])Osition as complex as that of any barium
salt we have examined, the formula for it being (riO)3SriJvs(NSo07)j„
I6H0O, which may be expressed more simply as (HOSr)3XSo07,
8(SrKXS,0;, 2H,.0), and in other ways :—

-Calc. Found.

Strontium i^S-9G 1\S-G9

Potassiiun \)'oi) 9-3'J

Sulphur 17'oo 17'41

Mtroxy-rad., XO.. . . .S-12 S-lG

Water 8-44 8-67

Hydroxyl I'oo —

A strontium sodium salt crystallises out in hemispherical tufts
of brilliant silkv needles, some hours after mixing either the disodium
sidt witli warm concentrated strontia-water, or the normal sodium salt
with strontium chloride. Xo analysis has been made of the salt.

Ca lei um o.r im ii hmilp h on a tes.

Calcium salts, including calcium hydroxide, are not precipitated
bv alkali oximidosulphonates. When solution of the nonnal or the
two-thirds normal ammonium salt is mixed with pure soft calcium
hydroxide, one w\o\. of the salt dissolves ab(^ut one mol. of the hydroxide,
this not havins' lieen added in excess, and ammonia is liberated. The

78

E. DIVERS AXD T. HAGA.

solution evaporated on the Avater-hath gives off ammonia and leaves a
crystalline residue, which may be taken to be a compound or mixture
of the salts CaAmXS.O^ and CaliNSoO^. Treated with water some
nearlv insoluble calcium oximidosulphonate is left, while tlie greater
part of the mass dissolves up as calcium ammonium salt.

If after dissolving, as described in the preceding paragraph, two
mois, of calcium hydroxide in two mois, of the ammonium salt, a
third mol. of the softest moist calcium hydroxide l^e stirred in, it may
be seen to give place to a voluminous precipitate which most probably
is normal calcium oximidosulphonate. The precipitate has very little
solubility in water, and is so free from ammonia as to evolve none
when mixed witli calcium hydroxide, a test however which is not
(juite conclusive.

Lead oximidof^idplioiiaies.

Reactions of allrdi o.riiuidosulplionales trith lead acetates. — The
relictions of oximidosulphonates with lead salts are complex. The
disodium arid dipotassium sadts give no precipitate with normal lead
acetate, l)ut precipitate with the basic lend acetates. With highly
basic acetates in good excess the precipitate is mainly or wholly the
normal hydroxy-lead salt (numbered 2 in the description of these salts
which follows). With a basic acetate not in excess and in solution
not concentrated, the ]n'ecipitate is the salt numbered 3 ; while with
medium quantities of basic acetate and concentrated solutions the pre-
ci[»itate is approximately the salt 4. The precipitates usually contain,
especially when strong solutions are worked with, acetate and alkali
salt, only partially removable l)y washing with hot water.

Normal sodium oximidosulphonate and normal lead acetate show
no immediate precipitation unless the solutions are dilute. AVith
enough of the sodium salt present, two mois, or inore, to one mol. of

OXIMIDOSULPHONATES OR SULPHAZOTATES. 79

the acetate, and using concentrated solutions, the mixture in a few
hours becomes filled more or less with a soft mass of minute crystals
of a lead sodium salt (7). Using much less sodium salt the mixed
solutions, if not dilute, also remain clear, but very shnvh^ deposit
spherical crystalline hard grains of the salt 4. The motlier-liquor of
this salt i^recipitates with water. If less concentrated the mixed
solution? yield a precipitate at once, which however redissolves on
heating, while concentrating the mixture by evaporation on a water-
bath causes the precipitate in what were even dilute solutions, to
o-raduallv redissolve. Evaporation to drvues^, yields a o"um-like
mass soluble in a little water but decomposed by much.

Normal potassium oximidosulphonate and also the Kg salt give
ari immediate precipitate with normal lead acetate, soluble in excess
of either mother-salt in strong solution, but the solution remains clear
for only a few moments and then gives a crystalline precipitate unlike
tlie flocculent and voluminous one first formed {j'f. Fremy). AVhen
the oximide salt is in excess the precipitate is G ; with the acetate in
excess it is 5. Xormal ammonium oximidosulphonate is like the
sodium salt in keeping clear ïov a long time after being mixed with
normal lead acetate when in moderately concentrated solution, and in
precipitating when largely diluted. Its behaviour has not been
further observed.

The normal sodium saU and a basic lead acetate give no ])reci])i-
tate when the former is in the ]^roportion of one mol. or more to one
mob of hemihydroxy-lead acetate, even after a long time or upon
dilution with watei-. With the basic lead acetate in excess precipita-
tion of the normal hydroxy-lead salt (Î) is immediate.

The normal ]wtassirnn salt in excess and in concentrated solution
remains clear for a very slioi-t time after being 7uixed with basic
acetate ; it then deposits a crystalline flocculent ]irer-ipitate rcdis^olv-

go E. DIVERS AND T. HAGA.

ing on heating and re-forming on cooling (ß). The basic acetate
being in excess, instant precij^itation takes place. Tlie K.-, salt in
excess with basic lead acetate behaves as with the normal acetfite [cf.
Fremy). With the basic acetate in excess it reacts in the same way as the
normal potassium salt. The normal ammonium salt with basic lead
acetate behaves essentially as the sodium salt, but the solvent action
of ammonium salts prevents complete precipitation. Except when a
basic acetate in excess is used, the uK^ther-liquors of the lead precipi-
tates are ricli in the two-thirds normal salt of the alkali metal used, a
ficf pointed out by Fremy, but emphaticilly denied by Claus,

Description of tlie .•■ictlts. — Like other oximidosidphonates the lead
salts combine readily and in varying proportions with other salts.
Accordingly, several double salts of lead and S(xlitim, of lead and
potassium, And even <^f lead and hydrogen can be prepared. These
are ])artlv decomposed b\' water, ])ut washing even with luit water
never removes all alkali salt. In nearly all the salts the lead is
present half as livdroxide or oxide. An exception is the unstable
two-thirds normal load salt. With admissible and only slight
(pialifications, all the salts we have analysed may l)e expressed as
derivatives of one. two, three, or fiiir molecules of the acid.

(1) Tirn-lJiirils Hornuil lead n.rimidosHjplioNalc, PldTXS^O;. This
si'.lt can be obtained from tlie normal liydroxy-lead salt (2) which is
insoluble in water, by stirring it with water containing almost enough
sulphuric acid to deprive it of two-thirds of its lead. The mother-liquor,
which is slightly acid, when mtiderately evap<H-atc(l in the desiccator
gives a crust of minute crystals. These aj^pear nndci' the microscope
as trans])arent prisms somewhat hollowed at tlieii' middle to a dumb-
bell-shape. Tlie salt is oidv sparingly so]ul)l('. and is very unstalile.
We have not analvsed it.

(2) NoniidJ ]iiiilro.ri/-h'iiil oxiinidosuljijioiiatr, (1 lOPb)3NS.O;,

OXIMIDOSULPHONATES OR SULPH AZOTATES. g]^

3H2O. — Hemihvdroxv or still more ba.sic lead acetate in u'ood excess
of the calculated (jiiantity yields this salt in the pure state when
normal sodium oximidosulphonate is poured into it with stirriuir.
When the lead sohitiijn is poiu'ed ^-radiinlly into tlic S(xlium salt the
precipitate conhdns generally a little lead hydroxide, free or C(jnd)ined.
The salt is (^uite stahle and may he freely washed wiLli water hot ov
cold. It is a very vohmiinoiis tlocculent precipitate, not in the least
slimy or gelatinous, evidently crystalline under the microscope. The
reaction hy w^hich it is formed is expressed by the equati(jn —

A^asNS.O, + 3( PbÜH)OÄc = (HO Pb)3X6,0, + 3NaO ÄI-.

The salt can also Ije prepared, when sufticient care is taken, 1)V addin«-»-
to good excess of tri basic lead acetate a solution of the disodium salt.
The reaction then is —

Na3NS,U, + rb3(0H)/0Ä:^),-(HüPb)3XS,0, + i>Xa0i^+lL,0.

This lead salt decomposes when heated, as already described
(p. 4:'J) into sulphite and nitrite. It is soluble in acetic and other
acids, in sodium hydroxide, in ammonium chloride and otlier
ammonium salts, including ammonium oximidosulphonate. It is
easily and fidly decomposed in the cold b\' sodium, potassiiun. or
ammonium hydrogen carbonate. The last named salt gives as an
intermediate product a soluble lead anmionium salt (7). The results
of analysis agree well with calculation : —

Calc. Found.

Lead 67-70 67-57

Sulphur (y[)l} CrSD

(o) l'iro-tltiras noniial hijdwxij-lcad oxi)nidosuli>]ioi[alc, (HOPb).,
HNS2O;, 1I.,0. — Idemihydroxy-lead acetate added to a solution of
dipotassium oximidosulphonate, the latter in small excess, ^ives a

82

E. DIVERS AND T. HAGA.

precipitate of tili ïri salt, not quite pure however, but containing small
quantities of jtotassium, acetic acid, and lead in excess. Tlie corres-
])ündin,u- sodium salt could no doubt be used in place of the potassium
snlt, but has not been tried. The precipitate is voluminous and
üoccident. Its formation is represented by the equation —

K,HN.S,U, + iXPbOH)OAc=(HOPb)JiNS,0, + LnvOAc.

Analysis of a preparation freest from potassium gave the follow-
inf'' results, which are compared "with the results of calculation for the
above formula with one-ninth of an atom of PbO additional : —

Cale. Found.

Lead (U-10 63-92

Potassium — 0*23

Sulphur 9-39 9-3(;

The slight approach in composition to the normal hydroxy-lead salt,
which the small excess of lead indicates, is in agreement with what is
observed in preparing the dipotassium salt, which is apt to crystallise
with a little excess of potassium. Calculation for the pure salt gives
lead 63'01 and sulphur 9*74 per cent.

(4) Flvc-!<ix(hs normal accio-lujilfoxij-lcad o.ciiiiidosuliiltoNale.
(Ä;:OPb)(HOPb),H(NSoO,),, 24-H,0.— When to two mois, of normal
sodium oximidosulphonate in somewhat concentrated solution excess —
say three to six mois. — of normal lead acetate is added, there form in
some hours or days hard, almost opaque, spherical granule.-, of radiat-
ing crystalline structure, some at the surface of the solution, others
adherent to the sides and bottom of the vessel. These granules are of
the required salt not quite pure however and incapable of being purified
because slowly decomposed l)y water. The preparati(3n of which the
analysis follows, contained, it will be seen, acetic acid in excess ofthat
indicated by the formula. It would be possible to include this extra

OXlMIDOtiULPHOXATES OR SULPHAZOTATES.

83

eighth of acetic acid in a complex formula, but to do «o would he
almost vain, for while the acetic acid was in other ])reparations also in
excess, the excess was variable though the total acetic a^id Avas still
about 4 per cent : —

C'alo. Poiiuil.

Lead (iovlS (i.VH»

.Sulphur ,S-UG 8-14

Acetic aci d ;> • 7 (S 4"2i\

(5) Ficc-.'iixtlis noniKil hi/druxii-lead potassium oximidosidphomite.
(ilOPb),5K,H,,(NSoO;),.— To form this salt calculation points to the
use of 7 mois, of the Kg salt and 5 mois, of normal lead acetate. It
was obtained by us Ijy taking -S mois, to 6 mois, of the respective
salts, the former in warm concentrated solution and the latter also in
concentrated solution. Almost immediately after mixino- them
together a dense granular but still somewhat ciu-dy ])recipitate
separated, which was drained on a tile. Air-dried, it lost weight in
a desiccator equal t(j 5'o })er cent.; the loss of 2H.() would be about
that. It was the dehydrated salt which was analysed. The equation
expressing its formation is —

7KJ4(XS,U0, + olM,(O.U-), + 5HoO = (HOPb)Jv,H,<XS

+ 10K2HXS,U,.
This sah is somewhat soluble in its mother-liquor. On addin"- more
ot the potassium salt to the mother-liquor, a precipitate (jf apparently
the salt next to be described formed.

'ale. Fuuucl.

Lead ■19-79 50-04

Potassium 9*41^ 9-15

Sulpliui- 12-34 li^-4()

(6) Eiißit-niiiÜis iioniial oxij-lcad putassiiun o.ciuiidusididiunatc.
(OPb,)K,H(X.S^,0.)„, or the nonmU salt (HOPb)PbKe(XS20,)3.— Pro-

g4 E. DIVEES AND T. HAGA.

ceeding ;i.s for tlie last described salt", but taking" to two mois, of the
potassiiiin salt only one niol. of the lead salt, the clear solution
obtained after agitating- to dissolve the precipitate first formed, yielded
us when stirred with a glass rod a sandy crystalline precipitate of tlie
composition expressed by either of tlie above formula', but with nearly
^6" of the potassium replaced by hydrogen, as tlie calculation sIkjws : —

Calc. Found.

Lead ;^3-51 3o-46

Potassium 1.S-Ü8 IS'oo

Sidphur 15'54 lo'Si^

The mother-liquor proved to be free from lead and [»ractically neutral.
The following equation ex[)resses, therefore, the reaction which had
taken place: —

4K5H(XS20,), + 1^ 1 •b(OÄc), + H,() = U Pb,K,H(NS,U,X+ iKOAc
+ 5lv,HX\S,07.

From/ s .s(///.— -Almost in the same way, that is, by adding
normal lead acetate drop by dro]) to warm concentrated solution of
the Ks «idt, agitating to dissolve the immediate precipitate, and going
on until suddenly a second crystalline precipitate ajjpeared. Fremy
got a salt liaving nearly the same content of sulphur as our salt just
described, Init with more lead and less potassium, which admits of no
simple representation by a f )rmidî), although there are, closely enough,
4K to iPb. We give the calculation for P1jK,(NS2Ü:),,, along wdth
Fremy's finding : —

Calc. Found.

(Fremy)

Lead i>7-85 28-40

Potassium 1^-04 1^-44

Sulphur 17"2'2 lo'Go

Xitrouen 3'77 3*48

OXIMIDOSULPHOXATES OR SULPHAZOTATES. 35

from wliicli it will l)e seen tliat liis sulphur is about one-tenth too low
for this formnlü. one wliich is besides improbal)le from the fact of the
lead in it being- wholly eomlnned with the oximide rndieal.

(7) EKjlit-iiuitlis normal Innlyoxij-lrad soiliinn o.rimidosiiljilionafe,
(I-IOPb)ÄGH(NS,0;),, 14H,0.— Concentrated solutions of two mois,
of the normal sodium salt and one mol. of normal lead acetate are
mixed together and left, protected from the air, till crystallisation
occurs, when the solution becomes filled, or ])artlv tilled, with minute
crystals forming with it a soft magma. The crystals are drained
from their mother-liquor and pressed between porous tiles. Thev are
efflorescent and are dissolved and decomposed by water. Two
quantities were prepared and analysed, the one more efHoresced rlian

the other. Calculation A is for 14 Aq. and P. for 10 Aq. :

A p,

Ca]c. Foun.l. Calc. Fonud.

Lead i>i)-:',S 21)-:n 80-97 U'?y2

•"Sodium !)-79 9*69 lO'Si^ 10-05

Sulphur 13-n;] 13-7". 14-8(i 14-35

This salt, like tlie potassium .^:ilr. r-Au lie formulated as a normal salt,
(HOPh)Pliîsra«(Î^^S,0,),, 15 HA), or as an oxy-lead salt. (OI^b,)XaJl
(XSoO;),,, but with water of erystallisation. The mother-liquor of tlie
crystals rontains much disodium oximidosulphonate : —

4]Sra3XS,0, + 2 Pb(OÄc), + 2hhO - (H0Pb),N"a,H(NS,0,)3 + 4XaOÄc

+ Xa,llXS,(),.

(N) Xonnal iliammonium hjidm.rii-h'ml ariniidosulplionate, FIOPbAm.,
iSTSj)-. — A normal stilt with one atom of the livdroxy-lead radical to
Uxo of ammonium has not been isolated in the pure state, but can be
got in solution aluKist pure, and the solution can be evaporated in a
desiccator to dryness with the loss of a small traction otdvofits
ammonia. Xormil hydroxy-lead oximidosulphonate (2) is pre-

86

E. DIVERS AND T. HAGA ; OXIMIDOSULPHOXATES.

cipitated from a known (juantity of normal sodium oximidosulphonate,
waslied bv décantation, and its last wasliing-water closely decanted.
The precipitate even on long standing still occupies a large volume of
]i(|uid, and if to it there is now added ])OW(l('red ammonium liydrogen
carbonate in cjuantitv calcuhited to decompo.se completely barely two-
thirds of the lead salt and the mixture is well agitated and then left
to stand, lead carbonate, filling a very small space, settles down and a
clear mother-licpior can be decanted having only a scarcely noticeable
odour of anunonia and containing one atom of lead to two of ammo-
nium, practically all tlie oximidosulphonic radic:d, and no carbonic
arid. Much water added renders it milky. It can be evaporated to
dryness in the cold and the residue redissolved in water.

I'he disodium and dipotassium liydroxy-lead salts cannot be
prepared in a similar wav. but a solution of either of these salts, or of
the diammoniutn salt, along with acetate, is apparently obtained by
mixing too-ether in concentrated solution one mol. of hemihydroxy-
lead acetate nnd one mol. of n(n*mal oximidosulphonate of sodium,
potassium, or ammonium. The solution dries up to a vitreous mass
with a little confnsed crystalline matter, and is precipitated on
dilution with water.

Constitution of Glycocoll and its
Derivatives.

(Appendix : General theory and Nomenclature
of Amido-acids.)

By

Jöji Sakurai, F. C. S., Rigakuhakushi,

Professor of Chemistry, Imperial University.

Glycocol], the prototype of that large and important class of
compounds generally called amido-acids, is usually represented as
amido-acetic acid, HoN'.CHo.COOH, in spite of an abundance of facts
which show that it must be . considered as an internal ammonium salt
of the constitution,

ILC-XH3

I I
OC-0 ;

and the object of this paper is to strengthen the evidence in support of
the latter view and, at the same time, to deprecate the almost universal
employment of the open formula in describing the reactions of this
compound.

The view that o-jycocoll is constituted as an internal ammonium
salt was first suofofested bv Erlenmever and Sigel (Lieh. Ann., 176, 349
[1875]), in order to account for the perfect neutrality of tliis compound
towards litmus ; it may also be advanr-ed to explain the high melting

§3 J- SAKUEAI.

point of o-]ycGCo]l and its insolubility in alcohol and ether. A similar
view of tlie constitution of taurin had already been put forward by
Erlenmeyer.

The great analogy in properties between glycocoll and anhydrous
beta'in (triniethyl-glycocoll) on the one hand, and the close relationship
between the latter and choline on the other, a relationship which
establishes the constitution,

H,C-N(CH3)3

I i
OC - 0

for anhydrous beta'in, giye a further support to the correctness of the
closed formula.

Again, the obseryatif^n made by ^larckw^ald, Xeumark. and
Stelzner (7Vr.. 24, 327!) [1.S91]), that glyc.^coll does not readily react
with mustard oils to form derivatives of thio-urea, contrary to the
behaviour of all primai-y amines, speaks against the commonly adopted
open formula for glycocolh

All other ]>roperties of this compound are in perfect accordance
with the constitution of tlie internal ammonium salt, and there is not
a single reaction which needs to be expressed by the open formula ;
on the contrary, there are several, as will be shown later on, which
can not be ex])lain('d by it.

In spite of the evident claim of the internal ammonium theory
to be exclusively adopted, it is astonishing to find th:it there are most
eminent authors who do not recognise it frankly and make common
use of it, and others Avho eyen disregard it altogether.^'* AYhile insist-

(1) Beilsteiu [Handhuch : 2 Aufl , I, 1182-1183 [1893]) does not even mention the closed
formula. Roscoe and Schorlemmer [Treat iw : 2 Ed., Ill, Ft. 2 [1890]) say in one place (p. 20)
tliat glycocoll must be considered as an ammonium salt, but in desci-ibing its reactions they
expressly use the open formula (pp. 20 and 100) ; alkyl g-lycocolls are, on the other hand, repre-
sented by the closed formula. V. Meyer and Jacobson {Lclirhucli : I, 828 [1893]) seem to be
greatly in favour of the internal ammonium theory, but do not use it consistently ; in fact, like

CONSTITUTION OF GLYCOCOLI- AND ITS DERIVATIVES. 39

ing- upon the analogy between glycocoll and its trimetliyl derivative,
they represent the one by the open, and the other by the closed,
formula. This inconsistency on their jnirt is. I think, to ])e attributed
partly, at any rate, to the erroneous manner in which the modes of
formation of glycfjcoU have been hitherto represented.

Thus, the producti(jn of glycocoll by the action of ammonia
upon chloracetic acid is always regarded as if taking- place by the
direct re})lacement of chlorine by the amidogen group —

H.>C-;C1 + HjXIL ll.,C-XH.,

I " """""' = I +liCl;

OC-OH OC-Oli

it, therefore, requires a fLU-ther strain of mind to re[)resent glvcocoll as
an internal ammonium salt,

HoC-NH. H2C-NH3

I -> I I

OC-OH OC-0 ,

and it is this awkwardness, no doubt, which has had much to do in
making authors hesitate in adopting the closed formula.

The above universally employed representation of the mode of
formation of glycocoll is, h(jwe\er. errone(jus, inasnuich as it does not
take into account the evident fact that ammonium chhjracefate must
first be produced. In (jrder to obtain glycocoll Ijy this method, an
excess of anunonia must be employed, whicli not only goes to neutral-
lise chloracetic acid but also to form an anunonium chloride,

Roscoe and Schorlemuiei-, they only employ the open formula iu describing the modes of
formation of glycocoll. Mono- and dimethyl glycocoUs are represented by the open, and
trimethyl glycocoll by the closed, formula. Strecker and Wislicenus ('l'ext Book: English
Translation, 416 [1885]) are more decidedly inclined to this theory; but still they are not quite
consistent, iu one or two places using the open formula and in several others the closed double
formula. The only work I have seen where the internal ammonium theory is exclusively
adulated, though on an iasufScieut ground, is the article on Glycocoll, contributed by Hell to
the Xeues HamhàJrterhuch der Cheime, III, 446 [1878].

90

J. SAKURAI.

H.,C-C1 H.C-NH3CI

" I + i^XH3 = I

OC-OH OC-ONH, ;

the latter part of this change being analogous to that which occurs
between ammonia and alkyl halides. This ammonium compound
must, then, be regarded as decomposing into glycocoll and ammonium
chloride liy the action of heat :

HoC-NHs'Cr: H.C-NH3

I :-■ 1=11+ NH.Cl.
OC-OiNH^I OC-0

The conception of the mode of formation of glycocoll here ad-
vanced is not a matter of speculation, but is only an expression of
actual facts, and it ncccssarihj leads to the internal ammonium theory
of the constitution of this compound ; for, if we regarded glycocoll as
Ha^. CHo. COOH, we should liave to assume that the highly acid
group, COOH, remains unneutrallised by ammonia even in the pre-
sence of an excess of the latter !

Again, it may be observed that in describing various other modes
of formation of glycocoll and of the " amido-acids " generally, the fact
is almost always either concealed or forgotten that it is the hydro-
chlorides or other analogous compounds which are first obtained.
Thus, it is stated that " amido-acids " are formed by heating a mix-
ture of aldehyde-ammonias and hydrogen cyanide with hydrochloric
acid, the changes which occur being indicated by some such scheme
as follows :

R.CH-NHo E.CH-XH., R.CH-NHo R.CH-NH^

i ^1 -> I -> I

OH Cj^ CO.NH, CO.OH ;

or, that they are produced by reducing nitro-acids with tin and
hydrochloric acid :

CONSTITUTION OF GLYOOCOLL AND ITS DERIVATIVES. 9]^

R.CH-NO., R.CH-NH2
CO.ÜH CO.OH .

" Amido-acids " are. however, not produced in these reactions ; it is
their hydrochlorides,

K.CH-NH3CJ

I
CO.OH

which are actually obtained. In order to prepare the " amido-acid "
itself, giycocoll for example, the hydrochloride is digested with silver
oxide (or litharge), and the silver-glycocoU decomposed by sulphuret-
ted hydrogen. The changes which occur in these operations are in
accordance with the following scheme :

*ö

H.C-NH.Cl HoC-NAgH.,Cl^ H.C-NAgH.,

I +Ag,0- I FTOj1= I I +AgCl.

OC-OH 00-OAg OC-0

H.,C-NAgH, HX-NH3

2 .| I ^ +H,8 = 1> I I + Ag.S .
OC-0 OC-0

It may be thouglit, however, that the separation of silver chloride
occurs, not in the manner indicated above, but as follows :

H.,C-NAgH.,Cl H.,C-NH.,

I = I + AgCl.

OC-OAo- OC-OAg

Against this objection, it may be pointed out that the properties of
silver-glycocolJ cannot, as will be shown later on, be accounted for by
regarding it as an amido-acetate, but that all its reactions are most
satisfactorily explained by the formula,

92 J- SAKURAI.

HoC-NAgH,
OC-0

Moreover, there is a complete parallelism between the two equations :

HoC-NHsCl H,C-NH3

1 =-11 +NH,C1

OC-ONH, OC-0

and

HoC-NA-lioCl H.C-NAgH.,

"I ' -=11 + AgCl .

OC-OAg OC-0

The synthetic formation of anhydn^iis l)etaïn fi'om trimethyl-
amine and chloracetic acid (I.iehreich : Her., 2, 1()7[18(JU]) can be
expressed in a mannci- [»crfectly analogous to that indicated for
glycocoll :

H.,C-C1 1I.,C-X(C113)3C1 HX-X(CH3)3

I +2N(CH3),-: I '--II +N(CH,VIIC1.

OC-OII ' OC-ON(CH3)3li 0(-0

If, however, we attempted to represent the abcjve change in a manner
analogous to that by which the formation of glyc(Jcoll is usually
represented —

H,C - iCl + ' CH3 iN(CH3)., H.,C - N(CH3).,

I • - I +CH3CI ,

OC-OH OC-OH

H,C-N(CIl3), 1-LÇ-X(C1I3),II

I -> I 1

OC— OH OC— 0

we should misrepresent the actual fact, inasmuch as dimethyl-glycocoll
is thus made to be the product instead of betaïn, unless, indeed, -we

CONSTITUTION OF GLYCOCOLL AND ITS DERIVATIVES. 93

assume that methyl chloride, supposed to be liberated during the first
stage of the reaction, goes to interact with dimethyl-glycocoll, produc-
ing betaïn and hydrogen chloride :

I-LC-N(CH3).JH + CliCHa ILC-K(CH,)3

I I = I I +HC1.

OC-0 OC-0

This assumption, which tins no merit as an explanation over that
ah'eady given, is, inoreover, hardly warranted by facts. For, even if
w^e suppose that dimethyl-glycocoll is first formed and then changed
into trimethyl-glycocoll by interaction ^^ith metliyl chloride, it is
difficult to imagine that this interaction would be so complete that
none of the interacting bodies should be left unaltered. The fact,
then, that by the action of ti-imethy lamine upon chloracetic acid, betaïn
is produced unmixed with dimethyl-glycocoll goes to deprive the
ground of the assumption necessitated above of any probability and,
thei'efore, to corroborate the \dew here advocated as to the mode of
formation of o'lycocoll and its trimethvl derivative.

The consideration of the mode of formation of snrkosi ne (raono-
methyl-glyrocoll), a bodv which A ol hard synthetically obtained by the
action <^f methylamine upon ethyl chloracetate ÇLieh. Ann., 123, 2G1
[1<S62]), leads to the same conclusion. The reaction is expressed as
follows, according to the view liere urged :

H,C-C1 H,C-NH,(CH,)C1

+ 2 NHo(CH3) + H.O - I \+ CoH^.OH

oc-ocj-is ' oc-oxri,(CH3) — ~-

= H„C-XH.,(CH,)

^^^_I^ + Nri,(CH3)CI .

If, however, we attempted as before to represent tlie aboA'e change in

C)^ J. SAKURAI.

the way commonly adopted in the case of glycocoll, we should have

H„C-:CÏ + HiNH(CH,) HoC-XH(CH3)

"I - I +HC1,

OC-OCA 0C-0C,H5

ILC-NH(CH3) H.C-NH(CH3)(aH,) j

"I -> I I

oc-oaH, oc-o

and in order to avoid the misrepresentation here involved that methyl-
ethyl-glycocoll is produced instead of sarkosine, we should have to
make an assumption similar to that made in the case of betain, an
assumption which is not supported by facts.

It is, then, thnt property, characteristic of the class of compounds
under consideration, of producing internal ammonium salts, which
alone can satisfactorily explain the mode of formation of sarkosine
and betnïn. Tliis consideration gives a fresh confirmation of the
correctness of the closed formula for glycocoll, unless we disregard
the evident analogv wliich exists between it and its methyl
derivatives.

The mode of formation of hippuric acid by the action of benz-
amide upon chloracetic acid has also to be represented in a manner
analoofous to that above advanced for 2'lvcocoll and its methvl dériva-
tives, thus :

H«C-C1 HoC-NH^.COCeHs.Cl

I + NH.COCJi, = I

OC-OH ' OC-OH

H.C-NHo.COCeH,

= 'i 1 + HCl .

OC - 0

This way of regarding the formation of hippuric acid makes it also a
ring compound, a conclusion which is not contradicted \y\ facts ; on

CONSTITUTION OF GLTCOCOLL AND ITS DERIVATIVES. 95

the contrary, very sliglit solubility of this compound, in cold alcohol
and ether, its feeble acid character, and its almost neutral taste rather
o-o to support this view of its constitution. That it does possess an
acid character, feeble as it is, is easily accounted for by the presence of
the benzoyl group, which imparts to the amiclic hydrogen the pro-
perty of being more easily replaced by metals generally than in the
case of glvcocoll ; and it is scarcely necessary to mention that there
are many well known non-carboxylic compounds, uric acid for ex-
ample, which possess acid characters. Again, the fact that all at-
tempts hitherto made to obtain what may be called hippuryl chloride,
COCl.CHo.XH.COC6?l5, have been attended with failure, goes towards
upsetting the accepted view that it is a carboxylic compound, an
argument which, indeed, may be used for denying carboxylic con-
stitution to all tlie so-called amido-acids. In one word, the name
benzoyl-glycocoU, alreadv in common use. expresses its constitution
perfectly —

HoC-XH..CüCeH5

"I I
OC-0

Passino", now, from the modes of formation (^f o'lvcocoU and its
derivatives, let us consider some of its transformations and discuss the
constitution of the compounds thereby produced.

A. Addition compounds. That, although glycocoll and its deriva-
tives must be regarded as closed bodies, its addition compounds should
possess an open constitution is easy to admit, remembering that a
nitrogen atom is not yet known to be capable of combining with more
than five monovalent radicals. The ordinary or hydrated betaïn,
COOH.CHo. K(CH3)3.0H, and the corresponding chloride, COOH.
CH2.N(CH3)3.C1, are bodies of this kind. Their formation presents,
therefore, no difficulties with the closed formula for glycocoll.

9ß J. SAKURAI.

(a) Comhinalion inth acids :

H.C-NH3 H.C-XHsC]

I I + HCl = I
OC-0 OC-OH

H.C - NH3 ILC - NH3 0 - CO

2 I I + HCl = J I

OC-Ü OC-OH CIH3N-CH2 .

(b) Comhination with metallic salts :

H0C-NH3 H0C-NH3.NO3

I + Kîs'03

OC-O oc- OK

wherens witli the open formula for e'lycocoll, we must assume, in this
case, tliat a douille decom]:)o.sition first takes place with formation
of potassium amido-acetate and nitric acid and that, then, the latter
unites to the amido_o-en group, unless we o'ive the irrational formula,
C00H.CHo.NH.,.K.XO3, to this compound.

(c) Conversion into the InjJrochloride of an amido-acetic ester :

HX-XH, ILC-XHsCl

I I + ROh"" +■ EiCl -^ I + HoO .

OC-0 ' OC-OR

(d) Conrersion into (jhjcoJJie acid :

Altliough this change is usually regarded as consisting in the direct
re])]acement of the amidogen group hv hydroxyl and, therefore, may
appear as otfering a difficult}^ to the closed formula, the first action of
nitrons acid must he admitted to be anah^ofous to that which occurs
between glycocoU and liydrochloric acid, namely tlie formation of an
addi tion compound —

ILC -ML H.C-ML.XO.,

I I + H.XO,. - I
OC-0 " OC-OH :

COXÖTITUTIÜN OF ULYCOCOLL AND ITS DERIVATIVES. 97

the latter, then, décomposes, under the ronditions of the experiment,
into glycollic acid, nitrogen, and water :

liX-xiis.xu., ru:-oii

I = I + X., + H.,(J.

OC-OH OC-011

13. Melallic dcrivaUces. Amono- the deri^■ati^■es of n'lvcocoJl one. wliich
is best known and most characteristic, is the copper compound. The
deep bhie colour of this bodv and its sohiljilitv in alkalies distinüTiish
it from ordinary carboxylic salts of copper, and lead us to the con-
clusion that it is most probably a cuprammonium compound, a con-
ception which can be readily expressed by the following scheme :

H.,C - XH3 H.C - XH., - Cu - H.,X - CH.,

2 I I 4- CuO = I 1 II +H.,0 .

OC-0 OC-0 0-CO

If we attempted to represent the copper compound as an ammoniated
derivtitive by hel[) of the open formula and gave it the constitution
Cu(XH.CH2.COOH)2, we should have to make the baseless assumption
that the hydrogen of the carboxyl group remains unreplaced by copper
even in the presence of an excess of cupric oxide. If, on the other
hand, we regarded the copper derivative as (H2X.CH._,.COO)2C*u, look-
ing upon glycocoll as a carboxylic compound (an acid) because it
dissolves oxide of copper and some (^ther metallic oxides, we might
argue that ammonium chloride or even anmionia itself is also an
acid !

Reference must here be made to two important papers bearing
upon the ([uestion : one by Curtius and Goebel (/. pralt. Clieiii., 37,
150 [18.S.S]), the other by Kraut (Lieh. Ami., 266, i^9l> [1891]). By
digestion of the hydrochloride of an amido-acetic ester with an excess
of freshly precipitated cupric oxide, Curtius and Goebel obtained a

98

J. SAKUEAI.

co!)per Compound which they represent by the forniulu, COOR.CHg.
NH — Cu — XH.CHo.COOK ; whilst Kraut denies the existence of such a
compound and shows, by a careful experimental study, that the inter-
acticjn between the hydivjchloride of an amido-acetic ester and cupric
oxide occurs ac('ording to the ecjuation :

H.,C-NH3C1 H.,C-NH, H,X-CH,

2 "I +CuO= 'I I +Cua + 2R0H.

OC-OR OC-0 — Cu — 0-CO

Kraut concludes from tliis that it is the hydrogen of the hydroxyl,
not that of tlie amidogen, which is replaced by copper.

It must be observed, however, that w^hat Kraut obtained is a
double copper compound, having the composition, CuCl2.(H2^^.CH2.
COO),,Cu, or more simply, as he himself puts it, HäN.CHo.COOCuCl;
but he does not give any explanation of the formation of this double
salt in support of his views. I think a satisfiictory explanation of the
formation and existence of this double salt can be readily obtained by
supposing that cupric oxide displaces an equivalent of an alcoliol
from the hydrochloride of an amido-acetic ester, according to the
equation,

H.C-NH3CI H.C-NH^Cl

oi-OR '"'"''^ oi^_o>Cu + ROH,

H.C-NH3CI H^C-NH.Cl-Cu-CJH.N-CH^

or, 2 "1 +2CuO- I I +2R0H.

OC-OR OC-0 Cu 0-CO

The view here advanced as to the constitution of the copper compound
obtained by Kraut is in perfect harmony with his experimental data ;
at the same time, it shows that his conclusion that copper must be
wholly carboxylated is groundless.

CONSTITUTION OF GLYCOCOLL AND ITS DERIVATIVES. 99

Turaing" now to the silver derivative, the metal himmonium theory
ofives it the constitution,

H.,C-NH.Ao-
OC - 0 ;

and although the formation of amido-acetic esters by the action of
alkyl iodides upon it is brought forward against this view (Kraut ;
op. cit., 310), I see no ground whatever for this ai'gument. Repre-
.senting, as is usually done, the action of alkyl iodides upon silver-
gl vcocoll, as if direct replacement of silver by alkyl radicals took place,

H,C-NH,;Ag+IiR

"I I "'" ■■

OC-0

the formation of amido-acetic esters, HoX.CHo.COOR, would, indeed,
oifer a difficulty to the metallammonium theory ; but the action in
this case, as in other analogous cases, must be regarded as primarily
of an additi\'e nature, the unstable addition compound, then, decom-
posing into silver iodide and the ester, under the conditions of the
experiment :

H.C-NH.Ao- HoC-NH.,AoI IIX-NH.

II ^ + IR = I -- I + Agi .

OC-0 OC-OR OC-OR

The formation of amido-acetic esters, and not alkyl-glycocolls, bv the
action of alkyl iodides upon silver-glycocoll is a strong argument
against the metallamitlo-acelic acid thcurij (MHX.CtL,.COOH), but it
does not t(3uch the metallammonium theory.

The formation of hippuric acid by the action of benzoyl chloride
upon silver ((jr zinc) giycocoll is, on the other hand, a death-blow^ to

IQQ J. SAKURAI.

the metallic aumlo-acctatc theory (HojST.CHg.COOM) ; it is this reaction
which has led many eminent chemists to regard metallic derivatives
of glycocoll as MHX.CH..COOH, and to represent the formation of
hippuric acid accordingly :

COOH.CH,.^H!Äg + CliC0C«H5 = COOH.CH,.NH.COCeH, + AgCl.

As already pointed out, however, there are reasons to believe
that hippuric acid is benzoyl-glycocoll, and its formation is, therefore,
no more easily expressed Ijy the scheme above given than by

H,C-NH..iAg + Ci;COC,H, H2C-XH..C0C,H,

I I ■" ■ -- i I +AgCl.

OC-0 OC-0

It must be observed that an opening of the ring, as in the case of the
action of alkyl iodides upon silver-giycocoll, could not occur here,
because both the radicals, CI and COCJi^, are negative.

The metallic amido-acetate theory can furnish no satisfactory
explanation of the formation of hippuric acid by the action of benzoyl
chloride upon silver-giycocoll, nor can the metallamido-acetic acid
theory that of the production of amido-acetic esters by the action of
alkyl iodides upon silver-giycocoll. Both theories are imperfect, in-
asmuch as each considers only one set of facts ; the metalkuiinioniLtin
theonj, on the contrary, takes complete account of the reactions of the
derivatives under consideration and, at the same time, meets the
objections and requirements of the other two theories.

With regard to other metallic derivatives, such as those contain-
ing iiicrcunj, cadinii))!. :inc. iiHKjnc.^imii, and lead, they are no doul)t
constituted like the copper or the silver derivative. Those containing
the metals of the alkaline earths — barium, slruntiiim, and calcium — have
only Ijeen recently obtained in definite and crystalline states (Kraut:

CONSTITUTION OF GLYCOCOLL AND ITS DERIVATIVES. ^Q]^

op. cit, 299). Horsford (Lieh. Ann., 60, 33 [1846]) and, more recent-
ly, Cartius (J. prali. Clinn., 26, 159 [1S,S2]) attempted in vain to
prepare these bodies in the pure state; it is only by mixino- concent-
rated acpieons solutions of ^-lycocoll and an alkaline earth, pourino-
the mixture into alcohol, and leaving- the precipitated oil for some
days in contact with the mother hquor that Kraut was able to obtain
them in a pure and crystalline condition. The evident difficulties,
under which these bodies are formed, mark them off from ordinary
carbox3dic salts, a circumstance which shows that they are probably
also metallammonium compounds. The comparative ease wdth which
derivatives containing copper, silver, mercur3^ &<•. are formed and
the difficulties, which attend the preparation of those contai ning the
metals of the alkaline earths, and which increase still more in the case
of those containing alkali metals (these derivatiA^es, in fact, do not
seem to htive been isolated as yet), sj^eak again in fovour of the metall-
ammonium theory; inasmuch as metals like copper, silver, and mer-
cury are eminently characterised l)y the ease with wdiich they form
ammoniated derivatives. In opposition to Kraut's words, " Ich halte
daher sämmtliche Metallverl^indungen des G^lycocoUs für wahre amido-
essigsaure Salz," I would rather say that all metalJic derivatives of
(flijcocoU are metaUammonium compounds.

The above examination of the modes of formation and trans-
fcn-mation of glycocoll irresistablv leads to the internal ammonium
theory of its constitution; there is, however, one other ])oint which
must be considered before going further. Tlie point in question is
the relation l)etween glycocoll and the so-called diglycolamidic and
triglycolamidic acids. The latter coinpounds are usually represented
by HX(CH,.COOH), and X(rH2.G00H), respectively, and glycocoll
and these two bodies are compared to mono-, di-. and triethylamine;
this relation is even regarded as an argument for the open formula for

IQ2 J. SAKUEAI.

giycocoll (Kraut : op. cit., 30!)). It is to be observed, however, tliat
di- and triglycolamidic acids behave respectively as mono- and di])asic
acids towards alkalies nnd alkaline earths, the composition of their
anhydrous sahs containing :iIkaH metals nnd metals of the nlkaline
earths^^^ being-

Diglycolamiclates. Triglycolamidates.

Ba.(C,HeNO,), K, . CcH.NO«

Ba.CeH.NO, ^'' .

It is onl}?- those of their derivatives containing copper, silver, zinc, or
lead^^^-metals more or less characterised by the ease with which they
form ammoninted compounds — which may be regarded as dibasic
(MVCiHsTvjO.) nnd tribnsio (:\r3.C;HeX()«) salts respeotively. These
facts are unexjilicable by the accepted constitution of di- and triglycol-
amidic acids, but receive an ample and ready explanation from the
following formulae :

Diglycolauiidic acid. Triglycolamidic acid.

H,C-NHo.CH„.COOII H.,C-NH(CHo.COOm,

II II

OC-0 OC-0 ;

the constitution of their salts or metallic derivatives being-

H.C-NHo.CIio.COONH, H„C-NH(CH.,.COOK).,

■| I ' II''

OC-0 OC-0

(2) For the composition of these and other salts, see Heintz : Lieh. Ann., 122, 2G9 ; 124,
207 ; 156, 54 ; Lüddecke : Ibid., 147, 272 ; Beilstein : Hanâlmch, 3 Aufl., T, 1191-1192.

(3) Barium forms another salt of the composition Ba3(r6HßNOe)2, but it readily changes into
the above dibasic salt on addition of acetic acid

(4) A dibasic salt of lead of the composition Pb. CgHyXOe is also known.

CONSTITUTION OF GLYCOCOLL AND ITS DEIUVATIYES. \0l\

H.,C-XH..CH..COO^ H.,C-XH(CH„.COO)X'a

'II'' " II

OC - 0 OC - 0

H.,C-XAuH.CH.,.COOAg HX'-X^(CH.,.COU).,Cii .

'11^ 'II"

OC-0 oc-o

A word must no^\ be said with regard to the term " amido-
acids."*^''^ This term which is so generally employed slioidd, in
accordance Avith the internal anmionium theory, he dropped from
chemical nomenclature and re[)laced by the word (jlijcocolls, at least in
the case of the so-called amido-carboxylic acids. This designation is
to be recommended, because it is not new. and is to Ije preferred to
other difficult names which might be suggested as expressing con-
stitution. Moreover, a particular glycocoll may be easily specified by
prefixing (/6'('^/f, propionic, &c. ; alanine would thus ])v propionic (jhico-
coU. These expressions have to lie distinguished from those, which
are already in use to designate derivatives of a particular glycocoll.
such as acetvl-glycocoll (aceturic acid) and benzoyl -glycocoll (hippuric
acid). AVe may also use such expressions as " glycocollic constitu-
tion," " glycocollvl group — CO^XH;, — ." &c. without confusion or
other inconvenience.

Asparagin and aspartic acid, as well as all other similar bodies,
such as o-hitamin, glutamic acid, leucine, leucic acid, &c. must he
looked u])on as glycocollyl «-ompounds. Tims, as[)aragin which is
usually regarded as " amido-succinamic acid,"

CU.OH

C,H3.XH^.

CO.XH^ ,

(5) In the ne^\- edition of his Handbuch, Beilstein changes it into " amino-acids," but very
little is gained h\ the alteration.

104

.1. SAKÜR.U.

is exceedingly like glycocoll Ijoth in its ])Iiysi("il and chemical proper-
ties, and does not possess any acid character, it ought, conse([nently,
to be represented by the fornuila,

CO.O

I I

aH,.NH,

I

CO.XH, ,

and called mccinamic ijhjcocoll. Aspartic acid, generally expressed in
name and formula as " amido-succinic acid,"

CO.OPI

I
aH3.NIL

I

CO.OH ,

must in reahty be (considered as monobasic, inasmucli as its so-called
normal salts containing alkali metals are easily decomposed by car-
bonic acid, and the only salts known in a definite state are the so-
called acid salts. This body must, therefore, be regarded as succinic
ghjcocoU, and represented by the formula,

CO.O

I I

CO.OPI .

The fact that a mono-ethyl ester is obtained 1)y direct etherification
of aspartic acid also goes to support this view of its constitution.

'I'he above consideration leads to the prediction not only of the
existence of two chemically isomeric asparagins, as is predicted by the
current view,

CONSTITUTION OP CILYCOCOLL AND ITS DERIVATIVES. IQS

CO.O CO.NH,

CH.NH3 CFI.NH3

I and I

CH, CH.

I ' 1

CO.NHo CO.O

but also of aspartic acids —

CO.O CO.OH

I I I

CH.NPL Cli.XHg

I and I

CH., C^II,

r ! .

CO.OH CO.O ,

each of which is fnrtlier capable of existing as dextro-rotatory, laevo-
rotatory, and racemoid modifications. It must l)e frankly acknow-
ledged, liowever, that all the aspartic acids known at present appear to
possess one and the same (Constitution, the two active acids obtained
from the ordinary and the sweet asj)aragins being optical isomers, and
all the inactive aspartic acids (jl)tained in various ways being identical
among themselves and also with that produced on mixing the two
active acids in ei^ual quantities (Piatti : Bei:, 19, 1(!-S4 [ISNIÎ]; Engel:
Bull, 50, 150 [18S,S]). There are, however, certain points in the
chemistry of asparagin and aspartic* acid, which require further inves-
tigation. The singular fact that, of the two as])aragins which are
regarded to be opti(\al isomers, one is tasteless and the other exceed-
ingly sweet, the fact also that these two active asparagins, mixed in
equal quantities, do not produce an inacti\e racemoid com])Ound, are
some of the points above referred to. Moreover, the inactive aspara-
gin obtained from monoethyl a— aspartate (m. p. 105°) and which,
from its mode of formation, has to be represented by the formula.

lOG

J. SAKUKAI.

CO.NH, OO.NH,

CH.NH, CH.NH,

I or rather [

CH, CH,

CO.OH CO.Ö

has not yet heen resolved into a(•ti^■e eomponents ; hut since van't
HoflP's theory does not admit of the existence of an inactive anieso-
toniic compound in the case of bodies containinjj,- only one asynmietric
carbon atom, the mesotomism of the above inactive asparagin has yet
to be achie\ed. It is not unlikely that tlic new active asparagins
thus obtained would yield active aspartic acids ditferent from those
known at present.

It is to be observed that tlie inactive asparagin is here assumed
to be chcmicaUij isomeric with the two known active asparagins (Cf.
A^ictor Meyer and -lacobson, h>r. r//.), and since the former possesses
the constitution above ui\eu, the latter should be

CO.OH ro.o

CH.NH, CH.XH3

I or I'atlier |

OH, CIL

c^o.xa, co.xib, .

It is proljable, however, that the inacti\e as})aragin will prove to be
the racemoid compound of the two known actixc asparagins, in which
case an asj)aragiii possessing the latter constitution has yet to l)e
discovei'ed. Our iion-ac(juaintan<'(' of a second aspartic acid presents,
therefore, no greater difficulties to the accej)tance of the view here
advocated as to the constitution of aspartic acids than our non-
acquaintance of a second aspargin presents to the cui-rent thecjry. From
tliis point of ^■iew, the existence of new asp:irtic «acids is again, at

CONSTITUTION OF GLYCOCOLL AND ITS DERIVATIVES. 1()7

least, as probable as that of new asparagins. Time will sliow wlietlier
these predictions are verißed or not by actual facts.

Returniiii^ to the cjiiestion of nomenclature, the generic name
taurins should be used in the case of the so-called amido-sulpli(jnic
acids, and different taurins distinguished by prefixing the names of
the divalent hydrocarbon radicals. Thus, " ^?-amido-ethyl sulphonic
acid," or " amido-isethionic acid," terms which are intended to express
the constitution of common taurin, should be replaced by ethiilene-laurin
(or simply, taurin) in accordance with the formula,

H.C-XH.K

'I >.

tL,C - so/

The "^5-methyl taurin" and the " ^5'-j^-dimethyl taurin," described by
Gabriel (Ikr.^ 22, 2988 [1889]), would be respectively propijlene-taurin
and vietlujl propulene-taunn, and expressed by the frrmulae,

H.,C-NH3. H.C - XH.,(C H3)

I >0 and I \() .

CH3.HC-SO/ CH3.HC-S0,/"^ '

TN'hilst the " y-amido-propyl sulphonic acid " (Gabriel and Lauer :
Ber., 23, 92 [1890]) should be named triiiit'tJnjh'nc-taurin, and express-
ed by the formuln,

im,-so/

CH,< >o

The above examples would, I hope, suffice to show that the nomen-
clature of the " amido-acids " here suggested is capable of general
application.

I have nuich pleasure, in conclusion, in tendering my best thanks
to Dr. E. Divers, F.R.S., for kindly looking over this paper.

SAKUR.U: CONSTITUTION OF GLYCOPOLL. 2Qf)

Addendum to the preceding paper.

June 1894.

Til the Vrocccilinijs of tlie Chemical Society, London, issued ]\I:iy
2. 18!)4. tliere a[)pe:ired an abstract of the preceding' ])aper and, \m-
niedi:!îe]v followino- it, a note on the same subject but from a pliysical
stand-point, by Dr. James Walker. Dr. Walker compared the electrical
conductivities of g-lycocoll, phenyl-glycocoll, hii)puric acid, and aceturic
acid with that of acetic acid, and found that, whilst glycocoll itself is
an extremely feeble conductor, its derivatives are far better conductors
than acetic acid. "Consequently", he says, " if acetic acid contains
a carboxyl grcuip, ])henylglycocine (etc.) must, a fortiori, contain
one." lie then concludes with the words, " If we are to trust to
analoo-y, thereibre, the evidence aiforded bv the electrical conduct-
ivitv ""oes to show that "-lycocine has not the ring constitution,
but the ordinarily accepted constitution represented by the fornuila
mi,. CH,. COOH."

In comino- to this conclusion I cannot follow him. That
glycocoll derivatives, with their high dissociation constants, are to be
reo-arded as open-chain compounds when dissolved in water is a
statement from which I do not, indeed, dissent. ]>ut, then, by the
very same reasoning glycocoll itself must be regarded as a closed-chain
compound, even when dissolved in water. Dr. Walker has not a
word to say in explanation of the remarkable difference, which he has
obser\'ed, between the non-conductivity of glycocoll and the conduct-
ivity of its derivatives. Before attempting myself to find some ex-
planation for it, I wish to emphasise the point that, if the electrolytic
behaviour of glycocoll proves anything as to its constitution, it is that

IJ^Q SAKURAI: OONSTITUTIOX OF GLYCOCOLL.

this body is not. an open-chain Compound, — -is not a c:ii'l)oxyhc com-
pound, or acid.

A probable explanation of the facts, and one which gives every
support to the view of the cyclic constitution of the glycocolls seems
obvious to me. It is that while glycocoll itself appears not to form a
hydrate when dissolved, its derivatives do form such I13 drates, and
these aMition compounds will be, as pointed out in my paper, open-

chain compounds. Oi'dinarv betaïn, for example, is, bv universal ad-

X(CH3),-CH,
mission, NrCHOaOH.CIL.COOH, and anhv<lrons belain | |

^ ^ ■ " () — CO.

I am not aware whether the electrical conductivity of betaïn has been
determined or not, but it will doubtless prove to possess a rather high
dissociation constant. Glycocoll hy(h"ochloride, an undeniable open-
chain compound, was, some years ago, found by Dr. Walker himself
to be an electrolyte.

Electrical conductivity of glycocoll derivatives must, then, be
regarded as most prol^ably due to their forming addition compounds
^vith water and their subsequent dissociation, but there is nothing to
show that these derivatives are themselves open-chain compounds.
On the contrary, the non-conductivity of glycocoll rather points to its
being close-chained (even in solution), and by analogy we may regard
its derivatives as being also ring compounds.

I must, therefore, maintain with added force that all facts, 'physical
as well as chemical, so far as known, point exclusively to the cyclic
constitution of the glycocolls as the one in every way to be preferred.

Divers ; Mf. Calomel.

Jour. Sc. Coll., Vol. VU, PI. I.

Dwcrs ; Mf. Calomel.

Jour. Sc. Coll., Vol. VII, PI. II.

Fig. 1

Fig. 2.

•'/

Divers ; Mf, Calomel

5.

3.

Jour. Sc. Coll., Vol. VII, PI. III.

à.

8.

Swi

M

m

X

m

j^\
-#

T

gf fi f N

tT

*

m

m m

it if-

m L'y

M M

+ A

0 0

0

^

m

On the After-shocks of Earthquakes.

by

F. Ömori, Rièakushi.

I. General Considerations.

§ 1. A strong' earth(|uake is almost invariaMy followed by
weaker ones and wlien it is violent and destnictive the nnniljer of
minor shocks following it may amonnt to Imndreds or even thousands.
When after-shocks are not reported to have happened it is probably
because they were deemed unimportant to record. Or it may be tliat
the seat of origin of the earth([uake being very deep or fir out under
the ocean-bed, the after-shocks did not reach the observer.

Complete records of after-shocks were obtained, I believe, f )r the
first time in the cases of the three recent great earthquakes in Japan;
namaly, those of Kuinimoto in 1889, of Mino and Owari in 1891,
and of Kagdshimi in 1893. The discussion of these records forms
the subject of the present paper.

§ 2. The numbers, daily, monthly, etc., of the after-shocks of
these three earthquakes, together with other matters relating thereto
are contained in Tables I — XXII at the end of the paper ; the shocks
being distinguished as " violent," " strong," and '" weak " (or " feeble ")
according t<3 their intensity ; the total or aggregate intensity of a
number of shocks is ol)tained by multiplying each shock by one of the
coefficients 3, 2, 1, according to its intensity, and taking the sum of
the numbers so obtained. Tlie after-shocks of the Mino-Owari earth-

112

F. ÜMOET.

quake wero recorded at the ^Meteorological Stations in Gifii. Xnij-ova,
Tsu, Kyoto, :ind 0.-:ak;i; those of the Kumamoto earthqunke, first at the
Pr.^ectural Office, and subsequent!}^ at the Meteorological -Station in
KuiU'iinoto; :ind those of the Kagoshima earthquake at the Police
Station ;ind the District Office in Chiran in Kagoshiin;i Prefecture.

§ 3. The number of after-shocks are found to diminisli ra])idly
with distance from the origin of the iriitinl eartlujunke,' as may l)e
s?en from the fillowiiiü- list: —

After-shocks of the IvriNO-OwARi Earthquake.

Stations.

Gifu.

Nao'oya.

Tsu.

Kyoto. Osaka.

1

Tokyo.

Number of shocks
during- the first two
years.

3305'' 1298'

314

125 70

33

Distance in r?' from
the central jxart of the
Neo-Valley.3

7 15

25

25 36

68

Of the six cities mentit^ned in the above taljle, Gifu, Xagoya, and
Tsu !U-e situated nearly along the axis of the de])ression ^Yhich extends
tn^m the Sea of Is(' to the ])lain of Mino and Owari, while the remain-
ing three, Kyoto, Os-dca. and Tokyo are situated in a line nnighlv
perpendicular to thi.s axis. These two sets of places must be regarded
as distinct. The decrease with distance in the numlx-r of after-shocks
is. in the case of the secc^nd group, seen to be more ra])id than in the
inverse ratio.

In tlie case (^f the Kumamoto and Kagoshima earthquakes, the

1 The origin of the initial earthquake is not necessarily the same with those of the after
shocks, so that in the immediate neighbourhood of the origins the statement in the text is by
no means true.

2 1 )'( = 2.4t miles or 3.93 kilometres.

3 The earthquake of Octoljer 2Sth, 1891 was most violent in the Xeo-Valley. See Dr. Koto's
excellent paper, " On the Cause of the Great Earthquake in Central Japan, 1891." (Jour. Sei.
Coll., vol. V. pt. IV) ; also § 35 of the present paper.

1 These numliers do not include the shocks within a few hours inmiediately after tlie
t-^arhquake, which w^'re not recorded. See Tables XI and XIT, and also ^ 17.

ox THE AFTEK-SHOOK« OP EARTHQUAKES. 113

()l)servin<r stations, Kuinninoto and Cliinni. may ])ractica1ly be con-
sidered as Ijeing- on the respective epi-centres.

§ 4. OftheooGo after-sliocks recorded at (Jifn durin^u' the first
two years, 10 were •• violent." 97 •' strong," 180(S •• weak," and 1041
"feeble," while in the remaining 40!) only sounds were heard without
shocks. From Tal)les IV and V. it will l)e seen tliat. in the days
immediately succeeding the initial great earth(juake, " violent " and
•• strong " shocks occurred very often, but that these became rarer as
time advanced. Of the 10 " violent " shocks, 9 occurred Avi thin the
lirst four months and the remaining 1 in September, 1892, twelve
months after. The "strong" shocks occu.rred all within tlie lirst
thirteen months, and the '"weak" shocks all within the first twenty
months. It thus seems that a.t the earthquake-origin grea.t instabilities
are removed quicker than small ones. Again, the numbers oi '"weak '
and '• strong " shocks at Gifu during the first two years are greater
respectively than the total numbers recorded during the same time
interval at Nagoya and Osaka. Assuming that the after-shocks all
originated somewhere near Gifu,^ it may Ije concluded tliat nearly two-
thirds of these h-id radii of propagation less than 10 /•/ and that ordy
one-fiftieth had radii greater than 35 ri.

§ 5. The number of after-shocks of the Kumamoto earthcpiake
recorded at Kumamoto up to the end of 189o is [)2'2, including 1
"violent" shock which happ-ened on August 3rd, 1889, five days
after the initia.l e;rrth(piake, 76 '• strong " shocks, and 845 "weak"
and "feeble" shocks and sounds. These shocks all extended over
small areas, (jnlv the "violent" one reaching a distance greater than
20 ri from the origin.

§ (). In the cases of the Mino-Owari and Kumamoto earth-
quakes, all the after-shocks were very much smaller in extent than

1 See § 33.

114

F. ÖMORI.

the initial shocks. In the case of the Kagoshima earthquake, there
was a second great shock about four months after the first, but the
two had quite different origins.^

The frequency or activity- of after-shocks at a given distance
from the origin is different for different earthquakes, and depends
mainly on the magnitude of the initial shock, and at the epicentre,
of course on the depth of the focus also. I may here state that the
focal depths of the three great earthquakes we are discussing were all
comparatively very small, being probably between 2 and 4 ri.

§ 7. The tal)le following will serve for making a comparis(jn of
the frequency and activity of tlie after-shocks of the three earth(piakes
as observed at or near the epi-f^-i of the latter in the days immediately
succeeding the initial shocks : —

Comparison of Frequency and .Activity

3P After-shocks.

(1) MlNci-C

wxm Eij.

(IT) KlMAMdTO E(,i.

(HI ) Kaciisiii.ma y.(!.

(IViJi.vno-^^

(V; Ka-

-cSh

Date.

Freq.

Act.

Date.

rreij.

Act.

Date.

lTe.i.

Act.

Frerj.

Act.

I'rcq.

Act.

Oct. 29tli, 1891

318

333

ruiy 30tii, ise»

27

32

Sep. 8tli, 189:-'

45

49

12

10

7.1

7.0

„ aotii, „

173

193

„ 31st, „

15

16

., 9tl., „

41

49

12

12

4.0

4.0

„ SIst, „

126

129

Aug. Ist, „

11

12

,, lOtl), ,,

28

29

12

11

4..Ï

I..-)

Nov. 1st, „

99

104

„ 211.1, „

8

9

„ mil, „

23

23

12

12

4.3

4..i

Aver.ige

12

11

5.

5.

The Kumam«3to earth({uake took place on July :^<Sth, 1<S8!), at
lib. 4 Dm., p.m.; the Mino-Owari earth(|uake, on October 2.Sth,
1<SD1, at 6h. 37m., a.m.; and the Kagoshima earthquake on Se})tember
7th, 1803, at 2h. 4()m., a.m. Though the frequencies and activities
of the first three groups, (I), (II), and (III), in the al)Ove table, which

1 See § 32.

2 " Frequency " means the number, and " activity " the total intcusiti/ defined as iu § 2, of
shocks during any given interval of time.

ox THE Ai'TEK-SHOCKS OF EARTHQUAKES.

115

are put in the same horizontal rows, may not he strictly €<3mparable
^vith each other, yet the ratios in the columns (l\) and (\) are seen
to be in each nearly constant. This Avill show that the law of decrease
with time in the fre(|uency or activity of the after-sli(^cks was nearly
the same, at least in the days immediately succeeding the initial
earth<|uake, in each of the three cases.

The dependence of the frequency ov activity on the magnitude
of the initial disturbance is uKjre clearly shown by the following table
of total numbers and activities of after-shocks during the first 30
days and the first 2 years : —

Time

(0 MiNu-OwAlil Ky.

(IIj 1\L',\IA.M(J-|'0 Eg.

(III) KAr.dsiu.MA Eg.

(IV)IiATro^^jj^^

,vn,.,„,--!J;

Interval.

Nmiibev

Activity

Number

Activity

Number

Activity

Number

Activity

Number

Activity

First 30 days.
2 yearns.

1746
33G5

1821
34S2

340
833

385
905

278

292

5.1

4.0

4.8
3.8

C.2

6 2

Xow the tot'dl (land and .<ea) areas of disturliance of the Mino-
Owari. Kumamoto, and lvag<3shi]na earthquakes were respectively about
54000, (i500, and 5000 square /•/. The magnitudes of these earth-
quakes, therefore, if represented by their areas, are in the ratios of
ll:l.o:l approximately; i.e.. the Mino-Owai'i eartli([uake was
greater 1)v eight or nine times than the Kumamoto eartlKpiake, and
by eleven times tlian tlie Kagoshin^a earthquake. The frequency or
activity ïov the first earth((uake is .seen from the above tal)le to l)e
greater by fi)ur or five times than that for the second, and greater by
six times than that fov the third. If we compare together the two
earthquakes of Kumamoto and Kago.shima, we find that the ratio of
the frequencies or activities of tlieir after-.shocks for the first 30 days
is 1.3, which is equal to the ratio of their areas of disturbance.

Thus the proportionality of the frec|uency or activity to the area

llß F. OMORI.

of distiu'ljance is ii<jt .strictly verified between these tAVo e:irth(|uakes
and the Mino-Uwari earthquake, and tlie number of after-sliocks
for the last as recorded at Gifii seems to l)e comparatively too small.
It may be that the town of Gifu is not sufficiently near to the district
which is to be regarded as the proper centre of activity of after-shocks
of the Mino-Owari earthquake. The etfect that may l)e due to
difference of focal de})ths nuist also be kept in mind.

Anyh(jw the frequencv ov activity of after-shocks increases with
the magnitude of the initial earthquake and therefore can to a
certain extent be regarded as a measure of tiie latter. Thus, for instance,
tlie Kumamoto earth(piake of July -8th, 1<S(S1), was foll(jwed, on
August ord, five days later, by a second strong shock. Acc(jrding to
the above criterion, the first shock was doul)le or trel)le as «""reat as the
second, the nundjers of their after-shocks beini>' in the ratio of 1)0 : 30
approximately.

§ <S. To show the time relation of after-shocks, curves have
been drawn, (Figs. 1,5,(), ....). whose abscissae are equal intei'vals of
time, and ordinates the numbers or activities of after-shocks during
these intervals. The form of the curves drawn, in red, through tlie
ultim-ate mean positicjns will l)e seen to be very like that of a
I'ectangular hy])erbola.

§ 9. The time rate of decrease of the frequency or acti\ity of
after-sh«jcks is at first very rapid, but afterwards becomes slow and
asymptotic. As far as the rate <.)f variati(m is concerned, the after-
shock curve may be regarded as ctmsisting of two nearly straight
])ortions, during the first of which the rate is great, and during the
second of wliich it is small, the turning pc^int occm-ring so'jner with
smaller earthquakes. In the case of the after-shocks of the Mino-
Owari earth(piake this transition took place ten or twelve days after
the initial shock.

ox THE AFTER-SnOCK'S OF EAT^TIIQUAKES. l\J

§ 10. To deduce theoretically tlie time-relation of the activity (or
frequency) of after-shocks, it will he assumed, firstly, that the activity
at any moment is proportional to the magnitude of disturhance in the
o-eotechtonic conditi(m then existinü' at or near the (^rii^in of the initial
e:u'th(|uake; and, secondly, that the reduction of this magnitude
depends on the corresponding activity of after-shocks ; that is, it is
assumed that the after-sh<x-ks remove so many points of instahility or
Aveakness at the (origin. It must also be supposed that each of the
after-shocks has its own series of secondary after-shocks depending
on its magnitude.

I^et // = the aetivity of after-shocks at any instant of time ;r :
?)?, = the corresponding magnitude of distin-hance at or near the origin
of the initial earthtpiake. expressed in any arhitrnry measure; and /,', /.'
be constants. We then obtain the following two ecpiations : —

y = l'.in;
— dm. = h'.y. (It — l/'.y.dx,

whence, l)v integration, sup])osing /.:" to be constant,

y = n.h-\ («)

II and /) being constants.

Tlie logarithmic time-decrement of the activity of after-shocks
seems to l)e a likely one, when considered from the analogy of certain
physical phenomena, but the result obtained by applying equation (<i)
to the records of after-shocks of the three recent great earthquakes is
not yery satisfactory.

A formula wliich gives nearly satisfactory results is the f )llow-

ing, —

A-

in wliich // is the frequeney (or activity) at time .r, Ic and h being

118

F. OMORT.

constants. This equation wliich re]')resents a rectangular liyperbola
may be deduced l)y tidcing terms (^dy as far as tlie first ])ower of .r
from either of the equations —

and y

1

2) + qv + rx- + ,

where //, /,', // and p, q. r are constants.

Equation (/>) im]>h'es that the frequency varies nearly in an
inverse ratio to the time.

§ 11. 'Jlic j\flHO-Oirari EnrllKjiialt-e.

Let us apply eqnatioii (]>) to the record of after-shocks t:ds:en at
the Gifu Meteorological Station. As may be seen fi'om Fig. 5, the
carve of actual frequency presents a series of maxima and minima, and
in deducing the mean values of the c<mstants /,• and // from observations,
the latter nmst either cover a number of such fluctuations, or betaken
to 1)e those at an early epoch when the rate of variation in the frequency
is very great and little alferted by periods of long (bu'ation.

Taking the half-daily mimbei's of e:u'th([uakes (buang the first
five days, from October l^Dtli to Xovember ^nd, LSOl,' and a|)plying
tlie mcthoil of Least Sipiares, we obtain the equation

440.7

in Avliich ./: denotes successive intervals of 12 hours, l)eginning with
the first half of October 21)th, ISDl, and // the corresponding numbers
of eartlKpr.dves. Actual and calculated values of 7/ are tabulated below.

1 Soe Fiu-. 0.

ON THE AFTER-SHOCKS OF EARTHQUAKES.

119

X

Date.

11, actual.

y, calculated.

0

1891,10,29, 0— 12 a.m.

171

190

1

0—12 p.m.

147

133

2

30, 0—12 a.m.

109

102

3
4

0—12 p.m.
31, 0—12 a.m.

71 1 6«-

7G

5,
6

0—12 p.m.
11, 1, 0— 12 a.m.

5 Î) ] uioan,

581 •^'^•

56

7
8

0—12 p.m.
2, 0—12 a.m.

4b

4G

mean,

■44.

42

9

0—12 p.m.

46

Now n' = 493 repvcsent.s the linlf-day interval wliich is equally
distant from the heiiinning and end of tlie year 1892, and .t = 1225
that of the year 1898. The values of y/ corresponding to tliese x as
calculated from e(|uation (c) are respectively 0.9 and 0.86. In tlie
actaal case, there were (S67 and 808 shocks at Gifu daring' tlie alcove
two years, giving the average half-daily nnnihers of 1.2 and 0.42.^

It is interesting th;it equation (f), deduced from the ohservations of
only a few da^'s immediately succeeding the initial earthquake, thu.^
represents with tî)lerable accuracy the frequency of eartlKpiakes a
year ov two later on.

In the curve re[)resented 1)y equation (l>), the point of the maxi-
mum curvature occurs at .t — VÂ:— /'. Substituting the values of the
constants ?j and // adopted in equation (c), we find .t= 19, denoting
an epoch of time about ten days after the great earthquake (see § 9).

Again, b}' taking the monthly activities for the successive eighteen

1 Assuming equation (a) and using the same data as in deducing equation (c), we
olitain

log y = 2.25-.rx0.10G,

which gives satisfactory results when .c is small, (thus, when .r = 0, )/ = 175; when .r=l, j/ = 13S;
when .(' = 2, ij = 103, etc.), but fails when .v becomes great.

220 F. OMORI.

months, from Xovember, 1891, to April, 1893, we get the foUwing
equation, —

16.9

y-

.2- + 0.897 {(1)

in wliich x denotes time in months, the origin being November,
1S91, and y the mean daily activity during the month .r. Now x = l.b
represents the middle of the year 1892, and .t = 19.5 that of the year
189o. The corresponding values of y calculated from the above
equation are respectively 2.1 and 0.84, the actual mean daily activities
of after-shocks in the two years being respectively 2.4 and 0.84.

If we put y = Tiî, equation (d) gives a; = 5 10 months or 42 years;
and if we put y = T(r, equation (c) gives ;r = 27000x ^ days or 37 years.
The meaning is that the seismic activity or frequency at Gifu due to
the residual effect of the great earthquake after al)out forty years from
the initial date may 1)e such that one " weak " or '' feeble " shock
occurs in each month. ]\Iakin2: gfreat allowance for error of calcula-
tion and quartering the above figures, we may conclude that at least
some ten years will elapse before the disturbed tract about Gifu can
practically regain its stability, that is before the activity or fre(juency
of after-shocks at that place reduces to the state of having one small
shock per montli.

The above conclusion, though (^nly the result of rough approxima-
tions, seems a very likely one, when considered in reference to the
Kumamoto earthquake of July 28th, 1889, which was far smaller than
the Mino-Owari earthquake, and whose after-shocks are still occurring
at tlie present day, about 4|- years after.

§ 12. The variation with time of the frequency or activity of
after-shocks of the Mino-Owari earthquake is comparatively simple,
and none of these shocks was of a magnitude comparable with that of
the initial earthquake itself. A few of the after-shocks, such as those of

ON THE AFTER-SHOCKS OF EARTHQUAKES. 121

January ord and September 7th, LSOl, were pretty severe and followed
by their own (secondary) after-shocks, fifty or more in number. Their
residual eifects were, however, of very short durations, being sensible
only for a month or two. Besides, in deducing equation (c), all the
after-shocks durins: the first eiofhteen months, some of which were due
to the severe ones above named, were taken account of, and, therefore,
the conclusion stated at the end of the last paragraph will not be
materially atfected by the occurrence at future times of similar severe
after-shocks in the Mino-Owari district.

§ 13. TJie Kumamoto Eartliquake.

The district about Kumamoto is steadily settlinir down to e(iuili-
brium, and there has thus far been no new great earthquake.' The
rate of decrease of the frequency of after-shocks seems in this case to
tend finally to be a little quicker than according to equation (h). The
mean annual frequency at Kumamoto is very well represented by the
following equation, —

1

•^ 0.0048 -i-=rx 0.0021 4- 2-2 X0.U043 (e)

in which x denotes time, in years, 1890 being the origin, and ?/ the
corresponding yearly number of earthquakes. The values of y cal-
culated from this equation for the years 1890, 1891, 1892, and 1893
are respectively 208, 89, 38, and 20, ngreeing exactly with the actual
numl^rs for these years.

According to equation (c), the numbers of earthquakes for the
years 189-1, 1895, and 1896 would be respectively 12, 8, and 5.
Xow, before the earthquake of July 28th, 1889, the average yearly
number of shocks at Kumamoto had been 3 or 4. AVe may, therefore,
conclude that it would be about seven or eight years from the date of

1 See Fis. 4.

122

F. OMORI.

the initial earthquake hefore the number of shocks in the disturbed tract
about Kumamoto can attain its original yearly average, if at all.

§ 14:. We have seen (§ 7) that the number of after-shocks of
the Mino-Owari earthquake during the first two years is about four
times greater thfin that of the Kumamoto earthquake during the
corresponding interval of time. If we now assume that the seismic
frequency at Gifu after a given interval of time from the Mino-Owari
earthquake is always greater by this ratio than that at Kumamoto after
an equal interval from the Kumamoto earthquake, the result inferred
from the analogy of the latter is that some nine or ten years will
elapse before the yearly number of earthquakes at Gifu is reduced to
ten or twelve (compare § 1 1).

The great earthquake of üctol)er 28th, 1891, must have removed
the principal geotechtonic instability which had existed beneath
the Mino-Owari district, and in this part of tlie country the present
epoch of seismic activity will be followed ])robal)Iy Ijy one of rest.
The average yearly number of earth(|uakes at Gifu before 181)1 was
about 15.

§ 15. I shall here remark that the space distrilmtion of seismic
energy as represented by the relative number of earthquakes during a
given interval of time may vary from time to time, and particularly
with the occurrence of great earthquakes. In general there is a coin-
cidence between the distri1)ution of destructive shocks and that of
ordinary minor ones, and the latter may be the consequences of the
former. In Japan the seismic activity, as far as smtdl enrth(|uakes
are concerned, is much greater on the Pacific than on the Ja])an Sea
side, and so is it with destructive shocks, of which 57 ^'/o took
place on the Pacific side, 28 »/o on the Japan Sea side, and only
the remaining 15 ^/o remote from either. More especially the
great shock of the 1st year of Ansei (1854), which affected severely

ox THE AFTER-SHOCKS OF EARTHQUAKES.

123

almost the whole of Japan, might have had some share in making the
distribution of seismic energy in this country such as it is at the
present day.

§ IG. Bujnsslon on the Seismic Fre^iucncij in Töhjö.

In connection with this subject, it may be interesting to examine
whether there has been in recent years a secular variation in the
seismic frequency in Tokyo.

Tlie systematic instrumental observation of earthquakes at Tokyo
dates from the 8th year of Meiji (1<S75) and has now been continued
for more tlian eighteen complete years. At first the record was
taken by means of Palmieri's seismograpli, but since 1887, it has been
taken by means of the Gray-Mihie seismograph. The numl)ers of
earthquakes during eighteen years (1876-1893) are given in the
followimi- table : —

(I) Eecorded by Palmieri's Seismograph,

Year.

Total Number
of Eqks.

Tremors.!

Difference.

1876

56

88

18

1877

71

40

31

1878
1879

. 50
70

21

30

29
40

Average of
" Difference."

1880

77

32

45

= 34.72

1881

66

32

34

1882

46

19

27

1883

32

12

20

1884

68

25

43

1885

68

38

30

1886

54

4

50

1 " Tremors " mean those shocks whose direction of motion Avas not distinctly shewn hy
Paliiiieri's Seismograph.

2 The Central Meteorological Observatory, where the record has been taken, was moved
in 1883 from its old position in the southern part of Tokyo to its present one in the Castle
grounds. The observation during that year might be imperfect and is rejected in taking the
average number of earthquakes.

124

F. OMORI.

(II) Eecorüed by THE Gray-Milne Seismograph.

Year.

Total Number
of Eqks.

Tremors. 1

Difference.

1887

80

44

36

1888

101

64

37

1889

113

75

38

Average of
" Difference."

1890

93

60

33

= 33.3

1891

97-

49=^

48

1892

73

52

21

1898

59

39

20

As fur u.s the total iiumljer of recorded earth(|iiake.s is concerned,
the frequency appears to be greater in the later epoch, (II), tlian in
the earlier, (I). l>ut, as the record of the number of tremors or very
small shocks may have been modified l)y the change in sensibity with
the change of instrument we can .safely take only those shocks whose
motion was distinctly registered, and then we see that their average
yearly number is neiu'ly identical in the two groups (I) and (H)/

The Yedo (Tokyo) earthquake of the 2nd year of Ansei (LSoo)
to(jk place 21 years before 1(S7(), and 38 years before 1893. The
above table shews that the resichiai effect of this earth(piake had
ceased to be sensil)le l)efore 1876, and tliat the )nmn seismic frequency
in Tokyo has remained ever since practically constant. It may here
be remarked that the intensity of motion in Tokyo on the occasion
of the eartluj^uake of the 2nd year of Ansei was far less than that in
the Neo-A^alley and neighbouring tracts on the occasion of the recent
Mino-Owari earthquake, the former earthquake being also smaller in
extent than the latter.

§ 17. On ilic Estimaium of the Frohahlc Total Number of Jflcr-

1 "Tremors" here mexn those shocks whose motion was too small to be distinctly
measured by the Gi'ay-Milne seismograph.

2 See Note to Tabic XVII.

3 See § 29.

ON THE AFTER-SHOCKS OF EAETHQUAKES. ^25

slioclxs of a Given Earllupuûc. — If the time relation of the frequency of
after-«hocks of an earthquake be represented by an equation, we can
readily calculate from it their approximate total number. Thus, if
//(I' :'/ij !li---!l„ ^^t' values of 7/ corresponding to x — 0, 1, 2..:)), we see that

from which inequaliiy we can approxijnately estimate the total number

(or activity) of after-sliocks, namely, ^z/,,,, // being made suitably

I)
great. To take an example, we have, from e( [nation (c), for the Mino-

Owari earthquake, —

145S , 1459^0 3] \ i«.i

2^'« > ^40.7 X log J — ';,■>. ' X log.io > 2/y«

Höh 145'.1

01- Xum > 2840 > £//,„,

in which /^ denotes the number of shocks at Gifu during twelve
hours, from 0 to 12 a.m., on Octo1)er 29th, 1891, and is equal to
190, the total number of shocks during two complete years, from
October 29th, 1891, to October 27th, 1893, being denoted by the sum

1459

^//,„. The calculated value of the latter thus comes out to be 290t) or
II
2950. The corresponding actual number is 3257.

Similarly the calculated total number of shocks during ten years,

73m

or the sum ^y,„, is found to be some 3600.

0

Now the great earthquake of October 28th, 1891, took place at
0.37 a.m., and the record of after-shocks at Gifu was not taken till
about 2 p.m. of the same day, the number recorded during the
remaining 10 hours, from 2 to 12 p.m., being 101. The total
number of shocks during the above initial d ly was prob:d3ly not less
than 300.

From these considerations I conclude tliat the entire number

120 F. ÖMOEI.

of nfter-shocks of the Mino-Owari e:irthqnnke disturbing tlie r/r/z/zV^
of Gif II is about 4000.'

Some of the great earthquakes in tlie world had areas of disturl)ance
niMiiy times bigger tlian the Mino-Owari earthquake, and we nia}"
assume that 10000 is probabj}' tlie higlicst possible total number of
after-shocks of an earthquake.

The after-shocks of tlie Kum;imoto earthquake are now approach-
ing the end and their total number is 950 or 1000.

§ LS. Tlie KdijO-thiiiia Eartliqiial'e.

The Kagoshima earth<|uake took place on September 7th, 18i)o,
and is yet only a few months old. Making an estimate from tlie recoi'd
already obtairiod f)f its after-shocks (see Figs. 15 and l(i), we find
that, at Chiran, the e]>i-f()cal tract, they may continue for three or
four years, and that the total number may be some six hundred.

The earthquake took place at about 2.1() a.m. and the record of
its after-shocks was not taken till al)out Î) [).m. of the same day, there
bei no- durin^i' this interval probabh' some 100 shocks. Makinnf this
addition, the total number of shocks at Chiran during about five
months, iij) to tiie end of January, 1(S!)4, is nearly 480, being less
than the number of shocks at Kumamoto during the corresponding
interval of time after the earthquake of Julv j?<Sth. TSS!), Avhich niay
br taken in round numbers as (iOO.

//. (hi llie Verioillciiij of tlic Fn'ijiiriicji of
u'ifter-slwclcs, etc.

§ 19. What has been said so î-iv about after-shocks relates to
the ultimate mean time-relation of their frequency or activity. When

1 It must be remarked that the conclusions regarding the after-shocks of the Mino-Owari
earthquake thus far stated are supposed to hold for Giiu, and may not hold necessarily for
other places which are not sufficiently near to th(> j-rincipal epi-focal tract.

ON THE AFTER-SHOCKS OF EAETHQUAKES. 127

examined particularly, however, there are to he seen in the variations?

of the latter various .sets of periodic fluctuations (see Figs. 1, 2, 8, 5,

etc.).

Earthquakes being isolated or discontinuons events from the
nature of tlieir causes, we can from analogy readily conceive why
the frequency of those after-shocks which happen in close succession,
should present a series of well-marked maxima and minima.

The fhiduatlons in the decrease of the frequency of after-shocks
of an earthquake may he of two kinds ; namely, those which are
proper to the earthquake under consideration, and those whose
maxima and minima occur at fixed epochs. The former are of the
nature of forced oscillations and may disappear after a time; while the
latter are of the nature of free oscillations and may hecome finally
predominant.

The amplitude of the fluctuations would evidently increase with
the maiïuitude of the earth(iunke.

As far as I can ascertain from the records of after-shocks there
are, besides the diurn:d and annual fluctuations, six diiferent series of
periods in the variation of the frequency, whose lengths range from ;i
il'W liours to severa.l months.

These varioii;-. periods h:ive Ijccn obtained by drawing curves
tlirough the mean p(«itions of points whose abscissae are equal time-
intervals of 1 hour, 2 hours, (i liours, 1 day, 2 days, 5 days, 10 days,
or 1 month, and whose ordinates are the numbers of earthquakes during
the corresponding intervals. The results given below were obtained
by a direct measurement from Figs. 1, 2, o, 5, etc.

§ 20. The Kumamoto Earthiiiicéc.

The curves of daily and 2-daily earthcpiake frequencies (Figs. 2
and o) respectively indicate periods whose average lengths are 4.()
and 12 da vs.

2 9 (S F. OMORI.

Tlie curves of 5-diiily and lO-dnily earthquake frequencies seem
t.) indicate periods whose average lengths are about oo davs and
.'i montlis respectively.

The curve of monthly earthquake frequency (Fig. 1) indicates
distinct fluctuations, of which there are seven between August, 1889,
and December, 189o. The dates of maximum and mininunn fre-
quencies as given by the curve are as follows : —

Maximviui. Miniuinin.

September, 1889, (?)
February, 1890,
November, ,,
August, 1891,
February, 1892,
August,
between March and August, 1898.

The successive intervals between the earlier well-defined maxima
or minima are from 7 to 9 months.

§ 21. Tlie Miiio-Ou-ari Eartliqunlr.

The curve of hourly earthquake frequency for Gifu (Fig. 10)
indicates a period of 8 or 9 hours, and also a shorter one of about 4
hours. The former is clearly shewn in the curve of 2-hourly earth-
cjuake frequency (Fig. 11), which gives an average length of 9 hours.

The curve of hourly earthquake frec[uency for Xagoya (Fig. 12)
indicates regular fluctuations, whose average length is about 4^ hours,
and amongst which prominent maxima, marked a, h, c, d, e,f, occur
at successive intervals of nearly twenty-four hours, shewing an
evident diurnal variation. The curve of 2-hourly earthquake fre-
cpiency indic-ites a period of mean length of about 9 hours.

The curves of daily, 2-daily, and 5-daily earthquake frequencies

October,

1889,

:\lay,

1890,

Feln-uary,

1891,

Octobei-,

5«

May,

1892,

October,

55

55 5

1898.

ON THE AFTER-SHOCKS OF EARTHQUAKES. ]^29

for Gifu (Figs. 7, 8, and 9) indicate respectively fiuctuations, whose
average length are 4|^, 12.3, and .33 days. Longer periods are not
evident.

The curve of monthly earthquake frequency for Gifu (Fig. 5)
indicates four maxima and minima between November, 1891 and
December, 1893, whose dates are as follows. — ■

Maximum. Miniumm.

April, 1892, June, 1892,

September, ,, F'ebruary, 1893,

April, 1893, June, ,,

September, ,, December, ,,

The intervals between successive maxima or minima are from 4
to 8 months.

§ 22. The Kagoshima Earlluiuahc.

The curves of daily and 2-daily earthc^uake frequencies for Cliiran
(Figs. IG and 17) indicate respectively periods whose a^erage lengths
are 4.4 and 12 dnys. The curve of 5 -daily earthquake frequency
(Fig. 18) seems to indicate a period whose average length is about
33 days.

The Kagoshima eartli(]uake is not yet sufiicientlv old to give
indications of longer periods. Tlie after-shocks of this and of the
Kumamoto earthquake were not numerous enough to enable us to
draw curves of hourly frequencies.

§ 23. We have before noted that a few severe after-shocks are
likely to be followed by their own after-shocks. It might be sup-
posed that the maxima, which occur in the curves of monthly eartli-
<|uake frequencies (Figs. 1 and 5) represent merely the effects of such
shocks and not the real fluctuations in the residual elfect of the initial
earthquake itself ]]ut the fact is, on the contrary, that strong shocks

230 ^- OMORI.

occurred when the frequency was going to reach a maximum, as with
the earthquakes of January 10th, 1894, and of September 7th, 181)2
(see Figs. 5 and 26). In the case of the latter earthquake, a maxi-
mum frequency took place indeed at Gifii, which was quite near the
origin, in the same month (September); but, at Mitake, a town about
7 ri from the origin, greater numbers of shocks were recorded in the
following two months. The three other maxima of frequency for
Gifu which took place on April, 1891, and April and September, 1898,
were accompanied by no particularly severe shocks. Similarly with
the maxima in the monthly earthquake frequency for Kumanioto.

In the case of the Mino-Owari earthquake, some of the after-
shocks are doubtless to be reiifarded as " fore-shocks " of the stron"-er
ones which followed.

§ 24. The Diurnal Fluctuation of tlic Aftcr-slioch Freqiiencij.

The diurnal and annual fluctuations of the earthquake frequency
have been discussed by various European investigators. The results
obtained by Perry, Mallet, and others, however, are more or less
doubtful, as they were chiefly based on stiitistics of vulgar records.
In the present instance, after-shocks have 1jeen recorded at meteoro-
logical stations provided with seismographs, and the results here
deduced should have therefore fu- ij-reater weiii'ht than those hitherto
obtained.

The curves of six-hourly earthquake frequencies for Gifu and
Nagoya (Figs. 13 and II) indicate the daily fluctuation very clear]3^

The Mino-Owari earthquake took place on the morning of
October 28th, 1891, and during the next thirteen complete days, from
October 29th to November 10th, there were 1258 shocks at Gifu arid
572 shocks at J^agoya. The distribution of tliese shocks in the
twenty -four hours of the day are shewn in Fig. 19, (1) and (2) (see
Tables XIII and XIV).

ON THE AFTER-SHOCKS OF EARTHQUAKES. 13][

The curve of the cliurnal earthquake fluctuation for Gifu (Fi.ir. 11),
(1)) indicates distinctly tlirce ma.rima, which occur respectively
between 4 and 5 a.m., between 11 a.m. and noon, and between (> and 7
p.m.. and tliree iiiiiiiiiui which occur respecti\'ely ])etween 1) and 10
a.m., between o and 4 p.m., and at 11 p.m. The intervals between
successive maxima are 7, 8, and *J lioiu's, and those between successive
minima 6, S, and 10 hours.

The corresponding curve for Xag'oya (Fig. 11*. ('2) ) indicates six
more or less distinct oscillations, giving a mean period of 4 hours.

It is evident (see § 21) that both the 4-liour and the 8-hour
periods existed together, but that the longer period predominated in
the diurnal eartlujuake frequency for Gifu and the shorter one in that
for Nagoya.

The curves drawn in red (Fig. 11). (1) and (2)) seem to indicate a
very slight diurnal variation of the mean frequencv, the minimum
occurring between 1 and '2 p.m., and the maximum at about 1 a.m.

The hourly distril)utions of 148 shocks at Kumamoto durinir
fourteen days, from July olst to Aus^ust lotli, 1881), and of 233
shocks as Chiran during an equal time interval, from 8th to 21st,
September, 1893, (see Tables II [ and X\'I), are, as shewn in Fig. 20,
very similar respectively to those for Gifu and Xagoya.

§ 2d. The Annual Fliicinalion of tlte After-shock Freiaencij.

Curves (1). (2), and (3), Fig. 21, shew the annual fluctuation of
after-shock frequency at Kumamoto averaged respecti^'eIy f )r four
years (1890-181)3), three years (1891-1893), and two years (1892-
1893). The first curve indicates ilivee inaxima which occur respec-
tively in M:u*cli, May, and October; and tliree minima which occur
respectively iu April, between August and September, and in Decem-
ber. The curves drawn in red indicate the annual variation of the
mean frequency.

132

I'. OMORI.

Tlie Miiio-O\v;iri and the Ka<^()sliim;i earthquakes are not yet
sufficiently old to give positive results respecting the annual fluctuation.
§ 2G. We have, in §§ 20, 21, and 22, found various periods of
the after-shock frequency whose lengths vary from a few hours to
several months. Tlie five periods of about 4 hours, S or Ü hours, 4^
days, 12 days, and 33 days, seem to occur constantly in after-shocks
of différent earthquakes, lîesides these, there may exist also a con-
stant period of some three months, (see § 20). iîut the longest period
is different in the cases of the Mino-Owari and the Ivumamoto earth-
quakes, the average length being about G months in the former and
8 months in the latter.

To see whether these periods and the diurnal atid annual fluctua-
tions stated in §§ 24 and 25 can be identified in the frequency of
ordinarji earth({uakes, we shall next consider the seismometrical
observations in Tokyo and over the whole of Japan.
§ 27. (it) Seismometrical Ohscrraiioiis in Tökijö.
The hourly distribution of IKÎS earthipiakes rec.mled instru-
mentally at Tokyo during sixteen years, from INTG to 1S<)1, (Table
X\'III),' is shewn in Fig. li>, (4). l^he diurnal fiuctuation presents
tJirce mai-inia, which occur resj-ectively between Î) and 10 a.m., between
3 and 4 p.m., and between <S and !> p.m.; and the titrée minima, which
occur respectively between 2 and 3 a.m., between 11 a.m. and noon,
and 1)etween (> and 7 p.m. The intervals between successive maxima
are (i. 5, and 13 hours, and those between successive minima 9, 7, and
S hours. The mean curve, drawn in red, seems to indicate a
very slight diurnal variation, of which the maximum occurs in the
evening and the mininunn in the early morning.

The monthly distribution of 1300' earthquakes instrumentally

1 Tables XVIIl ami XIX are t iken, hj periuission, from the Earthquake Report or the
Central Meteorological Observatory.

2 See the uote to Table XVII.

ox THE AFTER-SHOCKS OF EARTHQUAKES. i;',p,

recorded ;it Tokyo during eighteen years, from 1876 to 1893. (see
Table X\'II) is sliewn in Fig. 21. (4). The curve presents three
mnxinia which occur respectively in ]\rarch, May. and December; and
three minima wliirh oceur resper^tivelv in Januarv. April, and August
or September.^

(])) Sn.wioinffricnl OJiAcrvatiouH over All JaprDi.

The hourly distribution of 3S42 earthquakes in Japan (hiring six
years, from 1.SS5 to ISDO, (see Table XIX). is sliewn in Fig. li). (?,).
The diurnal fluctuation presents again f/nvv iJUi.rinHi, wdiich occur
respectively between 2 and 3 a.m., between 2 and 3 j).m., and iDetweeii
10 and 11 p.m.. the successive intervals being 12, 8, and 4 hours; and
tlircc minima, whieh occur respectively between midnight and 1 a.m..
between 8 and 9 a.m., and between 5 and (! p.m.. the successive
intervals being 8, 9, and 7 hours. The mean curve, drawn in red.
seems to indicate a slight vtu'iation, having a maximum in the early
morning and a minimum in the evening.

Fig. 19, (5) shews the liourly distribution of 5333 earthquakes
in Japan, including the after-shocks of the Mino-Owari and the
Krigoshima earthquakes,, (see Table XX). The character of the curve
is very similar to that in Fig. 19, (3) above described.

Tlie monthly distrilnition of earth»pi:dves during the same six
years'- is shewn in Fig. 21, (5). The annual fluctuation indicates
tlirfr nui.ri !)}((, occurring respectively in February, May, and Xovember;
and tliire minima, occurring respectively in April, August, and
December.

§ 28. ConcIu4ons.

Titrée distinct maxima and minima occur in the diurnal fluctua-
tion of the frecpiencv of after-shocks as well as in that of ordinarv

1 The maximum in March and the minimum in April, wbicrli are liere very sli^^'ht, are
markedly shewn in the Kumamoto curve, Fi^. 21, (I).

2 See Table XIX.

134

F. OMORI.

earthquakes. The hours at v^'hich these occur seem to be ditterent for
different localities and theref«^re these mny not each he shewn,
■\vhen we mix up earth()uake records from distant places of the world
too-ether. AVhether there are more earth([uakes durinij;' the nig'ht than
diu'ing the day is not certain, as may be inferi-ed from the mean curves
drawn in red.

The 4-hourly and (S-(or 1>) hourly periods indicated in the curves
of hourly earthquake frecpiency (Figs. 10, 11, and 1'2) are evidently
due to the nhove diurnal fluctuation. We h:ive not at present suf-
ficient data to determine whether other three constant periods of 4^,
12, and 3o days occur likewise in the frequenc}' of ordinary earth-
quakes.

With reg;u'd to tlie annual fluctuation, it is to b:' noted that :dl
the maxima and minima of the monthly after-shock frequency for
Kumamoto (§ 2(J), with the tAvo exceptions next mentioned, occurred
in exactly, or nearly, the same months as the maxima and minima
of fre(|uencv for nil fTapan (§ 27. ('Aj ). Only the second and third
minima at Kum;imoto (§ 20) took ])lace res])ectively in February and
]Sroveml)er, (LSilO), which are generally montjis of maximum earth-
(juake frequency. These mny denote fluctuations due to the " proper"
period of the Kumamoto after-shocks.

Again, the four maxima (^f the mi^nthlv fre(juency i\)V Gifii
(§21) occurred in A|)ril and September, which are gener:dly uKMiths
of minimum e'irth((uake frequencv. These ni;iy denote the fluctua-
tions due to the "proper" period of the Mi no-Owari after-shocks.
Of the four corresponding minima, three occurred in months of
minimum frequency, and one in a month of maximum.

As already remarked, strong shocks seem to have a tendency to
happen when the frecjuencv is going to rc-u-h a maxinuuu, and it
is interestino- to find thnt of the four severest ai'lcr-slincl.-s fthree of the

ON THE AFTER-SHOCKS OF EARTHQUAKES. I35

Mino-Ow:!ri, and one of the Kagoshima cnrtliqnake) tliree took place
ill JannaTv and one in September. I believe that periodicity ])lay.s a
very important part in the frequency of e:u'thcj^uakes, and its attentive
study may be of help in the prediction of changes in seismic activity
•md other events.

I shall here confine myself to merely stating the ficts. Theore-
tical speculations on this interesting branch of tlie earth's physics are
reserved f^r a future occasion, Avlien more materials respecting after-
shocks shall have been obtained.

§ 29. Aijaiii of the Seismic Frequencii in Töl'uü, etc.

As seen above, after-shocks indicate many periods in the variation
of the seismic frequency whose lengths range between a few hours and
one year. It is of C(v.irse possible that there should exist periods of still
lonii'er duration. We shall here ao'ain consider the seismic record
taken by instruments in Tokyo during eighteen vea.rs from ISTl! to
1.S93.' The curve of monthly e:u'th([uake numbers, a portion of
"which is shewn in Fig. 22, indicates very clearly the existence of the
annual period. There seem to exist also fliictuati(jns of a few
UKUiths' duration, of which, for instance, T can count more or less
distinctly—

10 during .']!) montlis Ijetween Jainiary, 1<S7<), and April, 1870;

15 ,, 50 ,, ,, September, 18(S2, and Xovemljer, 1(S85;

and 15 ,, 50 ,, ,, November, 1885, and Janu:u-y, 18'.)lj

The average lengtli of the period is 3.5 months.

Tiie curve of yearly earthquake numliors at Tokyo (Fig. 23)
shews fluctuations, the maxima of which occurred in the years 1880,
1884-5, and 188!', and the minima in the years 187.S, 1883,
and 188(), there being besides these a slight maximum in 1891. The
average length of the period is -l 3'ears.

1 Seiî Tal)leXVI[.

136

F. ÖWORI.

Tlie mean curve (drawn in red), Fig. 2o, seem« to indicate the
existence of a period' whose length is some 12 years.

One interestino- fact is that the three severest of all tlie earth-
quakes recorded at Tokyo during tlie eighteen years, namely, those of
February 2nd, 18<S0, of October 15th, 18S4, and of February 18th,
18S1), happened respectively in the years which mark, exactly or
nearlv, the epochs of successive maximum frequency. Ft may likewise
Ix' the case that the maxima of the longer (12 years) period may often
be marked l)y great earthcpiakes.

It may herebv 1)2 remarked that the greatest monthly earthquake
number, namely 45, was recorded in October, LSDl. Of these, 2S
happened within four days immediately after the great Mino-Owari
earth(piake (which took place on the 28tli of the same month), and
tlie remaining 17 before the latter, so that the ])roper number of
earthcpiakes in Tokyo may be assumed to have been 17x-|f, or 19.-
Of the other 225 monthly numbers the next greatest namely 18,
occurred in April, LSS!).

A good exam])l ' of the «')ccurrence of destructive earthquakes at
rather regular intervals is found in the seismic fre([uency for Kansu, a
Department in the north-western corner of China, where the intervals
between successive paroxysmal epochs range from 15 to (U years,
giving the mean length of -îo years.''

To discover periodicity in seismic frequency Ave must treat records
of earthquakes for différent localities separately. Ijy way of example,
in Fig. 2o, (J), (i>), (C) and (7)), are shewn the curves of yearly seismic
activity during 11) years, 18(55 to 1883, in Swdtzerland, the Vesuvian

1 If the sensibilities of the seismofçraphs )jy which the record of shocks has been made have
been different in the earlier and later years, the chief consequence will have been to affect
slightly the amplitude.

2 At Giiu only one earthquake had been recorded in October, liefere the 28th.

3 See the present author's paper, •' On Chinese Eartliquakes," Seis. Jour. Vol. T.

ON THE AFTER-SHOCKS OF EARTHQUAKES. 137

district, Sicily, and in the Balkan Peninsula and neighl)ouring
Islands.^ The curves for the Vesuvian district and Sicily shew each
a series of periods of about 5 years; and tliat for Switzerland shews one
well-defined of 12 years besides some ill-defined fluctuations of shorter
averaofe lens'th. It will be observed that the maxima and minima of
seismic activity for the two Italian districts occurred simultaneously,
but at quite different epochs from those for Switzerland.

§ oO. To investigate the relations, if any, between earthquakes
and the phases of the moon, sun-sp(jts, tem[)erature of the atmosphere,
etc., seems not likely to lead to valuable results and w^ould be, as
Mallet remarked, a ^vaste of scientific time and labour. Willi atmos-
pheric changes of pressure it may be different and I sliall here, there-
fore, treat shortly of the possible connection of the barometric height
-with the frequency of after-shocks.

In Table IV are li'iven the mean barometric heiijhts and the fluctua-
tions during successive days from October 28th, 1891, to A[)ril 3üth,
1892. It seems that earthquakes happen equally often with low as
with high pressiu'es. Thus from an examination of the record of tlic
daily seismic frequency at Gifu, I can count, between the above dates,
55 maxima and 55 minima in the frequency, and the means of the
barometric heights corresponding to these two sets are respectively
7()2.6J: and 763.3-1 mm., which are »practically identical.

Ao'ain, bio* barometric falls of 10 or 20 mm. or rises of 5 or 10 mm.
were not accompanied by any marked change in the seismic frequency.

A single abrupt change in the atmospheric pressure is not likely
to be accompanied by any fluctuation in the frequency of earthquakes.
If, however, barometric changes, whether small or great, (jccur at
regular intervals, then the earth's crust may finally assume certain
corresponding oscillations. Thus the dail}'' and annual fluctuations in

1 The data are taken from Fuchs' " Statistik der Erdbeben von 1865-1885 " ; see Table XXI.

l^j^ F. OMOßl.

the seismic frequency may p:irtly be due to those in the atmospheric
pressure. Especially are the curves of the annual l^aro metric and seismic
fluctuations very similar to each other.

///. On the DislribiUion of Aj'ter-slioch, etc.

§ 31. We shall lastly consider more particularly the magnitude
of after-shocks and their distribution.

Earthquakes are produced when strains in the earth's crust reach
a certain limit, and, as a very great shock would remove a correspond-
ingly great underground instability, it is probable that such a shock
would not, for a long time, be followed Ijy anotlier of a magnitude
comparable to its own, in the same or a neighbouring district. When,
however, the initial sliock is not very great, it may be followed by
another like it, and, even in this case, the position of tlie origin of the
second shock would usually Ije quite distinct from that of the first.

The above statements can well i)e illustrated by the f)ur recent
destructive earthquakes in Japan, namely, those of i\Iino-Owari, of
Noto, of Kumamoto, and of Kagoshima ; the three last ones were
nuich smaller than the tirst, which was indeed very great a.nd viokait.

§ o2. The Kumamoto earth(piake of July !2<Sth, LSSD, was
followed five days later, on August 3rd, by a second shock, which
was, however, as we have already seen (§ 7) only one-half or one-
third as great as the first. All the other after-shocks were much weaker.

The Nolo earth<p]ake of December 9th, 181)1, was fjllowed, two
days later, on the 11th of the same month, by a second severe earth-
quake. These two had ne:irly an equ;d area of disturbance, and the in-
tensity of motion near the epicentres was almost the same. Their origins
were, however^ different, the epi-centre of the first shock lacing at a
point, latitude 37°-!' IST, longitude 130° 40' E, or in the sea at about

ox THE AFTEK-SHOCKS OF EARTHQUAKES. l^y

II or -1 li to .the SSW of the Togi-muni, ILigiii District (Xoto
province); and that uf tlie second «h(3ck at a point about 2 ri to the
SSE of that of the first, also in the sea and near to the town of
Takahama in the same district. All the shocks in Xoto which fol-
lowed these two were snitdl/

The Kag-oshima etirth([uuke of September 7th, l<Sl)o, was followed,
on Januar}' 4th, 18D-1, about four months later, by a second shock whose
area of disturbance was a little greater than that (jf the first. These
two shocks again originated from different centres. The epicentre
of the first shock was irdand and near Chiran-mura in the Kiire
District, at about 7 /•/ to the SSW of Kagoshima; while that of the
second shock was in the sea at about 2 /•/ t(3 the vrest of Xoma-
saki, the distance between the two epi-centres being about 10 ri. The
damage caused by the first shock was much greater than that by
the second. The effect of the second shock on the frequency of earth-
quakes at Chiran was merely to increase slightly the amplitude of the
period of seismic activity daring which it occurred (at the epoch
marked a, Fig. 15).

In tlie case of the ^Mino-Owari earthrpiake, all the subsequent
shocks, more than oDOO in number, were far smaller than the initial
one itself. It i.> to be remarked that the three se\-erest of these
numerous after-shocks, namely, the earthquakes of January ord and
September 7th, l.S<)2, and of January 10th, 181U, all originated in
the Mino-Owari Plain, and not in the Neo-\'alley, wherein indeed no
very strong shock has ever occurred since the date of the first great one.

§ oo. T]tc Mino-Ou-ari Eartlniualc.

It thus seems that the Xeo-\^dley, or the principal epi-focal
tract, is ste;idilv settling down to equilibrium, while the Mino-Owari
Plain, having probably lines of weakness under it not completely

1 Unfortunately the after-shocks of the Xoto earthquake were not carefully recorded.

140

F. ÖMORI.

removed by the e.iithquake of October 28th, ISDl, has been affected
by the severe shocks mentioned above. It may be that the latter tract
is still to be disturbed in future by a few such shocks, wliich, however,
will then be of only secondary magnitude and not so violent as to
destroy houses.

Soon after the great eartliquake of October 2(Sth, 1(S91, temporary
seismological observatories were established at Ogaki and Midori to
cooperate with the ^Meteorological Stati(jns of Gifu and Nagoya. I
have myself passed, in the latter part of Xovember and the early
})art of December, l<Si)l, several days at Midori and Okawara, the
latter of which is a small vilhige about 5 /■/ to the NNW of the former.
Again the town of Gifu is about 7 ;■/ to the 8ISE of Midori, Avhich is
in the central [)art of the Neo- Valley, where remarkable faults have
taken place. In the following table are compared the dailv numljcrs
of earthquakes I have recorded at Midori and Okawara with those
observed at Gifu : —

Gifu.

Midori.

Okawara.

November 22nd, 1891

12

7

28rd

28

9

25th

9

5

2üth

15

15

27th

11

10

28th

16

7

29th

19

6

80th

14

8

December, 1st, ,,

7

2

2nd

1(3

7

...

3rd

17

4

...

ox THE AFTER-SHOCKS OF EARTHQUAKES.

141

Gim.

]\Iidori.

Ökawara

December, 4tb, lS9l

18

11

5 th

32

3

...

(jth

7

6

7 th

8

5

8th

19

4

...

Dth

22

Ü

10th

22

8

11th

14

5

...

12th

23

12

13th

14

2

...

14th

23

11

The results contained in tlie nbove table are graphically repre-
sented in Fio-s. 24 and 25. Fig. 24 shews that the number of earth-
cpiakes at Ökawara ^vas a little less than that at Gifu, and that the
seismic frequencies at the two places did not synchronize in the occur-
rence of maxima and minima. Fig. 25 shews that the seismic
frerptencv at Gifu synclironized in the occurrence of maxima and
minima with that at Midori, l)ut that at the latter ])lar-e it was generally
less than half what it was at the former.

Fr(^m the ahove we may conclude that the after-shocks wliich
Avere felt at ]\lidori were in the main the same as those felt at Gifu, but
that the latter place was mucli nearer to the princi])al centre of these
shocks than the former'; also that the shocks which were felt al
Ökawara had mostly proceeded from a centre ditt'erent from and less
active than that of the shocks at Gifu.

Thtis it seems that the activity of after-.shocks at the principal
epi-focal tract was distinctly less than at the regi(^ns to its north and
south. Again, from wliat \ can infer from information or from my

142

F. OMORT.

own experience while travellinu" in the vnlleys of the Tbi-kawn nnd
the Mng'i-irnwn, whieh are res]x^ctive]y to the west and cast of the
Xco- Valley, the activity of after-shocks in these two valleys was less
llian in the lattei-.

It is difhcnlr to determine the exact position (^f the most active
centre of after-shocks, 1)ut it was situated somewhere to the sonth of the
Xeo- Valley for some time immediately succeeding' tlie great earth<pi:dce
of Octol)er 28th, LSDl, quite near to Gifu and ])robah]y in the tract
adjoining the town on the west.

l>el*)w are tabnhited the monthly numbers of earthquakes recorded,
from November, IcSDl, to February, 1894, at various ])laces in the
three provinces of Mino, Owai'i, and Mikawa : — -

Y^<ar, Month

Place, District.

(Mino.)

Hacliiuian, (;tujô

Kôzuclii, Mugi

Takatouii, Yauiag'ata ...

Kitag'ata, Motosn

Dii, ÖUO

Tarui, Fuua

Takata, Tagi

Takasu, Shimo-Ishizu ...

Ogaki, Ampaclii

Girn

Ota, Kanio

Mitake, Kani

Taka.yauia, Toki

Xakatsiigawa, Eua

(OWARI.)

Otajima, Hagnri

Inazawa, Xakajiuia

Tsushima, Kaito

Koori, Xiwa

IMwajiuia, Xishi-Kasugai
Katsukawa, Iligashi-Kasugai.

^"agoyi

Atsuta, Aiohi

1891

189

2

X[

.Xll

Sniii

I

11 III

i\'

V

VI

vn

VIII

IX

X

XT

\ii

13Î!

36

175

IS

0

9

3

2

o

3

1

5

3

»>

4

30J

91

395

21

11

11

7

3

3

5

8

9

10

13

6

318

98

411

22

11

18

13

2

3

3

3

7

14

17

7

151

40

191

17

l(i

8

4

2

2

2

1

2

4

2

4

152

40

192

13

IV

8

1

2

2

o

1

1

3

2

4

137

48

185

12

10

S

5

•>

2

2

1

1

4

2

4

23?

GO

293

16

13

10

7

2

2

2

o

2

5

■1

5

2GC

66

332

16

13

10

7

1

2

2

2

2

5

4

G

4ir.

128

514

17

17

12

7

2

2

O

2

2

5

4

6

1087

416

1503

1G4

114

87

90

54

30

35

52

107

47

48

39

428

143

571

62

36

22

21

10

8

9

13

17

21

22

13

4S6

166

652

82

55

32

26

17

14

9

17

IS

25

21.

14

310

78

388

38

24

19

11

6

6

3

10

11

11

12

10

182

31

213

20

11

11

5

1

"^

2

5

7

9

4

-

533

20G

789

SO

45

25

28

22

9

13

28

33

33

13

30

575

204

779

74

45

28

14

9

5

10

15

17

24

24

18

462

126

588

31

27

16

10

5

4

4

9

11

13

17

11

468

217

685

150

52

36

55

53

50

52

4G

61

57

69

43

591

215

806

141

51

36

48

44

41.

49

42

54

54

55

40

45G

178

634

139

49

35

47

44

44

49

42

54

53

55

37

41G

113

52!'

43

oo

IG

11

11

12

1

15

13

8

12

14

29 J

72

3G(;

31

21'

il

9

8

10

■'■

15

11

13

18

14

61
106
120

60

52

53

70

70

78

867

251

333

161

SI

389
273

158

727
658
648
188
176

ON THE AFTER-SHOCKS OF EARTHQUAKES.

U3

Place, District.

1891

Uaiiaa. Chitn

CM IK AW A.)

(hiryû, Aouii

Okazaki, Xnk;i(l:i

Xisliio, Hazn

Gon, Hoi

'J'oyohashi, Alsimii

Koromo, Nis' i-Kamo

Asuke, Higashi-Kauio

Shiushiro, Miuauii-Shidara ..

Touiioka, Yaua

Taj>-iichi, Kita-Shidara

XI XIT Slim

177

194
175
142

98
72
288
287
138
58
135

42

219

241

218
178
124

91
3G3
357
171

75
167

IS'Ji

I II HI IV \' VI Vll VIII IX X XI Kll Slim

15

17

19

14

8

G

29

37

14

5

13

79
79
(J9
40
25
149
11)3
50

Place, District.

(Mino.)

Hachiuiau, Gnjö

Ko7Aicln, Mugi

Takatouii, Yamagata

Kitagata, Motosu

Ibi, Öno

Tarai, Fuwa

Takata, Tagi

Takasu, Shimo-Ishizu

Ögaki, Ampaclii

Giiu

Ota, Kamo

Mitake, Kani

Takayauia, Toki

Xakatsngawa, Eua

(OWARI)

Otajima, Hagnri

Inazawa, Xakajima

Tsnshiuia, Kaito

Koori, Xiwa

Biwajima, Nishi-Kasngai
Katsukawa, Higasbi-Kasngai

Nagoya

Atsutii, Aichi

Handa, Chita

Maegasn, Kaisai

Toyohama, Chita

1S93

1 II in TV V VI VII Vlll IX X XI XII Sum I

31

32

13

^

4

11

4

3

23

35

2

G

4

S

ü

1
1

IG

10

1894

1
3

4
308

1-07
21
21

150
56
18

350
65
45

110

30

13

15

4

11

5

62

28

57

8

13

341
64
17

366
39
G2
89
25
4
14

1
1

14

14
1

49
7
3

95

10
G

29
7
3

144

F. OMORT.

P]ace, District.

1893

1894

I

1
3
1

1
1

»3
8
i-

1

IT

2
2
2
1
1
1
1
2
2
1
1
2

1

TTT

2

1

1

1
1

3
3

1
1
1

1

TV

3

l

1

1
1

Ô

2

1
1

l

V

l

2

1

4
5
4
1

1

VI

1

2

1
1

2

1
2
2
2

1
1
I

VII

1

1

1
1

VITI

IX

1
1

1

1

2

1

X

NT

2

1

1

1
1
1
1
2

1
2

1
o

5
3
2

1
2

Sum

15

15

5

9

2

9

4

32

29

IG

7

9

9

T

13

8

5

8

4

G

3

20

40

13

5

10

16

II

(Mikawa;.

Chiryfl, Aouii

Okazaki, Xukada

Saknshima, Hazn

Xishio, Hazu

Horikiri, Atsnmi

Gou, Hoi

Toyohashi, Atsuuii

Koromo, Xishi-Kamo

Asuke, Higashi-Kauio

Shinshiro, Minauii-Sliiilara

'J'omioka, Yana

Taguchi, Kita-Shidava

Shiuioda, Kita-Shidai'a

2
2

1

1

1

2

9

8

1

E:irt!i(|uake reports from District Offices ;iii(l other stations in the
Aiclii Prefecture, i.e., m tlie two provinces of Owarl and Mikawa,
liave been sent in satisfactorily, and tlie ninnl)ers of shocks for these
as tabulated above are practically correct. The nunil)ers as recorded
at the Gifu Meteorological Station and tliree District Offices in eastern
Mino, namely, those at Mitake, Takayama, and Nakatsug-avva, are also
accurate; the records at other stations in the province, however, ;n'e
imperfect and give only the numl)ers of stronger shocks.

During 1<S!)1 and 1S1)2 the greatest number of earthipiakes was
recorded at Gifn. It is likely indeed that shocks may have been more
accurately recc/rded at a meteorological station than at District Offices,
but I believe the number of shocks was actually greatest in the
vicinity of Gifu for some time after the great earth<[uake.

The records during 1892 for the six ]:>laces in the western part of
Mino, namely, Kitagata, Tbi, Tarui, Takata, Takasu, and Ogaki, were
month l)y month nearly alike, the highest seismic frequency having
occurred at Oi>aki and the lowest at Ibi and Tarui. Of the two
places in the central part <^f the province, namely, Takatomi and

ON THE AFTER-SHOCKS OF EARTHQUAKES. I45

Kozuchi, a greater number of shocks was recorded at the former than
at the latter. Of the remaining six places, NakatsugaAva, Takayama,
Mitake, Ota, and Hachiman, the greatest number was recorded at
Mitake and the least at Hachiman.

The records taken at several stations in Owari shew an evident
change with time of localisation of seismic frequency. In 1<S91,
the f'-reatest number of earthquakes was recorded at ]>iwajima,
Otajima, and Inazawa, these |)laces being situated in a zone where,
it should be remarked, the motion had been very strong on the
occasion of the great earthquake of October, 181)1. In 1892, the
gre:itest number was recorded at Koori, and the next greatest at
lUwajima and Katsukawa. Again, in 1893 the gre:itest number was
recorded still at Ko(jri, but the next greatest at Otajima and Xagoya.
The least numbers always occurred at Atsuta and the stations in the
Chita Peninsula.

The rumibers of earth({uakes at Xagoya, Atsuta, Handa, and
Toyohama indicate an evident decrease of seismic actiN'ity ^\ü\\ dis-
tance as we go s(juth wards.

In Mikawa, the greatest numbers of eartlujuidves occurred at
Koromo and Asuke in the nin-th-western part of the ])rovince. the
activity there being nearly the same as in the vicinity of Atsuta in
Owari.

The seismic activity in the Atsumi Peninsula was less than that
in the Chita Peninsula.

§ o4. The distribution of seismic activity in Mino. Owari. and
Mikawa during 1892, 189o, and January, 189-1, will be clearly seen
from Figs. 27, 2.S, and 29, respectively, in which the curves are loci of
places where equal numbers of shocks have been recorded during each
of these intervals. In drawing Fig. 28, the numbers of eartlupiakes
during January and September have been omitted as a se\ere sliock

U6

¥. OMORI.

occurred in the beginning of e:icli of these months in the district under
consideration, and where conse(|uently the seismic activity in some parti-
cular places was greatly increased. The residual effect of these severe
secondary-earthcjuakes, however, as before remarked, soon died away,
so that Fig. 28 will fairly represent the distribution of after-shocks in
1892 due wholly to the great earthquake of 1891. On the other
hand, Fig. 29 will shew principally the effect of the strong shock
which took pLice on January 10th, 1894.

Fig. 27 shews more or less distinctly four axial lines, wliich
radiate from the vicinity of Koori and aloni!" which the seismic
activity was greater tlian in the neighbouring tracts. In Figs. 28
and 29, two of these axial lines, one of which proceeded towards Ibi in
the north-western corner of the Mino-Üwari Plain, and the other of
"vvhich proceeded to the basin of the Lsé Gulf, became insignificant and
Ave have only two distinct axial lines which extend from the same
centre towards ENE and ESE, that is, respectively along the u['per
valley of the Kiso-gaAva and into the mountain districts of the
northern Mikawa.

The origin of the severe earth(|uake of January 10th, 1(S94, was
just at the point of intersection of the above two axial lines, and tliose
of the severe earthcjuakes of January 3rd and September 7tli, 1892,
were respectively in the viritiities of Gif u (near to the Kiso-gawa) and
of Katsukawa (near to Xagoya), so that all these three shocks htid
their origins n])proximately on the same axial line running from
WNW towards ESE.

§ OD. On the Cause o,/' llie Great Earlhqiiah- of Octoher â8lli, 1891.

The fact that the centre of the greatest activity of after-shocks is,
not in the Xeo-A^alley where the shock h;id been strongest on the
occasion of tlie earthquake of October, 1891, but in a locality near
to it, suggests the idea that there had existed a very extensive

ox THE AFTER-SHOCKS OF EARTHQUAKES. I47

instability under the Echizen, Mino, Owari, uud j\Iika\va provinces,
and that the great earthquake Avas caused by some big fractures
produced in this underground strained portion of the earth's crust.

The district adjoining the south-eastern extremity of the Neo-
Yalley is not yet on tlie way of steadily settling into equihbrium, and
the four axial lines in Fig. 27 proliably indicate the positions of f<nn-
weaker or deeper fractures, the peculiar violence of motion on the
occasion of the great earthquake in the zonal tract at the western part
of the Owari Tlain and in the regions to the south-east of Koori (see
Fig. 26) being probably due to the existence of such lines of weakness
beneatli.' The underground fracture along the Neo- Valley was pro-
bably sufficiently great and must have removed the chief centre of
weakness at that district.

Tlie two axial lines in Fig. 2(S seem to indicate the positions of
fracture lines which Ijecame gradually |)rominent, and along which
the activity of after-shocks is at present greatest.

The centre of activity of after-shocks may in future change its
position, l)ut it will very probably recede from, and not approach the
JN'eo- Valley.

The distribution of seismic activity before 1<S!)1 is shewn iu Fig.
oO (see Table XXIT). It indicates no such [>eculiarity as that seen
in F'igs. 27, 28, and 29.

§ 06. EarlJi inal^r Sonnls. — I take thi.» ()p[)nrtunity of making
S(jme remarks upon earth(|uake sounds.

Many of the after-sliocks of the Mino-Owaii earth(piake were
attended with s(Hinds, which were essentially of two types, being
either rushing feeble noises like those caused by winds, or loud
rumbling S(.ninds like those caused by the fdling(jf a heavy weight on
y^round, or Ijy the dischar^'e of a u'un.

1 See Fi"-. 31.

l^j^ F. OMORI.

The sounds of the «econd type, which were sometiines like detona-
tions of tliuiider, were most frequent and distinct in the Neo-A^illey,
where, as I beUeve, their main origin really was. It is remarkable that
tremblings of the ground accompanying these sounds were invariably
very feeble, and often not to be felt at all, while severe sharp shocks
were usually not accompanied by distinctly audible sounds. This
peculiar phenomenon was ascertained likewise to have been observed
with the after-shocks of the Noto and Kagoshima earthquakes. The
following may be one of the possible explanations.

Among ruimerous de[)ressi()ns of small pieces of ground produced
by the great eartlnpiake of 1<S1)1, there was one which took ])]ace on a
high inoiuitain flunk near the village of Higashi-Yokoyama, in the
Ono District, Mino, and which was cyHndrical in form, being about
9 feet in diameter and 10 feet in deptii. The top of the cavity was
just shewn on the surface slope as a circular aj)erture about 4 feet in
diameter. From examples like this, we may suppose that some
cavities have probably been formed undergrouiid which are not shewn
on the surface. Especially in the focal region there may exist big
vacaiil spaces, and the falling down into these of superincumbent
rock masses would give rise to sounds accom|)anied by small move-
ments of ground, as muffled s(junds are heard when we throw stones
into a deep well or IkjIc. On the other hand, shocks may originate
by fracturing of strata in distiicts near the focus which would cause
sh;u"p tremblings of the ground, but not be accompanied by loud
sounds.

The slight rushing sounds often accompanying earth([uakes may,
as usually supposed, be caused by (piick tremors preceding the
principal earthquake motion.

ox THE AFTEK-SHOCKS OF EARTHQUAKES. I49

List of Tables, I-XXII.

I._Djuly Numbers of Eartliqnakes at Knmanioto, from July 20tli, 1880, to

December 31st, 1890.
II.— :\[oiitbly and Yearly Numbers of Earthquakes at Kumamoto, from July.

1889, to December, 1893.
ir.— Average Montlily Numbers of Earthijuakes at Kumamoto.
III.—Hourly Numljers of l^^artbtpiakes at Kumamoto, from July 31st to August

13tli, 1889.
IV.— Daily Numbers of Earthquakes and Barometric Heights at Gifu, from
October 28th, 1891, to April 30th, 1892.
v.— Daily Numbers of Earthquakes at Gifu, from May 1st, 1892, to December

31st, 1898.
YI.— Daily Numbers of Earthquakes at Nagoya, from October 28th, 1891, to

April 30th, 1892.
YIL— Daily Numbers of Earthquakes at Tsu, from October 28th to December

31st, 1891.
YIIL— Daily Numbers of Earthquakes at Kyoto, from October 28th to December
31st, 1891.
IX.— Daily Numbers of Earthquakes at Osaka, from October 28th to December

31st, 1891.
X.— Monthly Numbers of Earthquakes at Gifu, Nagoya, Tsu, Kyoto, and Osaka,

from October, 1891, to December, 1893.
XL— Times of Occurrence of Earthquakes at Gifu, from October 28th, 1 p.m.,

to November 10th, 12 p.m., 1891.
XII.— Times of Occurrence of Earthquakes at Nagoya, from October 28tli, 1 p.m.,
to November 10th, 12 p.m., 1891.

150

F. OMOEI.

XIII.— Hourly Nninbers of E:ivtli(in;ilccs at Gifn, from October 28tli to Noveiiil)er

lOtli, 1891.
XIY. — Honrly Nnrnbers of Eartliqnakes at Nagoya, from October 28th to Novem-
ber lütli, 18!»1.
XY. — Daily Xumbers of Karthqnakes at Cbirau, from iSoptember 7th, 181)3, to
January 81st, 1894.
XYI. — Hourly Numbers of Earthi|uakes at Chiraii, from September 8th to 21st,

1893.
XYII. — Monti ily and Yearly Numbers of Earthquakes recorded instrumentally at

Tokyo, from January, 1870, to December, 1893.
XMII. — Hourly Distribution of ll(i8 Earthquakes recorded instrumentally at Tokyo,
from 1870 to 1891.
XrX.— Hourly Distribution of 3842 Earthquakes in Japan from 1885 to 189U.
XX. — Hourly Distribution of 5333 Earthquakes in Japan.
XXI. — Yearly Seismic " Activities "' in Switzerland, the Yesuvian District, Sicily,

and the Balkan Peninsula and neighbouring Islands.
XXII.— Yearly Numbers of EartlK^uakes at different places in Owari, Mino, and
Mikawa, from 1887 to 189U.

ON THE AFTER-SHOCKS OF EARTHQUAKES.

151

TABLE I.-DAILY NUMBERS OP EARTHQUAKES AT KUMAMOTO,
FROM JULY 28th, 1889, TO DECEMBER 31st, 1890.

July, 1889.

August, 1889.

Day

'? 'to

go

CO

le 5

0

! 3

0^

■2 2

•-r ÎC

II

CO

3

1

...

1

6

4

11

12

2

1

7

8

9

3

...

1

8

18

13

35

40

4

2

11

9

22

24

5

5

6

11

11

6

1

5

5

11

12

7

1

3

4

5

8
9

5
1

6
1

11

2

11

2

10

4

2

6

6

11

4

2

6

G

12

4

1

5

5

18

4

5

9

9

]4

1

5

G

G

]5

1

4

6

11

12

16

6

7

13

13

17

3

3

6

G

18

1

5

6

G

19

...

2

1

3

3

20

2

6

8

8

21

3

4

7

7

22

...

2

1

8

3

28

2

4

6

G

24

2

3

6

8

25

1

2

4

5

26

2

3

3

27

4

5

5

28

1

1

3

5

G

6

2iJ

23

14

38

70

93

2

2 4

4

80

5

10

12

27

32

2

2

2

31

1

12

2

15

16

...

3

3

3

Sum

1

29

86

47

113

144

1

18

103 126

243

259

l')l>

F. UMORI.

September,

1889.

October, 1889.

Day

be CO

II

Oi CO

a

"Tî CO g

3

>5

•rH
■J

CO 'm

'' CO

III

ce

3

ce

1

1

. . .

4

5

6

2

1

1

1

1

2

2

5

6

8

1

1

1

2

1

3

3

4

1

1

1

3

1

4

4

5

1

1

2

3

1

3

3

7

8

G

1

1

1

3

4

1

1

1

3

4

7

8

1

...

...

1

2

3

4
1

2
1

6
2

6
2

<)

1

1

2

3

1

1

2

2

10

1

2

3

4

1

1

1

11

...

1

1

1

12

1

1

1

...

18

1

1

1

1

1

1

14

1

1

2

2

1

3

3

15

i2

2

2

1

1

1

16

1

1

1

1

1

2

3

17

1

1

1

1

1

1

18

...

1

1

2

4

5

19

O

4

6

1

1

1

20

...

...

3

3

3

21

o

2

2

1

1

1

22

1

1

2

3

1

1

2

3

23

2

2

4

1

1

2

4

5

24

1

1

2

1

2

3

4

25

1

1

1

2

2

2

26

2

'2

4

6

2

2

2

27

1

1

2

2

28

1

1

1

1

1

1

29

1

1

1

30

1

1

2

2

2

2

2

31

1

1

2

3

Slim

14

16

11

41

55

9

30

38

77

87

ox THE AFTER-SHOCKS OF EARTHQUAKES.

153

XOVEMBER, 1889.

Deckmbeb, 1889.

Day

b£ M

il

œ

"ce "3
o O

..- o w

=2 o o

s

CO

II

œ 'Ta

ce o
? 2

s

1

1

1

1

1

1

1

2

1

1

1

1

2

2

3

1

2

3

1

1

1

4

1

1

1

2

1

3

3

5

1

1

1

1

1

6

•2

3

3

7

1

2

2

8

1

1

1

3

4

9

1

2

2

2

2

2

10

1

2

2

1

1

2

4

5

11

1

3

3

1

1

2

2

12

1

2

2

1

1

1

13

2

3

3

1

1

2

2

U

1

1

1

1

2

2

15

1

1

2

2

1(3

1

2

2

2

2

2

17

1

1

1

18

1

1

3

4

1

1

1

19

1

2

2

1

1

1

20

2

2

2

21

1

1

1

1

1

2

2

22

1

2

2

1

1

1

23

1

2

2

2

2

2

24

2

2

•

25

1

2

2

1

1

1

2Ü

1

1

1

27

1

1

1

1

1

28

1

1

1

29

3

2

5

5

1

1

1

30

1

2

3

31

1

1

1

Sum

3

23

25

51

54

2

20

19

41

43

154

F. OMORI.

January,

February,

March,

April.

May,

1890.

18'J0.

1890.

1890.

1890.

Day

<u o

3

Ol

^1

öl o

03

u

m

3

11

^2

IJ C

'' 05

ce

3

u
a) ° ^

03

be œ

'S

3

1

2

2

1

1

t2

1

1

1

3

5

8

1

1

3

3

4

O

2

1

1

2

5
(5

o

2

2

1

2

1

4

2
2

1

1

7

1

2

2

2

1

8

1

1

2

2

1
1

10

1

1

1

1

1

11

1

1

1

1

1

1

1

12

2

2

1

I

]8

1

1

1

1

]4

1

1

o

1

1

2

2

15

t

i

1

...

k;

1

1

1

1

...

17

2

2

1

18

1

2

1

4

4

19

1

1

1

1

...

20

1

1

1

1

...

1

1

21

1

2

2

22

1

1

23

2

3

1

1

24

2

2

25

1

1

2

1

1

26

...

4

4

1

1

27

4

4

28

1

1

1

1

1

1

29

4

4

...

...

1

30

1

1

1

3

4

1

2

31

Sum

"2

14

16

1

7

8

4

1

35

40

16

2

4

27

33

ox THE .AFTER-SHOCKS OF EARTHQUAKES.

155

June, 1890 Ju

LY, 1890

Aug.,
1890

Sep.,
1890

Oct.,
1890

Nov.,
1890

Dec,
1890

Day

a; c

te ^

w

u

o "t; ::
.^ o c

3

O »3

111

u

2) ® ""

3 So

•Ü c c

'S

1

3 Ï a

111

CO

^ O 03

*^ o a

î^ o o

3^ ä

Ol o =

■ii o o

1

2

•2

1

4

5 ...

1

1

2

1

2

1

3

1

1

2

2

4

1

2

3

3

3

5

1

1 ...

2

2

G

1

1

2

Ï

2

7

1

1

1

8

1

2

1)

1

1 ...

2

10

1

1

1

1

11

«^

...

1-2

2

2

13

1

14

1

1 ...

1

1

1

15

16

1

1 ...

1

17

1

1 ...

1

18

1

1

2 ...

1

lu

1

1

2 ...

20

2

2

1

21

2

...

22

3

3

1

1

1

3

23

1

24

1

1 ...

25

1

1

26

1

1 ...

...

27

1

1 ...

1

1

2

28

1

1

2

29

1

1

1

1

30

1

1

2

1

...

2

31

...

1

...

Sum

6

21

27 3

17

20

19

14

6

6

3

156

F. OMORI.

TABLE II. -MONTHLY AND YEARLY NUMBERS 07 EARTHQUAKES
AT KUMAMOTO, FROM 1889 TO 1893.

Year \^

I

II

III

IV

V

VI

VII

VIII

IX

X

XI

XII

sum

1889

118^

248

41

77

51

41

566

1890

1(3

8

40

IG

88

27

20

19

14

G

6

3

208

1891

8

19

6

9

7

7

2

1

12

9

9

89

1892

8

8

8

8

7

8

4

1

2

5

1

3

88

1.S98

2

1

1

J

1

4

7

8

20-

Siiui

29

81

50

29

47

38

189

264

Gl

107

70

56

921

TABLE II.'-AVERAGE MONTHLY NUMBERS OF EARTHQUAKES

AT KUMAMOTO.

^\_^^ ildiitli

I

II

III

IV

\'

VI

VII

VIII

IX

j
X XI

XII

1890-1898

7.8

7.8

12.5

7.8

12.0

9.8

6.0

0.0

5.0

7.5

4.8

4.0

1 «91-1898

4.8

7.7

8.8

4.8

4.7

8.7

2.0

0.7

2.0

8.0

4.8

4.0

1892-1 «98

2.5

2.0

2.0

2.0

8.5

2.0

8.0

0.5

8.0

6.0 1 2.0

1.5

1. The number for July, 1889, iy tliat for four days, from 28th to 31st of that uionth.

2. Those shocks at Kumauioto which were evidently due to the Kagoshima earthquake of
Sopteml^er 7th, 1893, are not included.

ox THE .\FTE]{-SHOCKS OF EAP/niQUAKES.

157

TABLE III.— HOURLY EARTHQUAKE NUMBERS AT KUMAMOTO,
PROM JULY 31st TO AUGUST 13th, 1889.i

^\ Day

.TuLY,

1889.

August, 18S9.

snin

lutei'val ^v^

31

1

1

1

2
1
1

2
1

11

2

1
1
1

1

8

3

1
()

4
5
4
1

1
1

2

1
2

5

1

34

4.
1

2

1

1

2
1

2
2

1

1
1

1

1
17

5

1

1

1

1
1

11

(3
1

1

1
1
1
1

1

1

1

1

1
11

7
1

1

4

8

1

1
1

1

1

1

1
o

1

10

9

2

10

1

1

1

1
1

1

()

11
1

2

2

5

12

1

1
1

1

.5

13

1
1

1

1
1

1

1
1

1

9

0— 1 a.m.

1— 2

2— ;■!
:\— 4
4— .-.
.■)— (•)
(J— 7

7— s

8— 9
9—10

10—11
11 — 12
U— 1 p.m.

1— 2

2— :\
8— 4

4— 5

5— 6 .
G— 7

7— 8

8— 9
9—10

10—11
11-12

i

3

7
13
18
12
11
11

3

(i

4
o

4
3
11
5
o

1
4
3
8
4
5

4

Sum

15

1482

During Night, or Ijetween 0 p.m. and G a.m., 100
„ Day, „ „ 0 a.m. and 6 p.m., 48

148

1. The first great earthquake took place ou July 28th, 11.48 p.m , and the times of occur-
rence of suljsequent shocks were not noted down till 3 1st of the same month.

2 The times of occurrence of shocks were lost in a few cases, and tiierefore this nuuil)er is
slightly less than according to Table 1.

158

F. OMORI.

TABLE IV-DAILY NUMBERS OP EARTHQUAKES AND BAROMETRIC

HEIGHTS AT GIFU, FROM OCTOBER 28th, 1891,

TO APRIL 30th, 1892.

OCTOBER, 1891.

Dny

violent
shocks

strong
shocks

weak
shocks

feel.le
shocks

sonnds

sum activity

uieaui
))av. lit.

flnotnation,
in uiui.

in uiin.

1

2

8

4

5

0

1

8

9

10

11

12

13

14

15

16

...

17

18

19

20

21

22

28

24

25

26

27 .

28

Ï

"s

94

103

IL

3

753.7

- 3.5,+ 3.7

29

15

303

318

333

54.9

- 2. ,+ 2.2

30

3

14

147

9

173

193

61.3

+ 8.1

31

3

116

Ï

6

126

129

65.1

+ 1.3,- 2.3

Sum

4

40

660

1

15

720

768

I. The " Mean Bai-ometric Height " was olitained by taking the mean of six observations
on each clay, at 2, 0, 10, a.m., and 2, G, 10, p.m. The " Fluctuation " is the difterence Ijetween
these observations, + when rising, and - when falling. For instance, - 3.5, + .3.7 means
that the pressure first fell by 3.,5 mm. and then rose by 3.7 mm.

ox THE AFTER'SHOCKS OF EARTHQUAKES.

159

XOYEMBEE, 1891.

Day

violeni
shocks

strong-
shocks

weak
shocks

feeble
shocks

sounds

sum

activity

mean
bar. ht.
in mm.

fluctuation,
iu umi.

1

5

94

99

104

762.7

-8. , + 1.2

2

4

88

92

96

68,9

+ 1.0,-1.0, + 2.2

8

81

81

81

68.6

-8.

4

2

75

1

78

80

59.1

— 5.

5

1

52

58

54

55.8

+ 8.7

ß

4

56

1

6

67

71

59.8

+ 8.9

7

8

42

45

48

64.2

+ 4.2

8

...

1

41

42

48

67.0

+ 1. ,-2.6

9

1

48

44

45

68.0

-4.8

10

40

40

40

58.5

-6.8

11

87

1

88

88

60.5

+ 7.()

12

1

88

1

40

41

68.8

-1.7

18

1

82

1

1

sry

86

62.8

-2. ,+ 1.6

U

25

8

1

29

29

04.8

+ 2.1,-1.4,+2.

lo

28

6

29

29

64.4

-8.2

16

27

1

28

28

59.9

-5. , + 2.2

17

18 !
19

15

1

9

7

5

9
10

21
18
17

21
18
17

61.8
61.9
62.6

+ 2.5,-1.4. + 2.
-2.6, + 1.6
-1.2, + 8.

20

1

18

14

88

84

65.1

-2.7, + 2.8

21
22
23

1

1
2

10

5

9

16

4
8

5

21
12
28

24
12
25

05.0
65.4
61.9

-2.1,+ 1.6

+ 1.8
-6.2

24
25
20
27

1

8
4

14
6
8

4
8
9
6

18

9

15

11

18

9

15

12

69.1
08.8
68.8
66.6

+ 4.6

+ 1.2,-1.2, + 1.4
-2.8, + 2.8
+ 5.2

28
29

1

1

4

10

8

10

6

10
19

19
19

72.0
70.2

-1.4,+ 1.1
— 2 2

80

12

1

1

14

14

6().7

- 8.6

Sum

2

29

852

106

98

1087

1120

1(50

F, OMORT.

DECEMBEI

i, 1891.

Dny

violent
shocks

stronu-
shocks

weak
shocks

feeble
shocks

1
sonn'U

SllUl

activity

meau
bar. ht.
in mm.

fluctnatiou,
iu mm.

1

3

'J

1

7

7

7G7.4

+ 3.8

2

3

9

4

IG

IG

70.5

+ 1.3,-2.1

H

1

1

9

4

2

17

20

1 Gl. 8

-14.4

4

1

4

11

2

18

20

j 51. G

-1.4, + 5.9

5

1

17

12

2

82

88

(')4.0

+ 10.5

G

()

1

7

7

71.5

+ 2.8,-2. , + 2.2

7

1

(5

1

8

9

72.4

-2.5

8

11

5

y,

19

19

G 1.9

-15.4

9

o

10

8

2

22

24

57.9

+ 7.4

10

17

8

2

22

22

G 1.7

+ 1.8,-1.7

u

10

2

2

14

14

G2.G

+ 8.7

12

18

2

8

23

28

G 8. G

+ 1.2,-4.7

18

5

1

8

14

14

1 Gl. 5

+ 5.5

14

18

3

2

23

28

()G.8

+ 3.5

15

14

1

15

15

G7.3

-3.

le,

2

17

1

20

22

63.7

-2.2, + 1.6

17

10

8

18

18

64.7

+ 1.2,- 1.4, + 2.8

18

5

1

1

7

7

68.4

+ 5.4

19

...

...

8

2

5

5

70.8

-2.6,+ 1.8

20

5

1

2

8

8

67.3

-5.8

21

G

2

8

8

65.5

+ 1.8,-3.1

22

8

5

8

8

61.7

-4.2, + 2.4

28

2

1

8

()

G

62.7

+ 3.2

24

1

1

10

8

15

18

66.7

+ 8.7

25

5

1

(*)

G

70.6

+ 2.4,-1.8

26

7

2

9

9

70.4

-2.6

27

2

1

8

8

69.0

+ 1.2,-2.5

28

1

C)

7

7

69.3

+ 2.1

29

]

18

1

1

IG

17 1

70.7

+ 1.4,-2.5, + 1.6

30

10

2

8

15

15

71.2

-2.7

31

G

2

8

8

63.3

-7.8, + 3.4

Sum

8

9

204

187

G 8

41G

481

0\ THE AFTER-SHOCKS OF EIRTHQU.VKES.

i(;i

JAXUARl

', 1892.

Day

violent
shocks

strong-
shocks

weak
shocks

feeble
shocks

sounds

sum

activity

mean
bar. lit.
in mm.

fluctuation,
in mm.

1

4

1

1

6

6

766.3

+ 1.8. -2.4, + 1.6

2

8

2

5

5

65.3

-2.2, + 2.7

o

1

1

2

1

2

7

10

68.7

+ 3.2

4

1

(3

8

2

12

13

68.0

-2.6

5

2

3

1

1

7

9

66.6

-1.8, + 2.5

6

2

2

4

4

66.5

-2.7

7

4

5

9

9

68.2

-2.8, + 3.

8

1

8

4

4

64.2

-4.5

9

...

2

1

8

8

60.9

+ 1.6,-2. , + 1.8

10

4

3

3

10

10

62.2

+ 2.6

11

8

6

14

14

62.9

-4.4

1-2

1

2

2

5

6

54.1

-6.2, + 3.5

18

1

5

8

2

11

12

61.4

+ 10.

14

67.8

+ 2. ,-2.6

15

2

2

0

62.8

-5.8

1()

8

8

8

58.1

-8.8, + 1.8

17

1

1

2

2

58.7

+ 1.3,-1.5,+ 1.5

is

57.3

-2.2

11)

4

2

()

6

59.9

+ 6.9

t20

8

1

4

4

62.9

-3. , + 8.5

21

2

2

2

66.6

+ 3.8

22

8

1

4

4

70.6

+ 2.4,-2.4, + 2.2

28

1

2

8

8

71.6

+ 1.7,-8.2

24

1

2

8

Ü

6

69.6

0

25

1

1

1

3

3

68.8

+ 1.2,-8.9

2()

1

4

1

6

7

62.7

-5.4

27

2

8

2

7

7

61.3

+ 4.5

28

1

5

1

7

8

65.2

+ 2.

2'J

4

1

5

5

68.6

-8.2,+ 1.2

80

2

2

2

65.2

+ 4.4

81

4

1

5

5

64.2

-5.1

Sum

1

8

45

74

8(3

164

174

162

F. OMORI.

FEBEUARY,

1892.

Day

strong
shocks

weak

shocks

feeble
shocks

sounds

sum

activity

mean
bar. ht.
in mm.

fluctuation,
in mm.

1

763.4

+ 2.1,-1.5

2

3

1

4

4

61.5

- 3.1, + 3.4

8

7

1

8

8

64.9

+ 2.2,-2.

4

1

1

2

2

59.3

-11.2

5

1

1

3

1

5

6

60.7

+ 7.9

6

2

1

4

6

69.5

+ 3.5

7

2

2

4

4

71.8

- 3.3

8

...

1

2

3

3

56.9

-l8..5, + 5.9

9

4

4

4

60.9

+ 7.

10

...

1

1

2

2

66.4

+ 1.<S,-1.7

11

1

3

3

66.9

- ^2.],

12

2

1

3

3

62.3

- 5.4, + 1.5

13

1

1

2

2

62.5

+ 1.8,-1.9

14

4

2

6

(')

55.7

-11.7, + 4.8

15

4

1

5

5

57.5

+ 3.('.

16

o

2

4

4

62:7

+ 4.

17

1

1

2

0

6

7

64.1

- 2.8, + 4.4

18

1

1

J

64.5

- 3.4

19

4

4

4

59.4

+ 3.7

20

'2

1

3

3

62.0

2

21

5

1

6

6

60.3

- 4.2

22

1

2

1

4

5

58.5

+ 3.

23

1

2

3

3

58.5

- 7.5

24

1

3

1

5

5

54.1

- 1.5, + 0.9

25

...

3

3

3

59.5

0

26

...

1

2

3

3

56.6

- 4.5, + l.G

27

3

3

3

59.4

+ 5.1

28

8

8

8

59.4

+ 6.5

29

3

3

6

6

65.8

- 4.3

Sum

5

5

()7

37

114

119

...

ox THE AFTEll-SHOCKS OF EARTHQUAKES.

Kin

MARCH, 1892.

Div

JJ.ij

strong
shocks

weak
shocks

feeble

shocks

sounds

sum

activity

mean
bar. ht.
in mm.

fluctuation,
in mm.

]

2

3

5

760.4

- 4.9, + 5

2

2

1

3

3

63.4

- 1.4

3

3

4

7

7

60.1

- 2.5

4

3

2

5

5

58.9

- 2.3

5

58.6

- 1.7, + 2.4

G

53.0

-11.3

7

1

2

3

3

53.1

+ 8.8

8

3

3

3

60.9

+ 5.3

9

1

1

1

67.0

+ 2.9

10

1

1

1

66.0

- 4.4

11

1

1

3

3

64.2

+ 4.

12

1

1

1

64.8

- 9.

13

4

1

5

5

51.5

— 5.8, + 10.6

14

5

2

7

7

58.5

+ 4.9

15

2

2

2

63.2

- 1.7,+2.9

16

2

2

2

66.1

+ 1.5, -1.5. + 2.7

17

2

1

3

3

67.4

+ 1.3,-2.6

18

58.0

-10. ,+2.8

19

57.6

+ 4.1

20

■ i

...

62.3

+ 1.6,-1.0, + 3.1

21

!

2

2

2

65.9

+ 2.2, -2.2, + 1.5

22

. ! .

...

61.8

+ 1.3,-2.4

23

3

3

3

60.8

- 4.6

24

1

1

1

55.9

- 3.1, + 5.3

25

3

3

3

63.4

+ 7.5

26

.

1

1

1

64.9

- 5.8

27

1

2

3

3

65.2

+ 3.5

28

..

2

3

6

7

64.5

- 3.9

29

3

1

4

4

58.9

- 4.9, + 3.6

30

3

5

2

10

10

65.2

+ 6.8

3]

•

2

1

3

3

68.8

- 2.1, + 1.3

i^^iim

1

4

52

30

87

88

164

F. ÖMORI.

Day

APRIL, 1G92.

shocks

weak
shocks

feclJe
shocks

sounds

SUUl

activity

mean
Imr lit.
in mm.

fluctnatiou,
in mm.

1

2

2

2

764.4

-12.2

2

5

5

5

58.9

+ 8.2

3

1

1

2

2

57.1

-1.8, + 7.

4

2

2

2

61.8

+ 1.8,-8.1

5

4

1

5

5

59.8

+ 8.1

G

1

1

1

60.9

-8.1

7

58.9

-2.8

8

1

7

1

'J

i)

57.6

-2.5

1)

2

4

4

57.2

-1.8, + 4.1

10

1

2

8

8

64.1

+ 7.8

11

6)6.4

-8.4

V2

8

8

8

55.9

-12.4

18

2

2

2

55.6

+ 9.5

14

•)

2

2

61.8

+ 2. ,-8.()

15

2

2

2

59.2

+ 4.

U)

64.8

+ 8.9

17

8

')

8

68.7

-4.6

18

4

1

5

5

58.4

-8.2

h)

8

8

8

()1.6

+ 6.2

20

1

8

1

5

5

65.9

+ 1.8,-2.2

21

1

1

1

58.8

-7.4

22

i . . -

1

8

8

8

61.9

+ 8.9

28

2

2

2

65.6

-2.5, + 1.6

24

i 1

1

8

5

6

59.9

-6.1

25

i

1

1

1

55.8

-1.4, + 1.5

26

i

1 • • •

o

1

8

8

60.9

+ 5.6

27

1

2

8

8

64.6

+ 1.4

28

'

1

1

1

65.2

+ 1.2

29

1

5

6

6

69.2

-2.7

80

4

8

7

7

58.8

-8.8

8i;m

2

1

10

65

18

90

92

ox THE AFTER-SHOCKS OF EARTHQUAKES.

i(;5

TABLE V. -DAILY NUMBERS O? EARTHQUAKES AT GIF
FROM MAY, 1892, TO DECEMBER, 1893.

CT,

Day

1

'2

O

4

(3

7

8

9

lO

II

1-2

13

14

15

1()

17

18

19

20

'21

2 -2

•23

24

25

26

27

28

29

80

31

weak
shocks

Mat, 1892.

fefble
shocks

(')

sounds

June, 1892.

ST rang
sliocks

weak
shocks

feeble
shocks

souuds

Sum

42

54

30

1G6

F. OMOKI.

July,

1892.

Au HD ST, 1892.

Day

weak
shocks

feeble
sliocks

sounds

SUUl

strono-
sliof.ks

weak
shocks

feeble ,
shocks ^'^•""^^^

sum

1

1

1

2

4

4

2

1

1

2

1

1

3

1

1

3

1

4

4

•2

1

3

1

1

5

1

1

1

1

2

0

1

1

7

1

1

9

1

1

2

10

11

1

1

12

I

1

1;]

I

1

1

14

.)

2

3

8

15

1

1

10

. . .

1

1

17

Is

1

. . •

1

11)

•2

l-

3

1 1 ...

1

2

20

1

1

1

8

2

G

21

1

1

2

2

1

8

22

1

1

'2H

1

1

'2 \ 1

8

24

1

1

1

1

25

1

1

20

1

1

2

1

1

27

3

4

1

2

8

2<S

1

1

2

1

8

29

4

1

5

30

1

1

2

2

81

o

2

1

1

Sum

3

21

11

35

1

3

81

17

52

ox THE AFTER'SHOCKS OF EARTHQUAKES.

167

Skptembeb, 1892.

October, 1S92.

November, 1892.

Day

2 2

^1

¥ 2

1

3

II

"5 c

w "to

CO

i

fj

^ a:

? 2

_2 J

w a;

ä

5

3

CO

1

O

2

1

1

2

'2

2

... '

1

1

.)

a

o

2 \

1

8

4

4

4

4

2

1

8!

2

2 j

8

8

5

5

5

• )

8

2

2

6

5

_

0

1

1

1

1

7

1

1

17

9

28

1

1

2

8

8

•>

8
8

l

8
1

4

1

1

8

4

10

5

5

8

8

11 '

8

8

1

1

1-i

o

2

1

1

2

2

l;i

1

1

'_)

4

1

1

2

2

1

8

14

o

1

8

8,

8

15

1

1

(J

(j

ic.

o

1

8

2

2

2

2

17

1

1

1

8

18

4

4

1

1

1

1

19

2

2

2

2

...

20

1

1

2

...

1

1

2

21

2

2

...

22

:)

2

2

2

1

1

28

o

1

8

2

2

2

2

24

25

1

1

1

1

2(')

0

2

1

1

2

27

1

1

2

5

5

1

1

2

■ 28

1

1

2

2

29

8,

8

...

"80

1

1

2

! ...

81

1

1
47

Snui

1

8

79

24

107

1

2

87

7

3

1

40

5

48

IGS

F. ÖMOPJ,

December, 18'J

>

January,

1893

February, 1893.

Day

^1

M

■ji

'' 'ai

3> 5

a o

s

'è.~

-, X

X
X

s

1

1

1

4

2

G

2

8

1

1
1

1
2

8

8

4

2

2

5

8,

8,

1

1

6

7

8

8

8

1

1

1

9

4

5

1

10

i)

2

11

2

4

1

1

12

9

2

1

1

13

2

2

1

1

14

'2

2

1

1

1

1

15

1(3

17

18

1

1

1

1

19

5

()

1

1

20

o

2

21

1

1

1

2

22

1

1

2

2

28

1

1

2

1

8

24

1

'.S

1

4

1

25

2G

1

1

27

1

1

1

1

28

1

1

29

1

1

1

2

4

;-]0

1

1

2

o

81

1

1

o

2

2

Slim

o

80

7

89

1

28

7

81

1

le.

8

20

ox TUE AFTER-SHOCKS OF EARTHQUAKES.

109

March

, 18'J3.

April, 1893.

May,

1893.

June, 1893.

Day

Ol C

^ i5

a> o

3

1D^

M

i o

^ "x

■r.

3

. cr.

M

P

z

1

2

2

2

2

2

2

1

3

2

2

3

2

2

3

3

1

1

4

3

3

2

2

4

1

. 1

5

6

6

1

. 1

6

2

1

3

3

3

1

1

7

o

1

3

2

2

1

2

2

8

5

.")

1

1

y

2

2

1

10

1

2

3

1

1

1

11

1

1

2

12

2

1

• )

,

13

...

2

1

3

3

14

1

1

8

S

1

15

4

4

2

2

1

16

2

2

2

o

1

17

2

2

18

1

1

2

2

4

4

19

1

1

1

1

1

1

20

1

1

...

3

3

1

1

21

4

4

2

.)

2

1

3

22

2

2

2

O

4

1

1

23

1

1

2

1

1

24

5

5

1

1

25

3

3

2

o

1

. 1

26

1

1

27

1

1

2

2

28

...

1

1

...

1

1

29

1

1

30

...

1

1

1

1

81

1

1

Sum

1

47

4

52

54

5

59

3

25

4

32

12

. 12

170

F. ÖMORI

Jdlt,

August,

September,

October,

1893.

1893.

1893.

1893.

Day

"^ ES
^ ü
9 ^

w

o

a

CO

tu o

-)H Cß

CO

'^
O

CO

CO

Ä 13

CD O
=" CO

CO

O

CO

3

to

0) CO

« o
Ol ^

"•-* CO

«3

o

CO

3

CO

1

2

1

1

8

1

1

4

5

2

2

6

1

1

1

1

7

1

1

1

1

8

1

1

1

1

9

. . .

10

1

1

n

12

1

1

18

1

1

14

2

2

15

1

1

2

2

1

1

16

2

2

1

1

2

2

17

2

2

1

1

18

0

2

1

1

1

1

1

1

19

1

1

2

2

1

1

20

1

1

1

1

21

1

1

1

1

1

1

1

1

22

8

8

2

2

28

2

2

1

2

24

1

1

1

1

1

1

25

1

1

1

1

1

1

2(3

27

1

1

1

1

1

28

29

8

8

80

1

1

1

1

1

1

81

1

1

i)

2

Sam

17

1

18

10

3

18

18

2

20

18

1

19

ox THE AFTEK-SHOCKS OF EARTHQUAKES.

171

Day

November, 181)3.

1

December, 1893.

weak
shocks

feeble
shocks

sounds

sum

^veak
shocks

feeble
shocks

sounds

sum

1

2

8

4

1

1

0

1

1

6

1 1 ..

1

7

1

1

8

1

1

9

.

10

1

11

1-2

18

1

1

14

2

■2

15

16

1

1

17

1«

2

11)

1

1

1

1

•20

2

2

21

]

[

1

4

4

22

1

1

2

2

28

1

1

...

24

!

25

L

1

2

]

L

1

26

1

1

27

28

1

1

2U

80

1

1

31

1

1

Suui

2

11

8

16

2

12

2

16

17-2

F. ÖMORI.

TABLE VI.-DALLY WUMBEH3 OF EARTHQUAKES AT NAGOYA,
FROM OCTOBER 28th, 1891, TO APRIL 30th, 1892.

October., 1891.

X

OVEMBER, IS'Jl. D

•;CEMBHB,

1801.

Day

P o

III

2 1

œ
O

p g

M CO

U D o

^=2 'S

2 §

id

±1 ^ ^

ri ,-: o
ï o C

S

on

1

4

52

5(3 .

2

25

30

5

5

8

80

81 1

)

0

7

4

18

20 1

9

10

5

17
15

20
17

(5
2

(')
2

7

27

29

4

4

17
15

18
16

8
6

8
6

10

12

12

4

4

11

4

5

8

8

1-2

(*)

7

4

4

18

12

18 .

4

4

14

12

12

4

4

15

12

12

8

3

le,

18

18

()

(3

17

...

15

15 .

8

8

Is

I'J

9
4

9
4

1
^2

1

2

20

y

i)

«2

2

21

8

9

8

8

22

..

5

5

5

5

28

8

9

8

8

24

9

9

4

4

2.")

9

9

•2

2

2(5

5

5

2

2

27

8

8

8

8

28

4

121

126

5

7

1

1

2U

7

178

185

5

5

8

8

80

4

89

98

o

i>

8

8

81

8

7(5

79

5

5

y urn

1

18

4(34

488

1

27

888

416

3

110

118

ox THE AFTER-SHOCKS OF EARTHQUAKES.

17:5

.Iaxi-ary, \S'J2.

February,
1892.

March,
1892.

April,
1892.

Day

'o 9

*>. CO

n

li

to "œ

. to

IS

m

1) !0

=<-l OS

3

to

? 2

a o

3

00

3 'S
D c

3

3 o

V o

s

1

1

1

1

1

o

2

2

1

1

1

1

8

3

4

2

1

1

4

2

7

9

1

1

5

0

3

5

1

1

1

1

G

4

4

1

1

2

1

1

7

1

1

8

o

■2

V»

3

3

10

1

1

11

1

1

I

1

1'2

4

4

18

1

1

1

1

14

1

1

...

15

1

1

1(5

17

1

1

2

2

1

1

18

1

1

1

1

19

1

1

...

20

1

1

21

1

1

22

1

1

1

4

5

23

1

1

2

2

...

24

o

2

25

1

1

2

1

1

1

1

20

1

2

27

2

2

1

1

28

1

1

2

o

1

1

2

29

1

1

2

30

1

1

1

2

3

1

1

31

o

2

1

1

2

Slim

1

1

4

37

43

()

23

29

3

13

16

()

5

11

174

F. ÔMOrîT.

TABLE VII. -DAILY NUMBERS OF EARTHQUAKES AT TSU,
PROM OCTOBER 28th TO DECEMBER 31st, 1891.

October,

isyi.

November, 1891. 1 December,

1891.

1 >n y

.2 2
'> 'S

i|

to aï

0) o

:Ä to

1
»

? o
œ to

to ^ ,

to «H "œ

3

1

7

6

18 ..

'2

1

8

4 ..

1

1

S

()

6 ..

8

8

4

1

7

8 ]

L 1

2

5

1

4

5 ..

1

1

G

2

6

8 ..

7

...

4

1

5 ..

1

1

8

8

2

8 ]
2 ]

...
1

1
2

]()
11

7
8

7
8

2

2

I'J

2

2

18

1

1

14

2

2 ..

2

2

ir,

4

4 ..

]()

2

2

17

8

8 ..

Iw

19

'20

1

1

• . •

'21

1

2

Cj

0-2

'28

2

2

'24

, . ,

2

2

'25

2G

1

1

'27

2

2

28

1

12

58

'2

7r

5 1

1

2 ..

1

1

29

82

7

8r

) ...

4

4 ..

1

1

80
81

8

5
8

9

7

17
ic

)

1
1

1
1

Suui

1

15

99

25

14

0 8

17

69

89 8

21

24

ox THE AFTETl-SHOrrCS OF EAT^TITQUAKES.

175

TABLE VIII.- DAILY NUMBERS OF EARTHQUAKES AT KYOTO,
PROM OCTOBER 28th TO DECEMBER 31st, 1891.

l)n,v

October, ISOl.

November,

1891.

Decebiber, 1891.

strong-
shocks

weak
shocks

fee):)le
shocks

sum

weak
shocks

feel.ile
sliocks

sum

weak
shocks

feeble
shocks

sum

1

8

1

4

2

1

• • •

1

1

1

b

1

1

2

1

3

4

1

1

1

. . .

1

5

1

1

7

1

. . •

1

8

...

1

1

1
1

10

1

1

1

1

11

1

12

13

■

• • •

14

15

1

1

IC,

• . •

..

17

1

1

18

...

...

h)

20

21

1

1

2

22

...

28

L

1

24

25

...

2G

27

28

1

11

26

88

1

2

3

29

4

13

17

1

1

80

5

8

8

. . .

1

1

81

o

4

G

...

Sum

1

22

40

()i)

U)

4

20

G

4 10

17(i

F. ÔMOET.

TABLE IX. -DAILY NUMBERS OF EARTHQUAKES AT OSAKA,
FROM OCTOBER 28th TO DECEMBER 31st, 1891.

October, IS'.Jl.

XOVEMBER,

1891.

Dkcembeb,

1891.

I hvy

-=1

3

3

M^

e Ü

^

2 o

Oj o

^ 2

^ 2

IB

0; Q

'S ,s

^

IJ

"^ 2

CO

ITj 'S

■^ V)

ÎW M

^H CO

CO

st-l CO

CO

=(i. CO

1

^

o

1

1

I

1

1

4

...

1

1

1

1

7

1

1

8

( \

10

11

12

18

14

15

10

17

LS

• • •

19

'20

21

22

24

1

or,

^')

2(;

27

2S

1

O

8

22

88

29

;}

O

5

1

1

1

HO

1

4

(•)

'1 1

I

I

2

> ) 1

1

1

Snui

1

8

18

29

4()

.5

2

7

8

I

4

ox THE AFTEK-SUOCKS OE EARTHQUAKES.

177

TABLE X. MONTHLY NUMBERS OF EARTHQUAKES AT GIFU,

NAGOYA, TSU, KYOTO AND OSAKA, FROM OCTOBER,

1891, TO DECEMBER, 1933.

YeaT,

Month

(tIFLT

Nagoya

'i'su

Kyoto

Osaka

IbWl,

10

7-JO

483

140

G9

4G

11

1087

41(5

89

20

7

VI

41(5

113

24

10

4

1SÜ2.

1

1G4

43

12

5

2

o

114

29

10

5

2

3

87

IG

7

1

1

4

90

11

4

2

5

54

11

1

6

30

12

1

3

7

35

4

1

8

5-J

15

7

1

9

107

13

«2

1

1

10

47

8

4

2

11

48

12

1'2

39

14

4

2

.)

iH'ja,

1

31

5

•2

2

2

'20

5

»2

1

3

52

10

1

1

4

59

11

1

1

(5

32
12

14

1
•>

2

1

7

18

14

8
9

13

20

14

8

1

10

19

9

1

1

11

16

4

12

IG

7

1

12G

8uui

3398

1310

31G

71

178

F. ÖMORl.

TABLE XI.-TIMES OF OCCURRENCE OF EARTHQUAKES AT GIFU,
FROM OCTOBER 28th, 1 p.m., TO NOVEMBER 10th, 12 p.m., 1891.

Oct. '28th, 1891

4.5G p.m.

8.41 p.m.

0.49 a.m.

8. 5 a. 111.

1.55 p.m.

4.58

8.42

0.58

8. 9

2. 8

5.25

8.45

1. 1

3. 9i

2.11

5.30

8.50

1. 5

3.13^

2.12

5.50

8.58

l.lo

8.19

2.16

('). 4

i). 3

1.12

8.20

2.27

(111

'.). 7

I.IG

8.21

2.82

(k14

'.».17

1.17

8.23

2.41

().15

9.21

1.20

3.24

2.45

6.17

9.22

1.29

3.35

2.50

G.18

9.34

1.81

8.50

2.51

G.2G

9.36

1.84

8.59

2.55

G.27

9.40

1.88

4. 0

3. 8

6.28

9.44

1.89

4. 5

8. G

6.81

9.52

1.48

4. 7

8.10

G.34

9.58

1.44

4. 8

3.22

G.42

10.14

1.50

4.10i

3.32

6.44

10.18

1.58

4.15'

3.51

6.47

10.21

1.54

4.1s

3.52

6.48

10.21

1.5G

4.19

3.53

7. 1

10.30

1.57

4.22

3.55

7. 2

10.40

2. 7

4.23

3.58

7. 8

11. 0

2. 9

4.27

4. 1

7.13

11.10

2.11

4.29

4. G

7.14

11.14

2.19

4.82

4.18

7.18

11.82

2. 2 s

4.88

4.20

7.26

11.11

2.88

4.85

4.22

7.88

11.41)

2.84

4.8C)

4.24

7.40

11.58

2.8(5

4.88

4.25

7.50

October, 29th

2.89i-

4.40

4.41

7.51

0. 1a.m.

2.42

4.42

4.43

7.54

0.12

2.49

4.45

4.45

7.57

0.14

2.54

4.471

4.48

s. 7

0.20

2. 56

4.48^

4.51

.S.22

0.28

2.57

4.51

4.52

8.28

0.37

2.80

4.52

4.51

8.82

0.46

o. 8

4.55]

ON TUE AFTEK-ÖHOCKS OF EARTHQUAKES.

179

October. 'Jyth

s.-i.'i a.m.

1 1 .27 a.m.

I
0.54 p.m.

4.11p.m.

5. () a.m.

S.24

ll.:5:5

0.50 !

4.17

5. 7

S.2C)

\\.:m\

0..501

4.20

5. 8

8.88

11.40

1-0

4.22

5. 9

8.851^

1 1 .42

1. 2

4.28

5.12

8.87

11.45

1. 2.^

4.24

5.14

8.40

11.40

1. 4

4.48

5.18

8.41

11.48

1. 41

4.51

5.24

8.48

11.49

1. 7

4.-52

5.26

8.5()

11.49^

1.14

4..58

5.28

9. 6

11.51

1.17

5. 5

5.30

9.13

L1.5U

l.l'.i

5. 0

5.81

U.lC)

11..54

1.21

5.15

5.52

«.).2l

11.55-1

1.21-^

5.23

5.58

9.88^

11.58

1.80i

5.29

5.54

9.34.V

0. Op.m.

l.:55

5.34

0.13

i).8(i

0. 4

1.8(5

5.41

().16

9.89

0. 0

l.:38

5.45

(5.41

9.45

0. 7-1

1.44f

5.51

().48

10. 1

0. 8

1..M

5.54

7. 4

10.27

0.10

2. '.'

5.59

( . (

10.29

0.18

2.10

6. 2

7.12

10.: 57

0.14

2.25

(3.15

7.1:5

10.42

0.1 4i

2.:S5

6.21

7.14

10.48

0.1 7^-

2.4:5

6.30

7.15

10.484-

0.18

[

2.47

6.32

7.23

10.4(5

0.21

2.51

6.34

7.29

10.48

0.22

2.57

6.41

7.80

10.49

0.281

8.18

0.42

7.88

! 10.50i-

0.2.51-

8.22

6.44

7.. 50

10..59

0.20^

8.87

6.46

7.51

11. 0

0.28

8.81)

6.47

7.52

11. 5

0.29

8.44

6.50

7.5s

11. 7

0.81

8.45

6.58^

W. 1

! 11.15

0.87

8.471

6.56

s.ll^

11.20

0.40)

8.55

(5.-58

s.l:;

11.21

0.41»

8.57

7. 7

s.-i2

11.221

0.58

4.10

7.18

180

OMORI.

October, '29th

10. 7 p.m.

2.14 a.m.

4.29 a.m.

11. 2 a.m.

7.17 p.m.

10.17

2.18

4.30

11.27

7. lu

10.21

2.22

4.35

11.32

7.28

10.28

2.36

4.39

11.48

7.25è

10.80

2.42

4.41

11.49

7.2()

10.81

3. 2

4.42

11.51

7.27

10.50

8. 8

4.46

11.54

7.3U

10.55

3.101-

5. 5

0. 0 p.m.

7.3Ü

11. 0

3.17'

5. 8

0.12

7.45

11.10

3.18

5.12

0.2()

7.45^

11.15

8.2Ö

5.32

0.43

7.4(;

October, 80tb

8.29

5.39

0.58

7.47

0. 4 a.m.

3.33

5.42

1. 3

7.55

0. 5

3.34

5.47

1.11

7. 50

0. 0

8.35

5.57

1.18

8. 4

0.18

3.85^

6.10

1.15

8. <4

0.14

8.39

6.18

1.19

8.18

0.80

3.40

7.26

1.24

8.141

0.37

3.40i

7.80

1.27^^

8.15

0.41

3.4(3

7.38

1.40

8.1U

0.4(3

8.47

7.40

1.48

«.80

0.47

8.48

7.43

1.47

8.8C)

0.48

8.52

.s. 0

2.17

«.41

0.58

3.55

8. 41

2.23

.S.42

0.5(3

3.57

8.10

3. 6

8.4'.»

1.10^

4. 0

8.18

8.50^,

S..5U

1.14

4. 1

S.14

4. Hi

'J. 5

1.17

4. 2

S.15

4.19

9.85

Lis

4. 5

.S.21

4.85

9.40

1.22

4. 7

8.30

4.8C.

9.45

1.82

4. 8

«..-)()

4.47

9.4C)

1.37

4.8^

9. 6

4..5S

9.47

1.89

4. 9

9.85

5. S

9.5(J

1.40

4.19

9.49

5.14

9.57

1.49

4.24

'.).57

5.19

•.K59

1.50

4.25

la 12

5.28

10. 1

1 .54

4.27

10.14

6. 4

10. 4

2. G

4.27i-

10.49

(').2()

ON THE AFTET^-SnorKS OF EARTHQUAKES.

ISl

(k-tolier, aOth

2.40 a.m.;

6.:U p.m.

2.4t)

G.35

2.48

().40

2.50

7. 0

2.51

7. 2 •-^••■■^«'

7. 5 ^--

7. 7 1 •-^•^'••'

7.25 i •^'- ^>

7.27

3.10

7.:U

3.22

7.12

3.25

IM

3.40)

7.49

4. 4

7.5:5

4. 8

7 . 5 (

4.10

s. I

4.14

8. 2

4.17

S. 7

4.19

S.12

4.31

8.14

4.37

8.22

4.39

8.28

4.43

8.30

4.49

8.47

4.55

8.50

5. 4

8.5:^,

5. 9

'.). 0

5.15

'.».17

5.17

'.».22

5.21

'.».25

5.25

9.30

5.29

10.30

5.30

11.47

5.34

October, 3 1st

5.38

0.20 a.m

5.40

2.20

5.49

2 38

5.58

('). 8 a.m.

(;.14

(',.28
().35
().48
().57
().59
I . I

7.1;i

7.22
7.24

7.40
7.47
8.18
8.2(;
8.38
8.51
8.5: i
8.55
8.57^
9.11
9.27
9.:i4
9.:^.9
10.23
1().:'.0
10.; ',5
11.14
11.19
11.27i
0. 5 p. m

0. 8
0.30

1. 2^
1.35
1.48
^.22
2.32

2.53| p-m-

3. r

;i. 8

3.29

3.42

3.51

4. 1

4.12

4.40^.

4.51

5.12

5.45

C). 5

G.12è

G. 14

G.IC)

Ü.22

Ü.24

6.57

7. 8

7.12

7.lr)

7.17

7.21

7.27

7.29

7.32

7.34

7.35

7.55

8.11

8.14

8.15

S.24

8.35

8.4(;

9.14

9.o5

10.1 G p.m.
10.29
10.31
10.:iG
10.4,s
10.59
11. G
11.43
11.52
Nov., 1st

0. 7 a.m.

0. 8

o.l;;

0.18
0.19
0.42
0.55

1.29

1 .30

1.34

1.35

1.44

1.45

1 .57

2.1 G

2.27

2.34

2.42

2.44

2.47

2.50

2.54

2.57

3.22

4.33

4.40

]Si>

F. OMOTJT.

Nov. 1st

3.10 p.m.

0.33 a.m.

10.45 a.m.

8.19 p.m.

4.50 a.m.

3.47

0.37

10.53

8.44

5. 7

3.51

0.40

11.19

8.52

5.10

4. ()

0.45

11.46

9. 5

5.17

4.55

0.47

11.. 50

9.26

5.19

5.28

0.55

0.11} p.m.

9.35

0. ()

5.38

1.15

0.24

9.. 52

G. 29

5.55

2.12

0.47

10.14

('..40

5.58

2.25

0.51

10.17

0.43

6.13

2.36

1. 7

10.37

6.44

6.28

2.58

1.20

10.42

0.47^

6.31

3. 7

1 .20}

11.43

6.52

6.45

3.15

1.37'

11.57

7.12-i

6.49

3.26

2. 9

Nov.. 3rd

7.21'

('..55

3.29

2.23

0.12 a.m.

7.30

().59

3.47

2.32

0.;M

7.41

7.10

3.50

2.34

1.47

S.31

7.16

3.52

2.:W

1.52

>S.46

7.45

4. 2

2.45

2. 7

9.17

7.50

4. 5

:;. 1

:î. 0

9.;;7

7.57.}

4.10

3. .5

10. 1

S.20

4.29

4.24

3.13

10.15

8.35

4.29}

4.50

4. 3

10.21

8.38

4.. 54

4.58

4.15

10.: IS

9.18

5.46

5.40

4.28

10.13

9.26

5.48

5.50

4.47

lo.ls

9.27

(). 0

6.29

4.55

1 1 .42

9.30

().43

6.36

4.58

11.45

9.33

7.11

7. 7

5. 2

] 1 .46

9.;!5

7.21

7.12

5.32

11. .56

9.3.S

7.32

7.16

5.56

0.46 p.m.

9.5S

7.33

7.19

6.48}

1. 5

10.20

8.53

7.29

6.53

1.1-1

10.21

9.17

7.45

6.56

1.47

Nov., 2ii(l

9.30

7.4,s

7. ()

2.4:;

0. 2 a.m.

9.46

7.50

7. 7

2.55

0.16

9.55

7.52

7.12

3. 0

0.21

10.20

8. S

7.37

ON THE AFTER-SHOCKS OF EARTHQUAKES.

isr,

Nov., 3i-a

7. 3 p.m.

5. 0 a.m.

5.31 p.m.

6.24 a.m.

7.41 a.m. '

7.28

5.29

6.19

6.33

8. 4

7.29

5.34

6.58

7.10

8.20

7.38

5.54

7. 0

7.28

8.32

7.50

6. 8

7.10

7.48

8.53

7.55

6.26

7.16

8. 9

8.55

8.12

6.35

7.19

8.46

9.15

8.15^

6.36

7.45

9.16

9.45

8.40

6.47

8.30

9.18

10.20

8.49

7. 4

8.40

9.58

10.21

8..52

7.47

8.51

10.24

10.32

9.23

8.15

9.24

11.10

11. 2

9.34

8.17

9.35

11.32

11.11

9.35

9.20

9.50

0.15 p.m.

11.30

9.47

9.22

9.-54

1.48

11.41

10.14

9.38

10.17^

2.51

0.38 p.m.

10.31

9.55

10.39

3.35

0.44

10.36

9.56

10.41

3.59

0.58

11.43

10.27

10.59

5.10

1. 0

11.50

10.44

11. 0

5.13

1. 8

Nov., 4tli

10.48

11.10

5.28

1.10

0.10 a.m.

10.57^

11.17

5.38

2.13^

0.32

11. H

11.32

6.48

2.5ft

0.52

11.38e-

Nov., 5tli

6.52

3.44

0.57

11.51

0.22 a.m.

6.54

3.. 54

1. 3

0.34 p.m.

1. L

7.39

3.58

1.18

0.35

1.14

7.50

4. 7

1.27

0.37

1..52

8. 8

4.47

1.40

1. 0

2. 3

8.16

4.49

1.57

1.17

2.22

8.23

4.50

1.58

1.49

3. 2

8.36

5. 7

2.12

2. 7

3.22

9.10

5.28

2.25

2.19

3.48

9.19

5.34

2.39

2.20

4. 4

9.25

5.42

3. 0

3.15

5. 0

9.28

6.19

3. 1

3.34

5.15

10.24

6.3G

4. 5

4. 0

5.22

10.58

6.45

4.20

5. 2

5.24

11.35

1.S4

F. ÖMORI.

Nov.. Atli

1.14 p.m.

1.45 a.m.

11.28 p.m.

10.20 p.m.

11.37 p.m.

1.25

3.29

11.30

10.58

Nov., 6th

2.16

4.21

Nov., 8th

11.38

0.10 a.m.

2.17

4.55

0.22 a.m.

1 1.39

0.12

2.39

5.52

0.53

1 1.45

1. 4

2.45

6). 7

1. 5

11.48

1.17

3.42

6. 9

' 1. 7

11.56

2. 7

:i.50

6.i7

•2.:V2

Nov., 9th

2.47

3.55

('».57

1.15

' 1.18 a.m.

2.49

5. 7

( ..■)5

4.35

1 .27

3.17

5.35

7.47

5.50

1.40

3.26

5.47

8.12

5.55

1.46

3.30

5.57

S.26

6.39

2.47

4. 4

(i. 7

8.57

7.26

3.25

4. 8

6.42

9.27.1

7.48

3.36)

4.20

7. 2

9.49

7.50

4. 0

4.25

7. 6

11.41

8. 3

4.14

4.32

7.11

0.14 ]).m.

8.24

4.20

4.39

7.2;!

2. 0

9.12

4.22

4.56

7.51

2.30

10. 7

1.32rV

5. 9

s. s

2.36

10.41

4.43

5.32

8.10

2.48

10.44

5.25

5.52

S. 16

.{.49

10.58

5.27

6.18

9 25

4.34

11. 8i

6. 8

6.. 57

10. 10

5.12

11.17

6.20

7. 7

10.30

5.23

11.27^

6.23

9.42

10.50

5.47

11.451

6.37

9.58

11. 2

6. (')

2.44 p.m.

(').47

10.24

11. 5

7. 8

4. 6

7.24

11. 2

11.32

8. 5

4.15

8. 0

11.15

11.35

8.10

5. 1

9.30^

11.22

11.41

8.43

().33

10.59

0. 9 p.m.

Nov., 7th

9. 1

6.36^

11.48

0.25

0. 9 a.m.

9.1;;

7. 2

0.34 p.m.

0.36

0.25

9.40

8.10

2.13

0.39

0.36

9.58

8.14

2.52

0.44

0.48

10.52

8.40

3. 7

0.52

0.58

11. 7

9.44

3.39^

ox THE AFTER-SHOCKS OF EARTHQUAKES.

185

Nov., 9th

1 0.421- p.m.

3.35 a.m.

9.39 a.m.

7. 8 p.m.

4. 4^ p.m.

11. o"

5. 0

10.42

7.12

4.40

U. 4

5.39

11. 7

7.21

.5. 8

Nov., loth

(3. 8

11.381

8. 1

5.33

0. 0 a.m.

6.23

0.4Ü-1 p.m.

9.13

6. 3

1.24

0.34

3.55

9.59

7.25

1.45

7.32

5.37^

10.10

'..). (■)

1.57

7.45

5.38i

10.43

'.». 5

2.48

7.58

5.40

11.10

W.41I

2.57

S.51

5.51

11.30

10.22

2.58

9.31

6.47

11.33

186

F. OMORI.

TABLE XII.— TIMES OF OCCURRENCE OF EARTHQUAKES AT N AGOYA,
FROM OCTOBER 23th, 1 p.m., TO NOVEMBER 10th, 12 p.m., 1891.

Oct. 28th, 1891

4.24.30 p.m.

(S. 7.40 p.m.

11.13. 0 p.m.

3.30.50 a.m.

1.9.45p.m.

4.24.52

8. 9.35

11.18.31

3.34.30

1.16.5

4.28.10

8.10.40

11.21.26

3.39. 0

1.22.45

4.33.59

8.23.24

11.40.30

3.40.50

1.43.14

4.44.32

8.32. 0

11.44.11

3.41.30

1.54.13

4.51.40

8.32.36

October. 29th,

3.46.55

1.. 54.32

4.56. 8

8.36. 6

0. 2. 0 a.m.

3.55.35

1.55.00

4.57.10

.s.41.49

0.19.45

3.57.28

2. 2.24

5. 0.30

8.46.41

0.22.30

4. 2.10

2. 2.50

5. 4.40

8.52.36

0.34.11

4.13.40

2.10.31

5.15. 2

8.53. 4

0.38.39

4.21. 0

2.19.33

5.17. 1

8.55.15

0.46.29

4.21.40

2.21. 8

5.31.25

8.57.49

0.58.26

4.28.30

2.22.30

5.33. 8

9. 3. 0

1. 1.43

4.35. 0

2.25.46

5.39.55

9. 4. S

1. 7.38

4.36.30

2.31.40

5.42.15

9. 8.12

1.16. 0

4.39.30

2.35.18

5.45.40

9.21.12

1.20.20

4.44.15

2.29.44

5.51.59

9.24. 9

1 .30.29

4.. 50. 30

2.41.37

6. 4.10

9.33.3()

1.49.33

4.55. 0

2.48.56

(3.14.35

9.41. 3

1..53. 4

5. 3.30

2.50. 2

6.16.39

9.46.10

1.57. 0

5. 7.30

2.56.13

6.19. 5

9.44.13

2.1s. s

5.12.30

3. 2. 5

6.26.25

9.52.40

2.24.25

5.23.50

3.21. 0

6.27.50

lO. 7.59

2.29.26

5.28. 0

3.22. 5

6.38.27

10.11.33

2.33.46

5.50. 0

3.23.35

6.47.53

10.17.45

2.35.10

5.53.30

3.24.54

7.12.45

10.30. 0

2.38.21

6. 4.30

3.31.48

7.13.40

10.30.34

2.46.58

0.13.30

3.49.55

7.16.40

10.39.14

2.48.24

0.13.59

3.51.23

7.17.45

10.44.12

3. 0.59

0.43.25

3.52.14

7.19.25

10.47.21

3. 2.-57

0.45.20

3.53.12

7.26.30

10..53. 2

3. 7. 0

7. 0. 0

3.55.22

7.29.31

11. 2.12

3. 4.20

7. 3.30

4.13.15

7.34. 5

11. 3.36

3.L1.40

7.12. 0

4.15. H

7.41.14

11. 7.20

3.19.10

7.20.35

4.21.25

7.42. 7

11. \).:\2

: 1.20.55

7.:i(\45

4.22.10

7.46.30

11.11.21

3.28.50

, 7.40. 0

UX THE .AFTER-SHOCKS OF EARTHQUAKES.

187

7.52.35 a.m.

0.59.22 p.m.

7. 7.8 0p.m.

0.40. 4 a.m.

11.31.25a.m.

7. 50. 10

1.19.89

7.30.20

0.47.12

11.50.25

7.-58.32

1.24.12

7.39.10

0.49.29

0.29.32

8. 4.45

1.84. 4

7.45.20

0.51.20

0.48.27 p.m

8.11.28

1.34.22

7.46.10

1. 7.-32

1. 2.30

8.14.30

2. 8.1^3

7.55.24

1. 9.50

1. 4.19

8.26.10

2. 8.. 50

8.18.21

1.13.31

1.47.21

8.27.15

2.10.15

8 33.22

J. 1.5.. 50

2.24.12

8.33.37

2.42.30

s. 8 1.50

1.26.15

3. 6. 5

8.36. 7

2.53. 0

8.41.15

L.28. 0

8.17.28

8.37.53

2.59.40

8.48. 0

1.30. 0

3.59.25

8.41. 1

8. 3.50

8.51.37

1.30.22

4.28.35

8..50.57

8. 5.32

8.58.2()

1.89.45

4.39. 5

'.). 7.20

8.14.44

9. 1.45

1.51.25

5. 8.29

y.l5..56

3.16. 5

9.34.36

1..54.10

5. 9.19

Ü.21.25

3.22.25

9.43.38

1.. 56.21

5.10.10

•1.24.18

3.33.10

9.51.49

2. 6.20

5.24.15

'.).40. 5

8.35. 0

10.29.25

2.27.1s

6. 1. 6

'.».44. 0

8.44.25

10.44.30

2.86.15

6. 7.32

<». 54.40

8.45. 0

10.48.87

3.35.15

().27.12

10.20.50

4.19..')0

10.53.25

8.40. 0

6.8C). 2

10.37.40

4.22.20

11. 2. 0

4.16.45

6.42.82

10.44.20

4.87.15

11. 9.1(i

4.35. 0

7. 7.22

lU.46.8l

4.48.40

11.14..50

4.53.42

7.17.17

10.49.42

4.49.20

11.32.29

5. 7. 5

7.49.22

11. 7.20

5. 8.10

11.44.19

5.42.17

7.58.10

11.12.80

5.25.50

11.46. 0

5.-52.22

8.49.45

11.22.86

5.2.S.10

OctobeL- 80tb

5.57. 2

9. 6.-52

11.27.40

5.80. 0

0. 4.l2a.m

(').l0. 0

9.25.51

11.34.45

5.43.45

0. 5. 0

6.17.81

9.36.32

11.51.50

5.47.25

0. 6.51

7.19.20

10. 4. 7

0. 6.25 p. m

5.57.27

0.12.35

7.-54.2()

10.10. 7

0.13.40

(•). 4. 0

0.14.40

8.17.20

10.21.82

0.17.15

(').82.12

0.16.10

S. 17.20

10.28.22

0.46.87

6.89. 0

0.29. 0

8.30.22

10.27.35

0.47.. 50

().45.80

0.29.40

9. 5. 5

10.30.50

0.51. 0

6.51.20

0.32. 0

10. 5.27

10.37.-52

0.56.40

7. 4.28

0.37.15

LI. 2.40

10.58.35

18S

F. OMORI,

1 0.54.54 i).iii.

9. 2. 5 a.m.

11.44.15 p.m.

9.41. 15 a.m.

3.20. 4p.ui.

11.10. 0

10.33.25

Nov., 1st

10.41. 2

4.58.31

11.10.30

10.44.10

0. 8.27 a.m.

10.52.20

5.30.30

11.10.21

11. 0.36

0.47. 0

11.12.55

8. 9.45

11.28. 5

0.25. Op.m.

0.48.46

11.33.31

8.14.25

11. .52.38

0.34.10

0.50.40

0.24..50p.m.

8.19.50

11.53.10

0.37. 0

1.18.55

3. 0.22

9.39.10

11.55.10

0.44..39

1 .28.40

4.34.45

9.46.55

October, 31 st

1. 4.45

1.36.34

5.18.21

9.18. 5

0. 1.30 a.m.

1.25. 0

1.44.32

5.57.51

10.18.40

0. 6.31

1.56.15

1.46.10

6.30. 8

10.42. 5

0.15.13

2.17.35

1.48.44

7..56.27

11.25. 0

0.29. 7

2.25.41

1.. 50.44

8.19.52

11.55.30

0.34.28

2.37.25

1.52.56

8.27.10

Nov., 3rd

0.50. 0

2.48. 6

2.13. 0

8.39.3S

0. 0.1 5 a.m.

0.50.50

3. 5.10

2.28.20

9.11.55

0. 2. 5

0.58.12

3.29.59

2.35.13

9.19.21

0.35. 0

0..59.24

3.51.35

2.41.52

9.28.45

0.51.45

1.11. 5

4.18. 8

3. 7.36

9.37.22

0.55. 0

2. 5.52

4.39.20

3.20.21

10.51.40

1.17.35

2.27.32

6.13.2.S

3.46.17

Nov., 2iid

1.27.30

2.48.50

(5.29.25

3.47. 5H

0. 7.50 a.m.

1.25.10

2.5<s. 0

7.1s. 0

3.52. 0

0.24.20

3. 5.35

3.10.14

7.37. 5

4.24.10

0.34.36

3.24.10

4. 0.55

7.55.52

4.27.15

1.53.10

5.33. 0

4. '.». 5

8.34.25

4.37.40

1.56.45

5.43.42

4.21.30

8.4S.20

4.44.25

2. .s. 0

.s. 16.25

5.10.41

9.12.30

1.51.31

2.27.45

8.22.49

7. 0. 0

9.27.50

5. 9.21

3. 6.15

8.32.56

7.15.25

9.43.14

5.15. 0

4.14.32

8.3(5.40

7.23.30

9.49.55

5.19.10

5. 0.30

9.16.59

7.49.30

10. 0.20

5.37.25

(5.12.59

9.32.34

7.51.27

10. 8.50

6. 2. 0

9.50.40

9.46. 8

8. 4. 2

10.28. 0

7. 7.15

10.52. 0

10. 4.45

8.22.14

10.49. 5

7.12.46

11.52.34

10.23. 2

8.27.58

11.20.35

8. 3.10

1.. 58.23 p.m.

11. 0.29

8.56.30

11.24.25

.S.30.15

2.12.25

11.43.10

8.58.31

11.43. 5

9. 4.40

3. 4.58

11.43.37

ox THB: AFTER-SHOr'KS OF EARTHQUAKES.

IX!)

0. 1.25 p.m.

Nov., 5tli.

11.19.41 a.m.

3. 0. 0 p.m.

4.20.10 a.m.

1.22.30

3. 1.30 a.m.

1.25.15 p.m.

7.39.59

4.34. 2

4.50.40

3.54. 0

3.31.21

9. 3. 4

5. 7.12

4.51.42

4.45.50

7.43.30

11.14.30

5.13.40

7.29.20

ß.33.15

7.53. 0

Nov., 8th

G.25. 0

8.51.20

G. 34. 5

8.35.15

0.54.40 a.D3.

7.17.25

9..54.20

7. 7.35

8.39. 0

1. 0. 5

8.20. 2

Nov., 4tli

7.49. 0

9.23.40

1 .24.25

0.38.14 p.m.

0.18.29 a..in.

9.58.40

10.11.15

1.52.40

2. I.IO

0.45.25

10.24.25

10.32.25

3. 8.45

7.32.10

1.20.34

11.57.12

10.41.. 52

:;.24.4:l

7.52.10

1.27.45

0.44.52 p.m.

NoV., 7 ill

5.2;;. 12

8.55. 0

1 .33.4(*)

3.35.1:1

0. 1.20 a.m.

9. 2.59

Nov., 10th

2. 0. 0

5. 9.5t')

0. 9.21

11.10.5,S

2.44.30 a.m.

4. 2.10

5.42.45

0.10. 0

1 .2«. 5 p.m.

5. 2.50

4.12.21

7.10.21

0.12. 9

1.30. 0

G. 3.45

4.21.50

9.29.5()

1. 1. 5

1.47.15

G. 24.27

0.37.45

10.1().15

1.4().12

G.37.50

8.34. U

8.19.15

10.41.18

2.45.30

9.34.20

9.-32.35

11.59.30

10.45.38

3.1G.40

10.38.50

11.58. 7

2.22.31 p.ni

11.43. 5

4.18.20

11.3(").50

5.10.24 p.m.

3.15.18

Nov., Gtli

4.18.50

11.4G.39

5.41.34

3.5().45

0.50. Oa.m.

' 5.47.55

11..-)8.15

G.14.19

4.18. 8

1. Ü.I5

5.55. 0

Nov., 9th

7.27. 7

4.39.20

1.19. 7

G.38..50

0.3().25 a.m

s. 4. 0

4.44.35

4.25.40

9. Is. 15

2.48.47

7. 1.20

7.26.45

1. 9.22 p.m

3.25. 0

8.22. 5

9.59.45

2.48. 5S

3.37.20

IDO

F. OMORT.

TABLE XIII.— HOURLY NUMBERS OF EARTHQUAKES AT GIFU,
FROM OCTOBER 28th TO NOVEMBER 16th, 1891.

^\^ Day

October, 1891.

XoVF.JfBER, 1891.

\.^^

sum

Interval ^-.^^^

28

•29

30

31

1

2

3

J.

5

ß

7

B

9

10

0 — 1 a.m.

9

13

1

7

9

2

4

1

2

5

2

1

56

1- 'J

19

12

9

1

2

6

3

2

1

2

4

3

()4

•i— .s

15

6

10

9

4

1

3

2

3

1

1

3

58

;{— 4

13

20

5

1

7

3

2

3

3

1

2

1

62

4— 5

•25

20

12

3

6

6

2

1

7

2

2

6

92

:■")— (')

15

8

13

4

2

4

4

3

1

2

i)

2

().•;

(')— 7

4

'2

7

7

Q

o

5

2

2

4

1

5

3

47

7— S

14

5

0

4

4

5

2

3

1

2

3

3

53

s— '.)

14

9

7

2

1

5

2

2

3

2

1

49

i)— 1()

9

4

4

2

4

2

5

3

2

2

1

2

41

10—11

12

3

3

6

3

3

4

1

1

4

1

42

11— 1-i

...

22

7

3

4

3

4

3

2

1

3

2

58

n — J p.m.

2ß

5

3

1

4

3

3

1

6

1

54

1 - li

17

10

3

3

4

3

3

1

2

1

1

48

•J— :'.

11

8

2

3

2

6

2

3

1

4

4

L

2

...

3,8

;{— 4

10

Ç)

'2

5

4

2

3

2

2

3

1

2

1

36

4— .-)

16

11

(*)

4

2

3,

4

1

1

2

2

36

.•)— (■)

3

11

4

2

4

2

4

2

4

4

3

1

2

4

47

(■)— 7

15

15

5

7

7

2

3

2

3

2

1

2

1

1

51

7— s

13

16

12

11

5

9

6

5

2

5

1

1

1

3

77

s-^ 1)

9

12

11

ß

3

4

5

3

4

•1

3

>>

1

58

'J— lo

11

9

5

2

H

4

4

4

4

1

4

1

3

2

51

10—11

6

10

1

6

2

4

3

4

2

3

1

2

2

2

42

11— I'i

7
101

3
318

1
173

3
126

99

2
92

2
81

4

78

2

53

5
67

3
45

5
42

2

43

3

40

35

Sum

1258

Note. h\ foniniij^ tlip hourly " sdinf?," the record for Oetolier 'iSth lias not l)een taken into
account.

( Dnrino- Nio-ht,
i „ Day,

or lietween 0 p.m. and (! a.ui.. 709
„ ,, G a.m. and (i p. in., o49

I2ÔS

ox THE AFTETl-SIIOCKS OF EARTHQUAKES.

191

TABLE XIV.-HOURLY NUMBERS OF EARTHQUAKES OF NAGOYA,
FROM OCTOBER 28th TO NOVEMBER 10th, 1891.

^-\ Day

October, 1881

XOVEMVER, 1891

^\^^

sum

Interval ^^^^

28

29

7

30

14

31

9

1
4

2

3

3

5

1
2

5

G
1

7
4

8
1

9
1

10

0— 1 a.m.

51

1— 2

8

12

1

8

2

;^.

3

2

2

•>

44

2— 3

8

3

4

4

2

1

1

25

3— 4

IC)

2

1

5

2

2

2

34

4— 5

11

3

3

r,

3

1

1

32

5— G

7

4

1

4

Q

1

1

25

fi— 7

5

0

1

1

2

2

IG

7— 8

9

o

5

2

2

1

22

8— 9

10

2

5

2

4

1

1

2G

9—10

7

1

1

2

'1

1
j

1

1

1

20

10—11

5

1

2

2

2

1

14

11—12

6

3

1

2

3

1

1

1

1

1

21

0— 1p.m.

8

2

4

1

1

1

18

1— 2

7

4

3

3

1

1

3,

17

2-3

14

G

1

4

1

15

3— 4

11

9

3

3

1

2

1

1

23

4— 5

12

5

2

2

1

2

o
t )

IG

5— ß

10

7

4

2

2

2

18

G— 7

8

5

5

2

1

1

1

15

7— 8

11

7

4

3

1

1

1

1

2

2

1

24

8 9

13

7

1

2

3

3

1

1

2

1

22

9—10

10

4

3

4

4

3

1

1

1

1

23

10—11

9

4

9

4

1

2

3

3

1

27

11—12

10
115

G
171

7
93

4
G8

5G

2

30

31

20

1

20

17

20

3
18

IG

12

24

Sum

i
572

Xote.— In forminii,- tlie lionrly " suiii.s," the rocord for Octobor 2Sth has not I^oen taken into
account.

j During Xij^lit, or Ijetween 6 p.m. and 6 a.m., 34G
[ „ Day, „ .. 0 a.m. „ 6 p.m., 226

572

1. The times of earthquake occurrence were lost in a few cases, and therefore this num-
l>er is slig-htly less than aceordin«- to Table VI.

192

F. OMOEl.

TABLE XV.-DAILY NUMBERS OF EARTHQUAKES AT CHIRAN,
FROM SEPTEMBER 7th, 1893, TO JANUARY 31st, 1894. i

Septembek,

1893.

October, 1893,

Day

gl
to "m

CO

II

c«

be "^
II

C O

^ 0 CO

r-) «i

0;

3

>3

>

S

1

2

2

2

2

2'

2

2

3

0

2

2

4

0

2

2

5

6

7

T

T

T

8

4

7

34

45

49

9

5

12

27

44

49

10

1

4

23

28

29

11

9

11

23

23

12

3

15

18

18

...

1

i'

ï

13

ï

1

14

16

17

14

3

<2

16

21

24

15

4 8

12

12

ï

1

2

3*

IG

8

8

8

ï

1

1

17

l'

'ï

3

5

6

1

1

1

18

2

7

7

19

5

5

...

20

3 3

6

6

21

4

4

4

...

22

3

3

3

...

23

1

3

4

4

24

1

1

2

2

ï

1'

'2"

2

25

1

2

3

3

ï

]

1

3

4

2G

1

3

4

4

1

7

8

8

27

2

2

2

28

ï

2

3

3

29

o

ï

3

3

1

1

2

2

30

2

2

2

31

Slim

15

54

194

263

278

2

9

23

34

36

1. The Kagoshiuia earthquake took place on SeptemJjer 7th at 2.46 a.m., and the record of
after shocks was not taken till about 9 p.m. of the same day. The figures in this Table are
the numbers of earthquakes during successive twenty-four hours between 9 p.m. of each day
and 9 p.m. of the next.

ox THE AFTEE-SHOCKS OF EAliTHQUAKES.

193

November, 1893.

Decembek, 1893.

January, 1894.

Day

0) o

O M

M

0) O

O œ

D o g
:k o o

02

M

o

ce

be ■«

||
CO "m

ci Ü

O Ü =
CO

3

1

1

1

1

2

3

4

1

2

8

4

5

1

6

7

7

6

1

1

1

3

8

3

7

1

1

1

S

2

2

2

1

1

1

9

1

b

4

4

10

1

1

1

n

2

2

2

12

1

1

1

18

1 1

2

2

2

2

1

14

2

2

2

...

5

5

5

15

• • •

...

2

2

2

16

- . .

1

1

1

17

1

1

1

18

2

a'

4

4

ID

...

8

8

8

'20

21

• • •

1

1

1

22

2

2

1

1

I

28

1

1

1

24

• • •

.

1

1

1

25

4

5

5

1

I

1

26

2

2

2

4

4

4

27

...

...

28

2

2

2

1

1

1

29

2

2

2

80

2

2

2

2

4

6

()

81

2

2

2

yum

7

7

14

14

10

4

14

14

1

6

47

54

55

194

¥. OMOKI.

TABLE XVI. -HOURLY IsTUMBERS OF EARTHQUAKES AT CHIRAN,
FROM 8th TO 21st, SEPTEMBER, 1893.

Day

September. 1893.

sum

luterval

8
1

9

10

11

12

13

11
1

ir.

16

17

18

19
1

20

21

0— 1a.m.

3

1— 2

5

J

1

2

1

J

4

1 ...

15

2 — o

5

2

2

.)

1

12

3— 4

1

1

1

1

1

1

2

1

1 '-^

4— 5

1

1

o

1

2

J

1

9

5— Ü

b

3

4

1

2

2

1

1

17

(3— 7

1

1

2

1

1

1

1

1

9

7— 8

4

2

2

1

2

1

12

8— Ü

2

3

1

1

1

1

]

10

Ij— 10

4

7

1

o

1

15

10—11

]

]

1

2

1

0

11—12

3

3

2

1

1

1

11

0 — 1 13. m.

1

1

1

1

4

1— 2

1

2

1

1

5

1

1

12

2— 3

o

O

1

2

1

2

1

1

13

3— 4

1

o

2

1

]

8

4— 5

3

4

1

1

1

10

5— 0

1

1

1

1

4

G— 7

3

2

2

1

1

]

10

7— 8

3

2

o

]

1

1

11

8— Ü

]

2

3

9—10

2

1

1

2

1

1

8

10— IL

3

1

]

1

1

4

1

...

12

11—12

42

41

4

30

1
21

3

20

1
14

1
25

6

7

(')

0

.")

(3

4

10

Sum

233

[During Night, or between 6 p.m. and G a.m., 1 19
1^ „ Day, „ „ G a.m. ,, G p.m., 1 14

Zà'à

ox THE AFTEK-SHOCKS OF EARTHQUAKES.

195

TABLE XVII. -MONTHLY AND YEARLY NUMBER OE EARTHQUAKES

INSTRUMENT ALL Y RECORDED IN TOKIO FROM JANUARY,

1876, TO DECEMBER, 1893.

-^lonth
Year ^^

I

II

III

IV

Y

VI

VII

VIII

IX

X

XI

xir

SVILU

1876

3

4

6

11

5

3

3

5

3

3

4

6

56

1877

4

5

6

5

8

9

G

4

1

8

6

9

71

1878

3

8

7

2

5

4

4

1

2

4

0

4

50

1879

6

7

14

9

4

3

4

1

7

6

9

70

1880 ■

9

9

6

(j

2

9

8

4

1

3

10

10

77

1881

13

8

8

8

4

3

3

3

2

3

3

8

m

1882

4

7

15

6

3

2

2

1

1

4

1

46

1883

6

3

3

6

2

3

1

1

3

4

32

1«,S4

5

o

8

2

9

4

1

4

2

8

8

15

68

1885

7

9

8

4

3

6

3

8

10

3

7

68

1886

3

o

3

2

8

4

2

8

7

^

2

8

54

1887

10

4

3

8

13

5

ö

2

10

5

14

80

1888

4

15

7

(

11

9

9

7

11.

4

13

4

loi

1889

5

16

11

18

13

7

5

8

7

8

9

6

113

1890

5

5

6

15

14

0

12

7

4

8

10

2

93

1891

1

4

(J

7

10

7

8

4

4

45

12

15

123

189-2

9

9

2

7

8

(5

7

2

4

10

4

5

73

1898

6

4

3

7

11

9

4

3

5

3

133

3

1

59

Sum

103

119

122

118

142

7.9

98

86

71

73

108

127

1300

Average

5.8

6.6

6.8

6.6

5.4

4.8

4.0

4.0

7.4

6.0

7.0

72.2

Xote. — The greatest monthly earthquake number, namely 15, oecurrod in October, 1801. Of
this however, 2S took place within the last four days of the month, being- due to the residual
effect of the great Mino-Owari Earthquake. We may take (15 — 28) Xl^'^lO as the proper
nimber of shocks for October, 18'JL. Making this modifie ition, the total number for 1S91
becomes 97, and the average monthh' number for October becomes (3. These latter values
have been used in drawing curves, Figs. IS, (4) and 20.

196

F. OMOiU.

TABLE XVIII. HOURLY DISTRIBUTION OP 1168 EARTHQUAKES

RECORDED INSTRUMENTALLY IN TOKYO FROM

1876 TO 189L

-„^Month

I

II

III

IV

V

VI

VII

VIII

IX

X

XI

XII

sum

Interval

a.m.

0- 1

4

6

2

3

3

3

3

2

4

3

7

7

47

1— 2

1

1

3

1

4

3

1

1

3

3

4

3

28

2— 3

1

7

3

5

5

o

5

3

4

4

8

2

49

3— 4

1

4

2

2

4

3

1

G

3

4

33

4— 5

o

2

7

0

4

o

O

4

2

3

3

7

42

5— 6

2

4

4

7

4

2

3

G

3

2

9

46

6— 7

3

6

8

2

5

2

2

2

7

2

7

46

7— 8

5

4

4

4

4

4

2

4

1

12

7

52

8— 9

3

4

2

4

11

6

1

G

2

3

2

44

9—10

4

5

4

3

11

G

4

5

1

7

8

G

64

10—11

5

4

2

4

(3

5

.)

3

4

4

4

2

45

11—12

4

3

4

3

4

3

2

3

1

4

1

G

38

p.m.

0— 1

2

2

o

4

!)

8

4

2

2

1

(j

7

49

1— 2

3

5

5

4

4

4

3

4

3

7

9

4

55

2— 3

b

9

G

G

3

4

C)

4

1

4

4

50

3— 4

7

12

G

8

G

3

G

2

4

4

9

0

67

4— .5

3

1

11

8

3

4

5

3

3

4

5

4

54

5— 6

4

4

4

4

3

2

5

2

3

3

2

o

o

44

G— 7

4

4

G

o

(')

()

1

4

1

5

2

3

44

7— 8

G

3

7

3

2

2

3

8

f-.

4

43

8— 9

4

4

4

4

4

7

3

4

13

7

11

65

9—10

4

5

5

7

G

5

o

4

1

4

(J

G)

58

10—11

8

4

7

G

')

.)

•")

2

1

G

2

9

53

11—12

5

88

7

7

5

7

5

3

2

1

3

3

4
121

52

Sum

lOG

117

104

124

83

75

GG

G3

120

101

1168

(Du

ring î

sight,

or Ijel

ween

G p. m

. and

6 a. m

, 560

1

„ E

ay,

„

••>

5 a m

• ,.

G p. m

, 608

1168

ox THE AFTER-SlTOrivS OF EARTHQUAKES

11)'

TABLE XIX.— HOURLY DISTRIBUTION OP 3842 EARTHQUAKES
IN JAPAN DURING 6 YEARS, PROM 1885 TO 1890.

I

11

rn

TV

V

YI

VII

VIII

IX

X

XI

XII

sum

a.m.

0— 1

8

la

.^)

11

11

18

10

11

7

7

28

9

188

ERRATUM AND NOTE.

Papje 150, line 7, fur September Sih to 2 1st rfi/nl ßth to 21st, September.

XoTE TO Table XIX. — The numbers given in this table include those of the after-shocks of
the Kumamoto earthquake of July 28th, 1889. 'WHien the .shocks which happened at
Kuuiamoto duriui? the latter half of 1839 are excluded, the montldy distribution of
earthquakes becomes as follows.

INIonth.

I

II

III

IV

\

VI

VII

VIII

IX

X

XI

XTI

Xuml)er
of eqkes.

51

57

51

49

65

43

43

42

43

45

57

48

The curve of monthly earthquake frequency [Fi*^. 21, (5)] has been drawn from the above
modified data.

11—12

16

21

18

10

11

10

11

19

12

16

19

18

171

Sum

808

810

308

298

891

259

270

892

298

307

869

312

8842

( During Xight, or Ijetween (3 p.m. and 6 a.m., 1962
t „ Day, „ „ 6 a.m. „ 6 p.m., 1880

3842"

198

F. OMORT.

TABLE XX.~HOUELY. DISTRIBUTION OF 5333 EARTHQUAKES

IN JAPAN.

Interval

GiFU.

Chiran

Whole .Tapan

su m

0— 1 a.m.

5G

8

188 1

192

1— 2

64

15

158 1

282

2— 8

58

12

211

281

8— 4

G2

'j

17G

247

4— 5

1)2

9

180

240

T)— (',

G8

17

iGi;

24G

G— 7

47

9

158

209

7— S

58

12

1C)2

227

S— 9

4î)

10

151

210

î)— 10

41

15

157

218

10— u

42

G

1;î'.i

187

11—12

58

11

1 55

224

0 — 1 p.m.

54

4

151

209

1— 2

48

12

18G

24()

2— 8

88

[H

188

28U

8— 4

8G

8

1G2

20G

4— .^)

8G

10

144

li)0

5— G

47

4

182

18:>.

(')— 7

! "'1

10

1

185

lUG

7— S

77

11

15G

244

s— '.)

58

8

153

214

'.)— 10

51

8

184

248

10—11

42

12

185

289

11—12

85

10

171

21 G

Sum

1258

283

8842

5833

1. 'I'll.' (lala in this laMe arc collected from Tables XIIT, XVI, and XIX.

ox THE AFTER-SHOCKS OF EArvTHQUAKES.

199

TABLE XXI.-YEARLY SEISMIC " ACTIVITIES" IN SWITZERLAND,

THE VESUVIAN DISTRICT, SICILY, AND THE BALKAN PENINSULA

AND NEIGHBOURING ISLANDS, 5'ROM 1865 TO 1883. i

Year ^~""-~

Switzerland

Vesuvian
District

Sicily

Balk. Penin.

s nui

1865

8

45

21

88

112 ■

1866

7

24

19

104

154

1867

1;î

26

8

100

147

1868

28

16

8

10

66

1869

14

62

12

28

116

1870

19

51

6

25

101

1871

25

27

(■)

28

81

1872

1

8

9

18

1878

1

1

58

55

1874

. )

28

16

9

51

1875

8

8

5

5

21

1876

7

55

88

12

107

1877

7

17

6

4

84

1878

9

11

8,

6

2î)

1879

11

8

18

11

48

1880

88

1)

2(')

14

82

1881

()5

8

7

17

97

1882

40

5

17

62

1883

24

8

21

28

71

Sum

818

896

221

517

1447

1. The data are taken from Fuchs' " Statistik der Erdbeben." It is difficult to count ex-
act numliers of earthquakes from the Catalogue, and the figures in the table, wliich are intend-
ed to represent seismic " activities," are merely the nuuiljers of days in successive years on which
one or more shocks have lieen recorded.

1^(1

F. OMORT.

TABLE XXII.-YEARLY NUMBERS OF EARTHQUAKES ITM MINO,
OWARI, AND MIKAV/A, FROM 1887 TO 1890.

Plac'e, District

1887

1888

1880

1800

gnm

(Mixo^

Gifu (Met. Station.)

7

20

10

10

02

Xakatsugawri, Ena

11

11

7

10

89

Mitake, Kani.

8

9

7

18

87

Ota, Kamo.

(')

7

14

10

48

Takayama, Toki.

10

11

12

0

42

Kozuchi, IVInoi.

7

4

(')

7

24

Kitaj^-ata, INIotôsu.

5

7

7

C)

25

Ibi, Öno.

7

8

7

4

2()

Ög'aki, Ampachi.

13

5

S

(J

82

Takata, Taghi.

8

9,

7

4

10

Takasu, Shimo-Isliizn.

15

8

8

4

80

Hachiman, Gujö.

5

5

0

0

18

Kasamatsn, Hag-iiri

1

:}

7

(')

17

'rakatomi, YaDia^vita.

o

()

5

0

19

Tarui, Finva.

')

;}

•")

18

(OWAEl)

Xag-oya (Met. Station)

8

5

0

18

35

Atsiita, Aichi.

:î

7

(')

8

24

Katsukawa, Hig-aslii-lCasiig-ai.

4

.")

0

18

81

Shimo-Otai, Xislii-Kasngai

10

10

8

10

88

Handa, Chita.

1-i

14

(')

ir,

47

Tsushima, Kaito

8

4

;")

7

24

Inazawa, Xakajiuia.

(')

8

8

.')

22

Koori, Xiwa.

7

s

i")

4

24

(Mikawa)

Toyohashi, Atsumi.

2

.)

8

o

9

Shinshiro, Minami-Sliidara

s

.")

■i

7

28

Okazaki, Xnkada.

'.)

10

12

s

39

Chiryu, Aouii.

5

11

9

8

33

Koromo, Xishi-Kaiiio.

'2

1'2

0

9

82

Xishio, Hazn.

'■'

10

1 •)

10

29

Tomioka, Yana.

•2

i -^

.')

3

11

Tao-uchi, Kita-Shidai-a.

._)

8

8

5

18

Asukc, Higashi-Kaiiio.

5

11

12

14

42

Gon, Hoi.

4

7

7

7

25

Jour. Sc. Coll. Vol. VU. PI. IV.

Pig. 3. — Frequency of Earthquakes at Kumamoto.
(between Aug. 5tb-6tb and Dec. 3ôth-.3l6t, 1889).

Time (in uonihs).

ia99. 8 1880, 1

o

e3

ci

CÇ

CO

-ä

'Y^j

. cr
tc ^

00

■^ 1

"3

rjl

Ä

-fSm

C

O

>.

»— <

c

^

^

S

r-*

m

&■

^

•8îjb9 JO jaqmnn jf^a^aj;^

-^ >

CJ

.

,

»

c3 C

c

£

3

S

S

eS

^

c

à

«e 5

s-i

M

-M

IM

r- »^

~ eS

cri

1

1

1

1

'S -^

c

■^

c

O

T. ->^

1—^

'

'

*

K ^

ci
>

5+H ^

g

c ^

•"

3
O

i

(-■ n

rC

_

^

*

S S^

(M

*

*

-

,

X

o

2 t

O

d

■*

fe e

'

1

1

1

o

^

(M

«

îC

II

11

li

II

o

»4

11
X

X

ao

•sjnon 21

ti

^ôAe •s:qlja

jo aacimnjy

S

sjibe jo aiK{mnii Xlq;^Hoy4

a

Ü

tZ.

eu

S

<

^^

TS

s3

^

CO

OQ

03

r-(

J^

a>

a

00

o

r— I

o-

J3

r "^

r-H

o

00

p^.

I—I

o

c

>^

0)

o

r*

Iz;

i'

a

)-4

1

^

1

-u

00

be

S

iûteAS •sqbe Jo aaqnra^

-|j sjjba JO aequinu j^i^uq

Q.
^

-^ A

2 ^

IÜ .s

-H — ■

V

::i

^ CO

»

C G5

S^. 00

1 ^.

15 'S

e c5

T-H

.^" -Ô

^s ^

6 *

i-H

^ o

a

(K

§

^ "^

-H

e3 2

-

C 03

g

CT'

à

,£3 CM"

t= CS

6

^ -a

c ^

?^ -^

© ^j

S 00

8^ pj

u ST

p^

fe cc

•sjj[)o JO aaqoinn <f[anojj

^J8A8 -sîiba JO aeqraa^
o

ES

s^be jo aeqranu Xianofl

X

S:

o

Binoq 9

Jour. Sc. Coll. Vol. VII PI. X.

Fig. 18. — Seismic Frequency at Cbiran, (between
Sep. lStL-2i2nd, 1893. and Jan. 2Ctb-30th, 1B94|.

Fig. 17. - Frequency of Earthquakes at Chirau

(between Sep. l-2tLi-1.3th, 1893. and J
30th-3l3t, 1894).

Time (in
2-day intervali).
18S>4. 1.30-31

Jour Sc. Coll. Vol VII PI Kl.

Z&^ Fig- 19

I l-S ^ (1) and ('ij-Diiii liai Flurtiiatinn of Seismic Frequency soo %,

at Gifii and Nagoya. ' -g Fig. 19.

(51 — Diurnal Fluctuation of Seismic Frequency

s g ,'/ \ (1) Oifu (1258 shocks). A. S ■ t "1 J

" "■ (2) Na^;o.ya (572 ., ). / \ .1 '" "'*Pa°-

stocks)

ïifui»)

(2)
Xag..ya (X)

M Hour luterTsJs.

80 180- ,•■,■..■,•.•■,■■•■,... I

0-1 a.m. 3-4 a.m. 8-9 a.m. 1-2 p.m. 6-7 p.m. 1MÎ p.m.

, . . , Hour Intervals. ^^ Fig. 19 , (4)— Diumal Fluctuation of Seismic Frequency

0-1 a.m. 3-4 a.m. 8-9 a, m 1.2 p.m. 6-7 p.m. U-12 p.m. || in Tok/O. ,

0 ■.2 210 « 60 l'a <"^* shocks)

J Fig. 19., (3) — Diurnal Fluctuation of Seismic Frequency -So / \ / \ (♦)

in Japan.

(3842 shocks)

(3)

Hour mterrols. H""' Interval».

■ ■ ■ r ....... r- 20 . I ................ ,

0-1 a.m. 3.4 am 8-9 a.m. 1-2 p.m 6-7 p. m 11-12 lim 0-1 a.m. 3-4 a.m. 8-9 a.m. 1-2 p.m. 6-7 p.m. 11-12 p.m.

Jour Sc Coll Vol VII. PI. XII,
1

Fig. 21

(1), (2), (3), and (4; -Annual Fluctuation of Seisaiic Frequency
>. „. at Kumamoto and Tokyo.

,Fig.21 . (5) —Annual Fluctuation of Seismic Frequency in .Japan

20 ■« 40

^11 I II III IV V VI VII viu IX S XI XII

Fig. 20.— Diurnal Fhictuation of Seismic Frequency
ftt Kumamoto and Chiran.
(148 shocksl

0-1 tt.m. 3-4 a.E

Hour IntervalB.
1-2 p.m li-7pm U-12pm.

Jour. Sc. Coll. I. VII. PI. XIII

Fig. 24. — Frequency of Earthquakes
at Gifu and ûkawara.

1891.11.25 1891,11.29

Fig 25. — Frequency of Earthquakes
at Gifu and Midori.

Fig. 23.— (B).
, Seismic Intensity for the Vesuviau
District, (between 1865 and 188.3).

Fig. 23.-(C).
Seismic Intensity for Sicily, (between
1865 and 1883).

Fig. 22.— Seismic Frequency in Tokyo, (between Jan.
1876 and Dec. 1882).

Fig. 23.
Seismic Frequency in Tokyo,
110 (between 1876 and 1893).

Time (in months).

1876 1881

Jour. Sc. Coll. Vol VII PI. XIV.

160

Fig. '26. — Frequency of Earthquakes,

(between Jan. .1892 ,and Dec. 1893).

for Gifu.
„ Mitaké.
„ Koori.

100

0 I .
1892. 1

1892, 9

1893, 1

1893, 12
Time (in months)

Jour. Sc. Coll. Vol VII. PI. XV

Jour. Sc. Coll. Vol VII. PI. XVI

!

Pig 28. — Bistrihution of Earthquakes
during 1893 in Mino, Owari, and Mikawa.

( o olnserrivg station.

t ® metewological utation.

Curves are lines of equal earthquake numbers.

:Ui"-J

Jour. Sc. Coll. Vol VII PI. XVII

A 0

1.1 7

Fig. 29. — Distribution of Earthquakes
during January, 1894. in Mino, Owari, and Mikaica.

observing station,
meteorological station.

Curves are lilies of equal earthquake niimhers-

Jour. Sc. Coll. Vol. VU. PI. XV/II

I
jrig, 30.— Distribution of Earthqitxikes

during 4 years, 1887-90, in Mino, Oioari, and Mikawa.

I o observing station.

\ ® meteorological station.

Curves are lines of equal earthquake numbers.

0/i<i/i/iyfi

J/e</or/'

//nf/iù/>ff/t

mm\®

\

\

\

^

w

/

\ r v.^-^

*'^ /wrorno
/, . -

OIÄMÄ

Shin s'htro

_I^

SaAns/mna

>t

1^'^

Jour. Sc. Coll. Vol. VII PI. X/X.

1.1 7

pjg si,—The Great Earthquake of Oct. 28th, 1891.

fo observirifi station.

I (ï meteorological station.
(1) and ( Z) '"■^ isoseismal lines along ichich the maximum accelerations
of earthquake motion xcere respectively 2000 and 800 mm. per sec.

per sec.

The area most strongly shaken is indicated by red shades.

Mesozoic Plants from Kozuke, Kii,

Awa, and Tosa.

Bv

Malajiro Yokoyama. Rié^kushi, Riéakuhakushi.

Professor of Palioontology, Imperial University.

General ReiTiarks.

In ISIH). Prof. A. G. Xuthorst, of Stockholm, published a very
valnaljle pa])er on the fossil plants of Shikokii entitled Bchviiije zur
mcso:oi^r]ien Flora Jnjnuis'^ The present treatise dea.ls Avirh the same
sul)ject, and indeed partly Avitli tlie fossils therein described.

AYhen I ^vrote my memou' on the ^Middle Jurassic flora of Kaga
and its neigddDouring provinces,-^ I thought that the ])lants occurring in
Sliikoku also l)el(^ng to the same geological epoch ; but as meanwhile
the investigations of the Swedish palœol)otanist had shown tliem to be
decidedly younger tlum the Kaga flora, namely, either Upper Jurassic
or Wealden, I deemed it advisable to extend the im'estigations further
than he had carried it. and if ])ossil)le. obtain a more ])recise knowledge
as to the age in which the Shikoku plant-beds were deposited. AVith
this view I have been eno-ao-ed for some vears in üatherino- as inany
specimens as possible, not only fr«:)m the localities whence Xathorst
obtained his material, but also from several other places where similar

1) Ik'iiliscliriftcn do- iiiat]icmati:ich-natunris^en>tc]iaftlichi'ii Clnm^i' (h'r ludi^erliclicn Aladcmie
der lVis>'cn>:rhnft('ii, Wien, IJd. LVII, 1890.

2) Yokoyama. — Jurnsxie Phnit!^ fro)n Kaga, Hidn, and Fj'hi:en. Journal of the ddlccjc of
Scient'f. Tiiipi'riitl I'liiri'r.'iitii, J<i]jini, ml. Ill, ]f<S9.

90i> M. YOKOYAMA ; MESOZOIC PLANTS

fossils occur. These s])eciniens, to2;ethei' Avitli those nlrendv in the
inu.seiiin of the Iinperia] University, Tokyo, form the snl)ject of
the following pages.

]i>efore entering, however, into the consideration of tlie general
character of the above fossils, it may be well to give a brief account of
the geological nature of eacli locality in which plants are found.

Tlie places whence I obtained my material ai-e the following :

1. Kagahara, province Közuke.

2. Yuasa, province Kii.

3. Sakamoto, Fujikawa, and Tannö, in the Katsuragawa Imsin,
province Awa (Ashü).

4. Kataji, Ishiseki, and Tôgodani, near Ivvoseki, province Tosa.

5. Kaisekiyama, not far from SakaAva, province Tosa.

6. Yoshida-Yashiki, near Sakawa, province Tosa.

7. Chöja, in the Shiraishigawa-valley, province Tosa.

1. Kagahara.

In a long and narrow ]\Iesozoic depression in the northern part of
the Chichibu Mtmntains, commonly known as the Sanchu-Ditch,
there occurs a tliick series of shales and sandstones with suliordinate
layers of conglomerate. The greater pai-t of tliis formation was found
to bel(Mig to the (laulto-Cenomanian epoch. From beneath these
Cretaceous rocks, there peeps out, in the \'alley of tlie Hachimanzawa,
a set ot strata consisting of conglomerates in the lower part, and of
shales and sandstones in the upper. The .shales and sandstones
contain in their lower horizon innumerable remains of fresh-water
shells, of wdiich Cyrena forms the most important part. Plants occur

1) Yokoyama. — Versteincninpcn auif der j(vpanitichen Kreide. Fahrcwtoçirapliica, vol. XXXVI,
1890. On some Cretnceoux Fossils from Shilcokic. Joxirnal of Coll. Sdeiire, lnqi. Cnir., Jajxiii,
vol. IV, lit. II, 1S91.

FROM KOZUKE, KU, AWA, AND TOSA. 9()3

in a horizon higher than the shells, and are imbedded in a dark, soft,
often sandy, and at the same time micaceous, shale, easily splitting
into slabs. The state of preservation of the fossils leaves much to be
desu'cd, most of them being changed int(3 a black coaly substance. The
number of species which I could distinguish is 7, among which
Ciiparissidimii (?") japoïiicuni is by far the most abundant.

2. Yuasa.

In 18.S1, while I was engaged in reconnoitring the geology of
Kii, I discovered fossil plants on the northern shore of the lîay of
Yuasa, Yuasa being a town about 7 ri south of AVakayama. The
spot is locally known as Mizutani, and lies between low and high water
marks. Here a well stratified sandstone crops out from beneath the
water, steeply dipping towards the north. This sandstone passes ab(3ve
into a conglomerate overlaid by a dark-blue shale. The rock in which
I found the fossils is the sandstone. It is soft, fine-grained, greyish
to yellow, Ijrittle and often argillaceous, and easily splitting into thin
plates. The f jssils are generally in good preservation, but (jwing to the
brittle nature of the rock, large specimens are difiicult to obtain. The
number of species found is 13, among which recopteris Geijlerianü is
the most frequent. I obtained here also a species of Esther la.

3. Sakamoto, Fujikawa, and Tanno.

These three places all lie in the valley of the Katsuragawa, in
Awa, and very close to one another. The discoverer of the fossils was
my lamented colleague, the late Prof. Y. Kikuchi, who surveyed the
district in 1885. According to his report,'^ the valley is composed of the
Mesozfjic plant-bearing series, overlaid by the Cretaceous sandstone

1) GcoJofnj of Aica, 1883 (M.S.).

-;(J4 M. YOKOYAMA; MEöOZOIC VLAXTS

ill which I had already recognised the occurrence of the Middle Cretace-
ous Tritjoniac.'^^ The plant-bed consists of sliales and sandstones, and is
sometimes in such close relation with the Trigonia sandstone that it is
difficult to distinguish the boundary between the two.

From Sakamoto I possess many small fragments of a dark grey
whale in which badly preserved molluscan shells are found, and a big
block of the same rock containing a single species of Zaiiii'ophijUum
Huchianuiii Ett. sp.

At Fujikawa the plant-bearing rock is a dark-grey, hard, line-
grained sandstone Avliich is sometimes clayey and passes to shale.
Owing to the more or le.ss rough nature ol the sandstone, the preser\'a-
tion is fur from satisfactory, the minute details of tlie plants being in
most cases obliterated. The number of species found at this locality
is 5.

The plants of Tanno are found in a dark brittle shale, exposed in
a valley called Kashiwaradani, deeply cut by a stream. The upper part
of this shale becomes sandy, and on it is superposed the Cretaceous
sandstone. Kikuelii found in pebbles probably derived from the
sandy part of the shale some remains of fresh-water molluscs. The
preservation of the plants is generally excellent, but we have only four
species from this locality and all in small fragments.

From Hiura in Mitimi, near Sakamoto, Xathorst obtained o species,
but in my collection there is none which comes fnjm the same place.

4. Kataji, Ishiseki, and Tögodani.

These l<:.calities all lie in the Ryöseki " Hügelland " of Dr.
Naumann. The geological nature of this district has been recently
studied by Mr. ^1. Yamagami, now a geologist in tlie Imperial

1) On Home L'n'taci'nii.'f Fossils frum Sliilwku. Op. cit.

Î'KOM KOZUKE, KU, AWA, AND TOöA. 205

Geulouical Survey. According" to lii.s re[)ort, '^ the lowest Mesozc^ic
stratum in tlie vicinity of lvyö:^eki consists of a black shale upon which
there is a coarse conglomerate occupying the greater part of the
district. The plant-bearing rocks are shales and sandstones which
overlie this congl(3merate. Immediately over the plant-bed, there rests
the Cretaceous formation witli its characteristic Trujoida pociUifonnis,
and in the midst of this formation there peeps out at Okuminodani
a dark c<jmpact limestone, exactly similar to the Upper Jurassic
limestone of Torinosu in Sakawa, and containing the spines of the
same o'landiferous Cidari^ as occurs in it. The relation of this

o

limestone to the plant-bed is not exactly known, but from its direc-
tion of di[), it seems t(j underlie the latter.

The fossil locality of Kataji is in a ri\'er bed. Tlie r<jck is a dark-
grey, hard, fine-grained sandstone, UKjre or less clayey and easily split-
ting into large slabs. Fossils are numerous, but rare in species. Fur-
thermore, owing to the smoothl}^ polished and carbonised nature of the
vegetable substance, the details of the plants ai'e for the most ])art
effaced. The number of species and varieties obtained is 13, auKJUu'
whicli I't'copfcris (.TCijlcriana and ZamiopJiijllnm Bacliiaiunit seem to [)lav
the most im[)ortant part.

The fos.iil layer at Tshiseki which is a place very near to Kataji,
is exposed in a little stream-bed, and is a sandstone similar to that of
the preceding place, although somewhat ligliter-coloured. The number
of plants is many, but of species few. Moreover, the state of preserva-
tion is exactly like that of Kataji, so that the tw<j localities seem to
belong to exactly the same horizon. The number of species here
obtained is (i ; ChladopJileliis Xniliorsli is the most fre(juently obser\cd.

In the upper p;u't o( Tôgodani, a small side-valley of the
Kasanogawa, is a cliff which is the fossil locality here. The rock is ;i

1) On tlic riant-bcai'nuj Ikds of the Rijôscki Basin, I8'J^ (M.S.J.

206 ^I- YOKOYAMA; MESOZOIC PLANTS

greenish-grey sandy shale which passes into a shaly sandstone. It is
fine-grained and tolerably firm, Ijiit where it is exposed to air, it is
yellowish and more l^rittle. The plants are very well preserved. I
obtained 8 species, but principally Pecopteris Broivuiana, Zamiopliijllmu
Buchianum and Nilssonia scliauvihurgensis.

The majority of the fossils from the above three localities are
those sent by Mr. Gyöken Otsuka of Kyöseki at the request of Prof.
Koto.

Nathorst in his paper described also some from Yakyö, Ueno,
Torikubi, Otani, and Haginotani, places all in the neighbourh<iod of
Kyöseki.

5. Kaisekiyama.

This is a mountain north of Sakawa, also called Kompirayama
from a temple dedicated to Kom})ira standing on its top. The spot
where plants occur is on the southern flank of this miKintaiu, in a
soft, yellow, sandy shale, wdiere there are nnmerous impressions of
plants from which collection has been made partly by Mr. T. iSasa,
and partly by myself. The preservaticm is excellent, but the rock is
so soft and brittle that unless the greatest care is taken, the s})ecimens
are apt to fdl to pieces. The number of species which I could
distinguish from this locality anunuits to S, besides 1 variety, among
"vvhich l^ecoptcris Broicniana is the most abundant.

6. Yoshida-Yashiki.

At this place close to the town of Sakawa a series of sandstones
and shales covers unconformal)ly a dark, compact, bituminous,
so-called Torinosu-limestone containing fossils wliich are referable
to the Malm.'^ T'he plants occur in shales which are ash-grey and

1) Naiiuiaun un<l Xeumayr — Zur Crcologic und FaUcontolouie von Jitpan. Ihiiksdi. </.
)ii'-ttlicin..-natitrw. CI. d. Kalserl. Akad. d. Wmeii^dt., LVll Band, 180U.

FROIVr Kü7A'KE, KU, AAVA, AND TOSA. 907

sandy, aiifl easily 8|)lit into thin ])latcs. Tho only species which
I could find in a collection made l)y Mr. Xasa is that already des-
cribed by Xatliorst, viz., l\'copten'>; Broirniatia. However from the
specimens, which are ])resent in large nnmbers, I could convince
m^^self how variable the form of the ])innules is in different parts
of the frond. In the nnderlyino' limestone we find also now and
then impressions of plants which have been very probably drifted
into it. I have been aljle to distino'uish only two forms, one a
Cliladojihkhis, and the other a Nilssom'a resembling y. ptcrophyUoidef^.
However, owing to their unfavourable state of preservation precise
determination is not possible.

7. Chöja.

Chöja is a mountain village in the valley of the Shiraisliigawa,
many kilometers to the west of Sakawa. I got only a single ])iece of
stone from this locality, collected ])y ^Ir. Toyama, a zealous collector of
fossils at Sakawa. It is an ash-grey sandy shale, quite similar to that
of Yoshida-yashiki, and contains also only a single species, Fccoptens
Broiniiana. According to Toyama and others, a limestone occurs in
the locality.

Conclusion.

The number of fossils which I liave been able to obtain from the
above enumerated places amourits to 24 species and 1 variety. Of
these 24 species, 28 are jjlants and 1 a ])h3'llopod. Adding to these,
3 species and 1 variety described by Xathorst, viz. Macrotceniiiteris
mcmjinata, Lj/copodites sp., rtilophijlhim cf. cutchoise and Podozamites
lanceolatus var. latifolia, the total number becomes '2ß species and 2
varieties of plants, and I species of animal, lîefore proceeding, however
directly to the conclusion which is to be drawn from the occurrence

208 ^^- YOKOYAMA; MESOZOIC PL.VXTS

of these fossils, it is very necessary to examine wlietlier tlie floras of
all the localities represent one and the same i^'-eologieal horizon. In
the first place, that the 8 localities around Ryöseki, namely, Kataji,
Ishiseki, Toaodani, Yakyö, Ueno, Toriknhi. Otani, and Haginohmi,
palœontologically belong to the same formation is not to he the least
dcmhted ; for ont of 15 species which were obtained from all the places
together, 12 are fonnd at Kataji, so that Ishiseki with 6, Tögodani
with <S, Ueno with 4, Yakyö, Toriknhi, Otani, and Ilaginotani, each
with 2, may be fairly considered as a part of the Kataji flora. Of
the o species which are not found at Kataji, 1 is from Ueno, 1 from
Tögodani, and 1 from Yakyo, the last of which, however, is represented
in several other spots out of the Ryöseki District.

Also that the four places in the valley of the Katsuragawa, aIz.,
Sakamoto, Fujikawa, Tarmo, and Mitani, belong to the same terrain,
although the nundjci* of species is not so large, and the ivlation of the
respective floras to one another is not so evident as in Ivyôseki, there
is every reason, both geological and palaeontological, to l)elieve. We
have therefore in all 7 distinct localities from which fossils were got.
These localities sliow close aifi.nitv in their floras, as will be seen from
the following- table : —

FROM KÖZUKE, KU, AWA, AND TOSA.

209

Name of
localities.

Total uo.
of si^ec.

found.

Number of species
in common with.

ce

râ

=

^

i'r.

S

>*

■

■)

5

■ —

4

4

4

0

8

G

0

0

0

0

0

0

2 ."S .*

i 1

8

0

6

0

4

1

7

1

1

1

1

1

0

0

I'otal no
of spec,
fouml
else-
where.

Kagabara

Yuasa

Katsuragawa . .

Eyöselvi

Kaisekiyama

Yosbida-yasbiki

Cbûja

Unknown locabt\

Ï
13
0
15
8
1
o

]

o

8
7
10
7
1
1
0

Tbii.-^, Kao'abara has 7 species, 5 in common witli others ; Yuasa
lo, (S in common witli others ; Katsuragawa 1), 7 in common ^vith
otliers ; llyôseki 15, 10 in common with others ; and Kaisekiyama
N, 7 in common with others ; while Yoshida-vashiki has oiilv 1,
represented in many other localities, and Choja 2, 1 of wdiich is the
same as that of the precedinu' place. Such heinu" the case, the floras of
all tlie places except the last may he safely looked iip(m as helonuin"'
to the Sîime epocli. Of Choja we shall speak more l-ater on.

Having' thus proved tlie close relationship existing betweeii the
floras of the respective b3ca]ities, the next question is their age. Except
Choja, and an unknown locality yielding MacrotrvNidpteris manjuKita^
the number of plants collected in various localities amounts to '2b
species and 2 yarieties. Of these 2o species, 3 are not determimdjle,
and 15 are peculiar to Japan and indeed to the formation in question;
so that what might be aya liable, if eyer ayailable, for the fixing of the
age would be the folio wino' :

210 M. YOKOYA^MA ; MESOZOIC PLANTS

1. ()n.iicliio2J>iis elouijata Gcijl. sp.

2. rccoptcris Broiruiana I)inil\

o. Pecoplfris cf. virijinicnsis Fout.

4. Podoznniites lanccolatus LiiuJL rar. mi)ior Heer.

5. Voilozamites lanceolatvs Lindh rar. JatiJ'nlia Heer.
G. l\jdo.:auutes pusillus Veloiov.

7. Zamiopliijlhnn Bnclvianum Hit. .sp.

(S. ZamiopliijJluu} Bnchiaiunn Ett. sp. rar. auqiisfifoJia Fout.

1). Xilssoiiia i^chaiiiiihurgensis Diiul:. sp.
10. XilssoNÏa Joluislriipl Heer.

\jQX us now s])efik of ench specie.s separately. Foilozamites
laucf'olattts has u very wide ,2'eo2:rapliical l'îuiîïe. hut its vertical
distribution is equally as wide, for it appears in the liluetic and goes
up as high as the Cenomanian, according to Aelenovskv. It is not
improhaljle that we have here to deal, in many cases, with forms which
are in re-dity specifically ditterent, but as long as we are left to rely
only on leaves in their determination, we can not l)ut consider these
forms as belonging to one and the same species. Therefore this cycad
only tells us that we have here a formation which is referable to the
Mesozoic group. ( hn/clu'npsis eloiujata has hitherto been found only in
the Dogger of Japan where, however, it f)rms one of the most adjund-
ant ])lants. On this account we do riot know yet how wide its
vertical range may be. The case is différent with the three species
of Fecopteris Broirniana, Xihsoiiia scliaunihunjensis, and Zamiojjln/llnm
Biichianum. The first two are characteristic AVealden plants in Europe,
and the first has been also described from the l^otomac Formation in
America which Fontaine considers as Xeocomian. The third one is
Urgonian in Europe and Potomac in America, and is ^■ery abundant in
the latter region. These three species are therefore very im|)ortant for
Japan, especially because they occur in many localities and sometimes

FßOM KüZÜKE, KU, A WA, AND TOSA. 911

also in great profusion. Fecopteris [ virgitiioisis and 7.amioph>jUum
Buchiduuiit aiujustifolia are also Potomac, and if the latter is really
identical with Dioonites ahietinus, a« Fontaine asserts, it occurs also in
the European Wealden. The two remaining f<n-ms, Fodozamitcs
IJUsiUus and Nilssoiiia JoJuistriipi are hitherto only Cretaceous, the
former being found in the Cenomanian of Bohemia, and the latter in
the Kome-ljeds of Greenland, considered by Heer as of Lower Cretaceous
age. Tluis, the greater part of the fossils point to the AVealden or to
the Lower Cretaceous. But as the AVenlden is generally looked u[)on as
the fresh-water e<piivalent of the lowest Cretaceous, I go a step further
than Prof. Xathorst, and sp.y that the plant-bearing beds of Közuke, Kii,
and Shikoku represent the wIkjIc Xeocomian series, corresponding to
the Potomac of America. This assertion is in jiccordance with their
geological position. On the one hand they show a close relationship to
the TrifjoiUK Sandstone of the Gaulto-Cenomanian. and on the other
to the so-called T(_)rinosu or Cidaris Limestone. Avhich has been consi-
dered as Up[)er Jurassic. As to Choja, which has yielded onlv I'ccop-
Icris Brutnüana and rtiloplniUuin cf. cutchense, it is ditücult at ])resent
t(j make any definite statement ; but the (occurrence of a Pecopteris so
characteristic of AVealden seems to sliow, in s[)ite of the other, that we
have here to deal witli strata which in all i)rob:ibilitv bel(jn<'- to the
same g-eoloo-ical formation a.s the (jther localities.

In contrast to the Middle Jurassic flora of K;ig:i. a marked fcatiu'e
in the one above discussed, is the comparative rarity of species alreadv
found in (jtlier countries. This is no doubt due, not only to the pre-
ponderance of marine deposits in the early })ajT of the Creniceous
period and conse(pient sctircity in rocks containing vegetable reiiinins.
but als(j to the fai/t that the already known floras of this epoch sucli as
the Wealden and the Potomac belong to provinces widely distant from

912 ^^- YOKOYAMA; MESOZOIC PLANTS

ours ; wliil.st in the Dogger flora we have a comparatively rich one
from Siberia.

The determination of the age of the plant-heels leads to the con-
clusion that on the outer or convex side of Japan, ns far as our present
investigation goes, no Middle Jurassic fresh-water formation is
developed. Ijut whether the converse liolds true, that no Younger
Mesozoic p]ant-heds have Ijcen deposited on tlie inner or concave side
of Japan is at present difficult to say and must be left for future
investigation to determine.

For C(3nYenience' sake I propose tlie name of L'ijüscl^i Sen\'s lor
all those strata containini"' Younaer Mesozoic plants in distinction to
those containing- Middle Jurassic ones. For the latter I also sufi'^'est
the name Tcluri Series, from the ^■alley (jf the river lV't»jri in K:iga,
where thev Avere first discovered.

FiT.ICES.

Thyisopteris sp

Dicksonia tosaiia

Diclisoniopteris Naumaimi

On yobiopsis elougata

elegaus

Adiautites y uaseusis

Pteria (?) sp

Sphenopteiis teimicula

Peonpteris Browiiiaua

., Geyleriaua ■. .

cf. virgiuieusis

Cbladopblebis Natborsti

MacrotiBiiiopteris (?) mai'giiiala

LycoI'ODIAC^.
Lycopo Jites sp

CïC.VDEAC.E

I'oduzamiteb sp

,, laiji-eolatiis var. minor

1. ,. ., ,, latifolia

,. pusillus

Zamiopbyllum Bucbiaimm

ai- ., ,, var. angustitolia.

Nauuiaiini

tilossozamites parvifobiis

Nilssouia Johustrupi

., schaumburgeiisis

pteropbylloides

i'tilopbylbim cf. cutcheuse

CuNlFEE^.

Cyparissidium (?) japonicuiu

Torreya veuusta

Phsxlopoda.
JOslliena rectan>:ula

■■■

+

+

+
+
+
+

+

+

+
+

+

+

Tosa (Ryöseki district).

Middle Jurassic of Kaga, Hida, etc.

VVealdeii of Europe. Potomac of America.
Potomac of America.

lEboetic, Jurassic and Cretaceous of Kurope.
f Middle Jurassic of Kaga etc.

Ceuomanian of Bobemia.

Urgonian of Europe, Potomac of America.

Potomac of America.

Kome beds of Greenland.
Wealden of Europa.

Eajmabal of India.

FROM KOZUKE, KU, AWA, AND TOSA. 213

Description of the Species.
Filices.

1. Thyrsopteris sp.
PJ. XXIII, Figs. 2, 3.

On u piece of .-^toiie, represented in fig. 2, tliere ure many elongated
pinuie apparently l)el(jnging to a twice pinnated frond. They taper
very a"i">duallv forward, and seem to be tolerably distant. The
pinnules are long, alternate, more or less directed forwards. cl<jse
together, and sharply toothed. Tlie general appetirance of the fern
reminds iis of the many species of Thyrsopteris hitherto described
frcjm the Jurassic and Cretaceous rocks. It also closely resembles 0/i//-
cliiopsis donijaia tleyl. (see further on), which possesses, howeyer, more
acutely directed pinnules. It is to be regretted that the lateral yeins
are so indistinct Jis not to permit of any stricter comparison with the
already known forms being made.

Fig. o })robably re[)resents an upical part oï the same fern.
Pinnules are here linear and not luilike those of iVc'a|'/6'r/>; ■i'//-j/////t7<.s/.!>
Font, described below.

Loc. — Fujikawa, in the Katsuragawa basin.

2. Dicksonia tosana ///.

PI. XXV, Figs. 13, 13(^

Frond tripinnated ; racliises of various orders rather slender ; priinarij
rachis lent someirhat zigzafi, others nearhj straifjlit ; primarij pinna'
elongated, distant, rising at nearJij right angles to tlie rachis ; secondarg
pinnic comparatirelii sjiort, alternate, close togetlier and a little overlapping,
tliose on the bach of tlie racliis heing more acatelg directed forward and

214 M. YOKOYAMA; MESOZOIC PLANTS

hcawuj more clonijatal pinnules than tltose on the front puuudes linear
to elliptical, acute at apex, directed more or less fonrards and close together ;
veins fine, few, indistinct, an evanescent midvein sendiwj off a feiv simple
lateral veins.

This slender and elegant fern is, I believe, to be brouglit under
Dicksonia, and indeed close to D. acutiloha Heer (Yokoyama, Jurassic
Plants, p. 24, PI. I. figs. 2, 2«, 1^, from Avhicli however it is dis-
tinguislied in having shorter ultimate pinna;, which on the back
of the sec(jndary rachis are more acutely directed forward than on the
Iront. The ])inna3 with the linear pinnules look also not unlike those
ol ünijc]iinj).sis cloiujata Geyl. Veins are in most cases not dislinctlv
visible, in one or two cases however they were observed, and then
arranged as in fig. loa.

hoc. — Tôgodani near Uyôseki. Onl}^ a single specimen.

3. Dicksoniopteris Naumann! Nath.
PI. XXV, Fig. 4.

THcksonuqileris XaiDinnuii. — Nathorst, Beitr. zur lucsoz. Flora Japans, ]). 11,
pi. V, tig. 4.

AYhat Xathorst descnl)ed from Ilaginotani under the above name
was found also at tw(3 other loc:üities cited below. The fern is rather
slender, with distant, (opposite piniicC rising at an angle of G()-(i5°
from a straight main (?) rachis. Pinnules are elongated, finger-like,
obtuse, UKjre or less crenulate and (jiiite close together, ahhough not
overlapping. The margins of these pinnules show on each side 4 or 5
round fruit-dots, most (^f which however have left only slight impres-
sions on the stone. Veins have been nearly obhterated, the onlv
thing which I can now and then observe beitjg a- very line mid\ein
which near the apex is indistinct.

Loc, — Kataji and Ishiseki, near Pyoseki. Pare.

FROM KOZUKE, KU, AWA, AXD TOSA. 215

4. Onychiopsis elongata Geyl.
n. XX, Fi- 8. PJ. XXI, Figs. 1, 4.

f")in/cJtinj)sis elmir/ata — Yokoyama, Jurassic Plants, p. 27, pi. II, fig. 1-8, III Od,
XII, 0,10. Natliorst, Beitr. z. mesoz. Flora Japans, p. 4, pi. I, fig. 1-3, p. 8, p. 10,
p. 13, p. 14, pi. YI, fig 5.

T/ii/rsnptcris eloinjuta — Geyler, Ueb. Foss. Pflanzen a. d. Jnrafonn. Japans,
p. 224, pi. XXX, fig. 5, XXYI, 4, 5. Schenk, in Pichtliofeirs China, vol. lY, part X,
p. 203 pi. LIY, fig. 1.

This fern, wliicli is so nunierous in the Middle Jiiras.sic of Kagn,
Hidn, aud F^cliizen. is also abundantly re|)resented in several of the
lor-'dities in Tosa and Awa. It is easily distin2:nished from the nearly
related forms l»y the charaeteristic long linear pinnules acutely
directed f )rward. Sterile as well as fertile fronds were obtained, the-
fertile ones being quite similar to those figured by me in the above
mentioned woi'k.

Nathorst's opinion that SpliCNopteyis j\f(intell/' Schenk of the
AYealden and Thiirscqiteris capsuJij'cni Velen. from the Ceuomanian of
l>ohenn*',i belong to Onvchiopsis, is, I believe, quite justified.

JjOC. — Xumerous at Kaisekivama near Sakawa, and Fujikawa in
the Katsuragawa basin, ]ess so at Yuasaand Kag;diara. It also occurs
at Kat.'iji. TOgodani. and Ishiseki. and bv Xathorst it has been also
desci'il)ed from Yakvo, Ueno, Otani, Ilaginotani and Iliura (Mitani).

5. Onychiopsis elegans m.

PJ. XXYIII, Fig. 7, la.

Frond tirice piinuifed ; pitimr elo/Kjdtcd, rdchin sîeiidi-r iritli a terDiinal
pinnule ; jrinmilfs tolcrahhi do>^e tofjetlier, t])in, opposite or alteniote, directed
fonvcmh^ lanceolate, ]>roadest at hase and (jraduaUii tapering] above, entire in
the hirer half and coarsehj toothed at the upper, iritlt apex olitusehj poitited.

21Q M. YOKOYAMA; MESOZOIC PLANTS

Veins rather nnmerovs n'itli distinct hut evanescent midrein : lateral veins
aciite^ simple or once forhed.

The g'eneral features (^f this graceful fern are those oï Onijcliiopsis^
although in some respects it also resemhles Diclsonia. 'JTie generic
determination is therefore not settled.

Loc. — Kaisekiyama. Only one specimen.

6. Adiantites yuasensis, m.

VI XXI, Fig. 15.

I have onl}' a single and not quite perfect specimen of this fern.
It shows a pinna with a slender rachis on which we see sul)o})posite,
distant, very thin, oval pinnules cuneate at base and furnislu^d with a
short st;dk. The npper margin is not well preserved. Veins are
very typical, being fine, verv numerous, equal, divergent, and repeat-
edly di'-hotomous. Therefore, there is not the least doubt tliat we
have here to deal with a fern closely akin to our recent Adiantum.

Schenk desci'ibed from the Albours Chain a similar form under
t]\Q ii-A]\\L' oï Adianlinn Tietzei (Die von E. Tietze in der Albourskette
gesain. foss. l^flanzen, p. 8. ]»I. TI, fig. 2) whicli possesses larger and
moi-e closely set ])iiinules.

JjOC. — Yuas;i in Kii.

7. Pteris (?) sp.

V\. XX, Fig. 9. PI. XXI, Fig. 6, 7. (?).

A fragment of a fern from Kagahara (PI. XX, fig. D) having
thin, parallel-sided, spathulate })innules, which measure a little o\qy 2
cm. in length and <S-9 mm. in breadth. One of the pinnules is seen
attached to the slender rachis by the greater part of the base.
\"eins are fine and numerons. The midvein is distinct. The lateral

FROM KOZUKE, KIT, AWA, AXD TOSA. 917

veins are ^■ery fine, once or twice forked, and rise at first acutely, but
soon hend outward so as to be nearly perpendicular ti:) the mid vein.

The frao'ments in Figs. 6 and 7, pi. XXI, from Kii with evanes-
cent midvein resemble in mariy respects the fern just described. But
the pinnules seem to l)e bnjader and the veins coarser.

The specimens represented in the figures aljove cited are at all
events too imperfect to allow even a correct generic determination.

IjOC. — Kagahani, Yuasa (?). Rare.

8. Sphenopteris tenuieula m.
PI. XX, Fig. 11. Fl. XXL Figs. 2, 2a. Fl. XXYIir. Fig. 6.

Fnvul trij)itnta(cd ; mcliises of variou!^ orders slender; ultimate
pinna' short; pinnides tliin, remote, ohtnse, reri/ aciiiel if directed forirard,
ciincatc at hase, entire or lohed or evoi pinnateJi/ parted ; lobes or partitions
fen-, also obtuse a}id some of il i em mai/ a<jain be furnislied with ;? or 3 coarse
crenations ; reins few, eipial. fine, direnjent, and sereraJ times forliCd.

This fern is present only in small fragments, yet is sufficiently
characterised to be treated as a new form. It seems to have been
toleral)lv delicate, as can l)e judged from the slender rachises and
thin pinnules.

Sphenopteris Aucrbachi Trautschold (Der Klin'sche Sandstein,
p. li), pi. XYIII, fig. 5) and Sphenopteris Goeppcrti Dunker (Mono-
graphie der norddeutschen Wealdenbildung, p. 4. pi. I, fig. 6, IX,
1-3) are near relatives of our Japanese species. Ihit in the f)rmer the
venation is obsolete, while in the latter the lobe- are single-veined.
Tlnirsnptcris breripennis Fontaine (The Potomac or Younger ^Ies(^zoic
Flora, p. 124, Pi. XXXIY, fig. ?,, XXXYI, 2) is also not unlike our
plant, l)ut possesses more luimerous veins.

21,1^ M. YOKOYAMA; ME.SOZOIO PLANTS

WIkiI Xathorsl calls Spliennpleri)^ cf. Gnrpperti Diuikor (IViträge,
p. 11, ])]. \l, Uli'. '2, o) ]>rol):ibly pertains to the same lern.

Lor. — \ uasa (numerous fragments), Kaisekiyama (rare) and
Kaoaliara (rare).

9. Peeopteris Browniana Dunk.
?1. XXIV, Fig, 2, 3. PI. XXVII. 1-4, ocJ.

Pecopteiis rf. BionnUiiui — Katliorst, Beitr. z. mosoz. Flora Japans, p. 13, pi. V,
fig. 5.

Fccnpteiis Birnvniana — Dnnker, Monographie d. norcldentsclien Wealdenbildung,
p. -5, pi. YIII, fig. 7. Schenk, Fossile Flora der nord westdeutschen Wealdenforraa-
tion. p. 215, pl. XXVI, figs. 2, %t. Fontaine, The Potomac or Younger Mesozoic
Flora, p. 88, pl. XXIT, fig. 10, 11, XXIII, 2-7, XXYI, 3, 13.

Ah'th()}iteiiH Broinndua — Schenk, Zur Flora der nordwestdeutschen ^Vealden
ormation, p. 159, pl. XXVI, fig. 3-5.

Frond tn'pinnated ; racliises of the first and second order comparativeh/
slender, hut r'lejid ; jtrimanj pinnœ sidiopposite ; secontlnrij jtiinuv eJonijoted,
stroiifUt or Jiiflithj ciirred, opposite, sidiopposite or alternate, more or less
directed f'orirard ; jtinnides leatlteri/, opposite or alternate, quite close or a
little separtite, fDojer-sliapcd, ohtiise or acute, sijuietiiia's Vaßitlij falcate at tlte
etui of tlie pinna\ ami attaclied to tlte racliis irilh tJie u-liole hase ; tlieii are
(jeneralhi entire in the middle part of the frond, hut in its posterior part, or
in its anterior part udiere secondanj pintuv are reduced to jiinnules, tlieij are
crenate or toothed, much more elongated cren to linear, atul non' and then,
stronghj fcdcatc ; reins in smaller entire pinnules indistinct, in larger and
crenate or toothed, ones, with a distinct midrein u-hich sends of acute, simple
or sometiuu's eren twice to thrice forled lateral reins.

This is the most almndant fern at Kaisekiyama and F'ujikawa.
Nathorst had already compared it witli Vecopteris llrou-niatia Dunker.
The shaj'C of the pinnules in this fern is very diiterent in different

FKOM KOZÜKE, KU, AWA, AXD TOSA. 219

parts of the frond. ;md this has been already noticed by Fontaine in
his American specimens. Fecopterù hreripmnis Font. (The Potomac
Flora, p. 88. pi. XXVI, û'^. 4) seems to be only a part of the frond of
this fern ^vith toothed pinnules. rccopterh sp. of X'athorst from
Haginotani (/. c. V\. VI, lig. 4) also appears to belong to it.

Ill one of the fragments apparently belonging to this fern,
pinnules were observed bearing sori (]»1. XXVII, fig. 1, \a) Avliich
are preserved as black round d<jts in number of 2-4 on each side of
the midvein. 41iis mode of fructification strongly reminds us of the
recent genus Aspidiitm.

Log. — Kaisekiyama (most abundant), Yoshida-Yasbiki (lumier-
oiis), Chöja, Fujikawa (frequent), Tôgodani.

lO. Pecopteris Geyleriana Nath.
PI. XXL Fb-. U. V\. XXIII, Fig. 1, \ü. Ph XXVIII, Fiu-. 5.

Fecopti'iis (Tf)jli'ii(Oiii — Natliorst, Beiträge zur mesoz. Flora Japans, p. 8, pi. IV,
tig- 1, VI, 1.

This plant first described by Xatborst from Ryôseki has been
since found in several other localities. As this author had already
pointed out. many of tlie pinnules are erired, not only on the anterior
side, but also often on the posterior side, so tliat they become more or
less triangular in shape. And it is a peculiar character of this fern
tliat the lowest pinnules on the front of the rachis are often falcate
backward instead of forward. The pinnules are in general blunt, but
in tbose wliich are elongated are often pointed, ^"eins are in most
cases indistinct save the evanescent midvein ; in some cases, however,
dichotomous lateral veins were observed (pi. XIII, fig. \n).

Xathorst's assum{)tion. that the specimens with sm:dler and more
pointed pinnules like those represented in fig. :?, (5, pi. Y\ of his work

22Q M. YOKOYAMA; MESOZOIC PLANTS

may represent a thrice pinnated portion of a frond, has been verified
in several specimens in my collection. That the pinnœ were not all
in one plane is also shown by having them more or less overlapping.

This species, which exhibits a close relationship to Fecpoteris cxilis
IMiilhps (Yokoyauia, Jurassic Plants, p. 35, pL I, tig. .S-10.), is dis-
tiu<>-Lii.ihed from it bv auricLÜate and at the same time more falcate
pinnules.

/.or/.— lùijikawa (very frequent), Yuasa (do.), Kataji (d().), Fuji-
kawa (less fVepicnt), 'J'ögodani. ßy Nathorst it has been also de-
scribed from Torikubi.

11. Pecopteris cf. virginiensis Fontaine.

n. XXIV, Fig. 1.

Peüoplcrh vinjinicnsh. — ■Fontaine, The Potomac or Youuger Mesozoic Flora, p. 82,
pi. YIII, figs. 1-7, IX, 1-G, XXIV, 2, CLXIX, 3.

A fragment of a pinna with ahernate or subopposite, distant,
\ou<^' linear, toothed pinnules no doubt belongs to a fei-n which, if not
identical Avith, is at least very closely allied to, Pecopteris vinjiiiiensis
Fontaine, from the Potomac Formation of America. The lateral veins,
owing to the thick consistence of the leaf are not quite distinct, but so
far as I can see, they seem to l)e acutely directed foJ'vvard and at least
once forked.

Loc. — Fujikawa. Tanno (?).

12. Chlaclophlebis Nathorsti vi.
V\. XXVIII, Figs. 3, 4, 10, 11.

C'hlfKhipItlrhis .s;/).— Natliorst, Beiträge z. mesoz. Flora Jai>aus, p. -1, pi. I,
figs. 1-8, p. 8, l:-i.

Frotiil hipinnaled ; rachises conipaycitivebi slender : piitme alleniate,
eluiujated, lutrrutred at hase : pimudes coriaceous, opposite or alternate, close

FROM KOZUKE, KU, AWA, AND TOSA. 221

toijetJwr, oflcti continent at hast'. triiDujular to lanceolate^ entire^ falcate^
obtuse or pji'nted, an.i sliorter ut the posterior than in the middle part of tlie
j)innœ ; reins distinct; niidrein disappearing near tlie apex ; lateral reins
directed acutelij forn-ard and once dicliotomovs.

I ha\'e no doubt that what Nathor.st calls Chladophlehis sp. in his
Beiträge reters t(j this tern. This author had ah'eady recognised its
difterence from the closelv aUied Aspleniain CCJiladophlehis) ndufhiense
Bro't. in liavini»' the ijinna3 narrowed at base. The reiison whv he left
it unnamed was the indistinct venation which characterises most of
the fossils of Kyöseki. At Kaisekiyanra, howevei-, where the state of
preservati(jn is nuich more fivonrabie, the veins are well preserved,
and are quite simihu" t(} those of Aspleninm. lioesserti PresL from the
Rhœtic QÏ Yamanoi. (Yokoyama, on s(jme F<3ssil l^lants from the
Coal-bearing Series of Xagato, pi. XXIV, figs. 1, 2, 5). The dif-
ference between the latter and A. wlu'thiense^ which species are often very
difficult to distinguish, has been given in the work just mentioned.
In trutli, ChladoplileJiis Xatliorsti shows sucli a. close affinity to
Aspleninm lioesserti tliat it would be often quite impossible to dis-
tinguish the Uxo, especially when the specimens are present only in
fragments.

Loc. — Kaisekiyama ; numerous. Yuasa ; frequent but in frag-
ments ; Kataji and Ishiseki ; numerous. According to X'athorst, the
plant occurs also at Ueno, Tôgodani, and lliura.

13. Macrotaeniopteris (?) marginata Nath.

M(t<Tnt,cni()i>teris f! ) mavijiwda. — Nathovst, Beitrüge, loc. cit., p. 14, pi. YI,
lig. G, ()((.

What jSTathorst has described under the above name from an
unkn(jwn locality is not re[)resented in my collection.

222 ^I- YOKOYAMA ; MESOZOIC PLANTS

Lycopodiacse.

14. Lyeopodites .<]>.

Lijcopodites sj). — Natliorst, Beiträge, loc. cit., p. 10, pi. II, fig. 8.
This species found by Nathorst at Ueno is also not represented in
my collection.

Cycadeacse.

16. Podozamites lanceolatus Lindley
et Hutton.

PI. XXIII, Figs. 4, 5.

I'odu^xijiitcs luiiceolatiis. — Heev, Beitr. zur. JuraHora Uwtsib. u. J. Amur]., 1876,
p. 110, pi. XXYII, tigs. 6, 7, 8, 5iih. Yokoyama, Jurassic Plants, I.e. p. 45, pi. V,
fig. 8. Yeloiiovsky, Die Gymnospermen cl böhm. Kreidefurm. p. 11, pi. II,
figs, 11-19, 24.

There are only two leaflets in my collection which are referable
to the well known l'odozitiiiiies Idnccolatus. One of them (tig". 5) has
the petiole preserved, and the other (fig. 4) the apex. l>otli seem to
belong to the variety minor of Heer.

Loc. — Tanno.

Xathorst mentions and describes in his lîeitiilge p. 10, pi. IV,
fig. 7, another variety hilij'olia (jf the same s])ecies ihnn Kataji.

16. Podozamites pusillus Velenov.
PI. XX, Fig. 2, oh, 4, 5, 7.

Puduzinitites jamllua. — Yelenovsky, Die Gymnospermen der böhm. Kreideform.,
p. 11, PI. II, figs. 20-22, 24./.

A number of small rycadeous leaflets, mostly isolated, but rarely
attached to the rachis, oval or elongate oval in shape, r<3unded or blunt
at apex, and when well preserved, furnished with a short [)etiole.

FROM KÖZUKE, KIT, A^VA, AND TOSA. 223

These nre no (lonl)t to 1)C identified Avitli Vodozamitcx piisillus, ;i species
founded by A^elenovsky on :i Cietaceous form from lîoliemia. Fig. 5
represents u leaflet ahout 22 mm. long and G. mm. broad belonging
to a longer form. It possesses 22 distinct elevated parallel veins be-
tween which a single interstitial vein is visible. Fig. 4 represents one
attached to a strong rachis. It also shoAvs distinct veins. In other
specimens they are more or less defaced. A leaflet represented in fig. 2
left, is mnch slenderer than others and resemldes Vo}o:<(witcs lauceohitiis,
so that it is not impossible that this s])ecies may afterwards turn ont
to be ordy a variety of the latter.

roiloKiiiiifcs hinccoJdtux nir., wliidi I described from Kaga (Jurassic
Plants, PI. A\ fig. .")), is not unlike the longer forms of roiloznnrifcs
pusillu.^, although more abruptlv tajiering above and acutely ending.

Loc. — Kagahara ; fi-equent.

17. PodozaiTiites sp.
PI. XXV, Fig. .S-li>.

^[anv frafrments of a leaf Avith small, elonofate obionff to lanceolate
leaflets which look very much like tliose of the ])receding species. In
most of them, however, the veins are completely obliterated, and when
slightlv visil)le thev seem to be much coarser than in Podozanu'tes
pusilliis. In this latter respect, the ])l:iiit i'eseml)les some species of the
genus XcKji'/opsis descril)ed by Fontaine from the Potonvac Formation,
e.g., Naficiopsis lictcrophißla (Fontaine, Joe. cit. pi. LXXXVI, fig. 6).

Loc, — Kataji.

18. Zamiophylluin Buehianuin Ett sp.

PI. XX, Fig. 1. 1^1. XXII, Figs. 1, 2. PI. XXIII, Fig. 6.
PI. XXVII, Fig. bah. PI. XXVIII, Fig. 1, 2.

Zamioplnjlhnn Buchianum. — Nathorst, Beiträge z. mesnz. Flora Japans, p. G,
pi. II, figs. 1-2, III, V, 2, p. 0.

224 ^I- YOKOYAMA ; MESOZOTC TliAXTS

rtrinpJii/IImn Buchianum. — Ettingsliansen, Beitrz. Flora. d. "Wealdoiiperiode, p. 21,
pi. I, tig. 1. Schenk, Die Fossilen Pflanzen d. Wernsdorferscliicliten i. d. Nord Carp,
p. 8, PJ. Ill, fig. 5.

]')i(>()iiitcs JJuchi'iiiii.s. — Fontaine, Tlio Potomac or Younger Mesozoie Flora, p. 182,
pi. LXVIII, tig. 1, LXIX, 1, 8, LXXI, 1, LXXII, 1, 2, T.XXIII, 1-3, LXXIV, 1-3.

Tliis plant first pointed out l)y Xatliorst as occuiTing- at liyöseki
is ])rofuse]y represented in my collection. A specimen shown in fi,u'.
(!, pi. XXIII is from Sakamoto. It represents a leaf belono-ing to its
upper pnrt. The leaflets are narrower than those fig'nred 1)y Xathorst,
the l)roadest being about 5 mm. with 15-LS veins. It looks more like
fig. 2, pi. LXX of Fontaine. Tlie specimens from otlua- localities
show no essential difference from those descril^ed l)y Xathorst. In
some, however, the apex of the leaflets were observed, and in o\\o from
Kaisekiyama it was Itlnntl}^ pointed (fig. i'yah. PL XX\']I), wliile in
that of Yuasa it was acute (fig. 2, PI. XXII.).

P(^ntaine in describing this species from the Potoma.r-. where it
seems to be very abundant, used the generic name oï J)lii()Niffs. but as
this name is now applied to a cycsid whose leaflets or segments are
attached in front of the rachis with their whole base, it would lie
advisnl)le to retain that ]iroposed by Xathorst.

JjOC. — Kaisekiv'uiia. A'^nasa, Sakamoto, Tanno. Kagahara, Tögo-
dani, Ishiseki, Kataji. According to Xatln^ii-st, the plant occurs also at
Otaiii. T(~)riknbi. and I eno.

18a. Zamiophylluni Buchianuni Et-i. sp.
var. angustifolia Font.

PI. XXII, Fig. 4. PI. XXY, Fin;. 5. PI. XXVIU, Figs. 8. !).

Piimiitcs Bitchicmiis ntr. (nuiii-stijoliiis. — Fontaine, Tlie Potomac or Yoniiger
Mesozoie Flora, p. 185, pi. LXYII, ilg. (>, LXVIII, 4, XYI, 2.

FRO:\r KOZVKE, KIT, AAVA, .VXD TOSA. 90.5

Xoiie of the specimens which I refer to this form are quite
well preserved ; still I think I am sufficiently justified in jJacing
them under it, as the leaflets exhiliit essentially the same chnracters
as the foregoin<j^ species. They are long, linear, 2-2.5 mm. l)road,
acute at a|)ex, and somewhat narrowed at base. They are tolerably
close together, directed acutely forward, and one of them seems to
terminate the leaf. Fig. -i, pi. XXII shows a specimen from Yuasa.
Its leaflets possess 7-9 parallel veins. Two fragments from Kaiseki-
yama show also 7-i* ^eins. A specimen from Kattiji (tig. 5, pi. XX\ )
has them quite (obliterated.

Fontaine considers Dioouitrs ahietinus Micj. (Schenk, F(X«^sil Flora
d. norddeutsch. AYealdenform. p. 32, pi. XA^I, tig. 1) as referable
to this variety oî ZamioplnjUum Bucliiamini, to which indeed it shows
a great resemblance.

Lot: — Kataji, Kaisekiyama, Yuasa ; rare.

19. ZainiophylluiTL Naumanni Nath.
ri. XXII, Fig. 3. ?1. XXYI.

Ztouiapltijlhim yaununuii. — Natliovst, Beitr. z. mesoz. Flora Japan?, p. 7, pi. Y,
fig. 1.

I possess a large specimen of a Zamioph^dlum from Ishiseki
which is undoubtedly referable to the species fonfided by Xathorst.
The leaflets are distant, opposite and not tiipering towards the ba.se at
which place liowever they are a little contracted. The breadth
measures up to 20 mm. Besides this undoubted specimen oï Znmin-
phijIlKui XaunuDini, there are many fragments of leaflets from other
localities wliich measure 12-15 mm. in breadth, and are broader than
those of Z. Biifliidiiinn tignred by Xatliorst and Font:iine, in which the
breadth seems never to exceed 10 mm. These are therefore to be

2-2C} M. YOKOYAMA; MESOZOIC PLANTS

considered as those of Z. Nnnmanni. However as the difference be-
tween the two species lies not in the l)readtli, lint in the form of the
leaflets and in their mode of attachment to the racliis, it is in many
cases impossible to decide which of the t\N'o species we are really
dealing" with, especially when the specimens are in small fragments.
More(iver tliere is much doubt whether Z. Nauinatnii is not a species
founded on the lower part of a leaf of Z. IhicltiaHuiii.

Loc. — Kaisekiyama, Kataji, Tögodani, Ishiseki, Yuasa.

20. Glossozamites parvifolius 7//.
PL XXI, Figs. 5, 5a.

I possess a small fragment only of a pinna in which the rachis
dichotomises, with small, distant, opposite or subcjpposite, entire,
obovate leaflets, about o mm. in length and attached to the rachis
almost perpendicularly. Veins are few, equal, divergent, those in the
middle part of the leaflet dichotomous, and those near the lateral
margin simple.

This ])lant reminds us of l\')iJo.:aniitrs Ç(Tlo-'<so.:a)nites) ohoratns
Scherjk (Die Foss. l^flanzen d. AVernsdorferschichten i. d. Xordcarpa-
then, ]). 10, pi. II, tigs. 7-10, III, l-o), wliich however possesses
much larger leaflets.

Loc. — Yuasa.

21. Nilssonia Johnstrupi Heer.
Fl. XXV, Figs. 1-4.

Xilssnuiti JohiistiKpi. — Heer, Flora der Komescliicliteii, Flora Fossilis Arctica,
Vol. VI. p. 44, pi. VIII, figs. 1-G.

yil.sso)tia cf. nrieiitdli.s. — Natliorst, Beiträge. I.e., p. 5, PI. I. tig. 4-5.
Xow and then there occur at Eyöseki oval or linear oval, entire
leaves of a cycad which exhibits a close resemblance to Nils.sonia ori-

FROM KÖZÜKE, KU, AWA, AND TOSA. 227

cntalis Heer (Yokoyaina, Jurassic Plants, PI. XIV, figs. 4-9). These
leaves are in ireneral larirer than those from Echizen, some attaining"
4 cm. in breadth and up to 20 cm. in lenuth. In this respect they
resemble more tliose of Siberia floured i)v Ileer. Ihit tlie veins in the
Ryöseki plant are doubly as coarse, tliere being- only '2 in a millimeter,
wiiile in the other we can count 4 in the same space. Therefore we
have here t(^ deal with nnother species of Nilssonia, and indeed with A"".
Joluistrupi Ileer from the Lower Cretaceous of Greenland, which agrees
exactly with the Japanese form both in number of veins and also in
the more oval shnpe of the leaves.

Nihsonia hohciiiica A^elenov. (die Gymnos])ermen d. bohm. Kreide-
form, p. 11, pi. II, fig. 25-28) seems also to be a very nearly related
species. It has however the leaves longer and narrower.

Loc. — Kataji and Tögodani ; rare.

22. Nilssonia schaumburgensis Dkr.
PI. XX, Figs. 12, 14. PI. XXI, Fig. 14. PI. XX 11, Figs. 5-7.

yH.'i.'KDii« vf. scluntmhuiijoisiii. — Natliorst, p. 5, pi. I, figs. 6-9(/, p. 9, 13.

Pteroiiltijllum scliaionhuiyetise. — Dnnkei", Ueb. d. Norddeutscli. Waldertlion. Progv.
d. liobereu Gewerbesclinle in Casse], p. IIG. Monogr. d. Norddeutscb. AVealdeii-
fonu., p. 15, pL I, fig. 7, 11, 1, YI, 5-10. Ettiugsbausen, Beitr. z. näbereii Keiiiitn.
d. Flora d. Wealdeiipeiiode, p. 22.

AiKnnozdriiitca scJudDiibidr/oisiv. — Scbenk, Die Flora d. Nordwestdeutscb. AVeal-
deuform. p. 231, pi. XXXIII, figs. 1-9.

X^athorst has rightly referred this })lant to the well known Weal-
den form Xilssoitta .%-]tauiiiJ)ur(jcnsis Dkr. sp. The only difference exist-
ing between the Japanese and European forms is the entire absence of
seg-ments with rounded at)ex, the se2:ments here beinij' more or less
pointed in all the specimens. But as there are also such forms in tlie
European plant (ex. gr., figs. 4, 7, PL XII of Schenk), so there is no

228 M. YOKOYAMA; MESOZOIC PLANTS

reason why we can not unite the two, especi-.illy when we consider that
the shape of the segments in this genns is genera]]}^ very variable.

All the specimens hitherto found in Japan belong to the form
Avith narrow leaves, that represented in tig. 14, Vl. XX being the
broadest.

Loc. — Yuasn, Kataji, Tôgodani ; numerous. K;igahara ; rare.
Tlie plant is alscj mentioned as from Hiura Ijy Xathorst.

23. Nilssonia pterophylloides m.
PI. XXII, Figs. 8-10 ?]. XXV. Fig. 7.

Leaf eloiKjatcd (jraduallii tctpcrüiij antcriorh/ and endiiuj in a terminal
segment, deeph/ and reijidarh/ segmented; segments alternate, directed slighthj
fortrard, linear, yarcdled sided, apex ohtusc or ohtusehj pointed; veins ßne,
paralled, simple, 8-10 in niimher.

Althongii this plant looks like rteroplujllum, the insertion of the
leaf on the rachis distinguishes it from the latter genus.

The segments are all regular, narrow, parallel margined, o-4 mm.
in breadth, nnd standing inclined at atjout 70° to the moderately strong
rachis. The a})ex is obtuse in those at the posterior part of the leaf
and more pointed in those at the anterior part of it, and their line of
meeting on the surface of the rachis appears to ]ia\e Ijeen zigzag, as
seen in fig. 10, ]>!. XXII.

I am acquainted with no form of the genus which ca.n be directly
compared with our species. Indeed, Zamile>< gracilis Kurr (Ueitr, z. foss.
Flora d. Juraf<:>rm. Wiirtt., p. 11, pi. I. fig. 4) and I'teroplujllum
pecten Lindlcy and Ilutton (Fossil Flora of Great Ihätain, vol. 11,
]). 10-) greati}" i-esemble the Japanese form, as far as their external
appearance is concerned, but according to Schimper (Zittel, Handbuch
der Palaeont(jlogie, Abth. Palaeophytologie, p. l2i^o) they both belong

FROM KOZÜKE, KU, AWA, AXD TOSA. 929

to Ctenoplijilluin in wliicli tlie veins are partly forked and those near
the niargii^. run unto it.

A specimen represented in fig. 7, pi. XXV is a s])lendid one from
Kataji, showiiiu' the under side of the leaf, although the veins are not
well visible. The other figures represent fragments found at Yuasa.

Loc. — Kataji, Ishiseki, Yuasa; not rare.

24. Ptilophyllum cf. eutehense Morris.

Ptilujihi/lluiit cf. ciitdu-nse.- — Natliorst, Beitr. z. mesoz. Flora Japans, p. 12, PI. IV,
fig. 8.

What X'athorst has described from Chöja under the above denomi-
nation, I have not been able to find in my collection.

Coniferse.

25. Cyparissidium (?) japonicum »/.
PI. XX, Figs. 3a, 6, Ga, 13. PI. XXIV, Fig. 4.

Branches copious. altciiuAtc. risbuj at on acute an<jlc, .slender, and cord-
like, straifjhf or verij dujlithj curved, ultiinale brandies ahout 1 nnu. in
hreadtli: leares tinJn-icaled, closehj oppressed, acute at apex, ni'tli a lonipJud-
inal ridije on their haclis.

Iij K:igah;u'a there are remains of a coniferous plant whose
branches thickly cover the faces of stones. Unfortunatelv however
they are so flatly pressed and smoothed that the minufe details of the
leaves are not well visible; l)ut in a specimen from Tanno which is
much better preserved, the imbricate leaves, as they are always partlv
covered by the two preceding ones, appear lozenge-shajjcd. The longi-
tudin;d ridge is distinct and tolerably elevated, and in external im-
pressions of le:nes, it is present as a deep groove. It is not possible to
decide as to what o-enus the plant belonu's, there beino- several o-enera
Avith a similar type of leaves. 1 have brought it therefore provisionallv

230 M. YOKOYAMA; MESOZOIC PLANTS

under the oeniis Cjiparissidium Heer on iiccoant of its great resemblance
to C. gracile Heer (Kreideflora der Arktischen Zone, p. 74, pl. XIX).
It is also not unlike some of the Arthwtaxopsis described by Fontaine
from the Potomac.

Loc. — Kagahara; in great abundance. Tan no, Fujikawa ; rare.

26. Torreya venusta 7^/.
PI. XXII, Figs. 11, 12, 12(1.

Sleiii slender ; lea ces tolerably close togetJier, stihoiiposite, direeial soiiie-
irliai fortrard, siiialL linear, straight or sliglttlg ciirred outirards, gradualhj
tapering in front into an acute apex, base alniipllg rounded ami disiiclious;
no distinct midrib.

There occur fragments of a little' conifer whose leaves measure
5-() mm. long and 1 mm Ijroad, and are directed forwards at an angle
of about ()0° to the stem. Ibey possess no midi'ib ; but in its stead, two
strong' cord-like loniiitiidinul lines were observed in one or two leaves.
Therefore it is n<:)t wliollv impi'obaljJe that we have here to deal with a
plant at least closely akin to the recent genus Torreija. The leaf-sub-
stance, however, as it appears on stone, seems not quite so firm as
might be expected in this genus. Therefore the generic determination
must be left still unsettled.

Fontaine descriljes a similar cfmifer from the Potomac under the
name of Torrega rirginica (I. c. p. 22o4, pl. CIX, fig. 8).

Loc. — Yuasa ; not frecpient.

Phyllopoda.

27. Estheria reetangula m.

PI. XXI, Fig. 13.

Carapace-calces s)nalL <iuadrate, moderatehj centncose. broader than
high with umbo terminal; tJie dorsal margin straight and parallel icilli the

FROM KOZUKE, KIT, AWA, AXD TOSA. ')^i

(Uih'rior portion of tlw veiifnil ninnjiii. flu' jwstt'rior part of flif latter ohJimie-
hj aso'irh'ini to tlic slr(i/(iht posterior u\arijiii, diid I'nrinimj at the point of
junction II rointiJe.l (onjlc: anterior inanjin truncate, parallel to tlie posterior
and at riijlit amjles to tlie dorsal cis u-ell as to tlie anterior portion of tlie
rentrai uianjin, so tliat tlie ralres assume a decidehj iptadrate shape; froiri tlie
uiuho run tu'o flat diaijojtal edijes, one to tlw postero-rottral a)i(jle and one,
n-liicli is flatter tlian tlie other, to tlic middle parr of tlie rentrai marejin^
u'hich at this point ascends haclarards ; surface with ediout tirelre sliort
concentric ridijes^ hetween ndiicli there are still finer interstitial lines, fjeneitli
7.5 nun. lieiijht 3.0 mm., thicJcncss of the left redre, 2 mm.

I possess only a single specimen of this species in the form of nn
external impression. ]\Iy figure was drawn after a clav r-ast of it.
The finr-sided shape and two obliqne edges are characters Avhich
remind ns more of the Carboniferous Genus Leaia Jones (A Mono-
graph of the Fossil Estlieria?, Appendix, p. 117, pi. I, hg. 10-21.
Palseontogr. Soc. London, vol. XY, 1(S62) than any of the hitherto
described Esther iœ.

Loc. — Yuasa.

PLATE XX.

Plate XX.

KAGAHAIIA.

Fi(j. 1. — Zamioplivllnm lîuchianniii Ett. sp.
„ 2, 3h, 4, 5, 7, ia— Porlozamites pu.silhis Yel.
,, 3a, 6, 6a, 13. — Cyparissidiuin (?) ja])oniciim Yok.
,, (S. — Onyclnopsi.s eloii^'ata GevI. sp.
„ TA— rtevis(?) 8p.
,, 10. — SpherK^pteris tenuiciila Yok.
,, 1:2, 11. — Xils.snnia sdiaiiinlinro-eiisi.s Diiiik. sp.

Yokovarna . Mr.vozoic PltifiLs-

Jour. Sc. Coll. Vol. VII. PI. XX.

AiK-tor ill l<ii>ùlctri (J)l

PLATE XXI.

Plate XXI.

YUASA.

F'kj. I. -i. — Oijycl)io|)«is eloiigala Geyl. s[).

,. 2, 2a, 3. — Splieiiopteri.s tenuiciila YoV.

,, ^, 5(<.— Glo.ssozamites parvifoJius Yok.

,, 6\ 7. — ^Pteiis (?) «p.

,, 6'-ii. — Chladoplilebi.s Xathorsti Yok.

,, 12. — l^copteri.s Geyleriaiia Xath.

,, 13. — E.stlieria rectaiigula Yok.

., 11. — XiKssonia scliaiimljurgeii!sis Dunk. ,s[).

,, lu. — Adiaiititus yua«eii«is Yok.

Yokoyafna. Mr.voxoic Plants.

Jour. Sc. Coll. Vol. VII. PI. XXI.

Aiiclor in Itipidern del .

Seikwada TakyC Japan

PLATE XXII.

Plate XXII.

YUASA.

Fig. i, 2. — Zamiophyllum Biichianum Ett. sp.
,, o. — Zamiophylluui NuLiiiumiii Nutli.
,, i. — Zamiophylluin Buchiatiiun var. angustifolia Font.
,, Ô-7. — Nilssoiiia .scliauinburgeiLsis Dunk. «p.
,, 8-10. — Nil.ssonia pterophylloicles Yok.
,, 11, 1:3, l:ài. — Torreya veim«ta Yok.

Yokovarrui. Me.S'Ozoic Plant.s-.

Jour. Sc. Coll. Vol. VII. PI. XXII.

Aiwlor ÛI hzf^ide-fn deV.

Seikwado Tök^o Japan.

PLATE XXIII.

Plate XXIII.

FUJIKAWA.

Fig. 1, la. — Pecopteris Geyleriann Nntli.
„ 5, 5. — Thyrsopteris sp.

TAXXO.

Fifj. 4, ô. — Podozamites Innceolatns Lind, et Tlntt. 8]i.

SAKAMOTO.

Fi(j. G. — Zami(^|)l\yl!iuii Uiicliiannin Etf. sj).

> h ko \ a tiui. Mr. s-ozoie PioJits .

Jour. Se. Coll. Vol. VII. PI. XXIII

. h/ili/r- III Iii/ikIiiii il.'l

Scikwadô reWô Japn

PLATE XXIV.

Plate XXIV.

FUJIKAWA.
F'uj. 1. — Pecopteris cf virginiensis Font.

«

,, 2, 3. — Pecopteris ]^)rowniann, Diink.

TANXO.

Fi(j. 4. — Cyparissidiinn (?) japoiiicum Yok.

Yo/xoya/iui . A'/r.so^o(c PhinLv

Jour. Sc. Coll. Vol. VII. PI. XXI V.

Auffar- in l<ffn'rUtn <■//■/

Seikwadô Tok_yc Japa

PLATE XXV.

Plate XXV.

RYU SE KL

Fi(j. i-i. — Niissonia Jolmstrupi Heer.
,, 5. — Zamiopliyllum Buchianum var. angustifolia Font.
5, 6. — Dicksoniopteris Xaumaimi Nath.
,, 7. — Xilssoiiia pteropliylloides Yok.
5, S-1:L — Pcxlozamites «p.
., 13^ loa. — Dicksoiiia tosaiia Yok.

Yokova/iui . M^AOTofr Phm/.y.

Jour. Sc. Coll. Vol. VII. PI. XXV

Seikwado Tôk_/G Japa

Plate XXVI.

ISHISEKI (RYOSEKl).
Zamiophyllimi Xaiimanni Nath.

yokovanui. Af^'S^ozou- Plafil.s-

Jour. Sc. Coll. Vol. VII. PI. XXVI.

Ain;/or in Uipùleni ch'l .

Sspkwadô.TSkjia.Japa

PLATE XXVIT.

Plate XXVII.

KATSEKIYA^fA.

Fig. 1, la, 3, Sa, 4 ôcd. — Pecopteris l^»rowniann Dunk.
,, ôaJi. — Z:nnio])bylIum l'urhinnum Ett. sp.

SffllUTSlirGAWA.
Fiij. 2. — Pecopteris l^)i*owiiinnn Dnuk.

Yoküvarna . Me:yozoic PUtn ts

Jour. Sc. Coli. Vol. VII PI- ^XVII.

Autzlor Ù1 LxpCclem^ deù.

Seikwado. Tokyo. Ja

PLATE XXVIII.

Plate XXVIII.

KAISEKIYAMA.

Fig. i, '2. — Zamiophyllum Buchianum Ett. sp.
„ 5, 4, 10, ii.— Chladophlebis Nathorsti Yok.
„ 5. — Pecopteris Geyleriana Nath.
,, 6. — Sphenopteri.s tenuiciila Yok.
,, 7, 7a. — Onychiopsis elegans Yok.
,, S, 9. — Zamiophyllum Bachiannm var. angnstifolia Font.

)'oko\ <iiiia . M<.yOzo('r I*hin/.y

Jour. Sc. Coll. Vol. VII PI. XXVIII.

A//rh>/- I'ft /f//?ù/('f/é t/f/ .

Seikwado-Tokj/O-Japa

On some Organic Remains from the Tertiary
Limestone near Sagara, Tötöroi.

By

Kyugaku Nishiwada.

With Plate A'A'/A'.

[In tlie spring of tlds year I took advantage of a two days' stay in Sagava, Prov.
Tötömi, to examine the Tertiary limestone developed near the town and to collect
some of the organic remains ahoanding in it. In this way I came across some
nail ip ore limestone, the occurrence of -which in Japan has not yet been
recorded. Since then I have made another short visit to this locality and have
studied the collected materials, under the supervision of Professor M. Yokoyama, in
the laboratory of the Geological Institute, College of Science, Imperial University.
Prof. Yokoyama has placed me under great obligations to him for his kind suggestions
and for the loan of Giimbers paper on the fossil Lithotlutnuwin. — Hongö, Tokyo ;
November, 1894] .

So far ns I am awavo. there are few ]iniest<~)nes in the Japanese
Tertiary tliat can <^\\'c us more g-eon-tio^^tic interest tlian tliat under
consideration.

Upwards of 3 km. noi-tli-west of the town of Sairara, there are
two limestone hills, Ivini^' one on either .side of the Ilagimngawa, a
river which, after risinii' ii' ^lie environs of Xakanishi, farther north-
west of the present site, takes its .sinuous course in a south-easterlv
direction between these hills and empties itself into the sea close to
the town. That on the right of the stream is lor-ally known as
Mekamivama and the other on the left of it as Okamivama or

2P,4 K- XTSHIWADA; ORGANIC REMAINS FROM THE

Okamid'ii, in allusion to the nrunes of the villa^>'es ]\Iekaini and
Okami. According to the Topographical Survey of the Army De-
partment of the Empire, tlie former attains an elevation of 111.3 m.
above the sea, ndiile the latter is less high, or about 60 m.

The limestone occurring in the form of such isolated hills is of a
quite uni(|ne character. A glance at the rock brings before ns a con-
geries of the C'dcareous fn'ins of some organic remains. The colour
is crenm-wliite or greyish-wliite, sometimes grev. It is pretty hard
tlirough the cryst:dlisation of some of its organic inclusions. Chemical-
ly considered it is almost free IVom im])urities, — ferric (^xide, magnesia,
etc., being present in ordy very small quantities.

As regards structure, its stratiücatii^n is f(.)r the most part in-
distinct. The Okamiyama limestone is simply massive, there being no
structure suggestive of 1)edding, except fissure-planes of which it is full
and through which nnich solution has been effected. ]\luch of the
Mekamiyama stone, hidden as it is by its covering of soil, may be in
the same state ; but on the east flank of the hi lb where extensive
quarries afford an opportunity of studying its structure, it is certainly
stratified and strikes X. '20°-o\)° E. witli a very high inclination
towards SEE. j\Ir. Xakashima' of the Imperial Geological Survey
believes that the limestone of these hills suffers from an anticlinal
folding, extending to the overlying strata, and tliat what is seen on
the eastern fiank of the ^lekamiyama shows that the l)eds di[) awav
from a central axis very steeply to the south.

During my second visit to the locality, an attempt was made to
determine the relation of the limestone to tlie other sedimentary rocks.
The evidence then obtained Ijearing upon this is, however, not quite
decisive, although the rocks ma}^ in places be observed in association.

1. Shizuoka Zufaku Chisliitsu Setsumeisho. jx 12.

TEKTIARY LlMEiSTOXE NEAR SAGARA, TÖTOMI.

23-

Aloij''' il brook, ou tlie north-western side of tlie Mekamiyanin, there is
ex[)Osed a Tertiary formation of sandstones and shales, to which
reference will be made later on, which strikes nearly x\-S. and dips to
the east at an angle of o5°. On the north-eastern side of the hill is also
seen an alternation of shale and sandstone quite similar to the others,
the strike of which is nearly XE. and the south-easterly dip very high,
^iloreover, to the west of the hill. bey<Mid a, very narrow rice-iield is
laid bare alouu' a In'ook another .similar alternation Avith uorth-westerlv
inclination. At the Okamiyama, on the other hand, any such ex-
posures as the above are concealed by the talus of limestone blocks
on all f>ides of it. There is, moreover, l)ut little outcrop of rocks
in the neighbourhood, with the exception of a very limited patch
laid l)are at Oiwa. about 150 m. east of the hill, Avhere sandstone
and shale are found in a fragmentary state in association with the
limestone. So far as my observation goe.>, it seems most pro-
bable that a series of sandstones and shales rests directly upon
the limestone, and that the latter appears sporadically ihmi
underneath that series and still vourj^'er strata, as in the hills here
under consideration.

The accompanying hgures will perhaps render the mode of
occurrence of the limestone clearer than any description.

NWW

%/>

See

Fij,'. 1. Ideal section of the Meljamiyama from SEE. to XWW. Scale 1 : 7,000.
^ limestone quarries. s. sandstone and shale,

a. rice-field. , p. much younger Tertiary.

2'S6

K. XISHIWADA; ORGANIC REMAINS FROM 1'HE

Fig. 2. A soutlieru view of the Okamiyauia or Okauiiclai, wIk illy made up of limcstoue.
Sketched Ijy the uriter, Nov. 5th, 1S'J4.

'The oru'aiji(' remains occmTHiy pleiitilully in the limestone
are but few in genera. The I'uUowing are some of the species
wliieh I ha\e Ix-en al)le to recognise :

Litholhaiiniioii.

Corah.

Millcpord.

Furdiiiiiiiù'ra.

Turbo.

I'ccd'it.
The abundance (jf the remains referaltJe to eacli of these is
generally in the (jrder slateib

Description of the Fossils.

Lithothaninium ramosissimuni Reuss.
Vl. XXIX. Fig. 1, i>, :;. and 4.

yiiliiliura raiiUKsi.stiiijKi — ^Reiiss, Natarw. Abb. v. Haidiiiger., lid. II., 1818, p. 29 ;

T. III., Figs. 10 and 11.
yiilli]>nr<( r.'iiDo.sisiviunï — F. Ungor, Donks. d. k. Aka. d. W. ia Wien, Bd. XIV, 1858.

p. 28. T. Y. Figs. 18-22.

TERTIARY LIxMEöTü.XE XEA.R «AUARA, TOTÖMl. 9;.^ 7

Lithnthdiiniiuin r(niiosi.s.siiiiu))i — C. W. Guinbe], Abli. dur k. ba^'ei". Akad. der W., IL.

Cl. Xr, Bd. I, Abth. Manchen, 1871. p. 2i. T. I.
Litltot/iainnitiin naiumi-'itiiijitiui — A. PioÜipletz, Zeits. dor dent. Geol. Ges., Bd. ■18.

1891. p. 320.

Sy^temutic knowledge of lAtJiothaiiiniuii i.s not as yet in a sati.s-
factory condition. Tlii« Ibsi^il wns lornierly relegated to a con\\ under
the various nanie.s oi' Ccllcp mi. SjioiiJik's, XtiUipura, M('L>he.si((, and .1//7-
L'pcmhy Linné, Lamarck, Laniauroux, Cuvier. Ellis, Solannder, Heu.ss,
etc., among Avhom the last author gave the name of Xidlipora nuiiosissi-
vta to the irregularly ramified, coral-like calcareous lorm, occurring in
tlie limestone of Leitha near \ ienna. Kiitzing proposed the name
of Spoiiditt'!^ stalactica for the tinv stalactic form of it. Ilaidinger'
offered tlie explanation that X raiiiu-sissiiita is a sedimentary body. In
185<S, Franz Lnger proved for the first time that A', raiiiusissiiiia is
neither an animal nor a stalactic body, but a plant. In LS72, C. \\ .
Giimbel wrote an excellent paper, entitled Die m-ijcnannUn Xiilliporcii,
etc., in the above-quoted AbliandluH(jcn, in which he end.odies sys-
tematic descriptions of the fossil species and announces that ihey are
to be distinguished from one another only by tlie relative dimensions
of the tissue-cells. Solms-Laubach, on the contrary, said as
foll<j\vs : "Jt is extremelx" difticult to distinguish the species in
the li\ing re[)resentatives ol rliis group, and it may l)e re;ulih'
conceived that the difiiculty of dealing with the fos.'sil forms is
still greatei'. We shall do well to follow Unger in this matter,
and to put them all t(3gether as Litliolhainniiim rdiiiosissiiiiHiii.'' "
Still more recently, A. lÀothpIet/, in Miuichen, accepted (ji'nnbel's

1. Berichte über die Mittheilnngen von Frenmlen d. Xultinr., Bd. IV., ISIS, p. -112 — cited iu
Uuger's iiaper.

2. F(is.<il Botany (Euglisli translation of Einleitung in die ralaeop]i<jtolof/ie). Oxford,
1891, p. i'j. I am greatly imlebted to Mr. Kenjiro Fujii, liigakushi, of the Botanical Institute
of the College of Science, for the loan of this work.

^3J^ K. NISHIW.A.DA: (»IKiAXIO REMAINS FROM THE

opiiiifjii as to ideiitiücati<3n of the fossil species, in conformity with
Avhich he deseribed 14 s[>ecies. In the determination of tlie present
species I also am inchned to lolhnv Giindjeh

Tile remains of our Lithodiaiiiiiloii play such an im[)ortant
rule in the l)uildin^- up of ihe limestone, as to warrant the de-
signation of J.ilJiotliaiiiiuoti-Liiiicsloitc (^Xiilliiion'Nlialh-). The frac-
tured surfaces have a porcellan<His aspect, and are cream-white
in colour.

The thallus is nuich ramitied into tinv stalactic forms or
shrubs, Avhich are rounded at the ends, variable in length and breadth,
measuring from 1 to .') nun. across, and having smooth surfices.
(Fig. 1, PI. XXIX). Generally it i-esembles either that of L.
hi/ssoidcH (Lamarck) Phil., a hving species in the Adriati<; Sea,
which was d('scril)ed and figured by Dr. Ilauck in his ^lecroiühjcn
Dcutschlaiuls und ( fcslerrt'icJis.^ or (he lower <;)ne of the two fii^'ures
given in Prof. Zittel's llanihucli der ValaeutilLdnijic."

I*]xamined under tlie microscope, tiie cells composing the
outer p>art of a bi'anfh of the thallus are G-8 angled (not unfre-
rpiently 5-7 angled according t(j the impei'fection of tlie slides),
as may be seen in a ti'ausverse section. A vertical section, on
the other hand, shows that the tissue-cells are mainlv of rectangular
shape and regularly arranged in layers lying one on anotlier as
concentric shells. In certain slides, sections of the cells show the
walls as either njund or sinuous ; they can then hardiv be dis-
tinguished {vom those seen in a tangential section of Solrmtpora^^

The cell-division is active. In the '' hvpothallium," so-

1. Lijipziy:. 1S85, p. -llô, Tai. II.. Fig. l.

2. IL Ahth. PahieopJii/tologif, Leipzig. 1890, p. 38.

3. A. Nicholson^aud EtheriJgo, Geol. Mart , Dec. Ill, Vol. II, p. ö'^'J.
, Geol. MiKj. Dec. III., Vol. V., p. 15.

A. Brown, „ „ Dec. IV., VoL I., p. 115 and 105.

TERTIARY LIMESTONE NEAR SAG AR A, TÖTOMI. 239

called by Areschong-, or " Markstrang-" of Solms-Laubncb, tbe cells
are divided and multiplied principally by nie:!ns of dicbotomy,
or " snbdicbotomy ", according- to ]>ornet, and sometimes tricbo-
tomy or " subtricbotomy "; Avbile in tbe " [»eritlmlliiim," so-called
bv lu)tbpletz, tbe cell-increase t'dces ])lace mostl\' by tbe process of
transverse fission or ** Qnertbeilung'."

Tn tbe slides prepared, ti-ares of tbe pores sng"g'estive of wbat
are known as " tetraspores "' are .s()metime.s seen in tlie ])eri-
tb:dlinin ; no remains of cvstocarps bave been recog'nised.

Tbe approximate dimensions of tbe ])eritbaHic-cells are
12-20 n in lengtb and 12-10 n in breadrb, wbile tbe by-
potballic-cells are of still larger .size, being n(^t unfreqnently
25 p. broad and 37-50 p. long-. Tbe former dimensions ap-
proacb nuicb more closel}' to tinsse of L. nnno^issiimiui Eeuss tban
to tbose of anv described bv otber autbors. lielyiiig npon tbis fact
onlv, and putting- aside anv point as to f )rm, as Giimbel sugg-ested, it
will not be far from tbe trutb to class our species witb L. rainosissiinuni,

Stylophora sp.

PI. XXIX. Fig. 6.

Coral remains are also abundant, Ijut all of tbem found as casts
and conse([uentIv indeterniinabje. One of tbem, bowever, mav belong-
to tbe section of ^Lidreporaria, and ])erba[)s to tlie genns Sfiilophora,
in so fir as it sliows traces of tbe fully develo|''ed six septa, etc.

Millepora sp.

Pb XXIX. Fig. 7.

Besides tbe abov^e cond remains, tbere occurs anotber coral-like
form wbich may be regarded as belonging t(j Millcpora of tbe

240 ^- NISHIWaDA ; ORGANIC EEMATXS FROM THE

HiidroconiJlt'Ncr. A part of tlie hranclies of tlie cœnostenm entirely
converted into rrystalline caleite is .shown in V\. XXIX. Its
tangential .section exliil)its (races of the ü'a.stropores and dachdopores.
A vertical section shows that tlie creani-wliite cdcitied tal)es are
intersected I)\' transver.se partitions or '• tabnla?," a structure sn^'o'estive
of what are called zo()idal tid)es, which traversed the calcareous skele-
ton of the animal, and contained the ü-astro/o(tids and dactvlozoinds.

Xo Htcratnre relatinu" to fossil MiUrpnrn hevond fhe text-hooks of
Professors Zittel and Xicholson. is accessihle t(^ nie : and it is therefore
inipossil)le at ])artition to identify thes pecies or stiidv the details of
this douhifnl form.

Foraminifera.

The microscope reveals the presence of manv of the simpler forms
of Foraminifera, in the limestone, hut few of them are well preserved.
On tliis account no o-ood .sections for examination luive been u'ot hut
f^o fir as they can he identiht;d in .sections, thev appear to belong to
GluoKjernui, XüJosariit. Miliula, Ilotalia (^?), and ^linphistegina.

Peeten sp.

For (his specimen I am indebted to ^Ivx. ]\rive Atsumi, of Okami,
who was kind enough (o su1)mit it to me for examination. Its
species ])i'oves to Ije indeterminable through imperfect presei-vafion.

Loc, Okamivama ; rare.

Turbo mekaniiensis m.

ri. XXIX. Fig. ") a and h.

Shell tu.rlfmated nv ovate-pointed ; composed of 5 vsdiorlsl
convex and separated by subcanaliculated sutures, upper two wdiors,

TEETIAEY LIMESTONE NEAR SAG AR A, TÖTÖMI. 94 1

nearly smooth, lower tliree spirnlly sciili)tnro(l with h'ra-, wliicli
number 5 on the pennltiin;ite and ll^ on tlie last wliorl, and are
generally wider than their interstitial furrows: anions- the Tirai on
the l)ody whorl the siiUstitnral is tlie Jar^-est. Apertun; indistinct,
but nearly ovate (?)

Height of tlie shell -I.Smni.

Width 4:^mm.

Spiral angle 8G°

Approximate ratio of body whorl to entire shell 70 : 100.

T. melrimii'iisis is allied to some of the living species. In the
form of the shell, the nund)er of whorls, and the liras <^n the body
vrhorl, it resembles T. arh'Nsis Montronzier LSHO (Tryon, Mtouial of
Concholofiti^ vol. X, p. lOlî IM. 45, Figs. 9(), 1)7), from which, however,
it is distinguished in its sculpture, the living one having spiral ribs
Avhich are narrower than the interstices. In the last point it coincides
with T. ayijijro^toniiis Linn, 1758 (The same bo(jk, ]>, lî)7, PI. 40,
Fig. 18 ; ri. 46, Fig, 8) ; but not in the other characters of the shell
of this species.

It is mostly found as casts, of which parts of the limestone bed,
common on the eastern flank of the Mekamiyama, are full. The
specimen figured was kindly given me by Mr. Sadahe Yagi, of Mekami«

The Tertiary f<3rmati(jn, within which the limestone makes its
appearance in a local manner, subdivides into an Upper and Lower
series.^

1. Mr. Xakashium has given these divisions the names of Upper Gigawa Tertiary and Lower
Oigawa Tertiary, and considers the former to be probably Pliocene and the latter Miocene.

242 J^- >'ISHIWADA : ORGAXIC REMAINS FROM THE

n) The Upper serie.s is of vast extent, covering' most of the
soiitliern part of tlie province of Tötömi, and is overlaid discordantly
by the not-less-widely distributed Qiiateriiary forniati(jn. It consists
of shale, sandstone, and conglomerate, all of tnficeous nature, and
yields a number of fossil shells. Amongst them are species oï Xassa
japonica Adams, Lauipnnid :o;/rt//'.s' Lamarck, Ccrilhiuw, Chcmnitzia,
Tiissou, TcWnui iiasuta Conrad. Pefricohi, Corhula, Area tjranom Linn('^,
Ostca <ji<jn^ 'riiund)erg efc,'

Judging from tliese shell remains, the Upper series may be con-
sidered as contemporaneous wifh the Pliocene Tertiary developed in
the environs of Tökvö. a detailed account of which is found in Dr.
Ijrauns' Geolorm of tlic Eiir Irons of Tölnjö.

Overlaid by the PHocene Tertiary just described, there lies b)
the Lower series occupying a very limited area. It is m:iinly made
up of dark -greyish shale and brownish or greyish sandstone — ^^just that
alternation whir-h is found overlying the limestone, as already described.
With the exception of the limestone, this series has yielded so far
no harvest of any characteristic fossils for the determination of its
geological age. According to the order of superposition, however,
tliere ca.n be no reasonable doubt that this division is older than the
Upper one, so that it will b' admitted that the limestone may belong
to some older epoch than the Plioceije.

Xow. |)utting out of account the other remains in the limestone,
we know that /.. raDiosissimuni Reuss has hitherto been found in the
Miocene Tei'tiary and in no other formation. Hence, the recurrence
of this species in tlie limestone, togetlier with the f-.u-t that the rock
is overl'dd discordantlv by the above-mentioned Pliocene strata,

1. Loc. cit. p. IG.

TERTIARY LIMESTONE SEAR SAGARA, TOTÖMI. 243

seems to suggest tliat it may l)e assigned to the Miocene, together with
the associated sandstones and shales.

PLATE XXIX.

Plaie XXIX.

Fi(j. 1. — A macroscopical view of L. ramosissiinum iu the polished surface of a block

of the Limestone.
Fiij. 2 and 3. — Vertical sections of its branches x 80.
Fiij. 4. — Transverse section of the same x 80.
Fi(/. 5 a and h. — Tinho inckainienais Nisliiwada ; nat. size.
Fi(i. 6. — Casts of Sti/lnjdtora sp., greatly enlarged.
]''i(j. 7'. — Milli'pora sp. Vertical section of part of a branch of the cœnostenra.

Ai'sTvifvaola^ .Orgfunic Ti^mazns.
5a

5l3

Jour. Sc. Coll. Vol. VII. Fl. XXIX.
1

n-'

'^S

^i^^

\

-^.N^)

>^^

.... ^

^.-r^^M^^è

5r

\f5^.

«adô.Tôlyô.Japan .

Mercury and Bismuth Hypophosphites,

by

Seihachi Hada, Rigakushi.

College of Science, Imperial university.

Probably because of H. Rose's welJ-knowii observation in 182J7
of the reduction of mercuric chloride to mercurous chloride and of
rhi.s to metallic mercury by a solution of hypophosphorous acid, no
expectation of success has led to any attempts to prepare a mercury
hypophosphite. At the suggestion of Dr. Divers, F.R.8., to whom T
am much indebted for advice, f have tried the use of the nitrates ol
mercurv, and have therebv obtained the salt which I shall now
describe.

Mercurons nitrate ht^pophosphite.

This double salt is the only mercury derivative of hypophosphor-
ous acid I have been able to produce. It is pi-t^cipitated from a
solution not too dilute, and almost as free from acid as possible, ot
either mercuric or mercurous nitrate by a solution ot potassium or
barium hypO|)hosphite used in quantity small enough to leave some of
the mercury nitrate in solution. It can not be got by adding the
mercurv nitrale to the hypophosphite, or when .too much of the latter
salt is added to the mercury nitrate, because, in either case, it is at
once decomposed.

Since the formation of the salt from mercuric nitrate necessarily
involves the oxidation and waste of much of the hypophosphite
added, and also yields a mother-hquor very active on the precipitate,
mercurous nitrate is the proper substance to select in preparing it.

246

s. HADA.

Potassium hypophospliite is also preferable to the barium salt, for
when the latter is used the precipitate is liable to contain barium,
apparently as nitrate.

Mercurous nitrate, which must be frw fi-om nitrous acid, is best
prepared by dissolving- mercuric oxide to saturation in nitric acid and
shakin«' the solution violently with metallic mercury for a few
minutes, for in this way the mercuric salt is quickly and completely
changed to jnercunuis salt.

As the wbite precipitate obtained by adding the potassium hy-
pophosphite to the excess of mereuroas nitrate is slowly decomposed
bv its mother- liquor, it must be (piiekly ;-emoved and drainerl on a
tile without pi-evions washing.

Mercurous nitrate hypophosj)hite is unstable, but when dry it
onlv slowlv decomposes, becoming grey in the course of some days.
It is a white micaceous powder, slightly soluble in water, I)y which it
is soon decomposed with separation of mercury. Its composition is
expressed by the formula —

HgH,?02,HgN03,H20
It loses its water in a vaeiuun desiccator with scarcely any further
decomposition for some time. Heated, it turns slightly grey above
90^ and exjDlodes a little above lOO* yielding mercury and nitrous
vapours. Any quantity of it can be exploded at the common tem-
])erature by touching it with a hot wire.

With hydrochloric acid, it first gives mercurous chloi'ide and then
metallic mercury. With cold dilute nitric acid it yields metallic
inercury, while hot strong nitric acid dissolves it completely with
escape of nitrous lûmes. Sodium chloride converts it into mercurous
chloride and sodium hypophosphite, which only very slowly react to
give metallic mercury. Potassium hydroxide blackens it by forma-
tion pi'obably of mercurous oxide at first.

MERCURY AND BISMUTH HTPOPHOSPHITES. 247

The mercury and phosphorus in the .sait were determined by
dissolving it in nitric acid, evaporating- to drvnes.s. dissolving in
hydrochloric Mcid. precipitating mercury by hydrogen sulphide, and
phosphoric acid by magnesia mixture. The mercuric sulphide was
freed from any co-precipitated sulphur, and dried at 105-110°. The
nitric acid was estimated by treating the salt with strong sulphuric
acid in Lunge's nitrometer. Loss of weight in the desiccator served
for the water determination. The analytical results which follow
refer to three separate preparations of the salt.

Calc.

I.
Mercury 78-39 78-35

Nitrogen ::^-57 :?'78

Phosphorus 5*^8 5*56

Water 3-30 3-01

This salt is of interest ;is a doulile salt of univalent or quasi-
univalent mercury, since it pcjints to mercurous salts being salts ol the
radical (HgJ' rather than of (Hg)'.

A salt has been descriloed by H. Rose (Poyy. Ann., 40, 76.), as
produced by reaction between mercuric nitrate and phosphine and to
which he has oriven the formula — -

o

l^Hg;,3[Hg"(N03)„Hg"0]
I'his explosive body* would seem to be related to the salt 1 have
described, for it is not remote in composition from 3Hg'NUs,Hg'H.jP0j,
and is more probably a mercurous than a mercuric salt.

Bismuth Hypophosphite.
How it comes thai this salt has hitherto escaped notice it is not

* It was açfain examined by Aschau in 1885 (Chem. Zeit.) but not quantitatively.

Found.

II.

III.

73-01

73-04

2-76

2-82

5-25

5-50

2-45

—

248

s. HADA.

easv to understand. It is prepared by mixinq; a solution of bismuth
nitrate, free from any unnecessary excess of nitric acid, with barium
or potassium hypophosphite, avoiding excess of the bismuth nitrate, in
which it is soluble. Vhe bismuth hypophosphite precipitates as a
white crystaJiine powder, slowly decomposed by its mother- liquor, and
soluble in bismuth nil rate solution. Filtered otf and dried on a
porous tile, it can be preserved for days unchanged.

The analysis of this salt was carried out much in the same way as
that of the mercury salt, except that the bismuth sulphide was dissolved
in nitric acid and the solution precipitated by ammonium ciirbonate as
usual. Water was determined as loss in the vncuuni desiccator. In
the following table the calculation is for Bi(H2FU2)3,H20, which
therefore expresses the compqsition of the salt : —

Calc. Found.

I. II. 111.

Bismuth 19-41 49*49 49-40 48-77

Phosphorus 22-09 21-67 21-34 21-19

Water 4-27 3-11 3-26 3-39

Bismuth hvpophospViite decomposes very readily by heat, becom-
ing black and giving oif phosphine at temperatures only a little above
100°. At a stronger heat metallic globules and bismuth phosphate
are obtained. The globules washed with dilute hydrogen chloride to
clean.se them from adhering phosphate and then dissolved in nitric acid
prove to be metalli(^ bismuth free from phosphorus. As about two-
thirds of the bismuth is obtained as mt-tal, the decomposition of the
bismuth hypophosphite by heat may be expressed by the equation —

3Bi(H2PO,)3 = :^Bi + Bi(P03)3 + 6P+90H,
This hvpophos{)hite is noticeable for yielding metal instead of phos-
phide, but this fact is in accordance with the experience of Berzelius
that bismuth phosphide fully decomposes when heated.

The Acid Sulphate of Hydroxylamine.
Edward Divers, IVl, D., F. R. S. Prof.

Imperial University.

It its .somewhat reirmrkable that although several hydrochhjrides
of hydroxylamine ha\e })eeD described by Lo.ssen. the acid .sulphate
.seems never to have been obtained.

It is well known that if more .sulphuric acid is }>re.sent in an
aqueous solution of hydroxylamine tlian is suthcient to constitute the
normal salt, the addition of alcohol will cause this and not the acid
.salt to crystallise out, just as when added to acid ammonium sulphate
it will precipitate the normal sulphate. Without this addition of
alcohol, a too-acid .solution of hydroxylamine sulphate often refuses to
deposit anything. By attention, however, to a tew details, it can be
brouoht to yield crystals of the acid sulphate.

Solid hydroxylamine hydrochloride is to be tre;ited with, as near
as may be. the quantity of sulphuric acid calculated to form the acid
salt, (NH3Ü)HS0^. Ihc mixing is effected in a disli sufficiently large
to avoid loss by frothing over, and this is heated for some houi-s <jn
the water bath until all hydrochloric acid has been expelled. The re-
sulting clear solution becomes viscid when cold. It refuses to yield
the normal sulphate when a particle of this salt is dropped on its
surface, and slowdy dissolves it. But left to stand uncovered in a drv
cold atmosphere, and the vessel occasionally moved about, crystallisa-
tion suddenly sets in and the solution becomes traversed bv long})risms

250

E. DIVERS.

which almost fill it. Left in a desiccator for a couple of days more,
it becomes a translucent cake of damp crystals. The crystals are
very deliquescent and, after crushing and pressure between porous
tiles, yield results on analysis which prove them to be the acid sulphate
of hydroxylamine.

The analysis of the salt was effected by titrating it with sodium
hydroxide, with methyl orange as indicator, since to this the normal
sulphate is neutral. The sulphuric acid vsas weighed as barium
sulphate.

Calc Pound

Hydroxylamine 25'!^ 24'0'2

Sulphuric acid 74-81 72-68

ÛmïO 96-70

Decomposition of Sulphates by Ammonium Chlo-
ride in Analysis according to Fresenius.

By

Masumi Chikashige, Ri'éakushi,

College of Science, [raperial Uuiversity.

The accuracv in all respects of Fresenius' standard treatises on
analysis is usually so unimpeachable, that it seems proper to call
attention to a misleadino;- statement contained in a footnote to § 153
A of the seventh and latest English edition of the ( J udutitdthx Analysis,
and also to be found in the earlier editions of this work (4th, oth, Hth).
Having given in the text the method of separating fr<mi magnesium
and alkali salts the small quantity of barium left in solution by
ammonium carbonate, namely, by adding three or tour drops of dilute
sulphuric acid, Fresenius states in the footnote to it that the gentle
ignition there directed t^) be made in order to expel ammonium salts,
will also effect the removal of any small quantity of sulphuric acid
which may remain after ))recipitating the bai'iuni. It is this state-
ment which needs correction.

Even in § 6<S, ((. and § 74. a, of the same work we find enough
to cause us to doubt the accuracy of the later statement, for it is there
mentioned that magnesium sulphat^e is not decomposed by igniting it
with ammonium ohloride. arid, on the authority of Rose, that potas-
sium and sodium sulphates, which are decomposable by this treatment,
need for it to be effective its rejxiated application at a red|heat inducing
effervescence. Obviouslv, such ignition as this is not that gentle

252 ^- CHIKASHIGE.

ignition directed to be used in § 153, and would entnil seriou« loss of
alkali chlorides by spirting and volatilisation. But to place the in-
accuracy of the statement in the footnote beyond doubt, 1 have made
a few simple trials of the method.

Magnesia, 0.5 gram, was dissohed in a little hydrochloric acid ;
to the solution were added two dro))s of dilute sulphuric acid (1 to 10
Witter by volume), a solution of about two grams of ammonium
chloride, and ammonia in sniall excess ; and the whole w;is then
evaporated to dryness ;ind all ammonium salts expelled at a barely
red heat. Again, a solution of two grams of ammonium chloride was
added, and the evaj)(~)ralion and ignition re])eated. The I'csidue was
dissolved in dilute hydrochloric acid and mixed with barium chloride,
which gave a precipitate. On comparing this with that tlu'own down
by barium chloride from two drops of the same dilute sulphuric acid
in about the same volume of water, the ignited salts wei'e tbund to
have lost but very little, if any, of their sulphuric acid. The ex].)eri-
ment was repeated three times with fresh magnesium diloride. and the
same results obtained.

Similar experiments were made with sodiiun chloride, and no
Ijettei" removal of the sulphuric acid effected than when magnesium
chloride was taken.

It is thus quite evident that other steps must be taken to i-emove
sui])huric acid before it is allowable to calculate the weighed alkali
s:dts as chlorides or to resort to ways of separating magnesi;i from tlie
alkalis which require the absence of sulphates.

Ewart Johnstone's way to prepare
Nitric Oxide.

By
IVIasumi Chikashige, Rigakushi,

College of Science, Imperial University.

In IHH'2 D. Ewnrt Johnstone nnnonnced in the Chemical News,
45, 159, that cobalt nitrate and potassiiun rliiocvanate heated together
readily yield nitric oxide. Exce])t by Schertel, in the Befcrate of the
Berichte of tlie (h'niKDt Chemical Socirtji, thi« announcement seems to
have been received AviHiont criticism, and Michadis inserted it in his
edition of Graliam-( )rt()'.s J)tnr(j(inic Chnnistrii amon,u- the methods of
preparing nitric oxide.

The results of mv own testing of the method oblige me to con-
clude that Johnstone is altogether wrong. We are directed by him to
mix four parts of a solution of potassium thiocyanate with one part of
a solution of cobaU nitrate, such as are ordinarily in use in the
laboratory, and genrlv heat the mixture, when nitric oxide will be
copiously evolved. An equation is given of the action in which
four molecules of the potassium salt and one of the collait salt
appear, and since the quantities to be taken are so indefinitely
set down, I started my experiments wnth these proportions of the
salts, both practically in a state of purity. The result ])roved
that, in these proportions, as well as in many others which were
tried, whether the s<:)lutions are dilute (jr concentrated, only gently
heated or freely boiling, nitric oxide is not formed at all by

254 » ^- CHIKASHIGE.

them. Beyond the well known fact that the mixed solutions are
intensely green, no sign of any action was ohserved. When the solid
salts, a little damp, were heated together, nlso as recommended by
Johnstone, watery fusion occurred and the wnter boiled off without
o'as beinof o-enerated. When, however, the water being' Sfone, the
residue got much hotter, there occurred, as might have been an-
ticipated, an explosive reaction in which torrents of gases escaped.
These gases, collected over water and gradually mixed with oxygen,
proved to consist of nitric oxide to the extent of about two-thirds
of their volume, the rest being prin('i])ally nitrogen. Carbon dioxide
and ammonia were also freely given off. and condensed together to
form a sublimate, and a cloud, and a sohition in the trough-water, of
ammonium carbonate or carbamate. The residue smelt strongly of
ammonia and was black from the presence of cobalt sulphide, but did
not contain the sulphur and the carbon which eJohnstone supposed are
formed, at least not in quantities which I (^ould detect.

It is a matter of common experience that potassiiun thiocyanate
boiled with dilute nitric acid is decomposed with evolution of nitric
oxide and other gases, being partly oxidised and partly converted into
the insoluble yello\^' perthiocyanogen. N<3w, only a small cjuantity
or, rather, only a weak concentration, of nitric acid is needed for this
reaction, and the presence or absence of cobalt nitrate makes, I find,
no difference. Pi'obably, therefore, Johnstone's laboratory solution of
the latter salt contained nitric acid in some quantity, as Schertel,
indeed, has suggested.

The Acidimetry of Hydrogen Fluoride.

By
Tamemasa Haga, F.C.S., Rigakuhakushi, Asst. Prof.

and

Yükichi Osaka, Rigakushi,

Imperial University.

So far as we can find out, the acidimetry oï hydrogen fluoride
has not yet been particularly investigated. But it is well known that
as an acid it stands apart from hydrogen chloride and other strong
mineral acids. For while it even surpasses sulphuric acid in the
intensity of its reaction with water and many organic substances, it
yet shows such mild acidic characters, that its 'avidity' number places
it in this relation amonof veg-etable acids. Further, it not only decom-
poses the oxides oï some metalloids, such as boron, silicon, phosphorus,
and sulphur, forming fluorides possessing some degree of stability in
presence of water, but also gives with potash, soda, and ammonia,
salts which are alkaline to htnius.

To investigate tlie subject we have experimented with the follow-
ing common indicators of neutralisation of acids by bases : litmus,
iacmoid, methyl orange, j)henacetolin, Phenolphthalein, rosolic acid,
cochineal, brazil wood, and turmeric paper. These indicators have been
prepared and used in the usual way, for the most part as described
in ISutton's Volumetnc Änaljisis. In order to ascertain what personal
difference there might be in the estimation of the particular shades of
colour which indicated neutralisation we wf)rked separately, and upon

256 1'- HAÜA, AND Y. OSAKA.

materials aJl prepared by each for his own use. The first and last of
the tables appended contain the results obtained bv Ha^a, and the
second those obtained by Osaka.

For titration in the experiments recorded in Tal3le I and II,
decinormai solutions, in the experiments given in Table III twice
decinormai solutions oi' potassiiun and sodium hydroxides and of
ammonia were taken. I'hey were almost pure, containing only the
merest traces of si lien and alumina, but as these and also carbon
dioxide when present in noticeable quantity affect the indications of
some of the colour reagents, we were careful to determine for each
indicator the exact titre of the alkali solution in terms of a decinor-
mai sulphuric acid which hud been standardised gravi metrically by
barium chloride, thus rendering ourselves independent of the effects
of any impurities present.

The hydrofluoric acid examined was purified in the following
way :— Commercial ■" piu-e " hydrofluoric acid solution was distilled
in a platinum retort after adding a lew drops of a saturated solution
of potassium permanganate and a little potassium hydroxide. The
distillation went on at about loO°C. The product was found to be
free from hydr<jchloric acid and other foreign matters. Silica was
sought for by Jorgensen's chloropurpureocobaltic chloride test, and
not found. Before piu-itication the acid gave a small qufintity of
precipitate on standing with this reagent tor three days, but the
distilled acid «^ave none after standinü" for a week.

The purified acid was diluted and preserved for use in a gutta-
percha bottle which had for years been holding the acid before its
final puritication. For each titration, a portion was weighed off' in a
platinum crucible with well-fitting lid, and was then washed into a
large platinum dish in which the neutralisation was effected, except
that in some (.-ises the results recorded in Table I w^ere obtained bv

THE ACIDIMETRT OF HTDROöEN FLUORIDE. 257

transferring- the solution, when almost neutral, to a glass vessel, in
order to observe the shade of colour more accurately.

The strength of the acid was determined irra vi metrically by
mixing in a platinum crucible a weighed quantity of the solution
with excess of slaked lime made each time from a weighed quantity of
precipitated calcium carbonate, letting the mixtiu'e stand for a night,
drying, and then igniting till the weight became constant. The
solution of the acid used for the trials recorded in Table I was also
assayed by digesting it with precipitated and finely divided silica,
and then igniting with a little sulphuric acid. The results of the
two methods agreed well ; they are given in the table. That used
in trials recorded in Table III was assayed also by gravimetrically
estimating the acid as calcium fluoride. The silica method gave in
this case a low result and was rejected.

Jjitinna as indicator. Only historically and because of its universal
employment f<jr testing neutrality do we give consideration first to
litmus used in titrating hydrogen fluoride. On the addition of potash
or soda to dilute solution of liydrofluoric acid coloured by litmus, the
red colour rapidly deepens to violet and by the time «:)ne molecule of
alkali has been added the litmus has become an almost pure blue, that
is, blue almost free from any violet tinge. In fact the eifect upon
litmus in the saturation of hydrogen fluoride with potassium hydrox-
ide is much like that reversed of the saturati(jn ol jjotassium carbonate
by sulphuric acid. However, by practice and ])v keeping before one a
second vessel of water coloured by litmus of the right tint, — full bine,
as it would be called — it is possible to titrate hydrofluoric acid by
means of litmus. But its use is not recommended.

Blue litmus paper reddened with a solution of potassium fluoride
containing a. small quantity oi hydrogen fluoride becomes blue again
when left ex|)osed to the air. If the paper be wetted soon after it has

258 1'- H AG A, AND Y.JiÖSAKA.

become blue the red colour i« generally restored. If, however, it is
let dry thoroughly, subsequent moistening with water generally fails
to bring back the red colour. The colouring matter seems to be
modified by drying up Avith the fiuoride. Dissolved litmus added to
such a solution is also coloured red when much water is present, but
becomes more and more blue on evaporating the solution, and when
it has become sufficiently concentrated the greater part of the colouring
matter separates out as a blue poMder. This change to blue is not due
to loss of hydrofluoric acid, for on addition of water the colouring
matter dissolves a^ain oiving a red-coloured solution as before. A
solution in which litmus paper was turned apparently permanently red
and only on drying became neutral or very faintly alkaline, was in one
trial fcnmd to c<:)rrespond very nearly to HK3F4. Litmus paper is
known to be reddened by monopotassium orthophosphate in solution
and to become bkie again when dried, from which it may Ije inferred
that hydrogen fluoride, like hydrogen phosphate, is a poly basic acid,
H,F, or H3F, or H,F,.

Ijücuioid solution behaves essentially like Htmus, the diflerence
being that nmch less alkali is required U) produce a bluish violet
colour in a solution cohjured by lacmoid than in oue coloured by
litmus. By titrating to a distinct blue, good results may be got with
it, but it is not a desirable indicator. Lacmoid paper behaves also
like litmus paper

Phenacetolin changes in colour somewhat gradually when near the
neutrality point, but by titrating to pure purple or rose-violet free from
any tinge of yellow it may be used successfully.

Methfil oramje is quite useless for the ordinary titration of hydrogen
fluoride, although it seems to find neutrality in KoHF;.. The colour
changes are very indefinite.

PheHolplifhaleÏH is most satisfactory as an indicator for hydrogen

THE ACIDIMETEY OF HYDROGEN FLUORIDE. 259

fluoride, giving a. very sharp colour change at the point of neutraHty.
It cannot of course he used with aTiimonia.

Rosolic acid is ahnost equal to Phenolphthalein, i)ut not quite so
sharp in its indication. It can be used with ammonia as titrating
agent, although ammonia is a little slower than the fixed alkalis in
atfectinof this and other colours. In the case of this indicator and
Phenolphthalein the clutnge of colour is well adapted for being ob-
served in platinum vessels.

Cochineal and hrazil irood behave alike. Either becomes violet
gradually and indefinitely before the acid is neutralised, but finally it
turns bluish \'iolet. Tliis change of colour is fairly shar]) and is
satisfactory for indicating neutrality. lîrazil-wood paper becomes
bluish violet before all the acid is neutralised and is therefore unsuit-
able for use.

Turmeric paper is satisfactory but not quite as sensitive as some
of the other indicators and near the finishing point responds slowly.

If will be seen that, as in the case of the ordinary vegetable acids,
the best indicator is ])henolphthaleïn or. when ammonia is the titrating
agent, rosolic acid.

We have yet to mention that all these indicators give satis-
factory results only when the alkali solution is almost pure. The use
of an impure alkali solution or of a solution which has been kept long
enough in glass vessels to have taken u|) silica, carbonic acid, and
other impurities in appreciable quantity is liable to give ill-defined
and generally too-low results.

In conclusion, we liave to return our thanks to Dr. Divers for
his valuable suggestions in working out this paper.

260

T. HAGA, AND Y. OSAKA.

TABLE I.

Strength of the hydrofluoric acid fuj by lime method, 279%, thj by siUca
method, 278%, mean 2-785%.

Of this sohition from 1-2 to 2-9 grams were taken for eacli determination,
requiring from 10 to 39 ces. of decinorraal alkah.

Titndinn tritlt l'otasmim liydroxide.

Indicator .

< "olour when
neutralised.

Ces. ^ alkali

required by

one gram of

solution.

Mean.

Per cent, of
acid found
instead of

2-785.

Rosolic acid.

distinctly red.

14-07
14-06

14-065

2-825

Phenolphthalein.

distinctly pink.

14-16
14-05

14-105

2-821

Litmus.

pure litmus bine.

14-11
13-80

18-955

2-791

Brazil wood.

violet.

14-11
14-08

14-07

2-814

Phenacetolin.

purple.

14-22
14-10

14-16

2-832

Lacmoîd.

pin-e lacmoid blue.

18-99
14-10

14-045

2-809

THE ACIIMMK'I'KV <>F HYDRtxJIKX FLUOIUDK.

261

Titrai mil irilli SDiliinii h iiilrn.i-i'li'

Tn(lii':it'>r.

< 'olour win 'a
ncntvMliscil.

t 'cs- ^ alkali

required by

one o-1-a.m of

solution.

Mean.

Per cent, of
acid found
instead of

2-785.

Rosolic acid.

distinctly fed.

18-90
18-86

18-S,S

2-776

Phenolphthalein.

distinctly pink.

14-00
18-90

18-95

2-790 ,

Litmus.

pure litmus blue.

1411
18-90

14-01

2-802

Brazil wood.

violet.

l;i-s.-,
1401

18-98

2-786

Phenacetolin.

purple.

1404
18-92

18-9.S

2-796

Lacmoid,

pure lacmoid blue.

18-99
18-99

18-99

•2-79S

Titration icith Aiiniionia.

Rosolic acid.

distinctly red.

14-70
18-80
14-10

14-20

2-84

Litmus.

pui-e litmus blue.

13-62
18-68

18-66

2-782

Brazil wood.

violet.

18-85
14-20

14-025

2-805

Phenacetolin.

purple.

18-95
18-70

18-825

2-765

Lacmoid

pure lacmoid blue.

1 8-60
18-50

18-55

2-710

262

T. HAOA, AND Y. OSAKA.

TABLE II.

Strengt! 1 of tlio hydroünoric acid by tlie lime method. (1) 029%, (2) (}-ài%,
(S) 0BO°À, mean 6-82?^. Of this solution from 07457 gram to 2-l(;22 grams were
taken for each experiment, requiring from 23-8 to 68-7 cos. of deeinormal alkali.

Titratio)! irifli Potnamivi liiidro.riile.

Indicator.

Colour wh^-n
nf^utralised.

N

( 'cs. -jj- alkali

required by

one gram of

solution.

Mean.

Per cent, of

acid found

instead of

6-82.

Litmus.

distinctly l')]ue.

82 -2«
82-42

82-85

6-47

Litmus.

faintly l^lne.

81-84
81-47

81-41

6-28

Rosolic acid.

distinctly red.

81-65
81-49

81 -.57

6-31

Phenacetolin.

purple.

82-81
81-78

82-02

6-40

Phenacetolin.

faintly violet.

81-94

80-98

81-46

6-29

Phenolphthalein.

just pink.

81-89
81-57

81-48

6-80

Cochineal.

violet.

81-14
81 -06

81-10

6-22

Cochineal.

faintly violet.

80-79
80-79

80-79

6-16

Lacmoid.

distinctly blue.

81-01
80-84

80-93

6-19

THE ACIDIMETRY OF HYDROGEN FLUORIDE.

263

Titration with Sodium hydroxide.

Ces. -j^ alkali

Per cent, of

Indicator.

Colour when
neutralised.

required by
one gram of

solution.

Mean.

acid found
instead of

6-82.

Litmus.

distinctly blue.

31-93
32-24

32-09

6-42

Rosolic acid.

distinctly red.

31-87
31.93

31-90

6-38

Phenolphthalein.

just pink

31-92
31-86

31-89

6-37

Cochineal.

violet.

30-07
29-95

30-01

6-00

Phenacetolin.

purple.

30-52

30-6C.

30-59

6-11

Phenacetolin.

faintly violet.

•29-71
29-64

•29-68

5-94

Lacmoid.

distinctly blue.

30-07
•29-8.S

•29-98

6-00

'2iM

T. 1JA(4A. AX1> Y. ÖWAKA

Titratiofi iritli A)miionia.

Ocs. ^ alkali

Per cent, of

ludicatur.

neuLmlised.

r(>qiiired liy

one ^-raui of

solution.

Meaii.

acid found
instead of

6-32.

Litmus.

distiijctly blue.

32-37
32-80

32.62

6 -.52

Litmus.

violet slightly bine.

31-16
31-71

31-44

6-2'J

Rosolic acid.

distinctly red.

32 -UU
32 -().■)

32-03

6-41

Rosolic acid.

faintly red.

31-74
31-82

3l-7cS

6-35

Phenacetolln.

purple.

31-54

31-73

31-63

6-33

Phenacetolin.

faintly vicjlet.

31-17

;il-ül

31 -OH

(J-22

Cochineal.

violet.

31-26
3l-6<»

31.48

6-30

Cochineal.

faintly violet.

-30-77
30-85

30-81

C)-!«*)

Lacmoid.

distinctly blue.

31-13
31-12

31-125

6-22

I'HE ACIDIMETIIV OF HVIMKXIEX FLUORIDE.

265

Table III.

Streii.^'tli of liydrotiuorie acid ( aj by lime metliud, •24-98%'', H, j by .silica luotliod.
23"y8>, (c) by cak-iuiu tinoride method, 2511%; lueun of ('a j and Uj. 2oü2^.

Of tlii.s solution from 1-488 to -7457 gnmis were taken foi- eacb detevininatioii,
i-e(juii-ini,f froni !)H-28 to 45-7(5 ct-.s. of twice-deeinornial alkali.

Tit rat ion tritli Potassiiiin hi/dro.ciilc.

1 u< lica tor.

Rosoiic acid.

Phenolphthalein, distinctly pin

Colour wiieu
ueutriilised.

distinctly i^'ed.

Litmus.

Brazil wood.

Phenacetolin.

Cochineal.

Lacmoid.

Turmeric paper.

Ces- -^ alkali

required by

one giii'ii of

solution.

pure litmus blue

violet.

purpk

violet.

pure lacmoid blue.

reddish brown.

(■)|-i)()

(ri-bO
G-2-32

61-28
bl-54

(31-80
(jl-7(i

6-2-(J(3
(5 Ml»

Ü2-Ö0
6l-4i)

(jl-30
(31-li*

61-88
62-01

(■)2-().'3

62-49

6i-885

61-7N

61-925

(•)2-i4.-)

61-24:)

61-92

Per cent, of
acid found
instead of

25-02.

24-82

25-00

24-55

24-71

24-7-

24-8(J

-24-50

•24-77

i>6n

T. HAGA .AND Y. OSAK.4

TitratiDN /nth Sodimii Inidroxide.

Indicator.

Colour when
neutralised.

Cos. -3- alkali

required by

one gram of

sohition.

Mean.

t'tT cent, of
acid found
instead of

25-02.

Rosolic acid.

distinctly red.

61-81
62-08

61-92

-24-77

Phenolphthalein.

distinctly pink.

62-52
62-48

62-50

25-00

Litmus.

pure litmus blue.

61-86
61-64

61 -.50

24-60

Brazil wood.

violet.

62-78
62-28

62-48

'24-y'J

Phenacetolin.

purple.

62-72
61 -.52

62-12

24-85

Cochineal.

violet.

61-68
61-58

61-605

24-64

Lacmoid.

pure lacmoid blue.

61-10
(;l-82

61-21

24-48

Turmeric paper.

reddish brown.

62-70

62-80

62-75

25-10

THE ACIDIMETRY OF HYDROGEX FLUORIDE.

i>67

Titration with Aiiinioina.

Indicator.

Colour wlien
neutralised.

("es. -^ alk;ili

required 1>y

one g-ram of

solution.

Mean.

Per L-ent. of

acid found

instead of

25-02.

Rosolic Acid.

i

' distinctly red.

62-10
62-70

(•)2-40

24-96

Litmus.

pure litmus blue.

61-23
61-20

6l-2lr,

24-49

Brazil wood.

violet.

61-28
61-83

61 •2s

24-51

Phenacelolin.

purple.

62-66
61-19

61-925

24-77

Cochineal.

violet.

6110
61 -.52

61-81

24-52

Lacmoid.

pure lacmoid blue.

61-08
60-99

61-01

24-40

Turmeric paper.

reddish brown.

62-89
62-84

61-615

25-05

On the Poisonous Action of Alcohols
upon Different Organisms.

By
M. Tsukamoto. Nogakushi.

The physiological action of different alcohols has been investigated
by varions anthoi's, hut the conclusions reached haAe not been always
concordant. According to Dogiel,^ for instance, methylic alcohol is
less poisonous tlian ethylic alcohol, while Diijardin-Beaumetz and
Andigé^ state just the reverse as the result of their experiments. A.
Schneegans and -I. v. Mering" conclude that the primary alcohols
are less narcotic than the secondary ones, and tVie latter less so than
the tertiary, but this does not agree with the results of the experiments
made by W. Gibbs and E. d. Reichert.^ Xevertlieless it may be
concluded from many different experiments that on the wliole the
toxic action of alcohols in the methylic series runs parallel with the
inci'ease in the number of carbon atcmis contained in their molecule.
However, the subjects serving for these experiments were chiefly
warm-blooded animals, and very rarely tlie lower forms of life. It
seeuied, tlieretore, of S(jme interest to compare in this respect some
représentât trrs of all kinds uj licing organlsiiis. Of sjiecial interest also it

1. Pfliip. Arch., 8, liO.") : or Ampric. Chom. Jmtr., 13, 870.

2. Jour. Chem. Soc. Loud., 30, 539, from Compt. n'nd., 83, 80-82. In the Americ. Chem.
Jour., 13, 870, and Jour. Chem. Soc. Load.. 60, 1393, they are said to have found that
" methylic alcohol was /«'.•?.< poisonous than ethylic alcohol, but it is evident from the abstract in
J. Ch. Soc. Lond., 30, 539 that they have been misquoted.

3. Chem. Centrnlb., 1892. II., 367.

4. Americ. Chem. Jour., 13, 370.

270 M TSUKAMOTO.

seenierl to be to compare together the effects of methylic and ethylic
alcohols. I have examined altogether the action of nine alcohols :
methylic, ethylic, normal and iso-pvopylic, normal, iso-. and tertiary
bntylic, normal amylic alcohol, and, finally, allylic alcohol (CH :
CH.CH2. OH). These alcohols were dilnted with distilled water' to
the extent stated in connection with each exj)eriment, the percentage
triven being- that by volume.

Action of the Alcohols upon Lower Vertebrate Animals.

Foï" the experiments were chosen tadpoles of Bitfo nilgaris, Laui-.
in such a. stage of development that tlie hind legs had made their
appearance. Three individuals were put into 50 c.c. of the alcoholic
solution of a certain strength. (Jontrol experimeiits with plain water
were made in every case. In 0.1 'Yo solutions they became mc^tionless
after 1 '/.-^ hours in allylic alcohol ; after 10-2") minutes in amylic
alcohol ; and after one hour in butylic alcohol ; while those in the
other 6 alcohols of the same dilution were apparently not injured.
After 24 hours those in the butylic alcohol had almost entirelv re-
covered, but those in the amvlic and allylic alcohols were dead. The
tadpoles in the other 6 alcohols were alive aller 10 days.-

VVith methylic and ethylic alcohol the ex]ierinients were still
further extended. In O.H'Y,,, 0..^"/,,. 0.7"/,,. and 1 7„ solutions the
intensity of the narcotic action increased with rlie clegree of the
concentration, as was to be expected, but in all cases the action of
ethylic alcohol was stronger than tliat of methylic alcohol. Further,
in 1.5 "/o ethylic alcohol all the tadpoles showed great stu])or, while
in methylic alcohol of the same strength one only of the three was

1. <Jnly in those cases whoi-e infusoria served for the experiment, was ordinary well
water used.

2. In all cases some Spirofiip-a threads were added as food for the animals.

POISOXOUS ACTION OF ALCOHOLS UPON DIFFKREXT ORGANISMS. 271

rendered insensible, the other two were still tnoving after one hour.
After 24 hours, however, all of them recovered, those in ethyhc
alcohol beinir much weakened, while those in niethylic alcohol were
ahnost uormah Tadpoles in 'l^j^ solution of ethyhc alcohol ceased to
move in K) minutes and after 24 hours two of them died, while :i
third, although aj>|)arently recovered, died 40 hours later. A 2'^/,, s(jlu-
tion of methvlic nlcohol ])r<n^ed to be much wenker in acti<m upon them
than ethylic alcoliol. All were ap]>;n-ently insensible in fnmi 20-50
minutes, but onlv one of them died ;ifter 24 hours, while the other
two recovered, and were still alive 40 hours after. In 2.5 ^j^ solutions,
however, they were insensible in Ironi 4- LS minutes, and died soon
afterwards.

In furthei" experiments with propylic alcohols (norninl and iso)
we found that : — in 0.5 ^'/,j solutions the tadpoles stopped their motion
within 40 minutes in the normal and within 'M) minutes in the iso-
propylic alcohol, and after 24 hoiu's two of them recovered in each
case, while the others died. In 0.7 ^/^ solutions motion ceased after
17— oO minutes immersion in the normal and after 9-20 minutes in
the iso-alcohol. After 24 hours, however, although those in the
normal alcohol were dead, tw(j in the iso alcohol partly reco\ered,
only one (jut of the three dying. In P/^ solutions of these alcohols,
the animals became insensible within 10 minutes, and died soon
afterwards. In the case of the butylic alcohols insensibility set in
within 5 minutes in ".o"/„ normal butylic, after 3-!:) minutes in
0.5 °/o isobutylic. and after .'1-15 minutes in t).7^/o tertiary butylic
alcohol, death following in all these cases within 24 hours. In-
sensibility was also established after 9-14 minutes in O.o*^/^ isobutylic
alcohol and after 9-20 minutes in 0.5*^/,, trimethylcarbinol, but in
both cases the animals recovered. thouQfh those in the foriner alcohol
seemed much nKjre prostrated than those in the latter.

272

M. TSUKAMOTO.

It would seem from the precediug experiments with 0.1^/^, amvlic
îukI allvhc alcohols that the toxicity of amvlic alcohol upon the
tadpole is strontifei" tliaii that of allyJic alcoVi«)], coiiti'nry tcj what
has heen found in expefiîîseuts ti])i)ii other forms of life which will
be described later on, 1 mide. therefore, further experiments with the
two alcohols, of the same dilution (U.!*'/,,)? <>^ well ;is of higher dilu-
tions. In 0.1 "/„ solutic^ns tadpoles a^-ain died sooner in amyJic than
in allylic alcohol. When the animals however appeared to be in-
sensible they were tliis time removed into fresh water ; whereupon
those that had lieen in the amylic alcohol recovered, while those from
the allvlic alcohol did not ; the latter also had convnlsions, but not
the former. Tn ().()5"/„ iinivlic alcohol solution most of the tadpoles
were ]»aralysed in 1 '/-j hours, hut after 18 hours they recovered, and
were alive aftei- .5 days. In 0.01 "/^ solution of thnt alcohol hardly
anv action could be noticed. In other experiments, the nnimals were
killed in —

0.05 "/o allylic alcohol after 2 hours

53 M "^"^ h "

These observations show that allylic alcohol is a cerii strung poison,
far stron^rer than amvlic alcohol.

Action upon Lower Aquatic Animals.

Experiments were carried out upon Ostracodes (Cfums and Cypri-
dina) iind Infusoria (principally Paramœciuni) in the same manner as
those described above. Allylic alcohol was applied in solutions of

0.01

"A. -

0.005

'/. -,

0.002

1., .,

0.001

«/« V

0.0005

»/. -,

0.0001

%, „

POISONOUS ACTIOX OF ALCOHOLS UPON DIFFERENT ORGANISMS. 273

0.1-0.005 ''/„ and even in the Intter ca.se all life was extinct within 24
hours. Amylic alcohol in O.P/^ killer! all ostracodes and many in-
fusoria within one day. flioijo-h some infusoria were found alive even
after 5 days, hut by a solution of 0.5'y„ all aiiiîuîils without exception
were killed after one day. In biitylic alcoliol of the same strenu'th
most of the animals died after two days, ind in isobutylic alcohol after
3 days ; the feM" indixidnals that still showed signs of life Avere
evidently more or less paralvsed.'

In 1 "/o solntions of normal l)uty]ic, isobutylic, and tertiary butylic
alcohol (trimethylcarbinol). and of isopropylic and propvlic al(X)hols
infusoria were found dead after 18 h(jurs, ostracodes after 1<S hours
in isopropylir and butylic alcohol, after two days in isobutylic
alcoliol, and after three days in propylic alcohol. In tertiary bntvlic
alcohol most ostracodes were dead after two da}^^, but some individuals
still showed ccmvulsions two days later.

The observations made concernini;- the activity of methylic and
ethylic alcohol are contained in the following table : —

Alcohol.

1 7o

3%

Methylic
CH^- OH.

Infusoria and Ostracodes
were seen alive after 4
days.

Strong narcotic action ob-
served afte '20 hoars ;
no life after 4 days.

Ethylic
CHg-CHs-OH.

Ditto^

Strong narcotic action
after 4 hours ; all life
ended after '20 hours.

I. In another experiment the animals died sooner, the temperature, which, <.i course, has
o-reat influence, beins? higher. In the experiments with sohitions of normnl and isopropylic
alcohols of the same dilution, the animals were found alive after four days.
2. Compare O Lœw, Xntiir!. System dn- Giftwirkuvi/en, S. 26.

274

M. TSÜKAMOTO.

Action upon Phaenogams

The ürsit series of ex]}eriiiients was carried <nit with seed» which
were just begin ni iig to germiua te after having been soaked in water.
Tn each experinient 12 seeds were left in 50 c.c. of the diluted alcohol
for 24 hours in a closed vessel; after pouring otf the liquid the covered
A-essel was left at the oi-dinary Icniperature and the further develop-
ment (jf the Herni observed. The seeds treated were those of barlev,
soiia bean. Swedish turnip, Japanese onion (Alliinii n>>li(li>suiit. L.), and
sliungikii [ChnimntheMiinn corotKiriitiii, L.). In one set of experiments
sohiti(jns ol 0.1"/,, <^f the alcohols were used, while in another set
l)arley seeds were treated with i "/,, and turnip seeds witli <>.5'Y„ and
l"/o solutions. The results are shown in the following table: —

Alcohol.

0.1 7o

0.5 7o

i7o

Methylic
CIV OH.

Alive.

Ahve.

Alive.

Ethylic
CtVCHo-OH.

Alive.

Alive.

Alive.

Propylic
CH,-CH,-CH,UH.

Alive.

Alive.

Barley mostlj',
turnip all killed.

Isopropylic
CHtCHsVOH.

Alive.

Alive.

Barley mostly,
turnip all killed.

Butylic

Alive.

Dead.

Dead.

Isobiitylic
CH3-(CHsVCH-()H.

Alive.

Dead.

Dead.

POISONOUS ACTION OF ALCOHOLS UPON DIFFERENT ORGANISMS. 976

Alcohol.

0.1 'Vo

0.57,

i7o

Tertiary butylic

Alive.

Alive.

Barley mostly,
turnip all killed.

Amylic ^

Only turnip
mostly injured.

]:>pad.

CH3-(CH,),C'H.-()H.

Allylic

All seeds
killed.2

CH^: CH-CH^-OH.

In experiments with methylic and ethylir; alcohols of higher con-
centrations, of 1^*^'/,, and 3*^*/o- I found that while l^"/,j ethylic alcohol
had killed in l^ I honi's the ttn'nip germs, methvlic alcohol had not
killed them even when of tlie strength of o"/^,.

In other experiments the action of propylic alcohol was compared
with that of allylic alcohol on vouiig so/a l)eans Avhicli had reached the
height of 15 cm. in watei'-cnltnre. One plant was placed in 50(1 c. c.
ofn.l'Y,, solutions of each alcohol. That in the allylic alcohol died
on the ord dav, while that in the propylic alcohol was not injured. A
similar experiment was carried out upon pea plants ahout 85 cm.
long. The lower three leaves of the plant placed in allylic alcohol
(0.17,,) t'li-iied veliow and dried up in o davs. and its upper leaves
in 5 days ; and the ])l;int itself was found dead in 7 days. In a
control experiment, also no iiijui'ions effect was ol)served in the case oi
propylic alcohol.

1. In the case of amylic alcohol turnip seeds were al.so killed in the 0.3 % solution.

2. In a few turnip seeds the cotyledons showed some development, but in none did
the I'ootlets of the eiubrvo show anv.

276

M. TSÜKAMOTO.

Action upon Algae.

In 0.1 'Yo i^olntiori of:ilJvlic nlcoho] Spirihjtjnt (uvimiunis was killed
in 24 hours, while in the solutions of the other alcohols of the same
dilution it was found healthv even after 10 days. Tn other experiments,
allylic aleohol of 0.05 •'/„ also killed Spirofii/ni in 24 hours. Allylic
alcohol of O.OP/o killed it in three days, hut O.()05°/o had no effect even
in 10 days. In O.ö"/,, solutions, this alo^a was killed hy amylic alcohol
in one day, hy hut\lir in o days, hy isohufylic in 4 days, still later hy
tertiarv hutvlic, ju'opvlic, and isopropvlic alcohols. In dilutions oi'
1 °/o. methvli<* and ellivli«' alcohols had no injurious effect, hut
])ropylic alcohol killed the cells wifhin ,") days, and isopropylir, hul-ylic,
isobutvlic, and tertiarv hutylic alcohols killed it witliin '2 days.' In
'2^1^ solution, ethylic alcohol proved nuich more irijurious than metli-
ylic, for while most of the cells were dead aüei- H days in the former,
only a few were so in the latter, l)ut in either alcoliol all the cells
were dead after 5 (Liys. In a o"/o solution of ethylic alcohol all were
killed within o davs, while in that of methylic alcohol tliey lived
4 davs. In a 4"/o solution of methylic alcohol, however, all the cells
were killed within two days.

Action upon Microbes.

For these experiments principally methylic, ethylic, ainylic, and
allylic alcohols were used. In one series of experiments their toxical
actions were tested and in other series their nutritive qualities. On(^
drop of putrefying- hroth was introduced into the alcoholic solution of
a certain dilution, and then a sterilised solution of meat-extract \vas
infected from the alcoholic solution after this had stood 24 hours.-
The results ai"e shown in tlie folIowinL^- tahles :

1. [u auother case r;o isopropylie alcohol solution killed Spiruyi/ni cells ^rithin one clay.
:i. Of course, control experiments with plain water were also made.

POISONOUS ACTION OF ALCOHOLS UPON DIFFERENT ORGANISMS. 277

T.

Alcohol.

Aniylic

Allvlic

0.1 7o

0.5 7o

1%

Stroller flevelopmont i mi u\

, /V ^ .. The same as ^Ylth

at the same time

as in the control
experiment.

Developnient one clay
later than in the
control experiment.

0.1 7o

AVeak development ^
4 days later than in
the control expt.

Development two -
days later than in
the control expt.

Weak development 4
days later than in
the control expt.

II.

Alcoliol.

5 7o

l5 7o

20 7o

Metbylic
and

Ethylic

No difference from
the control experi-
ment.

Much less develop-
ment than in the
control expt.

No development
after 10 days'
standing:.

In ;i 27', >^<>lufion of iiniylic nlcohol hncferial life is evidently
f;-rentlv depressed, fnv tlio infection of meat-extrnet solution ^vith this
nicoliolic solution did not induce any development \vitlnu 7 days.

In a second series of exi^erirnents 0.1 7n l)otassiniu |)hi^s))hafe,
0.01"/,, mnoTicsiiini snlpliate. and the alcohol to I)e tested in a certain
dihition. ^v ere added to a O..")"/,^ meat-extract sohition. The infection
was made from ]»iiti'id meat. The results are as iollows : —

0.17.1 allvlic alcohol : after (! days, a weak hacterial development.

0.5 Y, y ., : nfter 8 flays, very slight development.

\. The liacterial vogetniiou eonsistoil priueipally of one Iciml i.f thin lont;- liacilli.
-. Hero principally micrococci were noticed.

278

M. TSUKAMOTO.

1

"A

amvlic

alcoliol :

10

%

ethylic

•

15

%

11

•5 •

20

"h

J5

.. '.

10

•%

meihyli

c ., :

lô

%

..

?' •

20

'%

..

•5 •

: nfter 4 days, considerable development.
: after 14 days, large development.
: after 14 daj's, quite clear, free from
bacteria.

no development,
stronii" development,
weak development,
no development.

In a third series ol" experiments no source of carbon for the
growth of bacteria was contained in th(^ solution except the alcohol
itself, the other constituents being only 0.5 'Yo ^h'^^t^ 0.1 ^/o potassium
phosphate, and 0.01 "/o magnesium sulphate. The infection was made
from putrid meat and the flask left to stand at the ordinary tempera-
ture. The results were as follows : —

O.P/o îdlylic alcohol : after 9 days, slightly turbid ; a few

bacteria.
: after 24 days, quite clear, no bacteria.
: after 14 days, no bacteria.
: after 10 days, slight tui-bidity ; small
oval ijeaü cclh^ were seen but no bacteria.
: after 5 days, slight turbidity ; small oval
shaped bacteria only were seen.-
In a further series of experiments I compared the nutritive effect
of the nine alcohols in dilutions of O.P/q with only the addition of
0.5*^/0 ammonium phosphate, 0.1*^/,) monopotassium ])hosphate, and
O.Ol^/o magnesium sulphate. These solutions were all infected from
the same source, viz : a Pasteur's solution that had been exposed in

0.5% .,

P/o amylir
10^0 ethylic

10% methylic

1. Compare O Lœw, Natiirl. System der Giftuirliiingcn, S. 20.

2. According to li. Brown (Chem. Soc. Jour., 1886.) Bacterium aceti utilises uiethylic but
not amylic alcohol.

POISONOUS ACTION OF ALCOHOLS UPON DIFFERENT ORGANISMS. 279

the open air and contained bacteria, yeast cells, and mould fungi. The
results are as follows : —

]\Iethylio alcohol : some development observed in one day, but

increase moderate even after 21 days ; small

and large yeast cells, and bacilli observed.

like the preceding, but the increase greater ;

besides yeast cells and bacilli some mucor-

like mycelium was present.

like the preceding, but a mould fungus with

white spores was observed in this case.

like the preceding (very little mycelium but

numerous bacilli) ; no spore-bearing mould

fungus .

after 2 days turbidity had set in, but the

increase was very small even in 21 days ;

Ijacilli and a trace of mycelium.

the quantity of fun "'i seemed a little largfer

than in the previous case.

the fungoid growth was here much larger

than in the last two cases, many bacilli.

small yeast cells, and iiiiicod<:rnia-]\ke yeast.

Ijut no mycelium observed.

after 2 days only very little development.

but considerable in 21 days ; here more

mycelium was observed tlian in any of the

other cases.

after 21 days, the liquid was perfectly clföu*

Ethylic ,.

Propylic „

Isopropylic ,,

Butylic ,,

Isobutylic ,,

Tertiary butvlic ,,

Amvlic

Allylic

and free from fungi.

1. Parbaps a developuieut might havu beou noticed if the ordiuary phosphate had boeu
used hero instead of the iiiouophospliate.

'2S0 ' ^^- TSUKAMOTO,

;.". These observations of the action of the alcohols upon microbes
lead to the conclusion that the common microbes of putrefaction are
killed by 0.5 "/o allylic, 2"/o iimylic, or 20 ^'/o niethylic or ethylic
alcohol (but probably not the spores'), if these substances are jjermitted
to act for 24 hours in absence of any nourishing materials ; further,
that in hio-h dihition the higher alcohols ai-e better food for the microbes
than niethylic alcohol, but as the dilution lessens, niethylic alcohol
proves a better nutrient thtin tlie higher alcohol, the latter then show-
ing poisonous action.

Furthermore, all ni}" experiments on ditferent organisms go to
prove that ethylic alcohol is a stronger poison than niethylic alc<jhol.
although Dujardin-Beaumetz and Andigé (/. c.) have asserted the
contrary, at least for higher animals. ]\Iy results, moreover, agree
well with those of Gibbs and Iveichert (/. c). who compared the action
of these alcohols upon the higher animals. Ft is further to be c<)n-
cluded that the normal and iso-propylic alcohols behave ne:irly ;dike,
and that of the three butylie a[colu)ls the normal is the most
poisonotis, and the tertiary the least. Allylic alcohol is not only
much more poisonous than the corresponding pro])ylic alcohol but
also more so than amyJic aleoliol. The toxic action of the allylic
nlcohol is evidently soniewlait dilferent IVoni that of the saturated
aleohols. whose toxical power generalh increases with the number
of carbon atoms in their molecule. According !<.> Meissner' allylic
alcohol acts 50 times more strongly than propylic. According to

1. Kucli has found tluit eveu a uuicli higher coucuutratiou would uot kill oortaiu
spores.

2. Meissner observes : '• Allylic alcohol damages the circulatory system, enlarging the
blood vessels and paralysing heart-action. Allylic alcohol lias none of the narcotic action of

the alcohols in the saturated series. Other important differences Ijetween allylic and other

alcohols are that, when inhaled, it attacks the mucous membrane, causes great loss of protein
matters, and acts .50 </«(t's .s7;oH//^r than propylic alcohol." (Chem. Oentralb. 1891, II 715).

POISONOUS ACTION OF ALCOHOLS UPON DIFFERENT ORG ANI6MS. -JH 1

inv observations, however, the diiference is much greater. Allylic
nlcc^hol seems ro attack }>roto})lasm directly, by chemic^il affinities
arising' from the double Hnl^inu^ of two carbon atoms, while the
saturated alcohols act merely cutalxtically Ijy transferring certain
kinds of motions. '

Livinii" protoplasm seems a very delicate indicator of ditferences
iij chemical constitution, and when we consider how indifferent dead
pnjtoplasm is towards hiOst of those pois<3ns that react easily upon
living protoplasm, we cannot doubt for a moment but that a great
chemical change uuist \vd\e taken place at the moment of death in
the proteids of the living protojtlasm.

In conclusion. I tender hearty thanks to Prof. Dr. 0. Loew for
the interest he has taken in my investigation.

1. Therefore those alcohols have here beeu conipared only in equal weights and not in
equivalent quantities.

Formulae for snQzr.

By

0. Sudo.

The folIoAving calculations of sn du were made to show the
advantao-e of the method of tindino; the multiplication-formula of
elliptic functions, given l)y Prof. Fujisawa in the second part of
his paper Bescarches on the Multiplication of EUiptic Functions (this
journal, vol. \l, pp. 151-226). For the notation adopted, as well as
for a full account of the method of calculation, reference is to be
made to that paper.

Considering tlie nmnerator and denominator of sn du as ex-
pressed in terms of a ( = 1+ -- J and ^{^^s/h sn u), and putting /? = 9,

(/ — 10 in the general formula^ for 7/^, lïf,_i .H',_2 i7,_o^ the vtdiies of
7^10, B^, Hg, H-, were found without much difficulty, and, thence,
by integrating ditferential equations (117) Qoc. cit. p. 202), H^, H^,
H^, H^, were successively obtained. (^n tlie other liand. H^ whicli
corresponds to H f)r /.•■^=— 1, was derived from the expressions of
sn 4 (h, i) and su 5 (u,i) by addition, and. then, l)y again making use of
the same ditferential equations, but this time in the reverse ordei*,
H^, H.J, Ho, were o-ot bv successive differentiations. The aoTeement of
tlie values of Hn deduced in two different ways verified the results.

The value of E is at once obtained from that of H in virtue of
equation (111) (loc. cit. p. 200).

284

(>. SUDO.

Finally, the denominator of sn 9« was transformed into its usual
form, where the variables are taken to be h and x ( = sn ?/). The
result was verified by coraparins: it with that derived from formuke
(146) and (150) (loc. cil. p. 215 and pp. 217-218).

In ever}' case the calculation was performed in duplicate, by
myself and Mr. Fujii, to whom I owe my best thanks. The results
are i>"iven in the accompanying two tables, in which the mode of
arrans*ement is obvious.

Table of H and E for sn9».

n„

Ill

n.

Ils

II<

Hs

Ile

M,

IL

Ils

IIio

—

tto

ç -

+

I

540

+

5544

2S512

p"

f°

// --

+

I I SSoo

4-

S236S

|70

ç'"

-757-4
^147jI

4-

207597G

5709G0
6262272

4-

1322496

13977Ö
1624320

4-

13S240

7372S

ç"

î"

_

-7475-0

4-

93004S0

4-

1019904

4-

163S4

s''-

:a)

+

70359S4

4-

2800396S

—

6773760

—

25S04S

-ra

;20

+

4010925G

739074^4
295009344

4-

300769920

9493S624
67397299-

4-

173209344

4-

1935360
20205 15S4

çM

}"

+

S3 5 14960

4-

957960000

4-

9761 Si 760

4-

165703680

c'-

?::

+

1353407^::

5623S460.S
999105408

4-

I 8077644S0

1SS4824064
349S5S252S

4-

25011002SS

9SS793S56

2240372736

4-

703217664

97517568
331776000

4- 3S92S384

- 94371^^4

fic

fi«

+

2Ö1 189624

_

1679225040

4-

31619S2336

_

5240563200

4-

4179S75S40

_

2992066560

4-

12S2842624

_

424673280

+ 93 192 192

— 1 I 796480

4- 104S576

ç*'-

il

+

366014340

_

1674487200

4-

41017G2240

_

4S I I 760000

4-

47655820S0

_

22S7411200

4-

I I S7020S00

_

201523200

4- 61931520

tl

v°

+

256239000

_

6S9919120

4-

2694880S00

_

1653696000

4-

2716070400

_

503331S40

4-

455270400

s

:w

+

4687S210

4-

276687360

—

56S51200

—

90316S00

î™

flo

-

5737S672
49155;6S

4-

4-

15S43340S
30554204S

-

721664640
4172S0S96

4-
4-

721073664
75SS6540S

-

7669S316S
3067S9120

4-
4-

409522176
25619S656

-

I 16895744
53968S96

4-

4-

22560768
20971520

- 353S944

- 1769472

4- 262144

Ç=0

c"

4-

I 1951 7120

4-

21 5205 120

4-

7SS52096

4-

530S416

î"

î

—

I354fc:!4

—

87023S0S

—

71723520

—

9326592

-■jo

-

153701 I

4-
4-

"S3435S4
36SS320

-

157593Ö0

4-
4-

41900544
4465152

-

8273664

4-

I 06444S0

i

^

—

432S2S

—

163468S

t'"

v

_

32106

4-

3574X0

4-

9120

—

4608

4-

256

çlO

î"

—

5616

—

576

■?"

c"

+

1044

120

4-

432

î *

'-"" î +

9

c °

L

l-o

K.

E. E,

F-,

Es

Kt

E,

Es

E,

E,„

The Common Denominator of sn 9», i-n 9», cln 9(i

.,■»

+

A-»

.1-'

-

54"

/>-'

a;'

,.f,

+

5544

/,■■- + 7,- '

x'

.!■»

-

JS512

jr-+k'

—

64746

A<

x'

.«■'°

+

,S236S

lc''+I('

4-

365904

A- '4- A«

x"

.-■' =

-

1397/6

Ic'+Jc"

—

1 1 30064

A '4- A«

—

2256300

A«

x'-

./'"

+

13S240

/.■-■+7.-'=

4-

2013696

A '4- A'«

4-

7425864

A'=4-A^

3:»

,.lr.

-

7372s

A:-+Ä-"

—

206668S

A'4-A'^'

—

1 3865472

A'-l-A'"

-

23959755

A^

x^

,.1S

+

16384

A - + /.■"

4-

1134592

A '4- A"

4-

14744064

A «4- A*

4-

35926400

A»4-A-'°

J.W

x-'

-

25S04S

/.■'4-y«

—

8322048

A «4- A"

—

2961792

A»4-A'-'

4-

1 7240400

A'»

Z-'

.r-

+

1935360

/.■»4-7,-'«

—

85261824

A »4- A"

—

339369696

A-'»4-A'-

x"

.(■='

+

173209344

Jc'+li'"

4-

993607296

A'°4-A-»

4-

1 6S0905 1 60

A"

X-'

.(■■-■'

-

20205 1 5S4

/.•• + /.'»

—

16S4230912

A'°4-A"'

-

4337444'6o

A'-4-A-"

x-""

/-'

+

165703Ö80

J,-+Jr'

4-

1970403840

A'» 4- A«

4-

734S242240

A'^'4-A"'

4-

1 1 170599 120

A"

x^'

r'

-

975'75GS

K-' + lr'

—

1671416832

A"" 4- A-'»

—

8876662272

A'HA"

—

19517910240

7.1. + /.10

x"

./■■-

+

3S9283S4

l-'+lr'

4-

1014644736

A'»4-A-^--'

4-

7810401024

;,,.+ ;,..,,

4-

24540420Ü96

A" 4- A"

4- 35546S11570

A'«

jT.

,r"

—

9437 'S4

1: '+!,-''

—

416710656

A"4-A='

—

4902543360

k'-+k^

—

22460465664

A"4-A'-°

— 46699S25536

A"'4-A"

x'"

,-:e

+

104S576

k'^+Ii^

4-

103677952

A">4-A=«

4-

2075566080

A-'HA-^'

4-

146121420S0

A"4-A-'-

4- 4456308S76S

A'" 4- A™

4- 6410S956408

A"

J.S6

./'"

-

1 1 796480

/.-'HP'

—

530S41600

A''- 4- A»

—

638945 2S00

A"4-A--''

—

30109939200

A'HA"-'-

— 63671 501 520

A"'4-A-°

X«

J-'"

+

61931520

Ic'-'+li^

4-

1682472960

A"4-A-=

4-

136217S9440

A'« 4- A"'

4-

444375676S0

A«4-A-'-

4- 65238653700

A=^

x"

J-'=

-

201523200

Ic'' + k^

—

3698073600

A'«4-A'^"

-

204S0S03200

A"+A='

-

46037191200

A-'»4-A--

x*^

J"

+

455270400

i'«+A:"

4-

5447692800

A«4-A-'

4-

2038S2 18400

A-=»4-A:='

4-

31047831000

A--

x"

j'°

-

50333 1S40

?i"+l^

4170355200

A» 4- A-»

—

106S4325520

A-'-' + A^'

x«

x"

-

903 1 6800

7."'+A-"'

59S752000

A'-'HA'-"

-

1 305469440

A~4-A:=»

-

1547190270

A-''

x"

y-

+

22560768

A'«+A:"^

4-

567447552

A--» 4- A»

4-

3242460672

A« 4- A"'

4-

7206503040

A='4-A^'

X»

.!■■-

-

353^944

A«+A"

~

145207296

A-»+A'=

—

1 567448064

A- 4- A™

—

5741214336

A='4-A-"

— 8688247920

A-"

X«

y

+

262144

A'»+A-»'

4-

23330S16

A-»4-A"

+

4124364S0

k--'+k'-

4-

2502280704

A"' 4- A»

4- 5911158176

A-" 4- A-"

x"

,.56

-

1769472

A-'+A-"'

68124672

A--4-A"

—

680147712

A--"4-A'=

—

2553061 24S

A»'4-A-™

- 3927693240

A-'

^.S6

r"

+

530S416

A^'-'+A»

4-

1 1601 1008

A^»4-A''

4-

720942336

A«4-A''=

4-

1 739448000

A^'4-A™

x**

a«

—

9326592

A^' + A^"

""

127683072

A'-*4-A--'

—

5 13s 16768

A»4-A^-

—

804466S0

;.:»

x"

.rf-

+

! 0644480

/.=«+/r»

4-

95122944

A=HA-"

4-

250490016

A'" +/.;-•-

x«

,,M

-

8273664

A='+A--»

—

4SS54016

A--'»4-A»'

—

82697715

Ir-

x"'

./■'■'■

+

4465152

po+7,«.

4-

W'^Siy/O

A-"=4-A-»'

x"

.(■'■'

-

1 6346S8

A-»-'+A«

—

3702204

A«

x^

J''"

-

460s

li'-'+li^

4-

343656

A-"4-A»°

r'

.(■'=

+

256

A-"+A'»

4-

10144

A" 4- A»

—

12330

A-™

x"

.(''

-

-^7^

A'" 4- A"

—

7344

A»» 4- A-»»

x"

,,7Ü

+

432

A« 4- A:'«

4-

190S

k'"

x"

J,;»

—

120

A-»4-A-'»

x"

,.«0

+

"

A-'"

.t'"

Formulae for sn 10?/, on 10?^, dn lOii
in terms of sn u.

By

E. Sakai, Student,

College of Seionee. Iniperiul Univovsity.

itn 1 0;/ —

v/i — .r- ^/i — k-x- A {x-\

D ix-)

en 10?/ ■-=

D ix-)

(In 10//. =

C{x^

Formiilfo lor sn 10?/, r-n 10?/, dn 10?/ in terms of ;r = sn 7/ nre
ofiven in the followinfr fonr tables, in which the mode of arrangement
will he apparent on inspection. Tliey were cah-nlated hv two entirely
di Heren t methods.

I'^irstly. they wei-c dcdiieed from the well-known equations :

S'-'2S-P- + JrP'
S'-lrP* '

S*-'2V-S'^P- + l-P'
Dix-) " S'--Jrr'

where A. B, C, ]> denote alo-ebraic rational intearal functions of .r- and
P=.r {r, - (20 + 20 Z"-^) .r-'+ (10 + 94 7.2+ 16 V) x'- (SO A- + 80 V) x'- 105 7.' x'
+ (8(-,0 Z-^+ :\m ]/') .7-1"- (240 7.-^+ 780 7.''+ 240 V) x^-+ ((U k*+ 560 7/'
+ 5(i0 7.' + 64 7.h x^' - (100 Ä-« + 445 k' + 1 60 Z^") ./"' +(140 k' + 1 40 7,-^") x^'
^:,Ok"'x-'+ l^ :>-'),
Ç:^^r^' -;i-12 ./■-+ (16 + 50 Z-) X*- (80 Z- + 140 7.^1 .7«+ (885 ZH 160 7/') x^
- (264 k' + 464 Z'' + 6)4 Z'^) x^" + (208 k* + 508 Z'' + 208 Z-^) x^- - (64 k'
+ 464 7.'ß+264 k^) .'ri^ + (l60 ZH335 Z-«) x^'-{\àO k' + HO k'"} x''+{r,0 k'"
+ 16 k'^) x^' - 12 Z'^- .r-- + 7.-12 x'-*] ,

ox (J E. SAKAL

.R = V'i-/r.r--' |1 - 12/;^/-'+ (507.2+ Ul')x'- (140 /c-+ 80/.'').r«+ (100 /r
+ 335 ^.4) 3,H_ ((34 7,->4G4 A^ + 264 7.") x'"+ (208 Â'*+50R 7^«+ 208 Ji") x'-
- (264 Ä;'''+404 7.'"+G4 1'') j''^ + (335 7cH 160 Z.-i«) x''-{80 Ä-^+140 Ä-^") ^r^«
+ (16 ^'^+50 7.'i").T-^-12 7.-i».T-- + 7c:^-^-*},

5f=l - 50 A- .7'' + (140 7r + 140 I') x''' - (160 Jc'+ 445 7.^+ 160 ¥} .t«+ (64 7.-
+ 500 7,4+ 5(-o 7,G + 64 JA x''- (240 A'H 780 V'+ 240 7.;«) .r^-+ (360 !/■
+ 360 A-«) .T^'- 105 Z-" •r^'''- (80 7,''+80 ¥') .t^^ + (16 7i-' + 04 7.-i"+ 16 A'-; 2-'
-(20 Ä-^'' + 20 7/^-) ,r"+ 5 A'^- •t--'.

To obtain P Q R S, P S ^vas first formed and the result multiplied by
Q and P. successively. P' was ,2:0t by multiplying P' by P twice in
succession. S^ was found bv squaring -S^ and. as a verification, they
were substituted in tlie well-known relation —

,<?C.r,A) - P(-JJ' ^•)^-'^"'' ^''^^

where 4/)= ;r — 1, «being odd. P^S'- was obtained by multiplying
P^ and N^ and also l)v squaring PS. In every case, the result was
further verified by putting /.:-=:l and /.:' = /.

Second I V, A, B, C, P) Avere calculated by a totally independent
method given bv Professor Fujisawa in his paper entitled " Researches
on the Multiplication of Elliptic Functions" (this Journal, vol. VI,
]>p. 151-22(3).

In the course of calculation, these functions were considered to be
arranged aceording to the powers of .r" as well as aecording to the
]i()wers of /,'.- 'I'he i'ormula^ to he used in the former case are given in
the paper just alluded t<\ and the corresponding formulae in the latter
case are as follows :

wlicre

7 - o y 'm ■= r

FOllMULJC FOK sn 10 «, culOu, dnlO« IN TERMS UF snw. 287

n (}i- — 4) (yi- — 1 6) hr — Am)

A„^, = 7l, ^2,H.o=(-l)

{•Itn+1)

'1.,:- 3 1 . -h,n,2--{ -I) (•2w + l)!

and. generally.

■lin rl/n + 1) J,,„ ,, + (n- (4/-+ l)-4;;r] .4,,,. „^ ,,,

- [ir [ir-3)-2ir (:2/u - i) + 4m-] A.,,„_.^ .„„.-(/r- 2;yi + 1) (yr-2m) Ä.„,^_i ,,_

r=0 \ TOT )

wliere

D

_ n D _ . _ 1 m-i o i"t-2 "-' 0^'- 1) (?r - 4) [n- - (m- 1)-]

•2;n !
X [ ['2m' — 0 {m — i)] )i'- — [m — 1) [1 m — 5; } ,
and. a'enerally.

'2m(2m-\)D,,„^,, + [4nr-4 (m-1)-] D,,,„_o,,,

=. [ ,^2 ^4;._ 4) _ 2,i^ ^2;;;, _ o) + 4 ^^/^ _ 1 .2 j £)_^^^___^_ ^^_^

-Or- 2m + 4) (/;--2m + 8) D._,^_,^o^..
wliere

1? —1 7-' _ ( ]Vn M-(».- — 4)(/r— 1(3) [„-'_4(yy;_l)2]

"Ii/i !

2<S,S

E. ÖAKAI.

X //z (yy; — l)Or — 1),
and. generally,

'2mC2m-\) B,,,,^,^ + [nH-^}'+\) - 4{ni-\f] I>\,^_,,,,

-=[;r\4r-4) - 2ir(2}n--I) +4>-l)-] L\,,„_,,5,_2

— (/r — 2w + 4) {)r — '2111+ 'S) L\.,„_i, 2,-2-

Numerator of sn lO »

VI -.r Vi-lc^x'

1 60

1024
512
1 1264
6391968
+ 21829632
516S4S64
+ 8704S192
10362SS00
8519ÖS00
46006272
+ 14680064
2097152
356515S4
2S9S00192
1499463GS0
+ 44395S394SS
35586310144
+ 19529072640
795S691840
2267021312

- 402653 1 84
+ 33554432

402653184
226702 1312

— 795S691840
+ 19529072640

355S6310144
+ 44395839488
1499463680
289800192
356515K4
2097152
+ 14680064
46006272
+ S5196S00
10362SS00
+ 87048192
51684864
21S29632
6391968
1 1264
512
1024
672

k "

k'+k-l

l-' + Jc*

k'+k'

k' + k^

k'+k'

k' + k'

k' + k"

k'+k'^

k' + k"

A- '+/.■'«

7.*+Ä-"

k'+k'-

k' + k-

k'+k"

k'+k'^

k'+k-'

k^'+k'*

k"+k->

k^'+k'"

k''- + k«

A-'HA-

k''- + k«

k"-+k»

k''+l^

A"+/v»

k^'+k"'

k'^+k«

k^+k"

k-+k^

r-' + k"

k'-'+k"

/.=*+/,'»

k'* + k"-

k"+k"

F" + /.-"

k^ + k"

P'+k"

P'+k"

k^+k"

A-=»+A"

k^ + k"

A" +7."

k"+k"

k"+k"

k'°-+k^

k"+k-^

A« + A«

'"

30S8
13920

22016
1 1 1 40S0

143170/-^
SI3524S0

292452864

72902860S
I2S2924544
I5S556I600
1346109440

747372544

2443 1 820S

1944059904

9S474393G0

19579715584

458321436672

502206251008

379297988608

2I371682S160

89306169344
26048724992
473II749I2
2604S724992
89306169344
2I37I6828160

379297988608

502206251008

+ 45S32 1436672

'9579715584

9847439360

1944059904

24431S208

747372544

1346109440
I5K5561600

12S2924544

72902S60S

292452S64
81352480

1 43 1 7072

1 1 1 4080

22016

1 3920

308S

k- + k'

k'+k'

k'+k'

k'

k'+k"

k' + k"

k°+k'-

/,"+A»

k'+k^'

k'+k""

k''+r-'-
k'+k-
k«+k-
k'-+k--

k'^+k-'-

k»+k''

k" + k«

7,"4-A°-«

k»+k"

7,-" +/.»-■

7,-" + 7,1«

7.>«+A"

7,"> + /,■•'

k'-'+k^'

l^-+k"'

k-'+k»

k^+k«

Ir'+k«

r-'+k"

A=»+A"

7r'» + A'-

A^'» + A'-

k^+k"

k'^ + l.'-

k^ + l'-

A:™+A''-

A-»+A'-

A"+A<--

A«

A'-+A"

/.■"+/.'«

A" + A'»

A«

4S794S7S0

1862949792

48992S6304

9021350912

1 1585656320

10163077120

5S035S9632

2S691251200

52561766400

205170696192

1 579107461 1S4

2381954S03712

2456797156352

1S8055S321664

109029924S640

4680717434SS

140039159808

4680717434SS

1 090299248640

1SS05583216G4

2456797156352

23S 1 9548037 1 2

157910746US4

205170696192

52561766400

28691251200

58035S9632

10163077120

I 15S5656320

9021350912

4S992S6304

1S62949792

487948780

A'

k'

k'+k^

k'+k^

A»+A"

A»+A'«

A-'+A"

A^+A-=°

A'»+A»

A'-'+A^

A" + A>'

A'M-A=°

A'« + A=

A-'^+A^'

A'"+7i?«

A'"+A:=»

A'^+A»

A"'+A»'-

A's+A«

A=°+A==

A-"+A-«

A--"4-A>=

A.-'^ + A'^

k-'+k'^

A='+7v"

A='+A'«

A='+A'"

k'-'+k"

k'^+k"'

A-'= + A"

A''+A»

A»«+A"

k'^ + k*'

k"

A"

_ 7552669520

+ 20953712480

_ 403807564S0

+ 53S95244800

— 48851558400
+ 133104640000
_ IOS7G9351680

— 5367225564S0

+ 234I496S66720

— 4979816492992
+ 706448 1 864704

— 7325749409792
+ 5760971530240

— 3444844216320
+ 1522507710464

— 3444S44216320
+ 5760971530240

— 7325749409792
+ 70644S 1 864704

— 4979S I 6492992
34I496S66720
536722556480
10S76935I6S0
133104640000

4885155S400
53895244800
403807564S0
209537124S0
7552669520

+♦

A'»

k"+k''-
A-'"+A"
A^+A:"

A'° + A>»
A-i= + A'«
A" 4- A"
7,ic+;.i8

A'HA""

A'«+A--=

A'«+A"'

A'-'H-A^»

A»»+A«

A"+A»

k^'+k"'

k'-'+k"

k-' + k^'

k^ + l"

k'^+k^

A=»

A'"4-A-»'^

k^+k""

A»»+7r«

A'^ + A»

A»=+A''

k'^ + k"

A-'» 4- A--»

A-'»

60719439320
I 20036973600
1 59288948960
269752754400
1 29401 SGS970

9956831666320
14145431403S40
15037580149696
1 22201 74200S32

7554798590976
12220174200S32
1 5037580149696
1414543 1403840

9956831666320

129401868970
269752754400
1592S8948960
120036973600
60719439320

A'-

A-'=+A-»
A" 4- A«
A»+A"'
A"

A™4-A~
A=»+A='
7," + A-=
A-^+Ar«
A==4-A:''
A^'+A^'

A-'»

A«

A-»^ + A'
7c»'- + A»
A"4-A^
A"=

232580171440

20583618969696
22629065444544
19073766466560
22629065444544
20583618969696

232580171440

A"'

7-4- A='
A~+A="
A-'H-A»

+ 25S34036525336

Numerator of en lO« =

4ÜÜ

1. "
k' 1

1I20

k'

I2SO

k'

512

h'

11520

ir-

I407SSSO

k'

40930560

k'l

S614I44O

k<

I3057=^SS

k'

141312000

k'

i UÛ49Û000

k'

53084160

k'

15728640

k-

2oy;i52

k'

^6700 1 60

k'

307363840

k'

1 64003 S400

k"

71975321600

Ifl

444S2SS7ÖS0

k"

23248X90000

k"

90439ÖS000

k"

2464153600

A"

419430400

/,"

3355443=

/.'»

419430400

/."

24641 53600

A'»

9043968000

J.«

2324SS96000

'■"

44482887680

/.■•'

71975321600

Ä"

164UO38400

;.=' 1

307363840

V

36700160

I-' 1

2097152

i-'

1 5728640

;.-•

53084160

A»

1 06496000

A-™

141312000

A-'--

130572288

''■"

86141440

''"

40930560

L"

14078SS0

k" 1

'- ,1

23760

2SI60

3659440

4 1 0S6480

I 780S7920

558046720

1243610368

197554176=

2223104000

17314611=0

SS76S5120

2694S4032

2208890880

11533025280

1 1 9826636800

923363205120

72539431731=

- 492S76595200

264757248000

105670246400

29491200OU0

5 1 33828096

29491200000

105Ö7024640O

264757248000

492S76595 00

725394317)1

9=3363=051 o

119826636S0O

■I5330=5"''0

2208890880

2694840 J ■*

887685 1 "O

17314611 0

2223104UO0

197554I/Ö0

1243610J6S

558046720

17S0S7920

3639440

A-.=

.

28160

III'

23760

k"

7920

k"

8=5

k"

\+ 136620

- 79965Ö

4- 7400380

- 33647040
+ 21 5 138570

- 1 048703460
+ 33786630SS

- 8507317=80
+ I4===557440

- 16682641920
4- 1344802S160

- 709759795=
4- 36134461440

- =03559393=So
4- 1092048583680

- 41774178441ÖO
+ 4l33945==l63=

- 3569667118080
4. 25400567S0800

- 14063579136QO
4- 577290240000

- 165365415936
4- 77290240000

- 1406357913600
4- 400567S0800

- J :,69667 1 1 8080
+ 4 33945== '03=

- 4177417844160
4- 1 092048583680

- 03559393280
+ 36134461440

- 709759795=
4- 134480=8160

- 166S2641920
+ 14===557440

+ 35786630SS
048703460

7400380
799636

A-.|

68640

A-'

X '

1034176

k'

_

256256

A-*

x"

4 6j6

+

5 S 400

J

x"

4
-1 4

+
+

4
4 4 r

+

+

4
44

4-

7rs
446

4-

G5536

A»

J:

.-4 4 4

J 4J 6

_

6 8

_

6 4 6960

-

1 3 1 0720

A»

X-

6 er

+

4 4

4-

f 96

+

r 6

4-

126156S0

A'"

x-i

64J 64/

_

J J 9J G

—

4S4 9 6

-

40 6

-

53S0341760

A"

x"

9

4-

SS 4S So

4-

46 rr 0

+

664 4 0

4-

052279=4480

A»

4-

64S0S55040

A-"

1 =

(•44 J

_

4J 4

—

69 4

—

-

2 1701 15072

A"

-

82463=93440

A™

—

607S595072

A=

x"

9 ''4 4

4-

r

4-

r s 949

+

4-

49=6336307=0

A-»

+

80103260160

A-"

4-

4289331200

r-'

x^

1 r 4

_

„

4 4

_

4

4 17330749440

A"

_

58330316800

A-'

—

2202009600

A»

x"

3 44 r

! +

4

4-

4 S9 4 f

4-

4

4-

4696591360

A"

4-

30S04541440

A'*

4-

775946240

A"

x"

8 4

-

) 4

J fS 4

—

4

_

3961139200

A»

—

11135S77120

A"

—

167772160

A-»

x"

09600

+

4-

4 4

4-

4 4^4

4-

S 0591 795664

A--«

4-

75754635^64

;.œ

4-

=464153600

k"

4-

16777216

A-»

x"

IIS J 4 4

_

4

_

94 ;f

_

6 j44 J440

r

:, :,^ } JS4O0

;■>

—

17175674S8Ü

k"

—

25165S240

A=

x'^

0 J 4 Qj

4-

1

4-

4 44

+

r

4

D 90 0

1"

+

S 1 104

1"

4-

75593154560

k"

4-

17930649S0

i'=

x"

s 8 J r

-

1 ; -

1 -1

/

s Ö800

n

—

2358050S1600

A-»

—

S074035200

A'

x"

46 J 4

+

;

4

4 S40000

A

4-

4-

672321536000

k"

4-

25790054400

A'

x"

ij jof 45 r r

_

1 _

t /

4 4

;

k^

—

204571574=7=0

k"

—

l=4354953='6

i=

x"

46 J 4

+

/

4 /

4

1 '

4-

672321536000

k-'

4-

25790054400

A»

x"

SoS j6 f

-

1 4 '

-

235805081600

k"

—

S074035200

A*

x"

10 9 4 C4J

1 +

t / +

44 +

4

' + .

*75593l 54560

A--"

+

1793064960

A"

J"

11 /4 J 4

} \ -

_

4

_

4

44 44

A

-

4

I '

■71756748S0

A-''

—

251658240

A-"

J.»

109600S

1 1 +

/

I 4

4-

4 9 s 464

4-

Oj9I 9j654

A'

4-

3 54Ö0, 64

! "

4-

2464153600

k"

4-

16777216

A-«

x"

S S4

; -

4 4 / -

4

—

-4 6

-

ojgriijj 00

;-

-

11I038 1

1 '

-

167772160

k'--

J. CS

3 44 r

4-

4

4

4

4-

9 46 OS

A*

+

4696591360

A--*

4-

30804541440

A"

4-

77594Ö240

A-"^

x"

1 66 r S4

-

b r J

-

4 4

—

r 9 04

_

497330749440

A^

-

58330316800

k"

-

2202009600

A'-

J.«

J ''4J 4 ■>

4-

5 er

4-

4

4-

J / 9

4-

49=6330307=0

A-»

4-

S0105260160

k"

4-

4289331200

A«

x"

'"''4 4 J

49 864

-

4

—

r er

_

2170115072

A-o

—

82463293440

k"

-

607S595072

A"

J*'*

J

+

s 4S S

4-

4 6 66 S

4-

9 6 4848

4-

652279244S0

k"

4-

6480S55040

k"

a;"

4 ''4

4-

J J)J 5Û

4 4

4-

*S4 r

f 96

;

4-

4 r

4-

5380341760
12615680

k"
A"

X-

4 4 4

4-

9 49 6
4 4SS r

4-

osro 0

S44 6

'

4-

1 ^

4-

13 10720
65536

A"

I-

4 4

—

J 8

-

4

_

x"

869 6r4

4.

5088640

4-

ß

A

x"

3 S

—

. I9-O0

X*

4236960

A'"

4-

582400

A"

I"

103417Û

1,1''

—

256=36

V

X

6S640

j

X»

Numerator of dn lO«

'^__ -

50

k-

il

X '

J ' ! +

400

k'

+

S25

A-.j

I '

JT ' 1 —

1 120

k'

_

7920

i'l-

10560

i =

z •

J. ! 1 +

I2S0

k'

+

23760

A-'

+

136620

A'

4-

68640

A- =

X '

J. 10

_

512

j.i.

28160

A-'

_

799656

A«

—

1034176

A'

256256

k'

x"

jU

+

115=0

k"

+

3659440

A-'

4-

74003S0

k'

4-

4236960

A'

4-

5S2400

k'

x"

z"

1407SSS0

i-l-

410S64S0

A-

_

33647040

k'

—

10252S00

A-'

S 19200

k'

x"

z"

40930560

*■=

+

17S0S7920

k"

4-

215138570

k'

4-

S6936640

k'

4-

1 50SS640

k'

4-

696320

A»

x"

z"

S6I41440

k"

55S046720

k"

_

1048703460

k"

_

665024640

k'

—

13391S730

A«

-

13271040

/;'

-

337680

A'

x"

T^

+

I305722SS

k"

+

I 2436 I 0368

k"

4-

3378663088

k"

4-

37S6222762

A"

4-

1404885056

A"

4-

121844736

A«

4-

6414336

A'

4-

65536

A»

X"

z-

I4I3I2000

k"

1975541760

k"

_

S5073172S0

A"

—

4 14347221840

A"

-

9937492160

k"

—

23830S60S0

A"

—

60456960

A»

—

1310720

A'

x =

^U

+

106496000

i=»

+

2223104000

k»

4-

14222557440

k"

4-

• 36722656S80

k"

4-

427433575S0

A"

4-

31637500960

k"

4-

3731362560

A"

4-

12615680

k'

x"

z"

53084160

J=

1731461120

A»

_

16682641920

k"

—

6490S2649Ö0

k"

—

119599991560

A"

-

1054S4397760

A"

-

41020335360

A"

—

5380341760

k'

x*

XS

+

I 572S640

k--'

+

887685120

A=

4-

1344S028160

A--

4-

79551371520

A"

4-

2283S5S4S0S0

A"

4-

324061 5667S0

A"

4-

219786648480

A"

4-

653379344S0

A»

4-

64S0S55040

A>

x"

J»

2097152

i»

2694S4032

A='

7097597952

A^'

_

6645S427392

A--

-

3049337 1 2S64

A"

—

679787735472

A"

—

73 156S5 16672

A"

-

372170115072

A'=

-

S2463293440

A"

_

607S 595072

k'

x*"

1 =

+

36700160

I»

+

2208S90880

A"

4-

36134461440

A=

4-

289564974080

A»

4-

1023576653520

A«

4-

16760S2159495

A"

4-

1319059633920

A"

4-

492633630720

A"

4-

S0103260160

A»

4-

42S9331200

k'

^c

x"

307363S40

k«

11533025280

A='

203559393280

A-

1166796195840

A->

-

2823637871790

A'«

—

324731053S240

A"

—

1836371937040

A»

—

497330749440

A"

_

583303 16S00

A"

_

2202009600

k'

x"

z"

+

I64003S400

k"

+

119826636800

A"'

4-

109204S5S36S0

A-

4-

373716SS44960

A-'

4-

5972935435935

A"

4-

4789004762880

A"

4-

1934627502080

A"

4-

374696591360

A"

4-

30804541440

A"

4-

775946240

k'

z"

X^

71973321600

i»

923363205120

A"

.^

4177417S44160

A-'

—

«784377578080

A»

-

9430334559360

A»

—

5390736855040

k"

—

1512473331360

A"

-

203961139200

k"

—

11135877120

k"

—

'67773160

k'

X"

l"

+

444S2SS76S0

7i=»

+

735394317312

A»

4-

4133945221632

A='

4-

I0g6c»S85773l2

A-

4-

15010S73077912

A»

4-

11019888419904

k"

4-

43059 123S2464

A>«

4-

850391793664

k"

4-

75754635364

k"

4-

2464153600

k"

4- 16777216

i-

X**

!•=

23248S96000

i»

492876595200

A^

3569667118080

k^

_

11S73296419S40

A~'

-

203210SS682320

A»

—

1S767855111040

A"

—

9430732239360

A"

-

2506234429440

A"

—

32501923S400

A"

_

17175674880

k"

- 25165S240

A"

z"

x"

+

904396SOOO

k^

+

26475724S000

A™

4-

25400567S0S00

k=>

4-

10897416491520

A-«

4-

23635664655S40

A-"

4-

37366147280440

k-"

4-

17193073638720

A"

4-

5779037790730

A»

4-

987231191040

A«

4-,

75593154560

k"

4- 1793064960

A"

z"

x"

2464153600

i"

_

105670246400

A--»

_

1406357913600

la"

—

80S 1396736000

k^

-

229292S8782400

A'«

-

34143S77303200

k"

—

2721928S969600

k"

-

11473036876800

A"

—

2437993267200

A»

—

235805081600

A»

- S074035200

A"

z"

x"

+

419430400

If

+

29491200000

A"

4-

577290240000

k"

4-

4627593830400

A-"

4-

17767399104000

A»

4-

35157394937300

k"

4-

36S497045S7500

A='

4-

20276745840000

k"

4-

5564509386400

A"

4-

672321536000

k"

4- 25790054400

A»

i"

I*

_

33554432

k"

_

5133S2S096

A-"

—

165365415936

A-»

—

1959063453696

k"

-

10565256379393

A»

—

28629943759872

k-'

—

40717498765944

A»

-

30540566594304

k->

—

1 1657673990144

A-=

—

3045715742720

A-»

- 134354953316

A"

I"

X*

+

419430400

k"

+

29491200000

A»

4-

577290240000

A^^'

4-

4627593S30400

A"

4-

17767399104OÜO

k"

4-

35157394937300

A"

4-

36S497045S7500

Ic"

4-

2027674584Q000

A"

4-

5564509286400

A"

4-

672331536000

A»

4- 35790054400

A-"

z"

if

_

2464153600

k"

—

103670246400

k"

—

1406357913600

A^

—

80S139G736OÜO

7(-"

-

229292S87S2400

l"

34143877303300

A-"

—

272192S8969600

A«

—

1 1473036S76800

A--'

—

3437993367200

A-«

-

235S050S1600

A»

- 8074035200

A»

x"

x»

+

904396S0OÛ

A-»

+

26475724S000

A-»

4-

25400567SOS00

A"

4-

10S97416491520

A--»

4-

23635664655840

JJO

4-

27366147280440

A'*.

4-

i7'i9307363S730

A"

4-

5779037790720

A='

4-

9S7231 191040

A-

4-

75593154560

A»

+ 1793064960

A"

I"

x"

-

2324SS96000

l-»

—

492S76595200

A--'

-

3569667118080

A"

-

11 8732964 I9S40

A»

-

203210SS682320

l"

18767S551 11040

k->

-

9430733339360

A-'

—

3506234429440

k"

—

32501923S400

k^

_,

17175674SS0

A»

— 25165S240

A"

x**

J"

+

444S2SS7680

k"

+

725394317312

A---»

4-

4133945221632

A-"

4-

109Ö00SS577312

A---'

4-

15010S73077913

A™

4-

1 10 19SSS4 19904

A"

4-

4305912382464

A-'

4-

850391793664

k-'

4-

75754635264

A=

4-

2464153600

k"

4- 16777216

A"

I»

X»

-

71973321600

k"

-

923363205120

A^

—

4177417844160

A->'

—

878437757S0S0

A-=

-

9430334559360

A-™

—

5290736855040

A-->

-

1512473331360

k^

—

203961139200

k-'

-

11135877120

A=

—

167772160

k"

i ""

l"

+

164003S400

/k"

+

1 19S26636800

A»

4-

10920485S3680

A«

4-

3737168844960

A-"

4-

5973935435935

p.

4-

4789004763S80

A-»'

4-

1934627502080

A"

4-

374696591360

;.=•

4-

30S04541440

A''

4-

775946240

*=

z"

x"

-

307363S40

k'-

-

11533025=80

A"

—

2035593932S0

k^

-

"1166796195S40

A^

~

3S23637S71790

;.•"

_

334731053R340

A»-'

-

1S3637 1 927040

A"

—

497330749440

A-»

—

5S3303 16800

A»

—

2202009600

A-**

[x-

I"

+

36700160

k" :

+

220SS90S80

A"

+

36134461440

k"

4-

289564974080

A™

4-

1023576653320

A»

4-

1676082159495

A^"

4-

1319059633920

A-™

4-

492633630720

k"'

4-

80103260160

A»

4-

42S9331200

A«

x°

x"

-

2097152

k"

—

2694S4032

A"

7097597952

k"

—

66458427393

A"

-

304933712864

A-"

_

6797S7735472

A»

-

731568516672

A-"

_

3721701 15073

l"

-

S2463293440

A"' 11 -

6078595072

i"

I"

x"

+

1 5728640

;,"

+

SS7685120

A"

4-

1344S0281ÖO

k"

4-

79551371520

k"

4-

22S3858480S0

A-"

4-

324061 5667S0

A-«

4-

219786648480

A--»

4-

65227924480

A«

4-

64S0S55040

A-""

x"

l"

-

53084160

;.'•

~

1731461120

A"

16682641920

V-

—

649082649Ö0 A"

-

119599991560

A»

—

1054S4397760

k"

-

41020335360

A-"

-

5380341760

K"

x"

x"

+

106496000

;.'"

+

2223104000

V

4-

14222557440

k"

4-

36722656S80

k"

4-

437433575S0

A-»

4-

2163750O960

k"

4-

3731362560

A"

4-

12615680

k"

x"

x"

-

141312000

A»

1973541760

k"

-

SSO73172S0

k"

-

14347221840

A»

1 -

9937493160

A"

_

23S30S6080

A»«

-

60456960

A"

—

1310720

Ar

x"

X"

+

1305722SS

A«

+

124361036S

A"

4-

3578663088

l'-

4-

37S6222762

A»

4-

1404885056

A»

4-

121844736

A^«

4-

6414336

A"

4-

65336

A-»

z"

x»=

-

S6141440

A" ij —

558046720

A"

—

1048763460

k"

—

6Ö5024640

A"

-

13391S730

A»

_

13271040

A-"

—

3276S0

A"

x^

x"

+

40930560

A» +

17S087920

A"

4-

21513S570

k"

4-

S6936640

A»

4-

1 508S640

A™

4-

696320

A--'

X "

X« -

1407S880

A-" —

410S64S0

A-"

-

33647040

A-"

-

10252800

A"

-

819200

A»

X*

x" +

11320

i"| +

3659440

A-"

4-

7400380

A"

4-

4236960

A'-'

1 4-

5S2400

A"

x"

''" 1 -

512

A»l-

28160

A"

799656

A«

—

1034176

A"

256256

A«

1

x^

X" +

12S0

A»

+

23760

A"

4-

136620

A-"

4-

6S640

A"

x«

X-' -

1120

k"

i"

7920

A-"

10560

A-"

X»

x" +

400

A"

r

825

ft"

j

x'"

:.: 7

^?

A-»

A"

:l

!

_

J

■^ ^

The Common DenommuLor ol' bU

cii lO», dn lO«

X ■

+

I
Ü

A"

X -

J- '

+

.S25

10560

A-^'
/.-•+*'

T '

X '■

X '

—

GSÖ40

Ic- + 7,'

1537S0

k'

X '

X '°

,. 11

+

+

256256

582400
s 1 9200

+
+

109S240
4382560
10457600

k'+k'
k'+k'
k'+k'"

+

8550620
3633S240

k'
k' + k'

—

696320

k-+Jc^>

—

15262720

k' + k'-

—

90S 1 7600

k'+k"

—

152392630

k"

I«

.f "

+

32;6So

X,-=+7.-«

+

13352960

A' + A"

+

1372S76S0

k'+k''

4-

318S53120

AHA"

x"

.r=»

_

6553Ö

k-+k'>

-

6430720

A' + A-'"

-

12345S560

k' + k"

-

154435520

A «4- 7,;'=

—

16616426

A'»

x^

X -

+

1310720

A ' + /.•"

+

60/S4640

k<'+k"'

-

931258S80

A» + Ä"

-

3S57933120

k"- + k''

x~

X-'

—

126156S0

Â.-«+y«

+

-9971955^0

AHA"

+

1 627075 1 200

A-'o + A"

4-

26972129580

A'-"

X-*

X -^

-

5224693760

A »+7.'»

-

39140076S00

A" + A"'

-

95711644160

k'^+k"

X™

r-'

+

64S0S55040

AHA-»

+

661566S2240

AW+A-"

+

226260775S40

A:«+Ä'«

4-

334446991 7S0

A»

«=»

,/ "'

-

607S595072

k'+Jr-

-

839S294220S

A'^+A^'

-

391379236S64

A'= + A"

-

809030S82880

A-"4-A"

x"

X ^-

+

42S9331200

k^+k'>

+

81175592960

A'» + A"

+

513329653760

A>M-7,»

4-

1448828156160

A" 4- A"

4-

2028073628295

A>=

x»=

.(■='

—

2202009600

k'+k''

—

58SS08 19200

A'-' + A^'

—

512257392Ö40

A>H7r=

—

1 970076774400

A-'*4-7r"

—

3776S71997440

A« 4- 7,"

x"

,M

+

775946240

A-«+A-»

+

3099S5 28000

k^' + I^'

+

3S251S96S320

A-'-- + A-='

4-

2033205790720

A" 4- A;"-'

4-

5334891769600

7.-"4-A-'

4- 7318608629345

A"

x^

.)■"

—

1 67772160

Z.-HA:»

—

1 1 177S20160

k">+k^

—

206771 3228S0

A'HA==

—

1565223157760

7.-"4-A-'

—

5702045050880

A'HA--

— 1072S905233280

A-'» 4- A™

X ^

.f "

+

16777216

k' + k"

+

2468347904

k"+ k"»

+

76373295104

A'-'+A=«

4-

869717442560

7,-"4-A-«

4-

453063735S080

A-«4-A=''

4- 1 1 843898229696

k'»+k~

4- 1624200S691672

A"

X "

.,■•=

-

251658240

Ä:'» + A:=2

-

17238589440

Ai= + A»

-

329352478720

k" + k^

-

2590202429440

k^'+k-'

-

9651975390720

A« 4- 7.-'

— 18331600645120

k''+k~

X '-'

.-■"

+

1 793064960

A-i=+A-'=

+

76041420S00

A" + /f"

+

1 0064096460S0

A-i'+A™

4-

5593290759680

k'>+k-'

4-

1499231321 1200

A™4-A-'

4- 20691061222280

A-"

x"

,. K

-

S074035200

k'' + k''

-

^37S23590400

k^'+k^

-

2140S26S2SS00

A» + 7.:='

-

8353239065600

k^'+r-"

-

16094921814400

k'^-+l"-'

x"

.(■"

+

25790054400

k"+k^-

+

443684454400

A'HA»

+

2645 74S0 19200

k^'+k"

4-

7245507984000

A^-'4-7r"

4-

1004S347400300

k-'

X "

X °^

-

25790054400

y«+p'

-

4436S4454400

r-'+l"-

-

2645748019200

A'-- + 7v»

-

72455079554000

A='4-A=*

-

1004S347400300

A"»

X ^

J. M

+

S074035200

7."'+^»«

+

237823590400

A» + A:»*

+

2140826S2SS00

7rHA'=

4-

!^353239o656oo

A- '4-7,^»'

4-

16094921814400

k'' + k^

X ^

rM

-

1 793064960

k'»+k^

-

76041420S00

7r> + A»«

-

1 006409646080

A-- + A-"

-

55932907596S0

7."'4-A'=

-

14992313211200

A--« 4- A™

— 206910612222S0

A^

X«

.r •■«

+

25165S240

^»+7,"

+

1 7238589440

A=»+A:»

+

329352478720

A" 4- A'«

4-

2590202429440

A"* 4- 7.:"

4-

9651975390720

7^=''4-A''=

4- 18331600645120

7.--'»4-7r^

X«

.,. ca

—

167772 1 6

7v«+7i:'=

-^

2468347904

A:™+7,:«

—

76373295104

k"--^+k"

~

869717442560

k'-'+h''

—

45306373580S0

A-«4-7r'

— 1 1S4389S229696

A=«4-A''-

— 1624200S691672

A'"

x"

.(■ ':-•

+

167772160

^■-" + 7,'-

+

1 1 177S20160

A:- + A"

+

206771322880

A« + A='

4-

1 565223 1577Ö0

k" + k''

4-

5702045050880

A^HA"

4- 1072S9052332S0

A:'"4-7r-

X "^

.,. Cl

-

775946240

7." + A:«

-

3099S52S000

A"' + A»

-

3S251S968320

A-=HA»»

-

2033205790720

A=»4-A»'-'

-

5334S91769600

k^ + k"

— 7318608629345

A"-

x"

.(• M

+

2202009600

k^' + k"

+

58880819200

A=«+A"

+

512257392640

A--^4-A»'

4-

1970076774400

A»» 4- A-'«

4-

3776871997440

7i»-4-/'"

X "

X ^

—

4289331200

Â.-°«+A«

—

81 175592960

A='+A"

—

513329653760

k^'+h"

—

1448828156160

A'HA»

—

202807362S29S

A-"

^C3

X'»

+

007S595072

k^+l«

+

839S294220S

A=" + A.'°

+

391379236864

k^' + k"

4-

809030S828S0

A" 4- A'«

X™

X '-'

-

6480855040

7.-»° + A'=

-

661566S2240

A'-H-A'"

-

226260775840

7.»' 4- A»»

-

3344469917S0

A»«

x'"-'

x-t

+

5224693760

A»-'+Â'-'

+

39140076800

A" 4- A'»

+

9571 1644160

A=«4-A»

X ^*

X ''

+

1261 5680

k'- + k"

—

-997195520

A»' + A«

—

16270751200

7v»=4-A'>

—

26972 1295 So

A»

X '"

X-»

-

1310720

A»= + 7.:*«

-

607S4640

7v'" + A''

+

93123SSS0

A»» 4- 7.'=

4-

3S57933t20

7^» 4- A*'

x"

a: 80

+

65536

k'- + k''

+

6430720

k'^+k'"

+

123458560

A-'«4-Ä"

4-

154435520

A-'"4-7.^=

4-

16616426

A"

X *

J.- »2

-

327680

7.« + A"

-

13352960

A»+A«

-

137287680

7,:=-' 4- 7.-''

-

318853120

A'HA:'-

x«'

X"

+

6(^6320

8 1 9200

A-'^+A*
7."»+ A"

+

15262720
10457600

7,-» + A«
A<' + A«

+

908 1 7Ö00
3633S240

7." 4- A«
A'=4-A-»

4-

152392630

A''-

x"

X™

x«

+

5S24OO

A" + A"

+

43S2560

A'- + A"

+

8550620

A"

J. SS

X»

—

256256

7.-*= 4- A«

—

1098240

A" +7,'"

X*'

XK

+

6S640

A"+A«

+

153780

A»

1

X«

j: ;i.

-

10560

A'^' + A"

x"'

X »e

+

825
0

1

ko
A"

X«

X '•*

The Diagram of the Semi-destructive Earthquake
of June 20th, 1894 (Tköyö).

By

S. Sekiya, Rigakuhakushi. Professor of Seismology,

•Alia

F. Ömori, Rigakuslii,
lijjpencil L liiversity, Japan.

rri. XXX.j

The eartlR|iiake of June Z^Otli. 1N94, was the inoat violent, that
lias shaken rökyö since the well-kncnvn Lireat catastrophe ol the l^nJ
Year of Ansei (IS55). The mean radius of the disturbed area was about
NO ri or 200 miles, and the total land area was 7.10U square ri, or
•4:^,000 square miles. The meizoseismal tract A\as a zone lying to the
east of Tök\o. and extending in a X-S direction from the vicinit\' uf
ihc town of Iwatsuki to Tokyo ]xi\', Xo house was absolutely des-
troyed, but iu the hjwer parts of Tökvö. many brick buildings receiwd
severe damage, and chimneys in particular were mostly thrown dcjwn;
some do.:û (godowns) h;id their plastered mud walls ver\- much cracked
and shaken down, tomb-st(jJies and islüdOrü (stone lanterns in gardens)
were overtiu'ned, small cracks were formed in the ground, and, in a
lew cases, ejection of water took place. The number of casualties in
the three Prefect lu-es of Tokyo. Kanagawa. and Saitama were 2G per-
sons killed and 171 wounded. In fact it was the severest shock that
the younger generation has lelt in this metropolis.

The diagram of the earthquake (PL XXX.) was taken by a
Large Motion Seismograph, set u\> in the Seismological Institute of

1^90

s. SEKIYA AND F. ÔMOKI.

the Univei'sitv. whicli was specially desigDed for recording strong
eartliquakes. The instrument is in principle the same as those often
described in the papers in the enrlier numbers of this Journal treating
of seismological subjects. The main differences are, ((/) the working
])arts are made stoutei' to withstand severe shakings, (h) llie writing
}>ointers are made longer so as to record large ranges of motion, and
(r) the p(jinters have no nudti[)licati<)n ratio so that the actual magni-
Uidc of the motion is givcü. This !-< the first time that a clear
instrumental reconl of a destructive earthfpiake was evei- taken in
this count rv : ]»r(jbal.)ly no such has ever been obtained in any other
countr}'.

The Seismograph records the motion decomposed into three
components. The wave lines on the two inner circles indicate one
tlie XE-ISW, and the other the SE-XW components, and that on
the outermost circle the vertical component. By compounding the
(•om|ionent motions we can und the resultant. The recording plate
revolved once in ILS seconds, and the short radial lines mark succes-
sive seconds of time counted from the start. AVe can thus determine
the mtiunitude and direction (^f motion at anv instant during the
earthquake, as well as the durarion. the intensity, and other elements
of the shock. ]>elo\v we give results deduced ïvoui the diagram.

Time of üccüiTcncc. lSi)4, June l^Otli, '!'■ V" W p.m.

Jlori'jonlal Moliott. The earth(pial-:e began as usual with tremors.
The duration of tremors, as indicated liy ordinary seismographs, lasted
about 10 seconils. The Lar^e ^Motion Seismograph dc^es not record
minute tremors, and the motion w;is already a few millimetres in
rarjue at the beij-inninu' of this diai>-ram. Howe\er. we shall take this
latter point as the beginning of the eartlujuake. The motion, already
strong in the Lst and '2nd seconds, became suddenly violent and
the ground moved o7 jnm. durinir the time interval Ijetween the ord

THE DIA.gr AM OF THE SEMI-DESTRUOTIYE EARTHQUAKE. 291

and 4th soconds. This was followt'd hy a oounter-inovemont of 7o
mm., Avhich was tlie maxiimini Jiorizontal motion «hivinu" tlie earth-
("(uake, and was a£:Tiin followed hy a motion of 42 mm. The m<jtion
during ahont 1 minute succeeding tlie ahove three most prominent
shocks was verv mucii -weaker, though still great in range. Some
large undulations ocrau'red hetween the 40ih and nDvd seconds and
again hetween the 7<)t]i aiid 78th seconds. Ihit the intensity of
motion was not so strong as hefore. The com])arativelv little damage
occasioned hv the prescrit earthquake notwitlist;uiding such great hori-
zontal movements is no douht dn.e to tlie sin;!!! numl)er of violent
oscillations.

reriod of rfnrl:())it<il Motion. 4"he rdi(Tve maximum liorizontal
motion was executed in 0-1) second, so that the complete period of
oscillation wouhl he InS second.

Direction of Motion. 44io directif^n of tlie motion changed as usual
during the earthrpiake. Ihit the maximum horizontal motion was
directed toward S, 70'' W., and the chief movements l^efore and after
were also directed nearly towards the same point, or else the opposite
direction. We have also examined the directions of overturning of
2ir) stone-lanterns Çisliiilôrô') in different parts of Tokyo. Their mean
direction of overturning v\-as toward S. 71^ \V. 44ius the direction of
overturning of columns is seen to i)e identical with that of the maxi-
uHun liorizontal motion.

1 ertical Mntion. The maximum vertical motion of 10 mm.
occurred in the 'îixl second nearly sinudtaneouslv with the iirst ])ro-
minent h(^rizontal niotion. AVrtical moti<uis o<:curred more or less
during the next ,')0 seconds.

Maximum Actrl erat ion. The maxitmun acceleration of the motion
of an earth particle, calculated from the ahove values of the maximum
horizontal motion and its period, is 444 mm. per sec. per sec.

292

s. SEKIYA AND F. ÖMORT.

This was the niaximnni acceloi'alion in tlie upper districts of Tokyo
where tlio u'roiind is composed of hard loamy soil ; in the lower
districts, where the ütoutkI is soft and marshy, it nearly reached
the value of 1,000 mm. ])er sec ycr sec. This latter has l)een cal-
culated from a similar rcmrd oiven l)y anotlier T>arixe ]\rotion Seismo
gra])!! set up in the lower ])nrt of tlie rity. Xow tlie maximuTu a'-œlera-
tion is the (juantitv which measures the destructive power of earth-
quake motion, and it may tlierefore he inferred tliat wluMiever it reaches
the aho\e values, chimnevs will he o-reatlv damao'ed and hiiildinii's
affected ns on tliis occasion. Tor the Mino-Owari ({reat Earthrpiake
of 1801, one of us has cah-nlaled, from ohscrNations of numerous
overturned and Iractured l)odies. the maximum ar-cclci-alion of (earth-
quake motion in the meizoseismal district to have l)een from o.OOO
mm. up to nearly 11). ()()() mm. j'cr sec. ]>er ^co.'^ These i-esu Its are
pr(jhal)Iy tlie iirst numerical estimation that has heen made of the
destructi\e ])0a\('1' of u'reat ('arth(|uakes.

]>iiir(fii)n. of ihr /\(irlliqinil-<'. Tlie sliakin^ti' lastefl ahout 4 miimtes
and .')0 seconds.

1. See l'\ Oiiiori : Tulilt- of tin- orcrtiiniiiifi (icccJeration, etc. Scisinological Jrmr. .lajian,
Vol. I, p. 143.

Jour. Sc. Coll. Vol. VII. W, XXX.

Beiträge zur Theorie der Bewegung der
Erdatmosphäre und der Wirbelstürme,

(dritto Altliandlunp;)
Von •

Dr. Phil. Diro Kilao.

Profossov fiir rhy>;ilc niul ^ratlicmatik an der lanihvirtliseliaftlielion Facultiit
(1er Kai>ci-liche]i Universität zu Tokyo.

§ Xrr. — ])0(liii2-iiiio- fiii* cinon vom hewo^'üclien Wirbrlu'cbicro cn-cicli-

iKircii Oit 1)(M zwcir-iclier WirlK'lbildiinu-. Voriiiidci'mio- dci*

^\ iiidstäflce. der A\ indriclitiiiio-. und des rjifrdrncks in ciiioiu

.'Solchen OrL

Es lassen sich aiicli Icidit AViridstiu'k(\ AVindriclitmiu- und

r.nftdniek in ('iiieiii ovondK'iicn Ort als Functionen von der Zeit (/)

(larstelkMi. wenn ein krcisfin-inio-es Wirlielo-chiet nntei" dem Einflnss

eines anderen nnendlidi fernen AVir1)el£i-el)ietes iilier den Ort In'n-

wegschreitet. Ehe wir dazn .schreiten, haben wir die l'edino-iing

Nils Ekiioliii (S^ftuska vct. akail. Hamllingav V. \'>. X' 11) hat eini>;v Ôleicluin^-on, die
ich früher in diesem Journal (N'.il. 1 pa^-. 13;?-U'J) entwickelt halje, ak unrichtig- l^ezeiehnet,
weil ieli die horizontnle d. i. zur Xorinnle der Erdöl lerfläche senkrechte Couipouente der
Geschwindii>-keit mit di'r restiltirfiidcn Geschwindi^-keit verwecliselt halte. Ich halie dao-eo-en
Folg'endes zu erinnern.

Pa die Erde dnreh Einrühriuii;- d.'r mit dm- Erd(> heweo-lichen Coordinaten, und der
Reihuno-skräite als ein un1>eweo-ter reilaino-sloser Kiirper lietrachtet wi-rden kann, ha1>en di.'
Componenten der Luftgeschwiudigkeit v, u, ic der Bedingung- zu o-enügi-n

0=11 {■()<! (n.r) + r cos rn>j) + )r cos (nz)
und zwar für jedes System .r // r, welclies der Gleichung der l'lrdoherfläche genügt. Diese
Bedingnngsgleichung, der die Componenten dor Geschwindigkeit einer bewegten Flüssigkeit
an einer stamm rei'mngslosen Flädie durrluw genügen müssen, sagt nun nichts anderes aus,

204 '^- ^'^'i^^*'-

niihor festzustellen, unter der ein o-egehener Oi-t von einem der
kreisförmigen AVirbel gebiete erfjisst werden kann. Es seien Jl^ und
Ro der Halbmesser der beiden Wirbelgebiete (1) nnd (2) Xj y, und
Xo, Yo die Coordinaten des ^[ittelpunktes derselben, mid x v die
des gegebenen Ortes in IVzng anf den Scliwer])nnkt der beiden
"Wirbel gebiete. ]>ezeir'hnen ferner fi nnd /'., die Abstände der Wii-bel-

als (lass (lie zur starren Körperfläelie seiikroehte Ccruiponente der Geschwindigkeit versehwindet
fiir j(?des System .r, //, z, welches die Olierfiäche des starren Körpers darstellt, d, li, dass die
Geschwindigkeit der bewegten Flüssigkeit entlang der stari-en Körperfläelie keine andere Bosnl-
tirende hat, als die zur Xormal der Körperfliiehe senkrechte. Ich hahe daher bei der Ableitung
der allgemeinen Beziehung zwischen Isodynamen und Windbahn durch die Definition der
isodynamischen Curven darauf hingewiesen, dass sie nur für Lufttheilchen gilt, welche sich
an der Erdoljerflüche liewegen. Sie gilt darum nicht für die Coordinaten, welche die Gleichung
der Erdoberfläche nicht befriedigen, d. h. für Tiufttheilchen, deren Coordinaten nicht die
obenstehende G renzbedinguug erfüllen.

Das sogenannte Hadley'sche Priucip hat sich in der Wirklichkeit nirgends bewährt. Wenn
aber Ekholm wegen des Umstandes dass die Gleichung 21, [pag 148 Yol. I dieses Journals]
zu demselben führen kann, glaubt, diese Gleichung selbst als unrichtig, und jede Schluss-
folgerung daraus als falsch bezeichnen zu können, so erscheint mir dieses etwas zu voreilig,
denn die Folgerung jener Gleichung (19d) [W. Ferrel (Recent advances in the Meteorology
Washington 1886 pag 205) hat üln-igens für den Winkel zwischen der resultirenden Gesch-
windigkeit und dem Meridiankreis eine ähnliche Beziehung aljgeleitet] setzt ja, wie ich pag.
1 to ausdrücklich betont habe, voraus ; dass die WindViahn die Isodynamen, oder unter Um-
stünden Isobai-en überall unter dem nämlichen Winkel durchsetze : (/. /;. das sog. Hadley'sche
Princip setzt eine liesondere Yertheilung des Luftdrucks und der Reibungskräfte voraus.
Wenn man daher auch durchaus berechtigt ist, dem Hadley'schen Princip jede schlechtldnnige
<iiltigkeit abzusprechen, so ist uian keineswegs berechtigt, demsellien jede Giltigkeit schlechthin
abzusprechen.

Der ganze Einwurf Ekholm's rührt daher, weil er nach dem Yorgang des Dr. Sprungs und
A. die Ben^egung der I.uft unter dem Einfluss der Erdrotation, als eines discreten Massen-
tluMlchens betrachtet. Da auf ein solches, das sich entlang der Erdoberfläche bewegt., die
-Milenkungskraft der rotirendon Erde

2 X COR e. U-. , —2/1 sin 9. i:

die Beweguugsliiilm sowohl in horizontaler als verticaler Richtung alileukt , glaulit Ekholm
zu der Schlussfolgerurig ))erechtigt, dass auch die Lufttheilchen. welclie die unterste Schichte
der Erdatmosphäi-e bilden, nur eine Bewegungsf<^rm annehmen können, l)ei der die Resul-
tircnde dei- Geschwindigkeit an der starren Erdoberfläche gegen die Yerticale nicht senkrecht
steht und die Richtung ihrer Projection auf die Erdoberfläche, von der ursprünglichen Rich-
tung, welche es auch sei, abgelenkt wird. Ich halte es indess für unzulässig, die Gesetze,
gemäss denen ein discretes Massentheilchen sich bewegt, so ohne Weiteres, wie Ekholm es
thut, auf die Bewegung eines Flüssigkeitstheilchens zu übertragen ; da die Bewegung einer
auch incompressiblen Flüssigkeit in ihrem ganzen Yerlauf n-e.wntlkh durch die Yertheilung
des Drucks, und die Grenzbedingung bestimmt ist, nicht aber diejenige eines discreten Massen-
theilchens, welche lediglich durch den anfiaujlkh'n Ik'ia'gumisxu^tnml bestimmt ist.

Beiträge zur Theorie der LJeweguug der Erdatuiuäphürc uud der AN'irbelstürme. 2*J5

gebiete von ilirein Scliwer[)urikt zur Zeit t, so ist. wie wir in dem
vorhergehenden L*aragraphen gefunden haben ((Tlfielumgen 105 und
106 pai;" 37 l Vnl. TT dieses Journals).

WO yoi I'v2 den WV'rtli Non />i und/':, zur Zeit l^-o bedeuten und

/^i — + l'n<(<-'Ji !h = + lïA^'^->

Die Integration ist iil)er den ganzen Querschnitt des Gebietes der
verticalen Strömung auszudehnen und das obere Vorzeichen gib für
die aufsteigende verticale Strömung, und das untere für die nieder-
steigende. Es ist, wie wir gefunden hal)en

bezeichnet man diese Grösse mit c. so hat man einfacher

,^1 = /'ui Vi + U [>,= -/>oiVl + tt. ( 1 )

Ist ferner /i der ^^'inkel, welchen die lîichUmg ,"i zur Zeit / mit einer
gegen Süd gerichteten Linie scliUesst, so ist

Zi = Zo-^% (1 + ^-0-/2 (-0

wo A=^i^^ ist. luid /j den Winkel zwischen der Hichunii* l>

und einer nach Xord gerichteten Linie bedeutet. Die Grössen

geben die Enlfernung des gegebenen Ürles vom Mittelpunkt des

290

D. Kitao.

Wii-belii'eljietcs (1). resp. (l^). Die Hcdingiui,!^-, «la.ss der Ort x ij von
(leui Wirbelgebiete (I) oder (2) zur Zeit / erreicht wird, \^t

-Ri --= Vi-J-i — .r)-' + (//i - !/)-
oder ' _^

Indem \vir die INjlarcoordijiaten einlidiren und setzen

.^1=^//! CO.S /i .,-.,= (K, cos '/y, x — <tcüsy^

so erlialten wir

Bi' = l' ' + !'''-- !'!>i ^■''■' (Zi-/)

B.f=o:-+<r + -2 l>l>., cas (/,.-/)

odei-, indem wir die Ausibäieke (1) und (2) einiiihren.

Ob der L»euel)ene Ort /' Z von einem der AVirbelgebiefe, oder \nn
keinem erreicht winb hiingt daA^on ab, ob diese transcendenten
(ileicliungen reelle ^\'erthe für t geben, odei' nicht. Es ist leicht
dieses in jedem gegebenen Fall ^'ernlittelst der graphischen Methode
zu entscheiden. Zu dem l'jude denken w ir uns den Anlang der Zeit
so N'erlegt, dass '/^.) — '/, wink d. li. wir ziiblen / von dem Augenblick an,
wo der gegebene Ort auf der liichtung der < ierade liegt, welche die
Mittel^HUikte der Wirl)elgel)iete mit einander \erbindet, und setzen

AV(jbei ''^ negativ genommen ^verden nuiss. \\enn ^ negafi\' ist d. /;,
wenn die beiden AVirbelgebk'te cycl()nal sind, oder die Masse des

ÜL'iträ^'O zur Theorie der Bewegung der Erdatmosphäre und der Wirbelst ürme. 2[)7

(i)

cyrionaleii Gebietes grösser ist, als diejenige des aiiticyclonalea
GelVietes. Wir erhalten dann für das Wirbelgebiet (1)

Äi" = fr + />oi C'*^ — - f'f'iM G eus A //
oder indem wir tlies nach c"^ atiflosen.

/-'Ol e^=!>cos Kö ± ^{B;- — fr s in^ KÖ)
Wenn wir die Gleichungen

U = /'üi e"^ U = P <-'"5 ^^'^ ± \/{B;'-fPsin' Kä)

als Cnrven construirt denken, so geben die Din-chnitts[)unkte dieser
Curven die reellen AViu'zeln der Gleichung (4).

Die erste Cm-ve ist eine exponentiale und die zweite Gleichung
stellt eine Ueilie elliptisch geschlossener Curven dar mit der Periode

-7^, wo }n eine ganze Zahl ist. Die Figur (1) veranschaulicht den

uno-efähren \'erlaul' dieser CLU've, indem — - = — anu'euommen, und

die Breite der elliptischen CurveJi i!Tüsser u'ezeichnet wurde, als sie bei

diesem A'erhtiltniss ^on

B,

hätte sein müssen.
Fig. 1

ZK

■it

•2a

•27t

K

'^7t

AT

3 TT

K

•>()^ I>- Kitau.

Die geschlossenen Curven «ind congruent und symmtrisch urn
die Ordinate für -j^ . und haben dabei den Durchmesser l^i?i . -wäh-
rend der Durchmesser (2^y), wehdier die (.'lU'venpunkte Â=^ ±^ mit
einander verijindet

'2 . (RA

ist.

A\ enn die beiden A\ irbelgebiete anticycl<jna] sind, oder die Masse
(\{:^ü anticyclonalen Gebietes überwiegt, so trilt't die Ciu've <>^i c^ keine
der ovalen Curven. wenn /' + /i*! < />„! i^^t. Für relativ sehr kleinen
\\ erth von /^)i kann die Ex[)onentialcurve wohl eine der ovalen Curven
durchsetzen, Avie es bei A der Fall ist, und die AVerthe von ä, denen
die Schnitt[)unkte 1, '1 entsprechen, geben die Zeitpunkte wo das
\V i rbelgebiet (1) den gegebenen Ort erreicht und dann verlässt, und
zwar so, dass der dem iSchwerpiudvte zugewandte oder von demselben
al)gewandte Theil des AVirbelgebietes durch den gegebenen Ort passirt,
je nach dem die Schnittpunkte unter-oder oberhalb der Punkte liegen,

in denen -^=±x d. //. das Kadical ^H' — o-^in- Kö verschwindet.

Sind daaeiz'en die beiden Wirbel ^'ebiete cyclonal oder überwiest

die Masse des cyclonalen (Gebietes, so schneidet die Ciu've /'„i t~^ keine

der ovalen Curven, wenn /v^/'— ^^i ist, wie bei der Cur\e C der

Fall ist. Ist hingegen /v,i > ," + it'i, was bei der Ciu've D der F^dl

ist, so kann die F^xponentialcurve wohl die ovale Curve treffen und

zwar, Avenn /'ui gross ist, mehr îds ein Mal ; d. L die Gleichung (4)

kann reelle Wurzeln haben, wenn /'„i>-/' + i4 ist.

Die Dimension der ovalen Curven vergrossert sich nnt wach-
p
sendem A'erlùiltniss — und die Curve verwandelt sich schliesslich in

r

die Sinuslinie, wenn — =1 wird. Je izrüsser demnach Tii ist u'eo-en

Beitrilo'e zur Theoino <\qv Bowof^'unp,- dm- Erdatmosphäre und der "Wirliolstürnio. •^99

den Abstand des «gegebenen Ortes von dem Schwerpunkt, desto (U'ter
Averden die ovalen Cnrven durch die Linie f^oi'"—'^ o-eschnitten d.li.
desto ()fter wird der Ort von dem Wirl)elgebiet erreicht.

Die Zeitpunkte, av(^ der gegebene Ort iu ein AVirl)e]gebiet geriitli
und wieder daraus liervortritt, sind demnach l)estimmt, wenn die reellen
"Wurzeln der Gleichung (4) l)estimmt sind. Es seien diese ''\ und 'Vo .
wo '^i<rj'î.2 sein möge. J>ezeichnen ferner ^i und t. die Zeiten, wo der
Ort i> y in das Wirbelgebiet ein. — und daraus hervortritt, so ist

]\Iithin

'1 —

Die Diitereiiz

t-ti =

o2t?l ^5l?

,^-"^1

giebt dann die Dauer, während deren der Ort in dem AVirbelgebiet
verweilt. Die Gnissen

giebt den Punkt der kreisförmigen IVgrenzung des Wirbel gebietes. wo
der gegebene Ort eintritt, und

den Punkt, wo derselbe austritt, und zwar in IVzug auf den gemein-
samen Schwerpunkt der beiden Wirbelgebiete.

So sind Zeitpunkte bestimmt, wo die in dem vorhergehenden
Paragraplien entwickelten Formeln für einen gegebenen Ort ungihig
und dann wieder giltig wer(h:>n. Innerhalb dieser Zeitpuidcte ^er-
iindern sieh Windstärke, AVindrichtung und Luftdruck narh and(M-en
Gesetzen, die wir jetzt ableiten wollen.

PjOQ D. Kitao.

AVir nehmen zanjlclist ;ni, class die heiden in l^ede stehenden
Wirhels'ehiete cvclonal seien, und setzen

r = VC''r-.^2)' + (//i-//o)-
i /Ci - -/■ ■'''< ^>) '1<'h ^ i p ^^'i '^"^1 = ^1

wo (?f'j ein Qnersehnittselement des AVirl^elii'ehiefe-j. nnd ila' dasjenin;e
des (Jehietes der verticalen Sti'()nnniinLî- Ix'deiitet. Da hei anf-
steio'ender Siriununf vom kreisfin-miü-cn (^)n('rschnirf das (ieljiet
derselhen mit dem Wii^helüchicte zusammenfüHt. liat inan. wenn die
lieiden AVii'belü-el)iete cvclonal sind, do -- d(o' zusetzen. AVii* setzen
fcrnei-

und hildcMi

^11. , SJxV ,,

-r =J>., =7jo

SO dass wir erliaUen

Beitrii^-e zur Tlieorie dor Bowegung der Erdatmosphäre und <1<t Wirlu-lstiinno. ^01

Es ist iriilier (pn.ü'. ool \o\. TI dieses Journals) nn el lU'e wiesen worden,
diiss der Einfinss der unendlich fernen AVirLela'ebiete auf ein \\ irlicl-
gebiet beriicksichtio-t ist. wenn wir der Function AV eine lineare 1- nne-
tion liinznfiiu'en. dei-en Coefficienten auf u'ewisse AVeise von den Coor-
dinaten der Mittelpunkte der AVirbelgebiete abhängen. IVi zweifaelur
\A'ii'bell)ildnng ist diese lineare Function fiir das Wirbelgebiet (1)

(7?/-/ii).r-(/i/ + I^iVv'
und fiir das AYii-bel gebiet (l^)

{B:-A.).r-{A:^B.Uj
so dass wii' für das Wirbelgel)iet (1) erhalten

Vi

-^ / '^'Y/ /'' '''''"'

wo

/''==V(.Vi-.r)H(//i-//;
und \V' der Differentialgleichung genügt

^-/TTV l AW{n TTV aj£i\ aJTTV _ /^ _ l TTV\

+ {n - ;'iU TTV + -^z- •'' /■ " ^^/'i =-- ^>'

tPag. 351 II Vol. dieses Journals ist, ein Druckfehler nul )eiiiprkt geblieben. Das Vorzei-
chen des Gliedes (B/— A^^) x ist positiv, wie hier angeo-el^en ist.

302 ^^- Kitao.

imrl i'iir das Wirhelgebiet (2)

p'^^/{x,-xT- + {y,~l/f

wo IIV der (rleicliung geniisl"

A\'ii' liaheii soiniMiIs ( ^)inponenreii der <ies«-li\viii<liLik('ir in doiii
A\'irl)(-'ln-ebio(e (1)

mid ill ilciu Wirbel ucbieto (2)

?r,-

l-'iir einen kreislVtrniigeii Querschnitt ist. \Yie wir l)ereits (png
ISS 1 dieses Jonrnnls) entwickelt Indien

Beiträge zur Theorie der Bewegimg der Erdatmosphäre und der Wirbelstürme. 303

wo zur Abkürzung gesetzt i.st

l^['-ï(i)^(K-Oj..y,

Eiitsprecheiidt' Ausdrückt' für das z\\eite A\'irl)e]gebiet sind durch
A ertauscluuju" des Iudex al)zidcilen.
Es ist nun

y --^ (>i + [>i

wo pi /ö- die absolut ecuomnienen Abstünde der A\'irbe]<:ebiete von
ihrem Schwer}.)Uiikte bedeuten, so dass

uni

Es ist ferner

■A = ,^'i cos /i = ,^n ^i — et. cos [^/o-T7% (i--^0 I
!Ji=l>isin /r-^A,iVl-î^ -S'" j_Zu- TT^ofj [i-st) 1
^2 = /'2 c'ü6- /,, - o„,VJ-s^ cos F/, - A/o^ ( 1 _ ./) I

304 ^^- '^'^tao.

l^v (//i — !/-2^ — !h ('■^-'i — ''è _ Pa ^'^ii- 7.1 — !h ''-Q-^ '/.i
)•- ~ r

P. slit \jj--^IOfj{l-St)J^ -/L,CO.S ^/, ^%(l-c^) J

1\ (.i\ — .i\,) + //.J ( y/i — //,.) _ P. cos /i + /t, si /I y y
r r

P., cos \y —-- %(!—;/ ) -^ ft. s ill X-2+ 'jlo;/{\-tt) I

f' ^ V/'f' + /'' - 'A"r-2 ^'Oii (z - zi)

^ 7/'"-' + 'M ( 1 - -^0 - 2//o„i vi^tt: ro6' 1^/ - /, + ^ kuj ( 1 - s/) j

Als Coiiij)Oiiciiteii der Windgesclnviiidigkeit iiidcui krei^füniiigcii
Wirbclgebiete (1) bat man

iil^ — ;- I /> cos / — />i cos /i + / (/' ) (," s ni y — fi, s tu y y) I = '-^ ' i^

(■>)

/'il- ■/ i\/ 1 P« c(js '/■, + n., sill y,
/'!= — — j^/y stii y-i>i sut yy- f {'/) {<> cos y-o^ cos y^) I +— ^ ^- ^. ' ^

woiiiil dif ('<)iu}MjiR'iiteii dvr W illd^■e.■>ch^villdi^■k^Jil iu dem gt'gcbentMi
Ui't /'i / zLir Zeir / be.'stiiiiint sind. AVir wollen zmvst die Grössen /^^
nnd /'■-' bestinunen. Es ist venmÜLie der GJeicbunii' für -l\V=--':l + -2Äsiii.O
(|):il;- 11)4. \o\. I dieses Journals)

wo — - — ?//., L^'esetzt ist. Die AusfilbinniL;' iler 1 ntei-ration eruiebt

Beitrage zur Theorie der Beweguuy der Erdatuiuspliüre uud der \Virl>elstiirme. ,')(),')

Es ist il'fiu'i-

!h -- T7±- / / p '^r '' 1 '- ^-^^

— " .' II •'' II —

Mithin isr

y\)

Xun ist "^^^ die Tanu'ente des Deviationswinkels im \virl»el-

freien (Jebiet. iJa derselbe im Allü'emeinen sehr ü'ross und "^

verm util hell >- 2 ist,* so Iblii't. dass im Allgemeinen

P, >> //,

sein mus.s. Als die Windstärke für den u'eu'ebeneu Ort zur Zeit t
erhUlt man aus (0)

+ ^y^-!' ■''■"' ^/i ~ Z^ + /'2 \!'i ~ !' '''->■'' *Zi - Z']) ( -" )

Die Vertheilunu' der Windstärke indem A\ irbelii'ebiete ist dem-
nach nicht symmetrisch in liezug auf d:is (Jenlrum (/' !>i /i="Z 1'' — ^^)'
wo. wie es sein sollte

^ =/l ^' + — -^. —

ist, vorausgesetzt, dass •2a'<;-i oder nicht -"^ = ri ist.

* Ver^ieiclie Auuierkuu;^' pa;^'. liS I di.}ses Jouruals.

;jQ(j D. Kitao.

AVi)' AvoUeü die Gleicliuno- (J^ etwas vereiui-ichen. Wenn der
gegebene Ort f '/ unendlich weit von dem Wirbelgebietc (2) entfernt
ist. lind sich innerhalb des Wirbelgebietes (1) befindet, so ist / — /
unendlich klein, da die Entiernunu' des AVirbeli>'ebietes xon dem
SchwerjHinkt. an elcher gleichzeitig der Pol des PolarsN stems ist,
unendlich gross gegen den Querschnitt des Wirbelgebietes sein soll.
Diese Ijemerkuno- eroiebt

(S)

ist, oder indem wir die lieziehung (<!) berücksichtigen

Diese (J rosse ist verschieden ie nach der Lnoe des ücoebenen
Oi-tes zu dem Cent rum des Wirbelg-ebietes. Wenn in dem Au"enblic-k
^V') Zi — Z i-^t, der Ort so liegt, dass 1>\-<.1> ist, so ist Q positix'. Wenn
der Ort aljer so liegt, dass f>\ -<-!>, so ist Q negativ.

Im ersteren Ort ist die AVindstiirke grösser, als in dem letztei'eii
Ort. Miui denke sich das kreisförmige Wirbelgebiet in 4 Quadranten
getheilt. ilieils durch die Verlängerung der Linie !>i . tlieils durch
den dazu senkrecht mit dem Radius Ci beschriebenen Kreisbogen.
Indem Avir den nach der Eewe^'unu-srielitunu' uelewnen Tlieil des
Wirbelgebietes den Aorderen, und den dem Schwerpunkt zugekehrten
riieil desselben den inneren Iheil nennen, so ist. ila bei den Cyclonen
/i mit i wii'-list

Beiträs't' znr Theorie der Bewe<,'ung der Erdatmospliiir.i und der "\Virl)ol.stüruie. 307

(1) im vorderen äusseren Qaarlrante !>i— f> -^-^ '/_i — '/_-<. 0

(11) im vorderen inneren Qnadi'ante ,"i — /'>0 /i — X-<0

(£IT) im liinteren iüisseren Qiiadranlc /'i — ,"-<;0 ;fi — />-0

(TV) in dem liinteren iniicren <^)irulrnii(e /'i— /'>-0 /i — 7::;-l^^

Diese \ ier <,'iiadranfen sind in der Fiijür '1 mir ents|»i-( clieiMlen

römis hen ZiH'ern i)eztiflinet.

Fig. 2

+ r Süd.

In dem Qiiadrante I ist demnacli
In dem <,)nadrante IL

In dem Qnadrante III

Qm - ~[ !' (zi - 7) (.ny) + A) + (o - /':) ^v'/C'^O-l )]
niid in dem Quadrante IV end lieh

9iv = ^'[/' i'/i-Z) (/(-') + /^')-(f'i-o) (A/C"')-! )J
C^ ist demnaeh in dem Qnadi-nnte II neo-ativ. aber in dem Qnadrante

808

D. Kitao.

TIT positiv, weil f if') von clem Umfang des Grenzkreises des Wirbel-
o-ehietes nach dem Lenrrun zu von bis wachst, luicl

■2/ sin ti

n

immer >- 1 ist. Es iolut hicrans, dtiss die AVindstärke F in

dem liinteren äusseren <,>ii:idraiite den grössten, und in dem vorderen
inneren Quadrante den kleinsten ^,Verth besitzt, wenn das cycloiiale
AVirbelgebiet diu'cb ein zweites cycloruales AYirbelgebiet bewegt wird.
AW'nn das zweite Wirbelgebiet anticyclonal ist, so ist fu negativ und
bat [wie weiter unten na(bgewiesen wird] denselben Wertb

Fj^ wird

(!) in dem vordei-en änneren (,)ua(h'ante ,^'i— /"<0 /^i — />0

(TT) in dem vordeivn inneren (^)ii;!drante /'i — /c^ 0 /i — />-0

(ITT) in dem hinteren äusseren <^)ua(b'ante f>i—f"<J'' Zi — Z-<*^

(T\') in dem liinteren iiuieren (^)uadrante /'i— /'>0 /i — />0

IL)

2;. sin ft

+ ,r Süll.

In dem (^)uadi-ante (T) ist (]emnacli

Çi- -'-7^^[,'^ (/i-Z) (./"l/O-l- A') + (,^'-/'i; (A7(/0-T)]
Tn dem Quadrante ([[)

Beitrüge zur Theorie der Bewegung der Erdatmosphäre und den- Wirbelstürme. 309

Tn dem Quad ran te (TT F)

Çni= ~ \j (Z-Zi) (/^/'') + A')-(o-,^^ (^^/(/0-T .]
lind in dem Qn-adrante (TV)

AVieder ist die Windstiirke in dem vorderen üu.sseren Quadrant
nm klcin'-ten. und iii dem liinteren inneren (^)nadrante am oTü.ssten.
Diese A.-^vmmetrie gemalmt uns lebliaft an die soo-enannte sfef^dir-
lielie Seite der Seefalirer im AVirbelsturm. und sie ist auch durch die
fortschreitende l^)e\vea-unî:;' des Wirlielsturras erklärt worden. +

Als ^\'indazimuth i'iir den gegebenen Ort zur Zeit t erlüilt man
aus der (ijeicluing

_ ^hl '^11 f^-i' ''i

indem man fiii- ii, r, ihre Ausdrücke setzt

/ ;'i r • y/ /\/ ~\ PoCosy. + u^sinvA

tiiga

( — I <> coü y — o, COR y^^f (o ) (o .<;■ iny— (\ .<?? n y^) 1 + — ^-' ' ' '— j

Da yi!'', wie r. als Functionen von der Zeit bekannt sind, sd ist
aur-h iiuj. a als eine Function von der Zeit dargestellt. Das Gesetz.
gemäss dem das AVindazimuth sich in dem gegebenen Ort mit der
Zeit ändert, ist gleichfalls sehr verwickelt. Indessen ; da das von
dem zweiten Wirbelgebiete herrührende Glied im Zäliler und Xenner

weo-en des F\actors verbal tnissma^sior klein ist, so dürfte die

\ eränderung des Windazimuthes wenig abweichen von derjenio-en,

t Vergleiche Ferrel : A nmi;il report of the chic-f signal officer of the army for the year
IS3.J. part (2) Recent advances in meteorology, pag 267.

310 I^- Tf^itao.

welche eintreten würde, wenn der i^'-eo-ebene Punkt in einem sonst in
liulie betindüchen Wirbelgebiete seine Luge zum Centrum desselben
in der Richtung der Sehne änderen würde. Für einen Ort, über den
d'is Centrum des Wirbel gebietes hinwegschreiten k-uui. kann man
angenährt setzen

O := ft,

t i 1

l>etindet sich ferner der Ort n:iliezn südHeli von den Schwerpunkt,
SU ist ■/ klein, und '/-i untersclieidet sich von dem geraden Vielfachen
von ~ um eine kleine Grösse. j\ran kann daher für fiin y^ — ^in y^ , y—y.i

und eo.^y—co.^yi = 0 setzen. Da ferner — ^ = ^ " = tarii gesetzt werden

kann, so folo-f für einen solchen Ort

4 r (■/-'/i') ^ sin (/ + /i)

2 ^- '• r cos i '■

tag a = : —

2 ' ^- ■'• r cof! i

So lange der gegebene (^rt sich auf der Vorderseite des Wirbel-
gebietes befindet, ist y—yj^"'^^- In der Xälie des Centrums wird

(1. Ji. <y~'i-\-yi

die U'indrichtinig rnaclit mit der Richtung /i d. i. des /'i den AVinkol
/. (îei'ath nun der Ort in «lie Hiiiterseite des AVirl)elgebietes, so ist
/-/i<:0. Mithin folot

'à^ 1' ^/i - /) + ^^ -^"^ (' + /i)

'1 '■ '■ r cos % '•

tag a = —

fU

"2 ' ' '' '• r cos i '•

tni] <^' ist demnach grösser als l'ifi '^ ', denn die Diffei'enz tag (y' — tag <y
ist ])ositiv f;U' denselben Werth von /'' und ['iy—yù

Beiträge zur Theorie der Bewef^ung der Erdatuiu.spliäre und der Wirl^elstüruie. ß]^ |

da der Xenner durchaus positiv isL Es iblgt hieraus, dass das
Windaziuiuth auf der hinteren Seite des AV'irbelgebietes grösser ist, als
M uf der ^'orderseite desselben, eine Thatsaclie. die man in der That
bei manchen AMrbelstiirmen beobachtet hat. j

Es ist auch leicht, den Luftdruck in einem gegebenen Ort als
Function von der Zeit darzustellen. Es sei ^f\' die Gleichung für die
Isodynamen im Fall, wo das Wirbelgebiet (1) tdlein existirt. Sie
ist bestimmt durch die Gleichung (Gleichung (-14) pag. 170 A ol I.
dieses Journals).

Es lässt sich hieraus die Gleichunff der Isodvnamen für den
Fall ableiten, wo zwei Wirbelgebiete vorhanden sind. Indem A\ir
für IFj setzen

oder kürzer

/^;(.?/i-?/2)+/'2Ù'i-aO 73 _ ^'-i^yi-u^-f^J-^i-^^

YV ; — A X — 15 I , A. = ;;; J-> — 7,

erhalten wir

ex \o/j 01/ / ÔIJ \d?j (*.r /

(9)

— (â W\ — 2/ sin 6) A IFj - « yi ^ o

t Verg-leiclie. K.. P. Benito V^iues : Apuutea relativos à los huracaues de loy autillas en
setiembre v octobre de 1875-1870.

3 1 2 ^^- Kituo.

AVir denken vins eine Function ^^\' , welche diese Gleiclumg in
dem Fall befriedigt, W(^ A=B^O ist. und setzen

wo ^i 5^2 gewisse Fuuctioneu siud. die nodv niihei- vax bestimmen
sind. Die Substitution dieses Ausdrucke in die (ijcicluiug (9) ergiebt

01/ \ dy ax / o-r \ c/f C.JC /

Hieraus kommt N"ei'm(')ge der lîei^lcutuug der 1' unci ion '^\'

^ ^ Cl/ ô.i

Avas befriedigt wird durch die Annahme

^ 01/ ^ ' ÜX

Im J'^all eines kreisllirmig begrenzten A\'irbelgebietes lassen si(;h
diese Gleichungen leicht integriren. ]\lan setze

X — Xi . //— //l
cos (S — Sill c = -^

JJa in diesem Fall ^ilFi eine Function ^ on ," allein ist, so kann
man schreil)en

Jy''i = — Jb — ^ — sin ip -»V 2~ ~^^ — 1 — '-'^^' 9

do
Xun ist

Beiträge zur 'l'huorio dur Uewe^iiug dor Erdutuiusphäro und dor ^Virbolstürulo. ß] "'j

Wenn Avir ^^otzen

5^1 — / sin Ç

wo 1 /'' allein enthalten soll. s;o haben wir

d-1 \_dJ_ I j,dJW^

d t)^ 11 d<> </ d;*

Die Ijeiclen Inteiirale der reducirten Glei<'hnn_u'

do- o dy (>'-

sind

1

Mithin erhalten Avir

wo (' . C" elie beiden Integrationscon.stanien «ind.

AVeil nun MV., von der Form a — /;//' i.sf, wenn nicht -»^^j'i i-"5t, so
haben wir

J)a '/'i in dem Centrum des Wirbelgebietes hei jedem A erhäUniss
— nieht imendlieh gross werden darf. >o muss C — 0 sein
\\ ir erhalten somit

h B i>" ■

/^ bB,o"\

Der Ausdruck für ^2 lässt sich hieraus ableiten, wenn wir hierin
für B, A und für sin ç, cos ip einsetzen.

'Ml D. Kitao.

iJa nun

— r- 7^ = —

(m—\) y I

isfj (lalicr

2/ .sn; ^^ , 2/ .s-//? ^ / 1 V-1'" ~ 1)

(M-1)

{m-\) \hi/

(la tl'rijer fiif ''/'i' diu Fiuictioii ((JO) pag 1S»3 N'ol. I dieses Journals zu
setzen ist. so haben wir als (Gleichung der Isodynamen.. indem wir
/'' für y setzen

Die Gleieluuig der lsod\nanien lilr das äussere Avirbelfreie Gabiet
ist (Gleichung (50) }>ag. 17- \'ol. 1 dieses Journals)

(fJ = - ( ;r+ ) C„, + (Jons/.

/ , -i/rsni-OX

Avo (f,^ eine particuliire Lösung der Gleichung -l<fa^-^^ i«t- bezeichnet
man die Abstände des Punktes .r [/ \ou den Centren der beiden
Wirbelgebiete mit ,"/ und f>-j\ so ist für unseren Fall

f a = — /h % f'ï—r-2 log !'-2

Beitrüge zur Theoiie der Bewo^'un^ dev ErJalmosphüre und dor Wirbelstürme. 315

Xennt man feruci' den Wertli von '/'a für eine selbst gegen /'/ uimI
f'i nnendlicli grosse Entfernung ^/\,o. so ist

*^ao =^ +( ^f -i ) (;-'i log f'„+ff..2 loq n^) + coiixf.

wo ,«,, eine T.änge ist. die so gross gedacht werden k;inn. dass die durrU
zweifache Wirbelbildunix verursachte Störung der Atniosphiire in der

Entfernung /'„ unniei-klieli, d //.. ~— unendiicli klein wird. :dso dass

man iiir />ü setzen kann

'K.^-^+ G

wo P(, den Druck, und <f (his l'ofcntinl der Scliwere und /'■ die Difhre
der Luft he<hnitet. Wir erhalten somit

(I) r=l,c-{- ) { n. Ion ^-^ + n., Ion ^— + - '- + G

\ « / V /'ü ' /'o / !'■

An dein < ii-enzknäs des Wirljelgebietes (1) ist f'i=I^i. und /'■_.' kann
= ;• gesetzt werden. Soll daher an dem Grenzkreis ,"'=^i 'Ai^^^-'i
sein. d. h.

/ 4/- silt- ti\ r ^ 11. , / r \~\ P,

4 inil-mr \ 'lrn — \/

^ / _ 1/.si,idB\ -n / r,' •2).f^inOA\'
+ sm c. BAC— r- + ms c BA C ; r- )

so wird dieses erfiUlt. wenn wir setzen

2m{in — \} 2}n{m—\)

116 D. Kitao.

con, = A + G + (. + i^*) [,. ,o,,( A) +,,, ,0,{j-)-]

4 ;/^ {?n — 1

jMitliitj orlialfcn wir

rV- ( )ir — '2+ ~ r I

■)n{)n— I )- \ 'i w — 1/

^o' ro.s c =^/' ro.<; ^ — _Oj cos '/i
ist. so li:it innii. iiidein iii:iii fill' 1' mimI A iliro AVcrtlio einsetzt

// (J3 fiin c-\-A cos Ç j= (j'.j 7' ''■"■'^' ('/.i~7^~f'v + f'l!' ■'^'" "/i^/' )

(In icnier

ist, wo /'i // (leii Di'iick 1111(1 die Lnftdielite in dem Wirbelg-ebiete
bezeiclinet, und F, die Resultante der (leschwindigkeit. so erliiilt man

Beitrajït" zur Tlioorio dm- Bowci^img- der Erdatmosphäre und dor WirbelstiiriiK". ;)] 7

/.-sin-dB^- / ., . , 1 \

r-, ( 7)1 — '2 + r )

m {m ■

+ llialL 1 r , _ {jLf"' - ' '"1 (p., , ra. iy, -y)-o,+ ,,.,;. .in (y, - y))
}>i{vt—\) r l_ \hi/ J|\ ■ /

- T D'" =•' + 4 ( ' + ^^ '■"''') ''"' + ^V + '^G ' ' ° '

womit (1er Luftdruck in dem Wirbelgebiete als eine Function von der
Zeit dargestellt ist.

Die Vertbeilnno- des Drucks in dem AVirbelgebiete ist nidit
8vmmetriscli in Ik'Ziig anf das Centrum. lîezeiclinet man die Glieder,
die nur /'' entbaltcii. mit H , und setzt man wieder cos (/i— ;{)=1 und
si)i i'/i—'/)'^'/.i~7,- ^" erhält man
P^ ^^ /.si)tß 1 r. //''\-('"-^'~l /t^ / N, ' A

-^~ L" ^''■■- " ''^ y " ' + "^r-) - ' '•■ - !" \—jr-f '■" ' - 1 ) J

oder indem wir die lîezielinnu' (H) beriicksicbtiu'cn. fiir/C"') ihren Aus-

T , . ,..., T -2 /.sin 0

(UMick oiijiuhreii. und = A setzen

//. '2mr tjn~\ \ \ ii 1 / )

_A'

818

ü. Kitao.

â.

2mr (?/i— 1) L ( \J^i/ ) J

Der Factor /iT^ ^1 + w -2 (-^V '" J-w+l ist positiv, so lange

m > 1 ist, da A- > 1 ist. Ist aber m < 1 , so wird der Factor für /'' = -Ri
= (A'-— 1) (;;i— 1), also negativ, und für ^«' = 0 . negativ unendlich gross.
Die Grösse

{m

^,["i>-«-<4r-"l-'-i]

bleil)t darum durchaus positiv, welchen AVerth vi auch haben mag.
Es folgt hieraus, dass der Druck Pi auf der äusseren Hälfte der Cyclone
{o>pi) bei derselben Entfernung des Centrums grösser sein muss,
als auf der inneren Hälfte {f'>pi), wenn /'s positiv ist, d. h. das andere
Wirbelgebiet auch cyclonal ist. Das Umgekehrte findet statt, falls
f^2 negativ, ist, d. h. das andere Gebiet anticyclonal ist.

Beobachtuno-en, an welchen man diese Asvmmetrie der Druck-
vertheilung in einer unter dem Einflass eines anderen Wiibelgebietes
wandernden Cyclone wenigstens qualitativ prüfen könnte, liegen, so
viel ich weiss, noch nicht vor. Ich bemerke hier nur noch, dass nach
Thom der Ort des tiefsten Barometerstandes in einem Wirbelsturm
sich sfeles'entlich etwas vor dem Centrum befinden soll.*

Vom Interess ist, wie ich glaube, ein System von Curven, welche
durch die folo;ende Gleichuna; definirt ist

F'^ = Cojist.

* "Wie Thom zu diesem Schluss (gelangt war, ist mir uubekanut. Eoye hat diese Bemer-
kung nur gelegentlich in seinem " Wirbelstürme etc." (pag 96) angeführt.

Beiträge zur Theorie der Bewegung der Erdatmosphäre und der Wirbelstiirme. 319

d. h. die Cur s'en ofleicher Windstärke, weil man damit eine Gleichunsf
für die Gestalt des sogannten Stm*mgebietes erhält; d.h. eines Ge-
bietes, innerhalb dessen der Wind als " Sturm " charakterisirt werden
kann. AVir denken uns zu dem Ende eine Richtung parallel dem
Meridian durch das Centrum des W^irbelgebietes gelegt, und be-
zeichnen den Winkel, den der Radius />' mit dieser Richtung schliesst
mit J^, und zwar so, dass J^ in dem Sinne von S.O. N.W. wachsen soll.
Wir haben dann die Beziehunçfen

o cos X = p' cos }ù + Xi

p sin ■/ = p sin p + iji

Da ferner

Xi

— sin y, = cos /i

ist, so folgt

p sin iXi — '/J = -*— \<^os t/j. 7/1 — sin tp. xA

p cos (/i-/) = -^ (^cos p.Xi + sin ^. y A ■\r ^/xj+yï'

da '\/x-^--\-y{- = P\ ist, so folgt aus der zweiten Gleichung

pi—p cos iXi — '/) = -^(cosp.Xi + sin i^. yA

Wenn wir diese x\usdriicke in (7) substituiren, und für p',p für
a-j und y. « und ß setzen, so erhält man nach einiger Umformung
unter der BerUcksichtigung der Beziehung (G)

F' = Ç (p) ^^^2llJL Los t/j [a + ßK+f(p) iß-ali)]
* Pi \

+ sin p [ß—a K—f(p) {a + ßK)]\
wo zur Abkürzung gesetzt worden ist

320 ^- ^^*^*'-

9 (f) = Tf ^^- + -^ (l + 1/ (/>)J') /^'+ ^

Die Grössen a + ß h ^ luïà ß — o. h . — Funetionen der Zeit f — hängen
îib von der Lao-e des Centrums des Wirbel oebietes in Bezu"' auf die
durch den Schwerpunkt gelegten Coordinatenaxen. Setzt man zur
Abkürzung

SO hat mau

F'^-^ {[>) + P cos yj [J + Bf (/.)] (10^)

Die Curve F"^ = const für ^ = 0 ist demnach kein Kreis und kann einen
Kreis /'= const in vier reelen Punkten schneiden, da diese GJeichung
für {i=^ const, vier reelle AYurzeln für î^ geben kann. Da p für keinen
AVerth von 5^ unendlich werden kann, so ist die in Rede stehnde
Curve eine algebraische geschlossene Curve. Avenn niclit m—^ ist. Sie
ändert ihre Gestalt, Avic ihre Lnge fortwährend, das erstere, weil die
Grössen -' und /^ durch Verlegung der Coordintitenaxen allein sich
nicht v.eo'schaffen lässt.

^^ ist ein System »jvaler um das Centrum das Wirbelgebietes
excentrischer Curven. Da die Function ^{p). und o '/l + !?/(/>}] mit

.'ibnehmendeni /' im allgemeinen abniinmt, wenn ^/i >- 1 oder "^ > -3-
ist, so wird die Gleichung (10a) in dem Qundrante. wo die Wind-
stärke am grössten ist, durch kleineres <> erfüllt als in dem Quad-
rante, wo die AVindstärke am kleinsten ist ; d. h. in dem Quad-
rante III [Fig. pag. (29)] ist der Werth von >> in der Gleichung
(10a) am kleinsten und am grössten in dem Quadranten II. Die
Curve der gleichen AVindstärke liegt in dem Quadrante III näher
gegen das Centrum des Wirbelgebietes hin als in dem Quadrante IF;

lifiträyo zur Thoorio der Beweg'uug der Erdatmosphäre uad der U'irljelstürme. 3^1

das Stiirmgebiet erscheint demnach nach der linken Seite der port-
schrei tun, usrichti mg verbreitert, wie es die Figur ungefähr veran-
schauUchen mau'.

Die Curven. sind in der Ellipticität übertrieben gezeichnet, denn
sie kann man in erster Annährung als Kreise betrachten, deren iMittel-
punkte verscliiedene Lage in Bezug auf das Wirbelcentrum G haben.

Da nämlich /'(/>) sicli von bis zu ^^^-^ ändert, während

f> \i)n li bis '* abnimmt, so kann man wohl annehmen, dass fil')
entlan^c der Curve F- — const., für die p überall doch einen endlichen
werth hat, nicht all zu viel variire, dass man daher der Function f{p)
in erster Aunährung einen gewissen mittleren Werth M geben könne.
Unter dieser Annahme wird die Gleichung (10a)

aO) F'-^^-(a^ M^'r + n cos t/, {.{ + DM)^o sin tp {B-AM)\ + ^-^^^^^^

was einen Kreis bedeutet, dessen Mittelpunkt durrh
_ 2U+BM) _ -liB-M A)

n\i + M-') ' ~ rAi + m

und dessen Radius durch

bestimmt wird. AVenn das Wirbel u'ebiet sich siid-östlich von dem
Schwerpunkt befindet, so ist ./ wie B p(3sitiv. vorausgesetzt, dass das

322 ^- l^itao-

andere Wirbel gebiet auch cyclonal sei. Das in Rede stehende Wirbel-
gebiet hat dann die Fortschreitungsrichtung nach N, NO, O. Ist
dabei /y> .7 M., so liegt der Mittelpunkt des Kreises (lOb) nord-
westlich \oui Centrum des Wirbelgebietes, in dem { — x, — y) Gebiete
der Coordinaten. Ist aber D^zAM, so liegt derselbe nordöstlich vom
Centrum des A\ irbelgebietes, — also ebenfalls auf der linken Seite der
Fortschrei tunu'srichtunu'.

Die Gleichung (10a) gilt indessen nur innerhalb des mit B-^ be-
scliriebenen Kreises. Da aber F an dieser Grenze der verticalen
Strömung durchaus stetig bleibt und jenseits dieser Grenze wieder
wachsen kann, in Folge der Abnahme des Abstandes zu dem zweiten
Wirbelgebietes, so sieht man ein, dass die Curve F-= const ausser-
halb des Gebietes (1) durch eine andere Curve continuirlich fortgesetzt
wird.

Es scheint in diesem Ero-ebniss eine Erkläruno; der Thatsache
zu liegen, dass die ^ ordere Seite einer Cyclone immer von viel längerer
Dauer zu sein pflegt, als die hintere Seite derselben ; einer Thatsache,
welclie l'iddington für eine so ausgemachte Thatsache erklärt haben
soll, dass sie keiner Belege bedürfe. In der That hat Piddin^ton den
vorderen Halbmesser einer Bengalischen Cyclone für den 12 und 13
October 1848 zu 140 und 115 Seemeilen bestimmt, den hinteren aber
nur 90 respectiv 65 Seemeilen.*

AVenn w < — ist, so dass die Luftgeschwindigkeit nach dem
AVirbelcentrum zu in's unendliche zunimmt., so verhält sich die Sache
etwas anderes. In dem Quadrante III herrscht nach wie vor das
Maxinuim der Windstarke. Da aber f (/O wie/(/0 mit abnehmendem
in's Fnendliche wächst, so wird die Gleichung (10a) in dem Quadrante
der minimalen Windstärke durch ein kleineres f> befriedigt, als in dem
Quadrante III. Es folgt hieraus dass das Curvensystem F- = const jetzt

* lîeje. " Wirbelstüruae etc." pag 96.

Beitrüge zur Thoorio der Bewe<>nmH: der Erdatmosphäre und der AVirbelstünne. 3^0

Dach der linken Seite der Fortschreit ungsrichtiing hin kleinere Kadien
aufweist, als in der hinteren rechten Seite der Cyklone. In diesem Fall
erscheint demnach das Stiirnigehiet eiförmig nach Hinten der Cyclone-
ausofedehnt, und der Punkt des minimalen Drucks das Centrum des
AYirhelgebietes Hegt somit nothwendig etwas vor dem Centrum des
Sturmgebietes, was mit der oben erwähnten Bemerkung Thom's, wie
mit der AValumiehmung lîedfield'sf zusammenfällt, dass das ILaro-
meter frewühidich kurz vor dem Windwechsel wieder zu steiofen

o o

beginne.

Indessen muss ich mich hier ausdrücklich dagegen verwahren,
als wenn ich srlaubte, diese Anomalien ledio-lich aus der Eeweg^uno:
eines AYirbelgebietes erklären zu können, da es durchaus nicht aus-
gemacht ist, dass die Bewegung allein zureiche, diese Anomalien zu
bewirken, und die Berilcksichtig'uno;' der a'asi^-en Xntur der Luft zur
Erkläruno- derselben nicht g-erade nöthig sei.

Wenn das Wirbel gebiet (2) oder die beiden Wirbel gebiete anticy-
clonal sind, so beofeofnet man einer eii^enen Schwierisfkeit in Folsfe
des ümstandes, dass das anticyclonale Wirbelgebiet nicht mit dem der
vertical niedersteigenden Strömung zusammenfällt, sondern strenge
D-enommen in's Unendliche hiniiberofreift. Die in dem vorio-en Para-
graphen aufgestellten Ausdrücke für ein Avirbelfreies Gebiet ver-
lieren eio-entlich ihre Gilti2:keit, wenn eins der beiden Wirbel o-ebiete,
oder beide anticyclonal sind. Indessen ; es ist daran zu erinnern,
dass die ausserhalb des Gebietes der vertical herabsteigenden Strömung
existirenden Luftwirbel, wie wir bei dem kreisförmigen Querschnitte
gesehen haben, ausserordentlich rasch verschwinden mit dem zuneh-
menden Abstand von der Grenze des Gebietes der verticalen Strömung,

t Redfield ; Observations on the Hurricanes and Storms of the West-Indies and the coast
of the United States. Sillmann's Journal Vol. 25 pag. 114.

324 T>. Kitao.

dass jeder massig entfernter l^mkt des äusseren Gebietes sclion als
völlig wirbelfrei betrachtet werden kann. AVir können uns demnach,
wenn die Begrenzungscurve des inneren Gebietes ein Kreis ist, einen
dazu concentriscben Kreis denken, dessen Radius B' so gross ist, dass

eine Function von der Form r~y[ir) unendlich klein wird, aber

noch unendlich klein ^e^en den Abstand der beiden Gebiete der ver-
ticalen Strömung. Dann gelten die für das äussere wirbelfreie Gebiet
aufofestellten Ausdrücke für AYindstärke, Windazimuth, and Druck in
dem Punkt ausserhalb dieses Kreises, h^ den Ausdrücken, welche
wir für einen Punkt innerhall) eines cvclonnlen Wirbelgbietes absre-
leitet haben, braucht man nur nur das Xorzeichen aou /Aj und ^^
zu wechseln, wenn das zweite Wirbelgebiet anticyclonal ist, da ^\ in
diesem Fall nur sein Vorzeichen wechselt, wenn auch dns Integrations —
gebiet grösser ist, als das Gebiet der verticalen Strönumg. ^Yir bilden,
um dieses einzusehen

/', - -4- / "'' / '"' (C-2 />. sin 0) odoily
2-. ,/ ü / 0

-T- I r C - '-^ / sin H) odndy

+

Das erste Integral ist nusgedehnt über das Gebiet der vertical
herabsteigenden Strömung und da:> zweite Integral über das äussere
rinofförmiofe Gebiet bis zu dem mit Ji' beschriebenen Kreis. Für das
erste Integral hat man zu setzen (pag 201 Vol. I dieses Journals)

„ . . ,, 2/ sin 0
r— 2 / .nn 0 = . r^

und für das zweite Integral

1 ä W ^ f ftV-

Beiträge zur Theorie der Bewe^'uno' der Erdatmosphäre mid der Wirbelstürme. 325

Die TntouTalion pr<j,-iel)f hiomiis

(In '1 7c o ' ' (>

d

wo C eine Constante ist. Es soll für y—B' der Wirbel verschwinden,
so dass

-^ ^ c^ = -^^ 7?\ Ine/ !> + Const.

(ff) H do ' ' 'I - '

Diese P>edin2"nn2' crii-ieht

. _ / sin li ^^

Es soll ferner an «Icr ^Trenze «Ter \('i"tie;ilen Striinuinii'

dWi dW,

do df

Dies o-ifljt

fii )• o = i^2

C = ; '-^e Y:>

Wir erhalten somit i'iir das äussere Uehiet

.- • . -11 Sinti -!^(4:\- ^{\

Es i'olot liierdnrcli

„ 'IXsintly.B-' -2/. Sinti / ^•■' ^2L(£L-i)

d. h.

., __ 2/ sin H ;-., B^- / . j^\ _ _ -Ik sin fl r^B£
womit das oben Behan])tetc bewiesen ist.

32 ß !>• Kitao.

Dil in (k'iu in IJedc stellenden Fall

r-2 = - -^if-

ist, so bleil)1' n:teli ^^■ie not

Der innnerhin willkiirlidie lîndius lï kommt demn:u'h bei der l>e-
tnielitnnfj- einei- dm'ch eine imendlicli fernen Anticyclone beweg-ten
Cyclone nickt weitei" in Betracht.

Wohl würde die l)estimmiint>- des Zeitpunktes, wo ein o-eo-ebenei-
( )rt in ein anti(n^clonali-s Wirbelfreliiet eintritt, odei- daraus hervortritt,
davon ab]üln_i>-en, Avelclien Werth man //' beilegt. Allein ; da die
Comj>onenten der (^eschwindiu'keit und die Fnnetion '/^ so wohl f'iii-
das äussere wirbelei-fiilhe (rel)iet, wie für das (iebiet der vertical herab-
steigenden Str()innng keiîi (ilied enthalten, das von dem Radius 7>'
abhängig wäre, so sind die Ausdrücke für Windstärke. Winda/imuth
und Luftdruck in einem anticych^nalen Gel)iet auch V()llig frei von
/.'', selbst wenn das aiidere Wirbelgebiet anch anticyclonal sein sollte,
und es hat somit keine Schwierigkeit, b\ <y, /' für einen in ein anticy-
clon;des Gebiet gerathenen Ort als Functionen von der Zeit darzu-
stellen.

F]s mag aber dainit geniigen, an den für ein cyclonales Wirbel-
gebiet entwickelten Ausdi'ückcn nachgewiesen zu halben, dass sie im
Stande sind, den \^^i'lai!f eines Wirbelsturans zu srhildern. AVohl
liegt der Gedaidœ nah, \\\y die iu'er auftretenden r\)nstanten willkür-
liche Zahlenwerthe anzunehmen und so die N'eränderung der drei
anemometrischen Elemente in einem gegebenen ( )rt numei'iscli zu
verfolgen, und so nälier zu ])rüfen, ob imsere in diesem und vorher-
gehenden l*aragraphen entwickelten Ausdrin-ke einigermassen richtig
den wirklidien N^erlauf eines Wirbelsturms darstellten. Der Haupt-

Heiträ<,'e zur Theoriu der Bewegung der Erdatmosphäre und der VVirbelstürmc. 327

zweck dieser Entwickeluij<>eii liegt indessen darin, um zu zeigen, wie
man hätte verfahren können, Windstärke. Windazimuth. und Luft-
druck für einen gegebenen Ort, .sowulil im Inneren, als Ausserhalb
eines fortschreitenden Wirbelo'ebietes. als Functionen von der Zeit dar-
zustellen, wenn man nur im »Stande ^^•äre, die hydrodynamischen
Dilterentialgleichungen unter Annahmen zu integrireu, welche viel-
geuauer dem thatsächlichen Verliältnisse entsprechen würden, als
dieieniiit'ii. ^ on denen wir ausgegan^'eu waren.

§ XIII. — Verticale Luftströminujen in der Erdatmosphäre.

Hisher war die Grösse y eiufich aJs eine Constante betrachtet
wurden, ohne nälier die Ijedeutuno' derselben festzustellen.* Wenn
/' eine Constante ist, und von der Zeit unabhängig ist, so ist die
Geschwindigkeit eines vertical auf — oder niedersteigenden Liift-
theilchens durch eine ExpcHientialfimction ausgedrückt. Offenbar
setzt ein solcher lîewegungszustand der Luft eine ganz bestimmte
verticale Tempera tur^■ertheilung in der Atmosphäre voraus, welche
dazu noch sich mit der Zeit nicht ändert. Wenn nun die verticale
Tem['eraturvertlieilung eine andere und mit der Zeit veränderlich ist,
so wird die Geschwindigkeit der v'erticalen Strömunii' eiiien anderen
Ausdruck haben, als Exponentialfunction und die Wirbelbewegung,
welche sie in der Atmosphäre hervorruft, eintan anderen Gesetz gehor-
schen, und kann mit der Zeit veränderlich sein.

Wir wollen jetzt die Annahme fallen lassen, dass /' eine Constante
sei, und den Fall zu hehandeln suclien, wo ;' eine gegebene Fimction
von der Zeit ist. Da die Geschwindigkeit jeder geradlienigen Be-

*■' Bei der üiichtigeu Feststellung der Grösse ;k (p^ig- 167 Vol. I dieses Journals) ist ein
^'erseheu uucorrigirt geblieheii. Es ist überall für g {/ti - /Li), g — — zu setzen.

weg-iHJU" immer in der Form /'■ i\ dari^'estellt werden kann, so behan-
deln wir damit Fälle, wo die Geschwindigkeit der verticalen Strömung
durch andere Function ausgedrückt ist, als durch Exponentialfunction.
Ehe wir aber zur Integration der F)ewegungsgleichungen für diesen
Fall schreiten, ist es nothwendig, uns ül)er die physikalische Bedeu-
tuno' der (rrosse y nähere Rechen-ichat't zu verschatfen.

Iveye hat* sdion näher die Ix'dinguugen les tu'es teilt, imter denen
die atmos]thärische Lutt sich im hd)ilen oder stabilen (rleichgewicht
Ijcündet, tuid einen Ausdruck für die Geschwindigkeit dei" verticalen
Strönuujg al)geleitet. und zwar unter der Aiuiahme. dass die veilicale
Al)nahme der Tempei'atiu' eine lineare Fund ion der Meereshühe sei.
Es ueht aus seinen Fnter>uchungen her\ nr. dass. so lanu'e die Ver-
theilung des Wasserdampfcs in iioi-izontalen Schichlen eine gleich-
förmiö'e ist, die Lufr sich im indifferenten frleichuewichte betindet
mid, wenn gleich die N<jn der Erdoberfiäche aus ei'wärmte Luft
fortwährend aufsteigt, die Bildung einer vertical :uifs teigenden
Ströniunt'" tuiterbleibt, dass aljer eine soh'he entstehn kann, wenn
eine mit ^^ asserdampf gesättigte Lultmasse \ un einem beschränkten
Gel)iet der Erdol)oi'tläche aus emporsteigt und die (hu'ch Condensati<jn
des Wasserdampfes frei gewordene \A änne den verticalen Temperatur-
o-radienten modiiiciert. Der Ausdruck, den Heve für die treschwindiu"-
keit einei' solchen \erlic;ik'n Strömung al)geleitet hat. ist angenährt
linear in Bezug auf die Meereshöhe, wenn wir le>tsetzen dass diese,
soW'cit AN'ir sie in Betra<'hr ziehen, nur eine juässige sei, und dass die
Geschwindigkeit an der fjrdol)erüäche verschwinde.

Wir wollen hier die Auinabe etwas allii'emeiner in Antritt' nehmen,
indein wii- Nor der ILuid keinerlei Annahme über die Beziehung der
Tem])eratur und der Meereshöhe maclien. Es herrsche in der Meeres-

* Th. Keye. Die Wirbelstüruie etc. 1880. Anhang pag. 224.

Beiträye zur Tlieorie der Buweyiiug der Erdatmosphäre uud ilor \\'irl)elstüruie. 329

höhe ^' der Druck p. und die 'renijiei'ütur H. Man hat dann, wie man
aii.s dem IJovle-Gav-Lu.-isacj^chen Geî^etze Jeiclit al)leiten kann

JL=e''JAT^i}lH^) (11)

wo p„ den Druck auf der ErdolierÜaclie. ll den mittleren ErdhaH»-
me.s>ier, H die ('onstante l'!l.l^74, welche jedoch hei dampfhahiuer Lult
von p und ^ , darum \ on : auf gewisse Weise ahhänut.. und tt
endUch die Zahl 'll'.\ hedeutet^. Wenn nun eine Masse Luft von der
Dichte iJ-i emporsteigt, und die his dahin im Gleichgewicht hetindliche
Luft durcldjricht. deren Dichte in der Höhe .: !',, ist, so erhält sie
einen Auftriel). dessen Grösse din'ch

(1-z (I li-

ât- {B +

?}- \ Pi /

hestimmt ist. wo ij die Schwere an der LrdoherÜäche hedeiitet.

Der (hiotient -^-^ ist offenhar nicht nur von : al)hängii>', sondern
a.ucli von den ( "oordinaten. welche einen Lunkt der durch : gelegten
Horizontak'l)ene l»estimmeii. da die Teni[feraturvertlieihing in horiz(jn-
taler Iviclituui:- audi im allgemeinen als eine tuigleiclimassige voraus-
gesetzt werden muss. Wir wollen hier indessi'n (K'ii (^uerschuitt der
verticalen Str<»mung so massig annehmen, dass wir uns - — als eine
Function \<)n ^V^v .Meereshöhe :, und eventuelle vmi der Zeit t allein
vorstellen dürfen. Ist ti,, die 'rem[)eratur ausserhalh des Haumgeltietes
der verticalen Strömung, und 'U diejenige innerhalh desselhen in der
Meereshöhe :, so folgt nach dem l)o\le-<T;i\-Lussac'schen Gesetz
in dem wir ferner annehmen, dass die emporsteigende Luttma.sse si<'h
unter dem Druck ausdehne, den sie in der Meeresliöhe ; voriindet

AVeiiii ^vil' Ulis ^'i unci ^a :>l.s Functionen von i so dara'estellt
(lenken. d;i.ss

wo ^^o die reni])ef;itur nut diT Erdubertiüclu; bedeutet, so erlialten wir

dt-

i)ie (^eseliwindiukeit —^ i.st hier positiv, und mu.ss so bestinmitr
at '

Averden. dass sie iiii- i — o verseh windet. So lanii'edeiunaeli fÇ'<:)n—f\z\'>-0

ist, ninunt -V" vu 0 aus zu, bis in der Meereshr)lie, wo / [■~\i=/' [■\ und,
dt

d"

-^ = constant wird. Wenn nuu ./\ï)a— ./"'U'.; von irgendeiner Meeres-

höhe ab, oder zu irg'end weleber Zeit seiu \ oi-zeichen wechselt. Avas nur

durch Null hindiu'ch uescliehen kann, so wird -^, da —j^ ueu^ativ

dt dt- '-

ist, entweder abnehmen, oder sein Vorzeichen wechseln, was wiederum

nur du.rch Null liijidurch ii'eseliehen kann. Da in dei' Meereshohe.

oder zu einer Zeit, wo fi^k—fi:^h = ^ wird, -~ = constant wird, und

diese Constante, wenn der Strom niedersteiij'en soll, nur IS'ull sein

kann, so niiisste — ,— von ï=0 aus li'eii'en den \\'erth Xull (ilhjcnniniiioi
dt ^- '-

haben : mithin miisste -rr^ neu'ativ oe\\eseii sein, als dei" Strom

dt-

emporgertiegen wai-.. was aber ein Widersprueii ist. Icli glaulje
hieraus schhessen zu diii'teii, dass eine Luftmasse, wenn sie einmal
eine vertical aufsteigende Stronuing gebildet hat, nie wieder vertical
abwärds zui' Ei'dobertläclie zuriickkeiu't, und dass, wenn in einer

Beitrüge zur 'l'lieorie der Bewegung der Erdatmosphäre und der Wirlpelstiiriue. ''9 1

U'ewissen Mcereslii»!)«' oder zu iru-end welclier Zeit /('?)a— /(■^)i = 0 wird,
die r^uftintissc Imiiicr wcircf (Mii}M)i'stcin-t jiltei" mir ciiicf ins niicnd-
liclie nbneliineiiden (u'scliwindiukcit. d;i die <Tr(»s^<e ''■ + ^L—f K^).x mit
WHclisciidei' Ilrilic ill's l neiidliclu^ ;il)ii<-lniu'ii l>:;iiiii.

Ist die Meereshölle, so weit Avir in ju'tnielit ziehen, eine nur
miissio-e. so k-mn nmn setzen

wo r den ^^ erth des verticjden Teinpeniturunidienten bedeutet, und

1

ijii :!ll<j('meinen < -r— - ist. ^hu\ erhäh diiini

cPz

V^ + ^,-r':7

da iVir einen niiissiu'en Wcrth von r — :, — 1 <rpsetzt werden kann.

(-R + •--)-

Weil — ^——- ein sehi- klcMner liiaieh ist. so kaiiii man auch so
schreiben

rJ-z
flf

^^"<> 'I . u ~ f^' U"es<'tzt ist. Die Ik'schleuniu'unu- ist demnach linear
• ,( + 1-1^

in l»ezno' auf ;. Es ist leicht liieraus eine rTleichuni;" liir ;' abzuleiten.

Wir differentiren die GJeiehunu'

dz

naeii t. indein wii- uns ;- ;ds eine Function von t vorstellen, so dass
wir erhalten

00 9 D. Kitao.

Ein Vevo'leif'li dieses mit (14) erij-iebt sofort

r + 4^=A-= (15)

Es erhellt znnäclist ;nis dieser l)eziehiin<>'. d:is8, wenn ;' die Zeit
iiiolit enflinlteii. il. h. die liewegTiiig eine stntic^niire sein soll, r' — r immer
])ositiv sein muss, d;i in diesem Enll

'' = //Lrf! = ^'

ist, wi(î schon \X . Feri'el ■ und Th. IJcvef ;d)U(d('itct h;dien. Der Fall
A' =r const, hedini^'t in dessen nicht notli\\('ndi2'. dnss auch y woi\ der
Zeil iin;d)li;inuiu- -ci. fnteuTii-t man lùi mlieh (lô) unter der AnnahmCj
dass A = const, sei, so erlùiJt man

^^Y^/v7(A>jg_(A'-;gr-A''

\7ÄT

(A'+ro)+(/*-ro^^^^^'^

wo /'öden AV(M-tli von y Wir ?'^ <> hedentet. In diesem Fall verändert
sich also ;- von mit wachscndin- Zeit von j'o ;nif A'. Die InteöTation
der Gleiehnnu'

d^ ^ 1^ ., /.AV(A-+;-„)-(A-;-,W-A7\
,// - • V ,. Kt , A' + ;-„) + i A' - ;-„) f - A7 )

eru'ieht

" = ^(^rKt iK + y^^MK-r.)'''''^

WO r^ die A[eercsh()he e\\\('^ Lnfttheilcliens znr Zeit t = 0 bedeutet.
Ans dieser (ilciclinnu' tlicsst weiter

* \V. Ferrel : Recent AcWances in Meteorolo<4-y. A\':is]iint;t(in. ISSr, pa«^-. 2'.»3.
t Heye : Wlrljelstiiruie &. Hannover 1880 pa«^-. 22U.

Heitrüyv zur Theorie der Bewe^'unt;" der P]r<1atiiiosphär(> und d<'r Wirlielstürnu'. |-)3o

dz

-^ =^0 (^rKt^K+r,)~^K-y„)r-K!^

D:is Liirtrlicilclicii be.sass zur Zeit ^^ = 0 die ^^Toscliwindiukeit /'„ î'o,
als (lie veiTicül aiitsrciuende StniinuiiL;' .sich 1)il(let('. Dieser Fall lüsst
sich (leiimach daiiin deuten, dass der \erricale 'rein]>ei-atiii'i;ra(lieiit
.sich ]iliit:Iic]i L;-eiindei-r liat. dass dalier die treschwindiukeit der aiil-
teiu'endeii Iddtst i'(»iiiiiiiiiiL;'. welche urs])iaiiiLi'lich den \\ ei'th ;'o - hesa.s.s,
sich alliuiilii!" dem \\ erth niilirt, der dei" neuen \ci'h<:'alen l'empei'atnr-
vertlieiliiiiL;' cnts|rricht. Ilaben die IdilHheiN-heii zur Zeit / = (> keine
durch Ditterenz des vertiealen 'reiujHa'aturLiradienkMi erzeuu'ten Ge-
schwindiiikeif. so hal ;nan j'o^'^ zu setzen, und so erhält man

. Kt — p- Kt

^-^■(S^7) O'O

Dieses enfsprichr dem Fall, wo die Dittei'en/ des xcrricalen lemperatur-
<j,'radienten sich u'ehildei und die bis dahin in Kühe l)etindliehe Luit
tilbiiiiVui enijioi" steil;'!, und so aueh aJliiiilIi(i eine xo'ticüle Luftsti'öniuniJ'
bildet.

Ihm demselben Werthe von ^* sind demnach zwei verficale
S(r(»mnnu'en Non wesentlich xerschiedenem (_'h:u"akrer m()L;lich. In
deui Fall. Wo die vca'ticale ."^n'omuni;" allmiilii^' sieh Itildet. ist

^. ,,(,..,_„-,.,)

soda.ss die [^ult in jeder .Schichte zur Zeit /^*) in Ivuhe ist. Anderes
hinii"(\Li('n. wenn ;'=A ist. Fn diesem Fall hat man liir ein Lult-
theilchen in der 1 liihe "o

dt "°

334

T>. Kitao.

SO (lass (lie Geschwindigkeit ties Liifttlieilelien zur Zeit f==0, den
Wei'th A Zo hat. Wenn wir nun näher die Gleichung-

Ix'trachten. so sieht man, dass in dem Fall, wo die l'.ildnng einer
vcrticalen Stromuii!"- eine allmjdiu'e ist, —r- iVu" '' = 0 einen cndliclien
Wci-th l)esitzt. aller l'uv t-^-x verschwindet, da ;' gx^üvn A couvergii'f.
[st ;- hino-egen für f=^0 der Constante A gleich, so kinmcn wir uns
/' lind A ;ils eine discontinuii'liche Function denken, wclclie für t = 0
verschwindet, aher für jedet andere t einer ('onstante gleidi wird.
Da in dem Fall, wo A = const, und y mit der Zeit vcrilnderlicli
ist, •/' erst nach dem W'i'ltmf eiiiei" unendlich grossen Zeit den Werth
A' erreicht ; da ferner in dem letzteren Fall nur eine verschwindend
kleine Zeit dazu nr)fhig ist; so sehen wir, dass y=h i'iir /-^O dem
Fall ents|)re(^hen kann, wo eine Luftmasse pliitilicli emporsteigt,
welche hisher sich im lahiK^i (rleicligewiclit l)efunden hat ; dass
/' = 0 für t^O aller den Fall darsU'llf, wo die I>uft heim l>ilden der
Ditferenz des vei'ticalen Temperahirgradienten sofoi't zu steigen
l)eginnt, und allmiilig das Afaxinuim ilu'ci' Geschwindigkeit erreicht.
In den Fällen, wo A- eine gegehene Function von der Zeit
ist, hahen wir auch zw(a Ai'ten d(^r veiiic-den Strömung zu unter-
scheiden je nachdem y für / ^<> einen endlichen Werth hat, oder
verschwindet. In dem ersteren F-dl entsteht die vertical aufsteigende
Luftströmung |)lötzlich. während sie in dem letzteren Fall allmälig
entstellt. W(dche l^^oriu der vertical autsteigenden Stnimuiig eintritt,
das hängt indessen wedei' von dvr Functionsfoiau der Grösse A-
nach von ihrem Wei'the zur Zeit t^O. Die \Vii'l)ell)ildinig, die sie
veranlasst, ist aber, wie wir weiter unten sehen werden, grundver-
schieden.

Beitrüge zur Theorie der Be\ve>^ung der Erdatmosphiire und der W'irl^elstüruie. oÔÔ

Es ist jedorli iiocli ZU l)emerken. dus.s der Fall, wo y für
t^O einei" eiidlicheü Coiistaiite uleich ist. einer uiidereii Deiituiif
fahio' ist. Mau kann sich aiicli die verticale Strönuing als Uinj4'st
entstanden denken, und die Zeit von dem Anuenl)lick zählen, wo
/' den Werth errei'-lit. welchen wir der durch die Integration der
rTleicljuiiii' (15) auftretenden Constante für ^=^0 gehen. Es ist daher
in ciuem soh'hcu l'all gleichgiltig, ol) wir uns die verticale Strömung
im Momente / (' plöt/lich entstehend denken, oder längst entstanden.

Wir fassen jetzt eine Luftströmung in's Auge, welche Aertical
ahwärts geht. In diesem Fail wird die Luft in dem Ivaumoebiet der
verticalen Strömung schwerer sein müssen, als ausserliallj desselben,
so dass

a ^- il

Da die (.Teschwindiukeit eines Lufttlieilcliens —f- hier nen"ativ

elf

ist und vermöge der an der Erdoberfläche zu erfüllenden Ijedina'unu'

\dt A=o

mit wachsender Zeit ihi'em Maxinnim (' zustreIxMi imiss. so muss

, dz
dt essentiell neuatix' u'esetzt werden. Mitliin follet

dt

d-,z _ R- / _ /^\
dt;' "^{B + zfK 'yj

]Jaher

i ^ . ^'-' //(4-/(^)a\

;^ 'J{R + zf\a + tl,-fy,)J

d-z
dt-

So lanue demnach /(■-•)i—/(':')a > 0 ist, strebt —t^ seinem l\Iaximuni

at

d-z
0 zu iiiif Avachseuder Zeit. D;i -^-^ positiv ist und i' seiuein Miuuinum

0 zustrel)t, ist- ~ also nei;"ativ. Kaiiii f {z)-,—/ [z\^ in irjj'end einer

Meeresliitlie, oder /u irgend welcliei* Zeit Null werden d. //. sein \<yv-

zeichen wechseln, so wii"d —f- noch weiter ahnelimeii. oder sein

dt

A'orzeiclieii wechseln. S(j dass die urspriiiiülicli niedei"steiu'eiide I^uft

nun mehi" empoi'sleiu'en nhisste, I)as erstere ist inui unniöulicli.

wenn /(-■)—/' C--') nin- einmal und nicht für -"=0 verschwindet, da ja

für -—- ein Ausdrucl-: resultiren würde, der an dei" Erdoherfläche

dt

nicht vei'schwindet. In dem zweiten Fall wird —^ . da wo t\z).,—f{z)^

dt

— 0 wird, einer Constante u"lei(,'h, und diese Ccjustante muss = i) sein,

weil die IviditunL!" der Stronuuig eine unigekelirte \vird. Da im

weiteren A ei'lauf der Zeit —r- von Null aus i)ositi\ wuchsen nuiss, so

dt '

, dz

müsste dt V(jn nun an positiv sein, was wieder ein W^derspiaich ist.

dt

Ich glaube hieraus schliessen zu diirlen. dass. wenn eine Luftmasse

vertical gegen die Erdohei'fiiiche herabsteigt und so eine nieder-
steiofende Str<)muni'- bildet, sie nie wieber rückwärts in die Höhe
steigt. Die Lufttheilcheii strinnen nur mit in"s Unendliche abnehmen-
der (resell windigkeit gegen eine Schichte, wo der Temporaturnnter-
schied verscliwindet. deren Höhe jeiloch sich mit der Zeit ändern
muss, wenn /V)— /(:)='* noch die Zeit enthält.

Jst die Meereshöhe, so weit wir sie in lietracht ziehen, eine nur
mässiü'e, S(j hat man an nährend wieder

d-'z
dt;'

V+^o/'"

Indem wir ir ^ —Jj=—y.z nacli t diüerentii'en und das

Beilrîiv;(j zur l'iieuric ilor Hüwui^-uug der Enlatumspliärc uml der WirljolsLüruio. 3o7

I{t'f?iiU:i! mit dem olieiLSteheiiden Aiisdi-iick verii'Ieiclieii. erhaltdi wir

' dt

(17)

indem Avir wieder

«etzeii

V + V

A-

Es Igelit :iiis (17) /.iiJiHchst herxoi', dass, wenn eine «Udionäi'e
verrical liera hstei^ende StroinnuiL»' entstehen .s<j1I, r > r' sein ninss.
Die Constiiiiz dei" (iriis.se A" Kedino-t hier auch niclit notliwendiu' die
Stationalität der xcrtiealen Striimnng. Fnteufirt man nündidi (17)
so erhält man

;Kt + (■ + (' - A7

I ~ * \pl<.t->rc — ,, -Kt-c)

(IS)

wo C die Integrationscoiistaiite ist. ;- ist also mit t \ erUiiderlieli,
und converii'ivt nach dem N'erfliiss einer luiendlich grossen Zeit

ireu'en A'. Die Jnteu'ration der Gleich un ii" — r- = —/'•'-' ei'uieht

•'■ '-^ ■- • ((t

C

Kt + c p —Kt—r

(IS)

WO C wieder eine willkiirliehe C(jnstante ist.

Es sei -'-n die jMeei'eshöhe eines Lull I lieil<'lK'ns /.iir Zeil t = ()
. und ii\, seine (Geschwindigkeit, so hat man

e'^ —e

woraus folu't

io^= —C A )

m- m- ^

0 0(S ^^ ^^^^'^"

aus welcher biqiiadratisclieii Gleichiuig C zu bestimmen ist. Ist j'o
der Werth von y zur Zeit ^=^Â so dass

S(j ei'liält mnii

IjS uiiiss liierundi zur Zeit t = (), als die Nerticile Strfunuiiu" sich
bildete, jedes Lnfttlieilciieii bereits die GeschAviiidiukeit 1!'= — y^z-
uehaljt iiabeii. und zwnr t'iiie solclic. dass /'o > A ist. Ist ahei' /'o < '*
oder = ^'' . so werden (■'' und (^ ijunginar. </■ li- die \erlieal lierab-
steiii'ende Sti-iinuiUL:' eri'eielit eine endliche jMeeresliölie ersi nach dem
A erfinss einer unendlicli ui'ossen Zeit, da /' gei>en /*. also gegen eine
reelle (rr<)sse doch convergirt, mochte (-' i'eell sein oder nieht. Wir
denken mis : ein J^ufttlieilclien befinde sicli zur Zeit t = <) in einer
unendlich grossen Meereshiilie und hal)e eine unendlich grosse Zeit
T gebraucht, um zur emlliclien Meeresliöiie -n zu gelangen. Aa(;li
(IS) ist dann

er

oder, da c-'^'^' verscliwindet

Indem wir dieses in (1<S) einfiUu'en, erhalten wir

Xun vei's.ch windet ^'-'^"(^'+0^ und t — T i,st die Zeit, welche da?

Beitrüge zur Thcorio der Bewe^iuij,^ der Erdiüinosiiluir.' und der Wirljelstüruie. 33«)

Lufttlu'ik'lu'ii o-e1)r:i licht hat. imi von einer ^[eereshiihe -o zur Meeres-
höhe ". zu »'ehniuen. iKV.eiclinet mau diese Zeit mit /. so hat man.

"r^r

- Jd.

Wir sehen somit, dass in der endliHien Meereslu)]ie nur eine
stationäre Strinniinu' vertical ahwiirts iMilstelien kann, wenn A eine
Constante ist, (/.//. mit anderen Worten, dass. wenn eine Luftinasse
allmidiu- nie<lersinkt. die so ^'ehildete vertical lierahsteigende Strinnuiio-
nur als eine stationäre die endhche Meeresh()he erreielien kann, wenn
die Diiterenz der TemperaturiiTadienten, die sie erzeugt, eine Con-
stante ist.

Ich lasse liier einige Werthe von r fol.uvn. und für zwar iiir die
Temperatur der untersten Luftselu<-lite 0° und 30°. indem ich für <j
die Schwerkraft der Tokyo- P.reite •).7!»S"^- annelime.

^^""^-^ '^"

[fC

:;o'-('

0.000001

0.000 1 SD."') ^

0.000 17!).S -^

0.00001

o.oooryjDl

o.ooonoc)«

0.0000-2

0.000847-2

0.0007 in?

0.0000:-]

0.000 lo:',S

0.000,S77S

0.0001

0.001 Ul(')4

0.001797:^.

(I.OOI

0.00lU)i)02

0.00.5r)8()6

Im Fall Ä'- eine u'euvhene Fiuiclion von / ist. liisst sicli die Tn-
teo-ration der ( rleichunu-en (lö) tuid (17) selten in g-eschiosseiier Form
ausführen; sie lassen sicli aher auf eine Cleichunu' zweiter Oi-dnunu-
von einficlierer Form zurückführen, ^[an schreihe

840

D. Kitao.

''/•

r±^^/.-'

uikI sotzo /'= -V , und l)ostiimne p so, duss
56

±<f-

so ci-lùilt iiinii ;ils ( 1 Icicliiiiiu' iiii' P

IC-tp

(1Î))

iiidciM wir (liosf's intciirireii, erliaJtcii \\\v

, (^^ _ <i ^oq ^

(If

WD (l;is olierc Zeielicii I'iii' die niifstcigeiidc, mid d;is imk-iv fiir die
Tiiedei-steiL;-eiide Sti'iniiiiiiLi' ü'ilt. Die Fiiiictiou 1^ \\\\\ eine eii^'eiie
r>edentiuio-. Ks is( iiiiinlidi

dz _ d log yj

IIiei';ins 1o|n-f nliiie weifere.s

oder

Couiit. p
Const

(2(1)
(1>1)

Nicht jede Aimtdune id)er die Fniietion A- fiilirt zii einer ii)()o--
lielieii l^)eweo-nnu-sfo)'ii). Für ein iiii;iùin;ii'es A' li:if iiinii /. li.

^ = . / cos li + n sill Id

wo A, />, zwei willkiirlielie Constanteii sind. Es wird dann

. . / licosM — .1 sin kf\
''"" * K-lcosJd + Bsinld)

Reiträcfe zur 'J'heorie der Bewei^iinc;,- der Erdatuiosphäro und der Wirbelst lirino. J-J^ |^

Wolil küiiii W —/'-'-■ an (](']• ErdoIierHäHie vei"sch\\ iii(l("ii, .so lano-c
y endlioli ist. ;- wii'd uhw iM'riodi.scli uncndlicli u'i-oss (iiid ir erliält
an der lvrdol)0]-flilr'lu' die inihesriiniiUc Form 0. x , was endlich sein
kann, und sonst tinendlifli i>-ross. z wird a))er ^ 0., d. Ji. jedes \a]\\-
theilchen, welr-he Meeresli()lie es zur Zeit ^ = 0 o-elia1)t lial)en iiiau',
iiiiisste in diesei- Zeit ziu* Krdol)erfläolie angelanu't sein. Eine solelie
lîeweu'iiML:,' ist nnmöolieli

Die Annahme A-= — - — - wo a eine Constante ist, IVihrl zu
a + f '

einer in(»L;'ii<'li(;n lieweuiiiiL;-. l'^in jiai'tienläres TnteLj'i'al der ^ deiehuntJ'

1 f/2j6

ist

eine unendliche Reihe, die für jedes endU(die t eonveru'iert. Es folut
hieraus

_ ^« )i(a + f)>>-^

V* (« + ?')"

1 ir\{n + \)

T convero-iert also u'ei^-en Xnll mit wadisender Zeit; die Ge-
schwindigkeit der verticalen Striunnnu' nimmt in jedei- Lnftschichte
Idrtwährend ab. Im Fall der anisteigenden kStrcJmung ist nach (-0)

'^ -"7 "^!(»+i:

und im Fall der niedergehenden nae1i (iM)

1 n2!(».+ l)

D. Kitao.

In dem crstcren I'^ill nimmt die Mcvrcslujhc eines Lult-
theilcliens rortwilhrend /ii, in dem lei/ieren hing-egen nimmt sie

o-eoen 0 alj.

Zu einer solehen 'I'miiscendente führt die einfache Anii.alime, dass
die Differenz der 'rem])ei-:itnroTadienteii sir-h wie verändei-e.

/- kann daher nnr dni-eli eine äusserst complicirte Transcendente ans-
o-edriickt werden, wenn man für K- eine Function annelimen wih-de
welclie dem wirkliehen \'c;i"]anfe einer verticakMi niclit stationären
Strömnnii- einigermaassen entsprechen konnte. Die Aufu'abe vvii'd
iridessen kacht, wenn wir uns /' als eine gegebene Function von f
denken, nnd nacli dem Gesetz fragen, gemäss dem die Differenz der
vei-ticalen 'rem])eratm-gradienten sicli verändern muss, um die ge-
geliene Gesclnvindigkcit der verticalen Strinnung zu erzengen.

Indess ; tuicli liier führt nicht jede Annahme zu einem mög-
lichen Aus(h'uck liii- die Differenz der verticalen Temperatui-gi-adienten,

da f + ^ liir ieden Werth von t durchaus positiv und endlich sein
dt ''

muss.

Der Werth N'on r ist, wie aus der oben stehenden Taliolle er-
sichtlich, sehr klein, rnid somit die Geschwindigkeit der verticalen
Strömung; sie würde z. 15. bei f^, = m°C und r'-r = 0.001C^ in der
Meeresh()he 100;;/. ei'st. C)m. betragen, und dies bei dem immei' hin
bedeutenden Untei-schiede Her 'remperaturgradieiitcMi 0.1 C" pi-olOO/;/.
So uerinii'füa-i^' diese verticale Gescliwindigkeit audi erscheint, so
kann sie zu stäi-ksten AVirbelstürmen A^eraidassung geben. Xach der
in pag. ISO. Vol. T dieses Journals gegelienen Entwickelung ist die
resiiltirende horizontale (leschwindigkeit des Windes an dei- Grenze
des kreisförmiü'en Wirbelgebietes

'2 V

1 +

ßciträffe zur Theorie der Bewcg'ua*^- der Erdatmosphäre und der Wirbelstürme. 343

wofür man, da — ^-^, iiiigefäiif gleich = 1 gesetzt werden kann,

schreiben kann

rB

CO =

V2

Wenn der DiirchiuesHer der Gebietes der verticalen Ströniiinu' 2
Kilometer ist, so ergiel)t .sich z, B. für r = 0.0()05()8H nur ().40l\

— — — . was einen \ollii' mifiilill)ai-en Wind 1)edeutet. Hat aber das

sec. '

Gebiet der verticaJen kStröniunu' einen Din-chmes.ser von iOO Kilometer,
so erreicht die*A\ indgesclnvindigkeit an der Grenze des Gebietes den

Wert h AO.'l '■ — . Wenn die oljenstehende Formel in der That

sec.

einigermassen der Wirklichkeit entspricht, so würde y für einen
grossen \\ irbeistnrm v<;n mehei'en Hnnderi Kilc^meter Durchmesser
einen ausserordentlich kleinen \^ erth haben müssen, voran siresetzt
die (rrenze der verticalen Stromunu" mit dem Ort der Maximalofe-
schwinib'ii'keit zusammenfalle, was indess durchaus nicht ausgemacht
ist. da die durch die obenstellende Gleichung gegebene Geschwinchg-
keit nur dann das Maxinuun ist, wenn j-^^;*- ist, (vergleiche pag.
'2i){) \'ol. I dieses Journals). 'V

Xun ist H nach dem Wert he des beobacB^ten Deviations winkeis
zu urtheilen (lÂandbemerkung pag 148 \o\. I dieses Journals) eine
(xriJsse \-on derselben Ordnung, wie 2/, das Doppelte der Winkel-
"eschwindif'keit der Erdrotation, und entschieden kleiner, als diese.
/' muss daher entschieden kleinei' sein, als 4:Äsiiitl, d.i. ().0000:^<SD(S(S

siiifi ( ), damit die Windgeschwindigkeit stetig nach dem Centrum

\sec.J'

des Wirbeh'"ebietes zu abnehme. Dieses verlangt aber einen äusserst
kleinen Werth der Ditterenz der \erticalen Temperaturgradienten,
und die Maximalgeschwindigkeit der Luft in dem ganzen Wirbel-
gebietcî würde Ijei massigem Durchmesser desselben so klein ausfallen.

ß^^ D. Kitao.

dass die Ijeweu'uiig der i^uft so üiit wie iiidiihlliar wäre. Xmi ist die
iiiechaiiische Kraft der kleineren Gykloneu. wie die amerieainsehen
Tornados, oft urösser, als diejenige der fre\vaiti<4en Wirbelstürme
Westindiens. Zur Erklärimu' der Tornados und derartiger Wirbel-
beweu'unu' der Luft müssen wir offenbar die Gesell windigkeit der
verticalen Strömiuig grüssei- anneliineii. als die Lngleichnng y^^-'ln
sie giebt ; also dass die Luftgescln\iiidigkeit statt nadi dem Cent-
rum des als kreisförmig gedachten Wii'belgebietes abzunehmen, in's
iinendliehe zuninimt. Wie wir bereits (])ag 1!)0 \ Ol. \ dieses

Jtjuruals) gesehen haheii, ist die resultirende horizontale Uesehwindig-

, . ... , . , . ;r '1a sind ,,

keit tui' /'>--:-". in dem wii' ^ in. = A setzen

oder

^■K-[^(fr'"-'-'

-^"[i(y)'"^'"-i]- (-J

Das erste (rlied in (liesem Ausdruck nimmt stetig mit ahnemen-
dem /' ab, das zweite filled aber in's Unendliche zu, und zwar um so
schneller, je kleiner der echte lirucli ///, //. //. je gi'össer ;- gegen ^f ist.
Um ein IJeispiei zu berechnen, setzen wir /' ^ '*.*"*15'S4. was iler
Diffei'enz der Temperaturgiadienten ().()()<M)777' C bei ^o = oO°(' ent-

s]»i'iclit, und schliesslich anniduainüsweise -^ — — — =1. Wenn wir

ferner annehmen, dass der llaihmesser des Wirbelgebietes 100');//.

betrage, so erhalten wir. da }ii^ -r— ist, folgende AVerthe ffir <«>

Heiträpe zur Theorie der Hcwc^niui; der Krd;itiii<)si>li;ire nud der \Virl>elstürme. 34/)

(0

(>= B

sec.

B
('- 10

±\U

B

''= 100

10.894

B

-"" 1000

■20. (Hj

Wenn wir weiter )' = 0,Ol584 setzen, .so (l:i>ss m = -^ ist, daiin

erhalten wir

(0

r= B

.,., „, '"^^^'-

sec.

B
'"= 10

79.50

B

''- 100

800.

B

'" ^ looo

sooo.

Wie n);in sielit. stei<i-t die (Tesdiwindiükeir der l.iiiV iivu'en das

Centrmn um so rnscljer. je kleiner der lii-iidi = y// ist. indem

ersteren Beispiel wächst sie \-on dem (irenzkreise aus in !)()0 ///. von
'J.'J [ m. auf nur 2.94 ;//. und in den weiteren 90 ;//. \(»n l^!' i ;//. auf
10. NS ;;/. um dann bei 1 ;//. Abstand xon dem Centrum orkanartiiz"
zu werden. In dem letzteren BeisjMele nimmt sie von dem Givnzkreis
aus in 90t) ui. von 'J'2A in. auf 79.50 ;//. um bei 90 »/. Al)stand vom
Centrum die uno-eheure Griisse SO'I ;//. zu en-ei'-lien.

Der Beii'ritf '' Darchniesser " eines Wirbel«truni.s oder eines Tor-
nados ist ein höchst va^'er. \'ersteht man darunter den Durchmesser
einer Lultgehiete.s, innerhalb dessen der Wind als orkanartig' bezeichnet
werden kann, s(j wiixl man dem AVirljel in dem ersteren Beispiele,
einen Durchmesser von nur ein l'aar Dutzend Meter zuschreiben,
ungeachtet der LuftAvirl)el durcli eine verticale Ströniung entstanden
ist, deren Quersrlinitt meherere (^)uadrarkil( -meter l)eträgt. Der Wirl^el
in dem zweiten l)eis[)iel liat hingegen schon einen Durciimesser von
1 Kilometer.

Wir sehen somit, dass der sogenannte Stucnidini'hme.-ser desto
grosse)' aust'älU. je grösser y gegen '2 m auslädt, (hiss derselbe mit dem
Durchmesser des Gebietes der verticalen Strömung zusammenfallen
nuiss, so bald ;'<2;f wird. Denken A\ir uns sonach ;' als eine Func-
tion, die stetig" mit der Zeit abnimmt aber so langsam, dass sie
während eines nicht all zu grossen Zeitinter\ alls als eine Constante
angesehen werden kann ; dass man daher innei-halh dieses Intervalls
die Wirbelbewegung als eine stationäi'e ansehen kann. AVenn wir die
(xleichung' (15) für um ein grosses Zeitintervall a.useinanderliegendc;
Zeiten an einem und demselljen Wirbel anwenden, so finden wir, dass
der Stuniidurchmessei' auch mit der Zeit aljuimmt, ao lange y>-'-i)( i.st,
und Non dem Angenblick an, wo ;'>2^f wird, wird der Ort der
Maximalgeschwindigkeit sich in dem Und'ang ([es ( îebietes der Ijer-
ticalen Sti-ömung herstellen, welclie den Sturm \eranlasst hat. l)a
diese Maximalgeschwindigkeit, w'enn der Werth des i( massig ist, nur
klein sein kann, so wird die Luftl)ewegung in dem Momente, wo /'= -^i
wird, längst seinen sturmartigen Charakter eingebiisst liaben. Anderes,
wenn /' anfangs nicht allzu gross aber ^'i-»-^, und B einen grossen Werth
hat. So lange y:>''2'>i ist, wird der Sturmdurchmesser kleiner sein
;ils B. Sinkt nun /' unter '-^^^ so wird wieder das Maximum der
Windueschwindigkeit in dem Umfanu'e der (lebietes der verticalen

Beiträcreznr Theorie der Bewef^imi; der Erdatmospliiire und der Wirlielstürme. ,-)47

Sti'ö!. luiiu- startfindcn. tind diese Maximiilge.sehwiiidigkeit, da B sein-
u-fos.s ist, und A- sehr u'l'o.^.-; werden k:iMii. noch eine Gni.s.se hahen, die
noch als siiinnai'tiy bezeiclmet worden kann. DeidvT man sich hierzu
noch die u'anze wirhchulc Afasse. din'cli irgend eine l i'saclie etwa von
dem 7ten iircit-enu'i'ade aus polwärt-s ucführr und den Ausdruck (-1")
nc,cli auueniiiirr uihi^' für eine niedere P>reite. so wächst 'D. ^m l> mit
der lu)heren l'reite etwa um das Aclitfache des Werthes, welchen es
1)eim 7ten l*reiteni>Tad ueli'd)t hat. es kann daher o) selbst, wenn y
weit unter tZ^f sinkeii würde, noch einen bedeutenden \\ ertli haben.

Der Sturnidurclimesser /ràY'//.s? demnach mit der Zeit, während die
Windü'eschwindiokeit uleiclizeitiu' abnimmt und der Deviationswinkel
immer mein- wächst, wenn der Luftwirbel von der niederen Pjreitc
polwärts wandert. Ein Zahleid)eisj)iel m(»ue dieses nälier erläutern.
Unter der Breite <S° sei ein \Virl)eI entstanden. Es sei dabei

•l'A sin. 8" _

so dass -2 H = 0.000040H

und r = 0.()0004■r^:^

ftec.
1

SPC.

sodass r> 2;r und — = m = 0.44

ist. Der iîadius des (rebietes der v(;rticalen Striimung sei 400,<HM) ;//.

T^ . T T^ ,, -D. Sinti - . , ,.., ^-IkslnH

Da m diesem rail = 1 d;dier A

r[^-f)

VI- . ^ ,

7-j r;r ist, SO hat man

( ! - m)-

liierans bereclinen sicli folsrende Werthe

348

D. Kitao.

B =: 400000 meter

OJ

o = 400000 m.

o = B

12 81 ■'"^^''''
sec.

o ^ 40000 ]n.

B

20.(;s

. 10

;> -^ 4000 w.

B

•27.!)

100

<> -- 400 m.

B

87.08

'" 1 ()()()

riinciii solclicii \\ ii-l)elstnrin wiii-de iii;iii einen Dnrelnne.sser \o\\
2x 10 Kilometer znselireihen. Wenn wii- ;inneliinen. dass, indem dieser
AA ii'l)el :illm;iliL;- in den 70ten I *>reileni;i';id ^elnnut. ;' etwa den Wei'th
O.OOOOi' angenommen ludje. so dass jetzt ;'<:'2« ist und das \[;ixinnnn
der AA indo-escliwindiükeit den I nifanu' des (^ebietes der vertiealen
StrcuniiiiLi erreicht, liat so erhält man i'iir dieses ^[aximiim.

400000 X 0.00002

2

2 / /2/ \-

-yi +(—./. 70°)

400000 X 0.00002

/, /sin 70^ V ,^ ,,.

meter
see.

Der Sturmdui-climessei- l)etr{igt denn jetzt iiniJ-eiHhr das Zelm-
lache v<jn demjenigen imter dem <Sten P)i'eitengrade. als wenn der
Durclimesser des Wirbels sich um 2 x oliO Kilometer = 720 Kilometer
aus_2;'edahnt hätte.

Die Windoescli windigkeit nimmt nach dem C^entriim in's [In-
endliche zu. wenn /'>-2;f ist; die von anssen her einströmende
Luft gelangt gar nidit nach dem (Ventrum, sie wird, da der Druck
nielit negativ in's Unendliche wachsen kann, in einer gewissen Entfer-
nung von dem Centrum zerlassen. Da innerhall) dieser Entiernuni;-

Beiträge zur Theorie der Bewegung der Erdatmosphäre und der Wirbelstünii«'. 349

die Luft nur eine eniporsteio-enrlt^ Heweo-iiiiii' li-ilx^n und an ilci- \\'ii-lH'l-
lieweçruiiL!' rinu'snin nidit mein" tliciliH^luncn kann, so niiiss. wenn
y^:>-2H ist. ein <Tebier dci- \)'inHstillcn luii das (entriini entstellen,
welehes aber verschwindet, soliald y initer 'In sinkt. Fudeni wir
dieses Gcljicf dci- W iiidsrillcn niilier /ii bestimmen versuchen, wollen
wii' die Fläelie. in der ('in(^ Fliissiokeit zerreisst, dni'cli die < 1 Icicliiniü'
detiin'ivn

p = Co)iüt.

wo (h'c ('(^nstajitc den Druck bedeutet, nntei" dem ein Gas anflu'ti't,
eine zusammeidiänoende Masse zu i)iJdeii. Für ideale incompressible
Fliissiokeit könnten wir mit AOn Helmholi/. * diese Constante = 0
setzen, wenn bei der P>eweu'iuiu\ weiche wii- betracliten, das Ge-
schwindio'keitspotential existirte ; 'allein, da ein solches hier nicht
existirt, wollen wir uns liiei- damit beii'niiiicn. anzunehmen, dass die
Fiift überall zei-iassen werde, in der Fläehe. wo die Druckvermin-
deruni;' einen o-ewissen dur<-li den Zustand der Fuff bcstinunten
Werth eiTeiche,

Wir haben als l)riick tin- das innere wirl)elerfüllte Gebiet ge-
funden [Gleichung (70) j)ag. l!'4. \o\. \ dieses Journals]

^^ = Coufif + G— -^ z- + Y^ jr

n '1 S

/.- fiitr ti ir

'Im [VI — 1 ]

,r r -1 ///Vi-'"-' / o \2w-2-| f

Es sei Po der Druck an dem l nu'linut' ^\*^v kreisförmigen Wirbel-
gebietes. Wir erhalten dann dnrch Elimination der Grosse Con^f + ^t.
indem wir zur Abkürzuni»- setzen

* Helmholz Über die discontinuirlichen f'lüssigkeits bewegimgen. Gesammelte Abhand-
lungen Band I pag. 14J*.

t An dem angeführten Ort ist inn Druckteliler un.Mrriù'irt gebliolxjn. Der liier an-
gegebene Ausdruck ist der richtige.

35 0 ^' Kitao-

«A-- L V/Ï/JV 2 A 4 ^ (m~l)V ™(»h-1)'(2ot-1- ^

wol)ei l^;//^l mir] ?// ^ 1 sein innss. Tin F:ill )» = 1 odor 7/? = -^ ist,

il'ill del' ()1)io-(' Ausdruck uiolit inch)- ; soiiderii .-inderc Aiisdrnnke, die
MiK'h leiclit eiit^'ickelt ^verden kiunien.

So liiiio'c i^/;/>l ist, ist /',, — /' d(Mnii:i(']i iilHn*;dl ])Ositi\' und ond-
lifb ; denn d;is erste Glied, dns den ui-össteii Wertb besitzt, ist positiv
und endli«'li. 1st nbcr 2 ;// < 1 , so \ ei*\\;indc!t sieb der oliiuc Aus-
druck in

^-['-(i)l('^)(T+a:i^)-

m (1 —mf{\ —2,7)1)

'j),,—p wii'd bicr in dem Centrum ])osifi\' unendlicli o'i'oss d.li.p iieir:iti\'
uneiidlicli üi'oss. Es muss d;iliei', da /> \n]\ p bis - oo al)ninniit, cineii
Wci'tb \ Oll f L;-eb('n. wo /» d(Mi Werlb annimmt, unter tlcm die Luft
zeiweisst. liczcicbncn wir diesen l)ruck mit /'„. so steUt «lie Uleicbunu-

// IV- L \ 7V / J V --^ A -i ( l - vy/J-7 "^ ( 1 - m?(:2vi - 1 )

die FliLebe dar. an <]ei' die Luft zerreisst. und innerliall) deren die Tjiit

Beiträge zur Theorie der Beweguag der Erdaluiuspliüre uud der Wirljelstürme. o51

nur eine aufsteigende l^ewe_o-an^- liat.* Setzt man bierin --=0, .so
erhält ineu den Radius eines Kreises, iniierliall) dessen Windstille auf
der Erdoberfläche lierrsebt. Dass ein solches (Tebiet uanz im innerer

des Gebietes der verticalen Stiiiiiiiiiii:' liei;t. ist leielit einzusehen. Da

p — P

" .j., " 'lositix- ist. iHiil del- rechter Hand siebende Ausdria-k tiir

-W =^ 1, : = 0 verschw indei aber tiir(-~j^H» unendlich gross wird,

s<j folgt, dass -W- ein achter I>ruch sein muss

Die durch die GleichunL: (21^7) dargestellte Fläche ist ein Hota-
tionshyperbolo'id höheren Grades; das Ivaumgebiet der Windstillen
Nerbi'eiterT sich mit dei' Meeresbiilie. und erreicht die Grenz der \er-
ri<'alen Strömung. (Li wo

Po — i'o ^ f ^

y. 2

d. h.

Die Geschwiniüu'keit der aufsteiofenden Luft wäre dann dieienife,
welche die Luft erhält, wenn sie aus einem Gelasse unter der Druck-
defferenz (/',, — ^'„) sti-öniT. Indessen, jede Schlussfolgerung hieraus
ist wertblos. «la. wenn auch p,,— l\, ein Paar Centimeter Quecksilber-
ilruck Ijetragen sollte. : eine solche Meereshöhe d;irstellen würde, dass
die <iilrigkeir unserer l^'ormeln als längst erloschen betrachtet werden
muss.

Im ein Paar Beispiele bei-echnen zu können, setzen wir /^,— /'„
= 7()0 ;//;;/. Qiiecksilljerdruck. was selbstredend in der Erdatmosphäre
innnöoiich ist. Im ball

* Williaui Ferrel (The Motions oï iluids aud solids ou tlie earth's surface. Wasliington.
1882. pag. 2Ü) hat durcli eigeuthüuiliuhe Behaudluug der hydrodyuauiischen Ctleichungen, als
die isolmvische Fläche für eine wirbelude Luftmasse, eine Kot itiousfläche vierten Uradcs
atigeleitet uud zwar ohne Berücksichtigung der verticaleu Strömung.

ot«) D. Kitao.

2 / sin ß , u I

r 10

E=1000??z..

haben wir d;inii für die Gren/e lier Winrlstillen.
273S67 = -2.62:î {^ + 1 5.482 (-^)^

+ ,..-„,,. (Ay

Hieraus Hndel innn aiiiiäliniugswei.se

IL =. 0.(K):S'.):-i. c]. i. n =-- 8/JH m.
R

Die <Te,s<'li\viüdi<i'keit der l.iit't i.st daKei

.^ , meter

(0 = ()^.2

sec.

In einem solchen Wirbel iiiiis^te der W^nd in der P^ntfernung 4
Meter von dem Centrum pl()t/.ii<di ;nifli<>ren.

Jii dem Fall -^ ^^ -y-r- hat m;ui hingegen

27(i.V.)7 - (d2.5 (^y + 2004.0 ( (—\

+ '"'(b)

Hieraus tiiniet m;ui auüenährt

-^ = O.OÔC)'.) d.h. o --^ 56.'.) m.

Die Gesdiwiudigkeit der Liit't ist dabei

meter

VMA

sec.

Das Gebiet der Windstille ist gi'össer, als in dem ersteren Heisjnel.
In einem solchen Wirbel würde die Windgeschwindigkeit allmälig

Beiträge zur Theorie der iJeweguuji' der Erdatmosphäre uud der Wirlielstüruie. 353

ofeo-en d;is Ceutnmi zuiieliint-ii. bis sie liei ö" ;//. Krifff^iTJUiii:' <'^

CO

en

iiictcy

Werrh loi. errei<"lii. inid ilic A\'iii(lsfillc pliit/licli (^inti-irt.

sec. '

Wie man .sieht, ist man im Stande, den u'anzeii \ erlauf der
kleiiieren \\ irljell)ililinigen in der I'^rdarmospliiire. w ic Wettersäiden.
und Tornado.s, vollständig zu x'liildci-n. wenn man mir auniminl. dass
j->.-2« sei. Bei den Wettersäulen ist so gut. wie sicher, dass die
< Geschwindigkeit der vei-ticalen Strömung, die sie veranlasst, eine
bedeutende Grrösse hat. wie denn mitemporgerisisene Kr)r|)er Schrau-
henhahnen heschred^en. deren (ianuhöhe gegen die Windung oft so
bedeutend ist. dass die letztere gegen die erstere i'ast \ erschwindet.
Dasselbe gilt auch tiir ilie Tornados ; denn die \Vindl)ahnen. wie sie
LoomisT und Andere nach der Kichtuni;' der umo'erissenen ßäunien
gezeiclinet haben, tragen ohne Ausnahme einen auffallend centri-
jxitalen Charaktei', was auf eine gnjsse fieschwindigkeit der aufsteigen-
den Strömung hindeutet. Wenn die Annahme y-:^'2.n fei'ner den
Umstand leicht luid imgezwungen erklärt, dass die sturmartige
Aiüregung tlei- Luft so wohl bei den Wellersäulen, als bei den
Tornados, auf ein kleines ziendidi scharf markirtes fTel)iet l)esc]iränkt
ist. dass die Lutt in einiger Entleriiung \on dem ('entrinn des AVirljels
wenig oder gar nieht in Mitleidenschaft gezogen ist. so glaube ich mich
zu dem Schlus- berechtigt, dass nur die Zulassung der Grösse ;'. Avelche
üTÖsser ist, als -2;/ im Stande ist. \\ ellersäulen. und Tornados in
ihrem ganzen Hubitus zu erklären.

Ijei den eigentlidien A\ irbelstüi-nien scheint der unter niederen
Breitengraden fast immer lieol)achtete schi-oife Übergang des Sturms
zur sogenannten "• fodtenstille daiaiil' hinzudeuten, dass man ;' auch
hier einen grösseren Wertli beizulegen habe, als 'In. Es Jässt sich, wie
wir oben sahen, nicht nur das Wachsen des M iirnidurchniessers mit

t z. B. der J'oriiado von Stow in Ohii) ;iiii :i()teu C>et<>l>er ls:57. (lii-y'. " die Wirlielstünnc '"
pa-i;-. O'J.

354 ^'- I'^itao.

der höheren Breite ungezwuuü'eii ei'klären, .sondern auch der Um-
stand, dass eine xöllige Windstille ini Innern eines Wirbelsturnis fast
nur unter den Iropen v(jrkuinjnt ; also da. wo er entstanden war. und
/' seine \ olle Tirösse Ijesitzt 5 dass in den höheren Breiten nur eine
Lull, niclit eine ('ahn im Innern eines Wirheisturms vorzukommen
pflegt. L nter den Tropen, wo die aufsteigende Strömung noch seine
volle Geschwindigkeit liat, wird die Fläche, innerhalb deren der
A\ ind orkanartig ist, nur eine lieschränkte sein und das Gebiet der
A\'indstille um so grösser, je grösser /' gegen -Ih ist. Wenn nun ï
allmälig abnimml. während dei' Wirbel polwärts wandert, debnte sich
das Sturmgebiet weiter und weiter nacli der lîegren/.ung des Gebietes
der \ertiralen Strömung aus, die den Wirbel \eraidasst hat. während
die Windgeschwindigkeit in dem Sturmgel )iete abnimmt, verschwindet
das (jehiet der Windstille in dem Grade, wie ;' sich -Ix nährt. Sinkt
/' nun unter 'l^c, so erreicht das Sturmgel)iet seine Maxinialausdeh-
nung, vorausgesetzt, dass der (Querschnitt der verticalen Strömung
sich mit der Zeit nicht geändert habe ; es tindet nur eine allmälige
Abnahme der Windgeschwindigkeit nach dem Centrum statt, eine
Lull, und das Gebiet dei' Windstillen ist verschwunden.

Die (hu'ch die Gleichung (ll^) bestinnnte verticale Strömung der
J.uft erleidet auch eine Ablenkimg durch die Erdrotation. Wenn
auch die Bewegunggleiclumgen eines \ertical bewegten Lulttheilchens
unter dem Eintiuss der Erdrotation leicht entwickelt werden können,
begnügen wir uns hier mit dei" Annahme, dass die BeweHuno- in
höherer Breite vor sich gehe, dass daher der EinÜuss der Erdrotation
auf die verticale Strömiaig als verscliwindend angesehn werden
könne.

'.Beiträge zur Theorie der Rewei^imi; der Erdatmosphäre und der Wirlielstiiriue. 355

Er ist :iher nötliio- näher zu nntersurhen. eineTi anderen Umstand,
von dem die Zulassigkeif dei- Ann.-ihmen wesentlich nhhans't, unter-
rlenen wir unsei'e Bewe_2'inio-so-l(:.irhnn<:on aho-eleitet haben. Da die
empörst ei_a"en de Tdift sieh aiisdelmen nmss. so kann, wie Reve* mit
Eecht heniei-kt ]iat. liie Oestah des Raums;ehiete8 der verticalen
Ströiniino' unniöalifli eine Cyhndrisehe sein, sondern eine konische
Avie 7nan hänfio' an den Wettersänlen und aiioh an den Tornados
beohaclitet liat. Es tVäo't sieli nun, wie Aveit inaJi das Ranmo-ebiet dei"
vertikalen Sti'Ömung als cyHndriscli anselien kann, wenn eine Lnft-
masse entweder theilweise oder völlio- o-esattiot von einem o-eoebenen
Gebiete auf der Ei'doberfläflie aus emporstei_o-t nnd so eine verticale
Strömnno- bildet.

AVir denken uns zu dem Ende das (lebiet auf der Erdöl verflache,
von dem aus eine dam])fgesä(tio'te Luft emporsteigt, durch eine ge-
gebene C'urA'e begrenzt, deren Gleichung.

I(x,ii, a,h)=() (22)

wo n. I). zwei J^ang(Mi sind, wcli'lic die Dinu'iisitni ilicses Gebietes
bestimmen.

Es sei !J der (^uersclmitt des Raumgebietes in der Meereshöhe
: und IJ^ v]]i solcher auf der Ei-doberfiäche. Man denke sich bei :
einen um (/ r entfernten Querschnitt, so dass das ^"olumen zAvischen
diesen beiden Quersduiirteii = i> .7: isr. Wenn M das specifische
Gewiclit der [.uft l)ei dei- IV'inperarur H, welche in der Meereshöhe r
herrscht, und unter dem Druck p liedentet. und Q die Dampfmenge
in 1 Cubiknietcr Luft bei derselben TemjH'rattu'. so geht dui'ch den
Querschnitt : die Masse

{M + QMJ (:>:\)

* Reye '■ dif Wirlxilstiiruu'. etc." pa^j. 2213.

sioh M mil — 7 — (/:. (i) vci'miiirlci't sidi iim —~-(l: dnrdi rotir]('iis;i-
(h: (tz

fj (J
tioii, iiiid CS \ oi-äii(]('rt" sich (J uni —^tlz. Die Müsso. welche (liir<-li
' a:

35 ß T). Kitao.

Tiiflem die Liift durch den (^)uei-schiiilt : + ilz o'eht, veraiiderf
'^^1 !.. n ;..^..... -,.1, 'W

fl:
den <^)iierschiiif1 : + (/; i^chl. isl deninncfi

wenn das condensirle Wnssci- mit emitorissen wifd. Soil (]i(» Masscn-
Ix-weuiini!; in jedej- I jiftschichte cine st(4iL:e sein, so nmss die Dift'ei-iMi/.
der Ansdriicke (i^'l) mid (24) ^ 0 sein ^/. //.

<1^^) "-("-.')#"-"

lei'aus lo|o-t

,. Const

M + i^

l»e/eichnet del' ln<le\ " den W'erlh d<M' hct l'cMeiidcii (iri'isse an del"
J^'dolxn-fliiche. so hat man

i^o ~ /»/ + Ç ^^''^

Et^^as anderes wiri] die Sa<'hc. ^^('nn das condcnsii-fc A\ asscr in
tropi'hîir-fliissiuxn' Foi'iii niederfallt, mid die Masse, welche dni-ch den
Qiierschnilt : + J: ij-eht. sich tliatsächlicli nni die Menue Dam pi' ver-
mindert hat. wclclier cjiideii-irt wnrde. Fn diest^n Fall ist -^r-d: in

r/r

(24) neuafiv zii setzen und die Diiferenz der Ausdrücke (-•>) und.
(21) uieich der c()ndensirteii Damptinenu'e, welche etwa als Keo-en

niederfallt. .1. h. -^ ^ U ,1-

dz

\\\

V ha Men s<»niii

Beiträge zur Tht^orie ili'v Hewegnnt;- <ler Erdatmosphäre und der \Virl)elstüruie. o57
\ dz dz \ ^ / dz dz

il. 11.

,,4^ + (,/+..)4^.o

dz \ ' ) dz

Hicnuis iolo-t.

/•z (IM , / ü \

Denkt man sieh f1 ans einer l'x'ziehnno- zwisehen ß und z

o
e]iininirr, .so erhält man. so wolil ans (25) als ans (2()) -^ als eine

Fnnction von r, etwa F(z). Da jedes Fläclienelement d iJ=d.r dy aich.
im A\^rliältniss 1 : F [z) verändert hat, so veränderen sir-h a nnd // in
('22) im AVrhältniss 1 : ^ F{z). ^lithin erhält man

als Gleichung der Gestalt der verticalen Strömung. Es findet eine
Verbreiterung nach Oben zu statt, weil iH+ Q gewöhnlicVi mit wach-
sender Meereshöhe aljnimmt. Da Q immer sehr klein ist gegen M,
wollen wir diese Grösse vernachlässigen. Es kommt dann sowohl
aus (25). wie aus {2Q) :

Nun ist

oder nach (II)

760 (rt + ^)

_ P^l R \2 dz

760 (a + fi)

858

D. Kit an.

WO {> das specifische Gewicht der I. lift im Norinnlznslrnid und
a = 278 ist, so folgt

n verändert sich mit dein D;unpfgeh;dt der Luft inid durum
mit der Meereshühe z. /i ist, wenn die Luft mit D:uni)l\2;erättio-t ist,
eine höchst comphcirle Function von .r.f Wir wollen uns indessen
hier auf die lîetrachtuno- einer mässig'en Meereshöhe beschränken
und setzen

B

i-Th)

1 fi.-O

Wenn wii- ferner für 77 den Werth o-ehen, den es an der Erdober-
fläche hat, so erhalten wir

\ a + Ool

Da, — —77- eewöhnlich ein sehr kleiner Bruch ist, so hat man
a + ^o

angenährt

^ ^ (2Grt)

wo

1 / 1 \ . ,

H = j- ( —77 7 ) ist.

a + ^o V // /

Die Zunahme des Querschnittes der verticalen Strömung mit ~
erfolgt demnach sehr langsam, und zwar um so langsîunei-, je höher
die Tempei'atur auf der Eroberfläche ist und je geringer der Dampf-

t Vergleiche- W. Thomson. Mathematical-physical Papers Vol. 111. pai^-. 255.

Beiträge zur Theorie der Bewe^uu-^- der Erdatmosphäre und der Wirhelstüruie. o59

o-eh-alt. da ' mit der Feuchtiijkeifc im Allfifemeiiieii abnimmt. So liat
man z.B. für /y„ = 0 77=80 7 = 0.0055.

für tio = 30

« = 0,UOÜ101.S
^ = 0,0000918

Für ï = 1000 ;//. würde si('li al« die mittlere Breite der verticulen
IStröinnng" ergeljen

V /4 " V 1-0,0918 ~ ^'^^-^"^

A\enn wir die Fläche als ein Kreiskegel denken, so erhält man
als die Tan^uente des hall)en Otfuungswinkels

für ^0= œ 0,0000552
flu- ti^=\\(r 0,0000498

Avoraiis sich so kleine Offniino'swinkel ergeben, dass man wohl das
Raumgebiet der verticalen Sti-ömung bis zm- Meereshöhe 1000»/.
als rein C}dindrisch betrachten darf, wenn - den AVerth 0,0055 hat,
wie es meistens ungefähr der Fall ist.

(ranz ähnlich verhält sich die Sache, wenn eine kalte dampfarnic
Luftmasse niedersteigt und so eine \ertical niedersteio;'ende Strömuna'
bildet. Sie kann dabei, indem sie sich nach unten zu erwärmt,
Wasserdampf von der verdrängten Luft aufnehmen. Allein ; wir
sehen davon ab, und erhalten wieder als Bedinuuno- der Continuität
der bewegten Masse

dz

360 ^- ^''^^"■

d. //. wieder

oder

i^V= MJJ^

o

<4 - M

wo !J^, M^ den Wertli dei- betreffenden Griisye in der Meereshöhe
bedeutet, von der hernb diese I.uftin;i«se sinkt. l);i .1/ mit nbneh-
mender Meeresliöbe ziiniiuint. so nimmt -^ niicb nncb der Erdober-

""o

fläche zu ab. d. //. das Kanmgebiet der vertical herabstei*i'eiiden
Strömung verengt sich nach Unten zu.

Teil bemerke noch, dass W. Ferrel * durch eine ihm eigenthiim-
liche geistreiche Combinahon einfacher Gesetze für die Gestalt der
Wettersäiilen ein Liotationshyperboloid. wie luisere Gleichung (-<>(')
erofeben würde, ü^efunden hat.

§ XI\'. — Slchislaùomirc l\ irliclbilduiujcn in der
Erdatiiiosphäre.

Xach dem wir die pliysikalische Bedeutung der Grösse ^ermittelt
haben, wenden wir luis zu der Aufgabe, Wirbelbewegung der Luft zu
linden, wenn r eine gegebene Function von der Zeit ist.

Die lîewefi'uno'sirleichuno-en für dus äussere Gebiet sind ;

+ nJ TFa = 0

und für das innere Gebiet

* W. Ferrel : Kecent Advances in the Meteorology Washington. 1886 pag. 299.

J?eiträ<4'e zur llieurie der Be\vi'<;iiii<4' der Erdatmosphäre und der Wirhelstüruie. o{^l
+ {n^^ r) J TF; ± -1 / sin H r -: 0 (2,S)

-Jfi = + r-

wo d;is obere Vorzeichen fiii- die vertical aufsteiuende und das iintei'e
für die vertical niederstei^'eude Sh'('ni)iiiiü' ü'ilt.

Da in diesen (Tleicliiinaen -^-v nicht vorkommt, so beeinträchtio-t

"er ^

die Abhä.ngigkeit der Grösse /' von der Zeit auf keinerlei AVeise die
specialisirenden Annahmen, untei- denen wii- diese Gleich uu'j'en ab-
geleitet haben ; sie gelten auch iiii- ein mit i\k'v Zeit veränderliches /'
ohne Weiteres. Da ferner die Gleichung der Massencontinuifiit für
incom)_)ressib]e Flüssigkeit xon / \<)llig unabhängig ist, so uilt sie
auch hier im unveränderter Form, in so ferne als wir fortfaliren, die
Luft als ineom[)ressible zu betrachten.

Wir wollen zunächst nälier iintei sui-lien. ob die Bewegung der
Luft auch eine nichtstationäre Form annehmen könne, wenn y zu
irgend welcher Zeit constant werden würde. Wir denken uns zu
dem Ende das unendlicli Luftg.-biet l)egrenzt. theils durch einen
Cylinder vom unendlich grossen Radius, theils in der Endlichkeit
durch das cvlindrische im übi'igen beliel)ig gestaltete Kaumgebiet der
verticalen Striunung. Es soll die Luft in der Fnendlichkeit überall
ruhen so dass

"bt ~ "bz ^ !J ^ ■'■ ^ U ~

■bJW, ôJT'Fa 5J1F,.

Die Function ç und ihre ersten Differentialquotienten sollen
überall stetiii' endlich und eindeutiü' sein. Von der Function II

362 ^- ^'itao.

setzen Avif voraus, «lass sie iihenill eindeutio- sei, dass ihre Differential-
(juotienteii iiherall sonst steti<2,' und endlich, aber in einem l\inkt
a, h innerhalb des Gebietes der verticalen Strömung unendlich gross
werden könne, aber so, dass das Integral

fj W

d,

über einen unendlich kleinen Cylinder um den Pnnkt a/ü aiisgredehnt
entweder verschwindet oder einer endlichen Constante gleich wird,
die aber noch die Zeit enthalten kann. Wir denken luis nm diesen
Punkt einen Kreis beschrieben mit dem verschwindendem Radius (e).
Da in diesem unendlich kleinen Kreisgebiete JW '\\^ eine Function
von dem Radius o allein bcti-achtet werden kann, so liat man

Soll daher 1 JW dco mit verschwindendem £ verschwinden, so
kann dies geschehen, wenn o—^ — mit verschwindendem £ ver-

schwindet. (/. //. —. — so uneiidli('h gross in dem Punkt a, h wird,

(I o '^

dass es ize^en — unendlich klein ist.

Es lässt sich nun lieraus schliessen, dass die Functionen

.... IJW y-JW • • 1 Dl, / 11-1 1

J vv , — r- — . ^ ,., wenn sie m aem 1 unkle a. h unencllicli weiden,

àt ot'

unendlich klein «ieüen — r- sein müssen, wo î eine verschwindende

Grösse ist, falls i>~^ — sich überall nur stetig" mit der Zeit Lindert.
Ditferentirt man die Gleichung (-7) nacli ?, unter der Rücksicht-
nahme der Bedingung, dass /' von / unabhängig sei

Beitrat?« zur 'J'hoorio der Beweguu«^- der Erdatniospliän' und di'r Wirlielstiirine. ,')(),')

7 1 ^ <^

d. h. ^rr- =0

ot

so erhalt innii. iiineiii wii- -r— - =^ ^ setzen.

ü r

ôjj^i ^JTT^ ?)}^i iJ]]\ Up, , /^^ir; , ^p,\^jp

t f ' c/ .r c^ // "by "h x \ "^ ]/ ^ ■'' / ^ t

+

Innerhall) des iindliclien klein(-ii um den l^iinkt a.h besr-hriebenen
Cylinders verwandelt sich diese Gleiehiinii' in.

~3r + 17^7- + '^^ + '^-'^*^^

Xon verschwindet -r^^ in einem solchen Gebiete, wie (î'>. Soll
daller diese Gleichung auch in dem Punkt a. h erfüllt sein, so müssen
die »Trossen — r-^-^ , -:r-^ ^ . von derselben (.Trossenordnunü" sein,

dt ü 1/ c y '-

wie Jp^. Da 7^' von (h:;rsell)en Ordnung ist, wie Jp-^, so muss

-~ r-^ audi unendlicli klein sein o-eoren — ^r- Mithin folo't, dass

-^r — - unendlich klein sein muss s^eo-en -^, dass daher das Tnteo'ral

ausgedehnt über (he verschwindende Fläche um den Punkt n. />,
verschwenden muss.

Für einen Punkt auf der Grenzcur\'e der verticalen Strömung
kann man die Gleichung (28) in eine andere Form bringen. Be-

3ß4 ^* Tvitao.

zeichnet man mit // die nach Innen i^'ezogene XonnaJo an der Grenz-

curve nnd setzt man

a — co.s(/?.r) ^i = cor in y)

so ist

IJW, -bJW, SJTF, SJTT^ .

: (J. ^rr i

Ml'i IW, MV, MV,

<i.

^ (:-;ü)

ä .r Ô ;/ 3 7/ ö ;? '

Wenn "wir nnnelunen, dass die Greiizcnrve der verticalen Strömnii^'
sich nicht Jindere, was wir, wie ich li-laulie, annelmien können, da die
Geschwindigkeit der verticalen Strömung sich nicht mit der Zeit
ändern d. li. die Temperaturvertheilung- in — und ausserhalh der ver-
ticalen Strönnuio- dieselhe hleilien soll, so halben wir ancli

3,r öy? ö // (in

(:!!)

Die Gleicliung- (l^!)) erhalt datier fni- einen Tnnkt der Grenzcurve
die Foi'm

Durch dieselbe lîehaiidlnng der Gicichting (-7) erhält man für
einen Punkt der Grenzcurve

Beitrag!» 7,nr Theorio der Bewocfiing der Erdatmosphäre und der Wirlieltrurme. o65

Er sei (1(0^ ein (^uerschnitteleinent der vertictlen Strönmiifi-. und
(hl ein Element des üinfanos der Grenzcnrve. Indem wir die
Gleichun«:' mit (ho^ mnlti)»li*'ir(MT. und uns der Formel erinnerii

— r ö.r?// S?/ S.r

crlmlten wir

+

Integrale von der Form / ^^ <?W; lassen sich nun leicht

in die Integ-rale umsetzen, welche über den umfang' der Grenzcurve

auszudehnen sind, wenn die Function wie ^^ — r— ^ überall eindeutig,

stetig und endlich Aväre. Da wir der Allgemenheit halber annehmen
müssen, dass es innerhfdb des Gebietes der verticalen Strömunsf einen-
oder meliere Un Stetigkeitspunkte geben könne, so denken wir uns diese
Unstetigkeitspunkte dadurch ausgeschieden, dass wir um sie Kreise
mit einem unendlich kleinen Radius beschreiben. Dann sind alle

,3 no D. Kitao.

die in Frao-e kominenden Functionen überall endlich in dem iihriof
bleibenden Gebiete. Da al)er der Betrao- der Flächeninteurale der von
den Un8tetigkeit.s|mnkten herrührt, nnt verscliwindendeni Radius der
umschriebenen Kreise verschwindet, so sielit man ein, dass die
Flächeninteo-rale (ooa) in Umfano'sintecfrule umsetzt werden können
ohne jede Rücksicht :)uf die l^anl^te, in denen die Functionen liätten
unendlicli gross werden können. Diese Bemerkung ergielit

„ S fhjp, , r- hJW, -, , , /•- ^JW; ,

du

~ ^'' + ''^ f'rit '^'' + Ai -' -' î^i "^'"^

Da entlang der Grenzcnrve die Gleichungen (30) und (ol) gelten,
so heben sich vier Glieder paarweise auf. Es bleibt nur, d;i a- + ß- = 1
ist

Es ist nun

und ferner

ö X

Beiträge zur Theorie der Bewegung der Erdatmosphäre und der Wirbelstüruie. 367

iy-' '^'V ^y »y

Hieraus folgt

= j ^''ÏX diO^ + / ^'T1J~ daJi - /jj6, Jj^i d(Oi
<i X ' 'h ij

Da ill dem Gebiete der verticalen Strüiiitiiig'

J^= + r
ist, so ist dieses Integral

= - /j^i -^ du T rf -iV>i doj,
Er kommt schliesslich

da /^r^ du = — J^, dco^ ist, so hat man

d. L, weil diese Gleicliung in jedem Punkt des Umfangs bestehen muss

|:^+,,,lï. + „M=0 (34)

ûnot on on

Wenn wir die Gleichung (i'7) auf dieselbe AVeise behandeln, so
erhalten wir, da hier überall

ist.

368 ^ J^it^^«-

Die Integrale .sind uuszudeliiien. theil« über die Grenzcurve der
verticaleii Strömurjo-, theils über einen mit unendlich grossem Radius
beschriebenen Kreis. Der Umfang der Integration ist also hier
unendlich gross, wie B,', \vel(.'hes den unendlich grossen lladius be-
zeichnet. Da aber die Luft in der Unendlicldveit überall und fügen
wir hinzu, immer i'uhen soll, so dass

JW^=^ Const = 0 -^ = Const = 0

Da ferner -r^ wie -^ verschwindet, s(^ bleibt von dem über B'
dn U

ausg-edehnten Integralen nur

„ /•-'/« du.
was dassellje ist

t2 - H R

In

Dîi nber 11.^ in nei* Unendlichkeit von der Zeit unabhängig ist
dass daher gesetzt werden kann

lAsiiiß

I^a=— V-.fa

(Vol. I dieses Journals |)ag. 171) luid ferner

^■^L=+ -ir R fo(j p + Const.
so dass

so fol Oft.

\ dn Jp^n- V df, ),j = n- ^ 'IB'

dn öt\ '2 X /

Beiträge zur 'i'heurie der 13u\vl'l;liu<4' dor Erdiituiosiiliäru iiml dor W'irbolsturuie. o()9

Mitliiii verschwindet (l;is iiljer den l infauif (\i's iineiidlifli uTo.ssen
Kreises ausgeführtes IiiteuraL Es koniiiit d;dier

wo die Iiitegration mia über deu Umfang dei' (îren/curve der ver-
ticalen Ströiimng aiiszufidiren ist. Da diese (Tieichnng wieder in
jedem Punkt der Grenzcurve stattiindeii iiiuss, erliaiten wir

-.r— ^ + jp^^ + H^ = 0 (35)

0)1 Ot 0)1 OH ^ ^

J)a nun an dei' (rreiizciirve -i~^ =- — r^ ist. so foln-f ans (ol)

OH du '- -^

und (35), da diese identisch erfüllt sein müssen

Geht man mit diesen (TJeichungen in die Gleichungen [r>'2) und
(3o), so sieht man, das« diese nicht anderes gleidizeitig erfülh werden
können, als durch

0 J{^i = Jj6^=0 ^» = ^it =0

3», In ^^ ^^

Es ist denn demacli

fjp.<'-'.=-f':t""-''

Hieraus folgt, da -IW=c~-lÄsiiifi nirgends und niemals sein

370 ^ Kitao.

\'(3rzeicheii wechseln .solj, du.ss in dem ganzen unendlichen Luftraum

<e\i\ muss (/. //, dass

li ^t

0 (H(5)

zu setzen ist. wenn y si(-h nicht mil der Zeit ändert, (jder zu
irgend einer Zeit einen endlichen Consta nten Wert h erhält ; dass,
wenn zur Zeit t = o y constant und endlich ist, der Bewegungszutand
der entstaricleiien Wirbel zur Zeit t = o^ durch die Gleichungen (ot!)
gegeben ist.

Wir sehen somit, dass wir wold berechtigt waren, als wir
mit der Annahme der l nveränderlichkeit des ;- mit der Zeit, (jhne

Weiteres — Ï— - — =0 gesetzt haben. In den nicht stationären Fällen,

wo y eine gegebene Function \on der Zeit / ist, jnuss der Anfangs-
zustand gegeben sein, der aber verschieden ist, je nach dem die verticale
Strömung, welche einmal entstanden war, von t=^o al), sich mit der
Zeit ändert, (xler von diesem Momente an zu entstehen beginrjt. In
dem ersteren Fall ist y eine endliche (V)nstante für t = o, es ist dabei
gleicligiltig, ob diese vei-ticale Strömung pUUzlich entsteht, oder längst
entstanden war. Tn dem letzteren Fall verschwindet ;- mit /.

Bezeichnet man mit /'„ den Werth des /' für t=^o so handelt sich
darum, eine Function W zu finden, welche in dem Gebiete der
verticalen Strönumg die Gleichung (-N), und ausserhalb desselben die
Gleichung {'27) befriedigt, und an der Grenzcurve der verticalen
Strömuuo' die Bedinuuni«-

für jedes / erfüllt, und . " , . " in der Unendlichkeit für iedes

Beiträge zur Theorio der Bpwt?criin<; der Krdatinnsphäre nml '\>^r Wirln-lstiirnie. o71

/ verscliwinrlen. Wenn die vcrticnle Striummu' znr Zeit f = o. die
Geschwindigkeit + j'o - li;it oder keiiie, so h;it 11 ', ferner in dem
ersteren F;dl die Bedinu-nnii" zu erfüllen

^àw, /^ç^ _^ yivr\ -djw, /\jç^ _ MVA

+ (,c + y^)û]\\'r-2À.^in^iro = o (:',7)

wo c'ir,— — -^ I Inf/ o (](,) ist.

nnd in dem l^tztei-en F:dl. wo /'o = " i>:t. soll

Ô ,r 3 //

= 0 JTFi = 0

indem wir annehmen, duss die Luft zur Zeit t = () überall u'eruht
habe.

fr, muss dabei der Gleiehuna' oenuo-en

Ô ,T

(¥ + ^) - ^(^ - 4!^)—- "(38)

falls j'o nicht verschwindet. Verseh windet j'o, so soll 1)'^ wieder der
Bedinguni»" o'eniio-en

TiX ä 7/

:0 zlTP\, = 0

Die Ditferentialo-leichuno-en für die Isodynnmeii [(44) und (45)
pag. 170 Vol. I dieses Jcnu'nals] bleiben unverändert, da hierin ein
Differentia l([Uotient nach t nicht vorkommt. Man kann sie durch
Integration der (.Tleichungen finden und damit den Druck, wenn IT
bestimmt ist, wie in dem Fall der stationären Wirbelbildung. Die
Gleichung für den Deviations winkel / (4()) pag. r70 Vol. I dieses
'Toiu'nals uilt al)er nur im Fall der stationären AVii'l)ell)eweguno'

37!2 D Kitao.

Für den vorliei^-enden F:ill finde?! wir leicht aus (9<:') (pa«-. 132
\'^ol. I dieses Journal.)

wo //'" + U- = r^/-

-7 ^ \--2/. am H= "

du- öj/

gesetzt worden ist.

Es lässt sich die Function IT für das wirbelfreie iinssere Gebiet
anch liier oanz allo'emein finden. Da auch hier ^(fa=^ ist, so lässt
sich die ( Jleichunii' (-7) audi, wie folo't, schreiben

Die (xrösse -IW^ ist ïnr das wirbelfreie Oebiet = Cousf = o. und da
terner — ^ + — -^ , und —iJl^^^l i,,^ äusseren Gebiete überall
und immer stetio- und endlicli bleiben sollen, so folo-f

J (-^ + n W,- 2/ sin ß çr ^ =0
was befriedigt wird, wenn wir setzen

^ + ;, w^- 2/ sm ß v^,= T cp, (39)

wo T eine willkürliche Function von t ist. Setzt man nun

W^ = e"""* F(xy f)

wo F e\i\e gewisse Function von den eingezeichneten Argumenten ist,
so dass

pH'iträiif /.\ir l'iiêiirii' der Bowe^junj^ fier Erdutinosphärc und dtT \\'irli>-l.stäriii''. ;j^ ;}

Kiiifiilu'iiMu' (licsf'.s in (8!') cru'ielif

Ilicf-iiis liiidet man

F= fr'' iT + -2/. .«i» H) c, fit -VC

wo (' eine Funetioii von .c. // allein s(-in kann, die a1)ei' dci" ♦"Tlciclinnn'
JC^o o-enüo-en muss, dn i-' t-incr solclien <^Tleie1innü- o-enüu'en muss,
was die Function f r^'^ (^T + ■lÀsiii f^^ç^fJt es in der That tliut.

]]\ = r " "fr " " (r + -2;, fil» tl) c, (It + r " " C. ( [())

kann als eint- allucnicine Liisuni;- i'iii' das iiussere wirbelfreie Gel)iet
lietraehtet werden, in so ferne, als zwei willkiii'iiclie Functionen 7'

und (' aiifti-eten. mit denen m;in ieder der Berlin o-uno-en, denen ^r — -

o'eniioen muss, o-ereclit werden kann.

Es soll r. B. zur Zeit f = o die Fuftbew^effiinii- eine stationäre
sein, so dass. wie Avir u'efunden haben

l^ao = ;; <f:.o f ao = + q^^ / ^"'' '' '^ '" '

Man o-eniiiit in diesem Fall <hu'ch die Annahme

^ 2/ sin H

SO dass die Bedino-una" ^C'=o erfüllt ist. und dnrch die Ausdchnnni»-
des Zeitinteo'i-als xon a his /. so dass wir erhalten.

374 ^- ^^'itao.

Die zweite willkürliche L'unction T bestimmt «ich ;ius der Be-
diiiüiiiiU', die nn der Grenze der verticalen Strönuiiig zu erfüllen ist",
und wird daher durch den IJeweuurji'szustand des inneren Gebietes
zur Zeit t l)e.stimmt. Soll /' zu jeder Zeit den Werth haben, den es
znr Zeit t = o o'ehaljt, d. //, soll die Ijewegung eine stationäre sein,
so hat man T = o zu setzen und <fa,=(foa.- Man erhält dann

T^a= e \^e - 1 J <f^, + e f„o

d. h.

W = —^ Uo

wie es sein müsste.

Für d;is Gebiet, wo der AVirbel niclit verschwindet, lässt sich eine
solche alJgemeine Lr)sung niclit tinden ; wohl aber, wenn wir uns d;is
Gebiet der verticalen Strüinuii"' kreisförmi": beç^renzt denken.

Wir fassen zuerst den Fall in's Ange, wo die verticale Strömung
eine aufsteigende ist. Wenn wir setzen

(f^ = '^ B- log p + Const.

<fi = '— ir + C on fit. o =^ X-+I/-

so erhnlten wir für das äussere Gebiet aus (:^7)
oder, wns dasselbe ist

( (

Setzen wir hierin

- Kf

JTK = e f

Beiträge zur Theorie der Bewegung der Erdatmosphäre und der Wirbelstüruie. 375

Avo ip eine gewisse Function von l und i tt) i-^t, so erlialten uir

S (S

was oreornücrt wird dui-cli

00 o

wo /' eine Avillkiirliche Function des Argumentes j y^^ "^ \-jr)
Ijedeutet. A\'ir erlialten somit

JW.^e-'-f\_fr.,^ + {-0] (41)

Eine Lösung, die als allgemein bezeichnet w^erden kann, da bierin
eine willkiu'liche Function auftritt. Die RotationsgescbAvindigkeit der
Lnfttbei leben ist somit

' = '2/ sin H — e

'V[/V'«+(l)l

Das Eigentbiimb'cbe bei der also bestimmten F^)rm der Wirbel-
beweuunii' ist, dass sie mit ae^visser mit der Zeit variabler (Tescbwindio--
keit sicli j'ortpßiDizt. " erbält nämlich, wenn }'=o wäre, denselben
Wertli, so oft die Grösse / ;-<:/;' + (^j denselben Werth erhält. Denkt
man sich die Function /" so bestimmt, dass sie für einen reellen AVertb
C verschwindet und dergleichen für jedes Argument, das kleiner ist
als C, so pflanzt sich die Wirbell)ewegung von Aussen her nach
dem Gebiet der verticalen Strönuuig fort, denn die Fortpflanzungs-
gesch windigkei t

376 ^- I'^itao.

ist negativ. Wenn deuinacli eine Nerti(:i,l aiifsteio-eude tStrömuno-
entsteht, so ^eratheu die entfernten Lufttheilelien zuerst in Drehunir,
und diese Drehung ergreift ein Liifttheilchen nach dem anderen
allmälig nach dem Centrum des inneren Gebietes zu. Avährend die
LufttheiJchen, deren Entfernung zur Zeit t kleiner ist, aJs // ^s/C—fydt
gerathell ikjcIi nicfit in Drehung, sondern bewegen siclj nur cent-
ripetal n;iclr dem inneren Gebiete mit dei- Geschwindigkeit, deren

bomi)onenten ~ — . und V-^ sind.
' à X 0 ji

Wenn die verticaJe »Strönuuig eine niedersteigende ist, so erliält
man, indem man — y, für /' setzt

Mitthin als lvotati(jnsgescli\vindigkeit eines I.ufttheiMiens

Auch in diesem Fall pfianzt sich die \Virbelbeue<nui<r fort und
zwar von Innen nach Aussen mit der (Tesehwindigkeit

(l^>^ yB

(Û ~ VcTfydF.

vorausgesetzt, dass der Aulangszustand der l>e\\egiujg so beschnffen
ist, dass die Function für (-^j —//'fZ^ = 6' verscli windet. Die entfernten
Lufttheilelien bewegen sich hier anfangs nur centripetal nach Aussen
hin gleichsam, als wenn sie zuerst durch die aus dem inneren Gebiet
herausfahrende Luft ^^ersclloben würden, elie sie in Drehung gerathen.
Es lässt sich -/ IF ebenso leicht für dass innere fJebiet finden.
Im Falle die kreisförmige Strömung aufsteigt, erliält die Gleichung
(l^S) die Form

Beitrüge zur l'heorie der Beueguug der Erdatmosphäre und der Wirbelstürme. 377

oder, indem wir dieses etwas iimlin'iiieii

W'eijii wii- liieriii setzen

-IW,=f{,o, t)-^2Àsln tie-^'^-"''' fr</''"^'-"dt

wo f eine Function von den ei nf»;esclii-i ebenen N :iriul)eln l)edcutet,
so kommt

. ' =-^ 2/. sin fl {}■ — }') c-^^^ re''^ -^ df —■lÄsiit tl y

d t dt ./

2/ .s/;/ ^ y = '2Äsin d )
Die Addition dieser Gleichungen ergiebt

Setzt man hierin weiter

wo ^ eine Function von / und (> ist, so erhäh man als IJestimmungs-
gleichung- für <p

1^7^ D. Kitao.

was l.)efriedi_i>t wird diu'ch

wo F eine willkiirüclie Function des Argumentes /yclt+Io(ji-'^\
ist. Wir erhalten somit als die allgemeine Lösung

JW,^ e - "' ^-^^ [i^'l fr dt + H (j^y] - -2/ sid 6fre"'~-^r'i< ^^ (4;.)

iVucli hier pflanzt sich die Wirl)ell)e\vegiing von Aussen nach
dem Centrum zu mit dei' Geschwindigkeit

'dt- ~'^^

lileichfalls luiter derselhcn Voraussetzuni;-. Avie (jhen. Für ein Gebiet
der vertical herabsteigenden Strönmng findet man hieraus dundi Ver-
tauschung des /' mit — /', so dass

JW.^e ~''''f^'^' Pi^"'/( 0'-/"/' '^^ + -^^' •^^■" f>f'/^'^' "" ''' dt'J (44)

Unter der nändichen \ Oraussetziing. wie oben, pflanzt sich die Wir-
l)elb(!wegurjg von Innen nacli Aussen fort mit der GescliAvindigkeit

Die in den Lösiuigen (41) (4:?) (4.">) uiid (44) auftretenden
willkürlichen Functionen sind nini geinjlss des anlangliclien lîe-
wegungszustandes zu besfiminen, welcher daher gegeben sein nmss.
Wenn wir nun annehmen, dass die Wirbel be weu-uno" zur Zeit t = o
eine stationilr<' gewesen sei, also dass /' zur Zeit t = o = Const — ;-„
U'ewesen sei. Es ist sonach für t = o

Beitnij^e zur Theorie der Bewet^un>;' der Erdatmosphäre und der Wirlpelstürun'. r)7 1)

It

0

Für (las Gebiet der vertic;ilaaf«teio'enden Striuniniu- isf domafli
zur Zeif f ^n

iL li.

m— —

/'o

was sieh ancli so schreiben lässt

T)i(^ r.(>siniü' (4.']) lässf sidi -uich so sclirciben

Setzt man hierin '' = 0, so hat man

jii-„=F[%(ty]

Ein Veruieich dieses mit (44r^) er^icljt sofort

^ ['""(ir) +/„ ' ^"J = - -^^^ L' - (ir) " /„ ' "Ü

Somit erhalten wir

+ l_(0'»'-",'»-'./>'"-] (44,,)

Wenn y nun für jeden Wertli von t denselben AVerth j'o behalten

380 ^- Kitao.

sollte, so iiiiisstc audi JW^ denselben Wertli behalten, den es bei t--=o
,i:eliabt bat. In der That verwandelt siHi (44/>) in di(\seni Fall in

'2À sill ft

iJi—y-^ I '" '''

_ 2/ sin tl

d. //.

JTFi = AW,,

W il- scbrcibcn zur Abkiiivimi;-

2/ ,s/;/ // = a

,=...'>■" /•'i../i''-rl.'V„ B

("-;•.)

''-To
so dass wir baben

Hieraus erbält man weiter
dWi o.e --X

dl'

_ ae -Jit p

Ist /' - Goiisf = ;',^ für jedes ^, so erhält man hiei-;ins

P>i'itra<4"' zur 'Iht'oric der ]^e\\t't!,-uii<4- del" Erdiituiosjihärc imil dt-r \\'irljtdstiiriiii\ ß.S |

do

-'■'I' r^''^''r.t " y^' .\ . ''.r^""'

['-:-(iry "]

,1. h

rlW

(1 I I (1 II

wie <'s sein sollte.

I']s li;m(lclt sicli iet/t (l;iiMiiii "'' t'lli- <l;is iiiisscrc (ù'l)iet zu

(III

bilden. Audi in diesem l'ail kfdiiieii \Vii'liei nidit \()rh!inden sein, d;i
[—^ — 1 fill- t — (i sonst in der 1 nendliehkeir iinendlicli lîtoss sein

V <h' )

miisste. Die LcVsiniLi' (40) uilf dnnuii in diesem lall.
Wii" setzen in (40) (' -n. inid

so (lass w n* liahen
Es isf daher

-T— Ii){j ;> + Const

EJîlLA.r^'frr^-n.inti^r/'dt

——^=: -—r (r + -2/.f<ni /hyr dt

d 11 •111 I

v\- i ' IT '^Wi dW^ .... ^ . 1 ,

Die (Trenzl)eaini>iinir — ^ ^— r-^ '"i* "==B ern'iebt"

'^ ' dp d II ' '-

e-^'iR r ^^, . „, rt .. U.C

- ' ['^^ l\T + 2}.si»H)yr''''df:

T^(-'--^)

d. h.

I {T + -2/. sin II) y/''df-=a (.l + B- —\
womit 7' bestimmt \V(^rden ist.

'-> v^ •; I ). Kitan

AVii- crluilteii somit

df> -
Fill* ;- = Consif — j-^ wird dieses

Brr-y> { C \

dp

_ _ alï-
d. h.

dW. dW„

do do

wie es sein sollte.

Hiermit ist die Aiifu;il)(' für den vorlieofenden Fall vollständio-
!i'elr»st. Ansdriickc iVii' die < Jeschwiiidigkeitscomponenten und Druck
können olmejede Scliw ici'iukeit aufgestellt werden ; die Defferential-
glei<'hung liir die Win(]lialin ist al)er auch l)ei der einfachsten Aniiahme
über die Function ;' äussert \erwickelt.

A\ ir beschäftigen uns mit der vertical lierai )steiçfenden Strömimo-
von kreisföi'migern (^)uerschnitte unter der Voraussetzung, dass die
Wirhelbewei'un«'- zur Zeit / o eine stationäre üfcwesen sei, d.]i.

¥-^^ + (-+ro)JTf',,-2/./.^^',^o

dp
woraus sich ergiebt

JÏÏA. = ''''"

^^+/'o

da -IWi^ für ,"--0 nicht unendlicli sein kann. Soll die F(»sung (42)
für t = ù diesem Lileich sein, so sieht man, dass man dieser P>edinsfun2"

genügt, wenn man setzt

Beiträge zur 'l'lieoric der liewegiiug der Erdatuiospliare und di-r W irlielstiirriie. otS3

und

JW.^ae-"''^^'^^'' (-i^ + f\r-''"^y^"dt\ (47)

Wird /' coiLstaiit iiiid=;'o, 5>o muss üir jedes /

sein. Av;i8 in der Th;it der Fall ist. Es ist näiidieli für y—Yo

d. h.

iSetzen wir zur Abkürzung'

so dass

AW:=a.l'r-''' (4cS)

Für das äussere Geljiet uilt die Lüsuni;'

und es soll f so bestimmt werden, dass

für t = o J W^ = A >Fi„

für o==B

dp dp

erfüllt sind. \']s ist inin l'iu- t ^^ o r = To.

3S4 1>- Kita

2f, df>

<L h.

+ ;./lIF =0

l'Jii \'er,u]eicli dieses Aiisdi'iicks mit ('!!') erii'iebt

folii'Iirl) ei-liiilt iiiiiii

à W^=^ -^ e - "' - "' [ (1) ' " ' -^f' /'"] (5(1)

Für /' = Coiust - ;v, koimiil

wie es sein sollte.

Xiiii ersieht sicli ans (4N) diireli Integration

ZWi a.l'('->:io

ôo

(50/i)

weiJ -^ — - nie für ,"--o nnendJicli liross werden darf. Es i;ruiebt sich
ferner ans (50)

ö^o '2iii, (1 + ?y; ) /^ /'

wo (,/ hier eine Function von / sein kann.

Bcitriij;f zur TliL'urie dt'r Bewi'^-iwij^- der Krdatuio.spliaro uud der U irlielstuniie. ^NÖ

Die rxMliiiniiiiL^-

S)i; Ml'i

ö,o

■R

eruicnt (lalili

Hieraus folu't

O-B ^),i + i,i i'^y,H C_ O.B -nt .,

'2m(\ + vi) '-' ' '^ B ~ 2 "^

„ ali- _ >./ / r - " '^ \

C^=^^ri— e ( -^ + i I

'2 \ iini + III )/

folglich crlialteii nil'

/"joy'i

II 'A 1)

[--^;;^S('-'■■'■t^-^■3)] ^-)

womit (lie Aufgal)e auch für den Fall der lierai ).steii:'endeii StrümiiiiL:'
vollständig' gelost worden ist

Tni Fall /' für jedes / constanj, = To i•'^^ l>:if nian

Mn aBri'-^< r /'o ,-/o^

ô.o

[

+ — ^^—

}'oi //' + yof _ |\

+

'"/o '

in (J +^/^)

(,_„ »:[(-^) = -0)]

uE^e-^'i I /-„ K/

[

2,o L^'+J'o in\\-{-iii

„ +j:î^(.-,.-"[(^j-0)]

i/. (i.

wie es hätte sein niiissen.

;^8()

Ü. Kifcao.

\\\v wollen ilif lÀeclimiiin in eiiii_ii'eii einfaclieii FiiJJen durch-
liilireii. Wir uelimeii zuiiäch.st an, dam

1

r

a-\-t

.sei, wo ;i eine Zeit ist und = — . Es i.st dies ein F;dl. der in so

/o

ferne vom Interesse ist, als es iingetUlir dei' Wirbelbewegung der Luft
entspricht, welclie entsteht, wenn eine Luftniasse von einem kreis-
i(>rnng l)egrenzten fiebiet aiifdei' Erdoberfiäehe aus durcVi ii'gend eine
Ursache emiJOi'iieschleudert wird und so eine \ertical aufsteiü'ende
Sti'()nunjg xcranlusst. Denn es ist

dr \

dt (a + ff

Mi thin

Die Beschleunigung ist demnacli Nidl ; mithin \ ers<dnvindet die
Diffei'en/v der verticalen Temperaturgradienten r' — r.

Da

_ z dz

ist, folgt

z = C{a + f)

wo (' eine Constante is(. ?vennt man die aniängliche <»eseli windigkeit
der empcn'gescldeuderten J^uftmasse /r,,

so ist (' = /r„

mithin " = "',,

Jedes lAifttheilchen steigt mit eonstanter Geschwindigkeit in die
Höhe, wähi-end die Geschwindiofkeit, womit die Luft durch die Schichte

: = Con.st i)assirt. in's Unendliche wie r abnimmt.

' a+t

Beiträge zur Theorie der Bewegimg der Erdaluiosphäi-e und der \Virl»elstüruie. 3, S 7

V.^ ist iiiiii in {li(\><07ii l-';ill,

(vi — 1 > \ (I /

wo bei die l'iinction ./ <iiircli tlicilweisc rnrco'i'.-itinn in dir l-'onn
a'ebracht werden knnn

.,.(ü±i)-.".,.«../;^

dl

Diese Functionen Avachsen in's Unendliche mit waehsendem /,
Die Fnn('ti(^nen nber

Be-'-' , Cr-'''

verseil winden mit wachsendem f. Xidit so Hnmittelb;ir eiidcnchfend
ist es. ob die Function A(:~'^' mit wachsendem / \'ers('hwindct. I»ei
dieser Fra^'e handche es sicli nur dai'ani. oii

C. , ^ ^y /•' r^'tdt-\

o'eo-(Mi Feinheit convcruiii. da

lim ( e jr = — 1

-^ r^'fdt
ist. AVeil nun 'o {a+t) mit in's Unendlicbc wachsendem / o-eo-en

gyt

/„ lä^

1

«(« + 0

convera'irt. so sieht man. dass

388

D. Kitao.

Inn I )c [It + f) (' I -, — r-Tz I = I

ist, (l.'is.s (liilior niK'li

//9W e '■' J =-- 0

isf.

^Iiiii findet niis del' l^^)rnicl

or cij

1 = ^

Mr

<)// Ô.7

;iis (lie C()in])Oii('iiteii del' ( icscliwiiKliukcit i'iir das änsserc wirljclfrcie
(_Tel)iet, da

B-

ist

f[^"« (— :^)]

llici'aiis crlijilt man als die i-csidfifciidc ( u'seliwindi^'keit

ti/' V (r^ + f)- \ 7^/^/

Die also diiiT'li ciiic (Mn])oi'o-es('hleiidort(' Ivuf'tiuasse entstelionde
\\'ii"lK']1»('\\('L;iiiiL;' ist d(-iniin<'li cine ('Vclonalc. mid die Wiiidiicscli-
wiiidiLi'kcit nimmt ^\',^^m mil wachsender Zeit ins Unendliche al).

lin den I)eviati(in\vinkel vax linden, könnte man die ^'efundenen
(iescli\vindiL;keitsc()m])onenlen in die Mleicdiirng" (•)'^) sid)stitnti<)n.
Man uelanu't abec kürzer dazn din-cli die Uberleoiinii". dass die Is<i-
dvnamen. iin«l auch [sohareii aucli liier mir con<'enti'ische Kreise sein
k()nnen, so dass

X II + !l '' =^ (I) o coü i

Iîri1r:i!4'' /HT Tlu'ovic ili'v liowc^uni;- der Krdatuiospliiirc mid diT Wiil'i/lsl iinin'. ^JSÎ)

ist. IIi('i':iiis tiiidct iii:iii für (l;is iiussêi'e Gebiet

^/r/ / = <i. (a -\-f)e '"' (a+B- —\
[m Aiifhiiii" (1er 1 >e\veL;-iiiiL;- ist

fii(l i ~ (/, (t

/ I 1 \_ ^/r^

\ y» — 1 ni (vi — 1 ) / 7/(

,lli

faq i

m (vi — 1 )
'2/ -s/n fi

un vi= — . niid n = — ist

/'o /'o

Für t =^ y: , wo die WirbelheAveüniiü' Ncrsclnvuiiden ist, ist, <l;i

Ac~'' (a + f) u'e^-eii — . iind (ii-rt)t'~'''B nud (a -\-t)r -''' (' ofeoen H
^ ' • ' ;f ^ ^ ^ -^ ■ '

eoi) vei'ii'ii't. wieder

'"{I I =

(/. //. d(M- Deviîiîionswinkel (MTeidit einiiinl ein Mnxiimiin oder
Miiiiiiimii iiM Verl;iiit dei' Zeit, uni diiiiii demselben Wertlie wie ani
Aniliiig der J]ewe,uunu' alhnäiii;' zuzustreben. Es ]ässt sieb nun obne
Weiteres iil)erseben, dass diese A^eränderlicbkeit des Deviationswinkels
diireb den Wertb von j'o bestimmt wird. Wenn /'o «i'rösser ist als
-"•, so trägt der Wirbel einen mebr centripetalen C'barakter : / muss
bier zuerst mit der Zeit abnebmen. Ist aber ;'„ kleiner als ^f, so
näbrt sidi der AVirbei einer kreisfiirmigen Drebung ; / muss daber
zuerst mit der Zeit waebsen. um dann n.it versclnvindendem Wirbel
wieder zu dem antUnglieben AVertbe allmiilig abzunebmen.

Als Componenten der (reselnvindigkeit im inneren Gebiete findet
man

X

•Iji + f)

-^'■-"[-'-«-^ay""""]

390 ^- Kitao.

Hieraus folo-t als die Re.snltirende

Als Deviationswiiîkel findet man weitet'

Die \Vii'l)elbeweo'niig- der Lnft ist auch im inneren Gebiete eine
cvclonale und ihre Geschwindio^keit nimmt mit der Zeit in's Fnend-
liehe ah. Der Deviationswinkel ist znr Zeit ^ = o

(7r-j'„) L m \Ii/ J

aber für / = x wieder

tag i —

dn C (^n + i') i' ~ ■'' l'iii- / = x verschwindet.

Naeli dem A erlauf einer sehr uTossen Zeit stellt sidi in dem
inneren Gelnet derselbe lîeweu-ungsznstand her, wie in dem iinsseren
Gebiete; die Windgesch^^indigkeit vermindert sich dabei nnausgesetzt,
bis die ganze Wirbel be weii'uno- verschwunden ist.

Es ist leicht Ausdrücke für den Luftdruck im inneren und
äusseren Gebiete aufzustellen, was wir aber unterlassen wollen, da
wir dal)ei nichts mehr ketmen lernen, als die Thatsache, dass die
Druckdepression mit der verticalen Strcinmng verschwindet.

Wir wollen einen anderen Fall in's Ange fissen den Fall näm-
lich, wo die vertical aufstegende Sti-ömung allmalig entsteht, so dass

r Hfiträs'' '-"r Theorie der Bewegung' der Erdatmospliäre und der W'irlielstüruie. 5!)1

■ Kt — f> - Kt

wird. Weil in diesem Full für t^o, y = o if>l, >s.) hat man aiicii

JWi=o für t = o

Dieser Bedingung wird u'enüu^'t, wenn man in (4.')) L'^o .setzt, iiik
se h reibt

Hieraus findet man darcli Integration

d.h. indem wii" fur /' seinen Ausdruck einsetzen

V ^ "i ' V ^' JJo [ei^t + e-Ktf

Fur das äussere Gelnet liat man wieder

"/iT + ^^y/'.

U

Die (irenzbedinu'unu" eru'iebt
Mithin ioiot

oî)2

D. Kita

^W., a h'R' - Kt / A7 , - Kl \ I '' Kl (c '''' - '' -^^"')f/A

Udei' ki"u-/er, iiideiii man «etzt

-;'^/ A7 , -A7\ /' Kl JC'^' -r-AQrfi

— r — ^ ~- <s \ t

an

-^^yt)

Mail liudcl somit als die Compoiieiiteii dor (jesclnviiidig-kcit tiir das
iiiiu'ic Geriet

(M

(f[t)

al.s die lîesuiti rende

A o / / e A7 — e - Kl \-'

Für das äussere Gebiet

A" ii!- r/rA'-r -A'\ ^ ^T

A- E=^ r-^r.Kt-c-Kt\ "1

als die Kesidtirende

)ie 1' iiiietioii ^ lilssl sich aiieli so schreiben

Beiträge zur Theorie iler Btswegiiu;;- der Er<latiJi<^si>hän' iiud der Wirl.elsturuie. oDo

/ ^ e + ><^ dt _ .-|

Wenn ;f> A' ist, so \ers<.'li windet das erste (ilied in der Klainuier
mil waelisender Zeit und das zweite Glied nimmt daltei die un-
bestimmte Furm t). X an. L)a nun

und

n— K

e(H-K)t

ist. wie man nach der gewöhnlichen Metliode linden kann, so hat
man

1

9^(0t=. =x(7;3r - 0 ^ "^^^^

Nach dem N'erlaufe einer unendlich grossen Zeil, wo j- -^ A wird
werden die Componenten der Geschwindigkeit im inneren Gebiete

A a A

h' a A

12 2 (;f — h )

imd im äusseren Gebiete

KR' / . a

A R' / , ^^ \

Ai?- / '/ \

^m D. Kitao.

AI;so .stationär, nhcv cine ünderc stationäre W irhelhiJdunu', als in
dem Fall, wo die \erticale laiitströmnug etwa pliitzlidi entstand,
lind sich mit der Zeit nicht ändert.

Wenn ->' = A odei' ;f <; A wird, wächst — r — so wold im inneren,

als im äusseren (lehiet mit wachsender Zeit in's l nendliche, während

~- oei>en einen <irenzwerth coii\ eriiirt, der durchaus endJich ist.

Ulf ~ ~ '~

Da nun die (rrosse A eigentlich keinerlei Einschiankiing unterworfen
ist und dahei- eine physikalische iJedeutunu' audi diesen Fällen ^^ = A
oder A >- ;f zuk(jmmen muss, so ist entweder eine solche Wirhel-
hildunu', \\'\e die 1 behandelte, un möglich (jder A ist hei der W'irljel-
hildiinü' s()l(;her Art (hu'ch die \ oro'äuü'e in der höheren Scliidite der
Frdatmosphäi'e eingeschränkt, um welche wir uns hei der Bestimmung
der Grösse /' eigentli«'h gar nicht bekümmert hahen. Da- aber die
Annahme ;^>> A' zu keinerlei Absurdität führt, s(j halte ich es für
wahrscheinlich, dass die Annahme Non der Constans des /\ unzn-
lässig- ist, wenn A bei der W irbell)ildung, wie die in Hede stellende,
einen gewissen W'erth übersteigt. Es ist indessen nicht zu ver-
gessen, dass ein solcher Fall, wo die Geschwindigkeit der vertical
aufsteigenden Strömung mit dem \ ertliiss einer unendlich grossen
Zeit sich zu einem Maximum steigern soll, nichts mehr ist, als
eine mathematische Fiction dass. wenn die fiii" eine solche Wirbel-
Ijildung entwickelten Ausdrücke besser einer solchen in der A^atur
angepasst werden sollen, y eine Function sein muss, welche mit
wachsender Zeit zu einem Maxiiiuim steigt, um dann \vieder zu
verschwinden.

Als Deviationswinkel hndet man für die beiden Gebiete eine und
diesel Ije Functicm

tau t^-lA sin II r - " ^ ,. ^ ,./ / c'^' -^- ~j(t

■RoiträoT" zur Thoorii' «lov T^owo^'uns:; iLt Erflatnios])li;iri' nu'l >li'r Wirl.i'lstiirnhv ^i*')

Es ist für t ^ 0
Am

/■t (pKt p -Kt\

Jim. L [r^^t^c-Ktf ^ 0

f^p Kt — c - Kt)

ist, so fola't

tag l = 0

Alle TjiftflifilHicii liofinden sioli zur Zeit / = 0 in rontviptnlcr
HeAveii'unfi-. VWv f = ^, f;vlls ;f>> A ist, wiivl ta(j i wiodor iiiil)Ostiinint.
IMnii erl^Hlt nlcr ii;icli der o'tnvöhiiliolieii Methode

taq i = T^

Der Deviîitioriswiîikel wadist demiiaHi von 0 -ins mit war-hsender
Zeit fort und l'ort zu dein diirrli die o1)ii:-e (Tleielirmo- l)estiinint('ii
Werthe. Die AVindlxdm verwandelt sieli dalier aus einer Tlei-adcu
alltnäliii" in eine loo-nritliniisehe S]»ii"al(*. Avelclic die Oi-enze des inneren
Oelnetes contiiuiii-licli durelisetzt und in unendlieber Anzalil dei-
Wiiidun,o-en das Centrum des iuiieren Txebietes erreieht.

Wir wollen als das letzte Beispiel mit einem Fall der vertieal
lieral)steiii'eiiden Striuniinu' l)eseli;i,ftiu'ên, indem \\\v annelimen

1 a = —

'' ~ ' <i+t " To

In diesem Fall ist die lîeschleunigun«:- uielil ^ull ; denn e> ist

.. clr -2

' dt {a+ff
so dass demnach

A'=

aofî

D. Kitao.

(l. 11. die Differenz dvv vei-tienlen 'renipernluro'rndienteii .sich wie

; — —T-r, vernnderl". Ans r]p)- nipichnno- --ii =y,z er^'ie^t sicli ferner
(»7 + ;^)- fit

G

[rt + f)

wns eitie inJ^U'liclie l)e\veo'nni:' nnzeiut, d;! : inif w;ielis(Midev Zeit u'eo-on
0 ('(^iiverii-irt. Die rie.sclnvindifi-keit eines rinftheilcliens ist

C

d"

rit {a+tf

Sie iiinimt nlso mit wneliseiider Zeit ah und Nci'scliwindet für / -- y:,
\V() die ïjifttheilclien erst die Erdoherfliielie en-eiclien. Es ist nnn

Geht man mit diesem Ans(h-iich in (o(\/) ein. so liât m;in für (L-is
il mere (ïehiet

nnd

S (T,

'<\" ' 2 {'I + 1)
niid ans (•") 1 ) inv das äussere Gel)iet"

"^-1 1 /a + t

?// ~ 2o \a + t)L

m + 1 a ot m\\-\- m)

mv-'~"<^ -'))']

1)1(1

Hierans erluilt mari als Componenten der riesehwin(hgkeit im inneren
Gehiet. indem wir znr Ahkürzunü" setzen

ficitviiüo zur 'l'lu'onf' df-r B(-we<jnnL;' (U-r ErilMtirmspliaic im«! il'-r W irl.t'lstüriii«-. /)i)7

1 r^'''-\

Ç ( M - -r-. +

+ 7)1 (I n

'2(a + f) 2{a + f)

y

".(( -ut ,,.

nur] für d-is äussere Gehiet, indem wir sct/ei

7» (

E^?/ aB

•iri'^ + t)

iF(-^)[-^-^'K'-'-"'^-'^)]

Die Fiiiictioii c~''^(f{t) verscliwiiider mit wnclisendei' Zeit und
derii'leiclieii <■"''* f/j ^ t). \v;is für einen \^('rtli a niicli lialicn niau". l)i('
A\ induescliwindiu'keir nimmt dalier mit der Va-ü ins I uendliehe :dt
lind iiir siAw o-rosses / werden ilire {'oini'oncntcn im innei-en (xehiete

'2i«.+ f) •![,(, + t)n

'l{ii + f) 'l{(l + t)n

lind im äusseren Geliiete

B- ./■
" = ~r^. r

aB'-x

B-

■2 >r {a + f) '2 nu- {(( + t) -1 /'- {a-itt)

("-^)

oi)(S D. Kitao.

J);i nun so wohl im inneren, als im äusseren (lehiete für o-ros.ses t

V

dt

dx ~
dt

du
:dx

O.X

a

Je

so sieht man, class die AVindbalm nach dem Verlanf einer grossen
Zeit eine Joi;arithmische Spirale wird, die nach Aussen anticyclonal
Liewimden ist und den Gren/.kreis der \'erticalen StiT)inunf>' conti nuir-
licli <hii'('liset/.t.

Fin- Deviationswinkel ündet man im inneren (rehiete, indem wir
it = — einführen

/'o

,1. h.

imd in dem äusseren Gebiete

'^ L n^Y. (/'o + ^O \ ^'' VJ

In ilcni inneren Gebiete wird demnach t'iii- / = y:

taq i =

lüid ancli indem äussei-em Gebiet, al)er tn.it dem Unterschied, (iass
dieser lîeweg-ungszustand li-iihei- in dem inneren Gebiete eintritt,
als in dem äusseren (iebiete, da c - ''^ schneller vei-sch windet, als

\ iel complieirter werden die Ausdrücke fiii- die (xesi-hwindigkeits-
comnonenten, wenn wir iil)er den anfänglichen Znstand der Beweu-unp-

Beiträge zur Theorie der Be\vei'niti<>- der Erdatuiosphäre und der \VirV>elstürme. 31)9

in dem kreisibrinigen Grebiete, wo die verticale. Ströniimg entsteht,
eine andere Annahme machen als diejenigen, tiii- die wir die hier
auftretenden willkürlichen Functionen erniiftelt hal)en. Dass wir
über den Anfanszustand andere Annahmen machen künnen, erheUt
daraus, dass für t = 0 /'o=0 JTFj nicht notliwendig zu verschwinden
brauclit. Denn in diesem F;dl fallen die Gleichunges (i^7) und (2<S)
in die eine zusnnimen

-hAW lAW IW ?JIF ^W ,„. ,, .... , ^

H ^ -T — r -r h n â\\ == 0 tur t =^0

0/ "hx ^ij "hij

was anzeigt, dass ^TF eine Function von x ij sein kann und zwar eine
Avillkiu'liche. da sie keiner anderen Bedino:uno- zu o'enüo-en hat, als
der, dass sie in der Unendlichkeit verscliAvindet. Es ist aber ein-
leuchtend, dass, was wir auch für eine Annalune über den anfangliclien
Bewegnngszustand machen mögen, die Aufgabe auf die (Quadratur
zurückgeführt ist, wenn das (Gebiet der verticalen Strömung kreis-
ibrmiir oder, was wir liier hinzufügen wollen, ueradlinio: beo^renzt ist.
AVenn in der Erdatmosphäre mehr als ein Gebiet der veränder-
lichen verticalen Strömungen vorhanden sind, so bewegen sich die also
gebildeten Gebiete der nicht stationären Wirbel und die Differential-
gleichungen für diese Bewegung können auf diesellje Weise, wie
wir sie für stationäre Wirbelbewegung aufgesteht haben, entwickelt
werden ; sie aber sind selbst im Fall, wo nur zwei unendlich von
einajider entfernte Wirl)elgebiete vorhanden sind, nicht integrirbar
da die Functionen (f und TF die Zeit ex}»licite entlialten. Indessen ;
in dem Falle, wo nur zwei um Fnendliches von einander entfernte
AVirbelgel)iete vorhanden sind, und eine von denselben stationäre ist,
<3der die diu'ch dieses Wirbelgel)iet allein in der Fnendlichkeir verur-
sachte Luftgesi.'hwindigkeit gegeben ist, lässt sich die Luftbewegung
in- und ausserlialb des anderen Wirbelgebietes finden, da wie wir

400 l>. Ivitao.

iresehen h:iben, der Eiiiüiiss des iiuendlicli fiTneii Wirbelaebietes ;iiif
das in Rede stehende Wirl)elaebiet jederzeit durch die Hinzufiio-uno-
einer gewissen h'nearen Function zu ^TT' I)eriicksichtiot wird.

Wir h:i])en somit durch conse([iiente lUduindhuig der unter
specialisirenden Annahmen vereinfachten h_v(h'odvnamischen Dif-
ferentialgleichungen die Avesentlichsten Eiuciiscluiften der C'vclonen
und Anticyclonen nhgeleitet und die Bewegung eines Wirbelgel)ietes
als eine Folge des A <»rli;indenseiiis eines andei'en Wirl)elgebiel:es
erkannt. Wir haben ti'rner den AVeg gezeigt, (he zeitli<-he A erän-
derung der Windstärke des W'indazinuitlis luid des Luftdi'ucks in
einem gegebenen OiT l)ei zweifielier AVirbell)ildiuig aus der F]igen-
bewegung der Wirbelgebiete ai)zuleiten und sind dabei zu manchen
Thatsuclien geführt worden, die in der Wii'kliclikeit beol)aehtet worden
sind. F'iir die ^ on manchen Meteorologen mit Autwand \on vielem
Schai-ftinn erklärte Ausdehnung der Wirbelst iirme in den liiiliei-en
Jjreiten haben wir Avenigstens den Weg gezeigt, sie als eine Folge der
A erlangsaimmg der vei'ticalen tStronumg und ih]'er \\ aiidc^-unu- nach
der höheren Jjreite abzuleiten. F)iese Hesult:ite unserei' bisherigen
Entwickelungen (hirften demnach wohl als Heweise gelten, class die
vereinf lebenden Annahmen, von denen \\\r ausgegangen waren, nicht
allzu unser Strel)en beeinträchtigt hahen. Fjnblick in den verwickelten
Mechanismus der Luftwirbel in dei- F^rdatniosphäre auf rein analyti-
schem Wege wenigstens für die untere J'artie dei- F]rdatmos])häre zu
gewinnen.

Indessen muss ich hier auf zwei Ungenauigkeiten ;uihnerksani
machen, wehdie aus den eben erwähnten (irundnnnahmen fhessen und
die Anwendbarkeit unserer bisher entwickelten F'ormeln eigentlich auf
die unterste ^Schichte der Erdoberfläche beschränken möu-en. Die

Px'iträifo zur ThiJorie «Ln- Kewfu'iiny; <1pi' P^nlMtmosphäro nn<\ ilcr Wirlielstiirini'. À() 1

Aiiiiiiliiiu', (lüss (li(' Liif'r sich il! dein Ik.-iiiui jiusscrlinU) (]('< <j\li?i(l-
visolieii lJ;niinii"el)ietes dei" verticaleii Stfittimiiu' iil)ei';ili piii'uHcl /.n
(1er ;ils eine iiiiciKlIiclie "Ebene oednrlircn l'ii-dolierflache be\ve£i'e, wäre
nui' dann znliissii;-. wenn die lîeibiinu' an <lei- Erdöl lerHäclie in jeder
Lnftschicbte dieselbe ^ erzöiivi-nnu- der beweu'ten F^iifr bervorbi'iiehte,
was wohl niclit der l'^dl isi. Im dabei- Ausdrücke /ii crlialtcn. die
sieb den Aoi-oäniien in dei* liTiber uele^'enen Enftscbiclite bei \\'ii-bei-
bildnrjo- l)esser anpassen lassen, unisste man die Reibunii- erst in der
.'Ulf der Erd(^l)erfläcbe zuerfiillenden l>edinuiinii' beriicksicbtiu'eii. indem
wii' die innei'e Heibunu der Enft in Hef^linunu' zieben. Wie weit aber
die in Rede stebende Annahme zulässig- ist. das lässt sich leider nielit
benrtbeilen. Als eine Folge der Annahme, dass die Wirbelaxe überall
senkrecbt zm* als eine nnendliehe El)ene gedacliten Erdoberfläche stehe,
haben wdr die (ileicbnno- erbalten

j^^, einer Constante

'- oder einer Fnnetions von / allein

Wie wir geseben haben, entsteht ans diesem (besetz der verticalen
StWunnno- keine analytische Sehwierio-keit. welche die T nznlässisfkeit
der Annahme, deren Folge das Gesetz ist. irgend wie darthnn würde.
Nichtsdestowenigei- i'esnitirt eine l'ewegnngsform. die wenig wabr-
sclieinlich ist. Jedes Lnfttbei leben, weiches sieb ausserhalb des
(rebietes der vertical aufsteigenden Strinnung horizontal bewegt hat.
wirbelt z. B. in einer Spiralbabn hinan!', so 1)ald es den Plante! des
cvlindriscben Ranmgebietes der verticalen Strinnung durchsetzt. Da
abei- die Geschwindigkeit der verticalen Strömung niebt etwa, wie es
in der Xatur sicher der Fall ist. naeb dem umfang des (\dindi-i sehen.
Gebietes zn alJmälig abnimmt, sondern nur von der verticalen Coor-
dinate abhängt, so wird jedes Lnfttheilchen beim Durchsetzen der
Gi-enzfläcbe der verticalen Strömung xiUlig unNcrmitlelt emporgeris-
sen und die Striununs'slinie erscheint dorr »laeh dei- II()lie g(d\niekt.

402

D. Kitüo.

während ihre horizontale Projection vcillio- continnirlioh :ius dem
äusseren Raum in den inneren hinein verläuft, und der Luftdruck
ändert sich dahei s])rünf>"s weise um den Betrag ' ^ .

So sehen wir denn, dass die Annahme, auf der jene;^ Gesetz der
verticalen Ströinuni»' herulit, sti-eng çrenommen niclit zulässiu' ist, dass
wir /' auch als eine Function von .r, ?/, zu denken haben, um Aus-
drücke zu erhalten, weldie genauer die Vorgänge in einiger Höhe iiher
dem Meeresniveau d;u-stellen würden. Die IVwegungsgleichungen
werden aber dann so verwickelt, dass sie sieb Ix'i dem jetzigen Zu-
stande der Analysis schwerlich integriren lassen würden, auch wenn
wir wieder die verticale Ströniiing kreisförmig oder geradlinig begrenzt
denken. Es ist indessen dagegen zu erinneren, dass /' im Gewöhn-
lichen eine sehr kleine ( «rosse ist, dass, wenn /'"-' für /' = 0,0001 5S also
vielleicht grösser als der lveiI)Uugscoeffirient auf der Ei'doberfläche
selbst in der Meeresh('>lie lOOO;//. erst den kleinen Werth 0,158

sec
erreicht, dass die Discontinuität des Luftdiaicks au der (rrenzfläche

d(M' verticale!^ Stnnnung in dieser ziendich grossen Höhe erst
0.000158 mal dem iiiisscrcn (,)uecksill)erdruck in dieser Höhe betragen
wüi-de. Wenn dabei- do- 1 )urr'hmesser der vei'ticalen Strömung nicht
unbedcntend und in Folge dessen die liorizontide Luftgeschwindig-
keit an der Grenze dcM-seliKMi liedeutend ist, ^o wii'd die Neigung der
Striunungslinic an dei- (ii-enze gegen die Hoi'izontalebene sellisr in
dieser Meereshöhe so klein sein können, dass ^vir wohl berechtigt
sind, die Annahme bis dieser ll()he als zulässig anzusehen, welche zu
jenem Gesetz de»' \erli<';deii Sti-(»muiig gefühi't bat.

MBL WHOI LIBRARY

UH njz R

:^^