f. d r^ f« .?«£ ^ m ^ » THE JOURNAL OF THE COLLEGE OE SCIENCE IMPERIAL UNIVERSITY, VOL. IL WITH 27 PLATES. '^ n ic ^ m ^f m \f. -«- - ^T- PUBLISHED BY THE ÜNIVEESITY. TOKYO, JAPAN. 1889. Publishing Committee. Prof. D. KikuChi, Rigakuhakushi, W. A., Director of tlie College (iw oßcio). Prof. K. Milsukuri, Rigakuhakushi. Ph. D. Prof. C. G. Knott, D. Sc, F. R. S. E. Prof. S. Sekiya. 2q :i% nUXTEI) AT XHE SiaSHIBUNSHA, TOKYO. CONTENTS. Vol. II. Page Ueber die Darstellbarkeit willkürlicher Functionen durch Reihen die nach den Wurzeln einer transcendenten Gleichung fortschreiten. von Dr. Pli. E. Fujisawa, llitjakushi 1 On the Coniioosition of Bird-lime, l^y Edward Divers, M.D., F.E.S., Professor, and Michitada Kawakita, I^LE., F.C.S., Assistant Professor ot Chemistry, Imperial üniversily 17 On Anorthite from Miyakejima. Dy Yasushi Kikuchi, rdnakmld, Assist- ant Professor of Geology, Imperial University. (Fleuch) 31 The Source of Bothriocephalus latus in Japan. By Isao Ijima, Uhjahnshi, PL. D., Professor of Z'oology, Imperial University 49 Earthquake Measurements of Recent Years especially relating to Vertical Motion. D.y Skikei Skkiva, Professor of Seismology, Imperial University ^' On the so-called Crystalline Schists of Chichibu (The Sambagawan Series). By Buxdjieo Koto, i:i;iaktiJiakiisJn, Ph.D., Professor of Geology, Imperial University. (Plaies II- V.) ^ 77 On the Plants of Sulphur Island. By Samurö Okubo, Assistant Profes- sor of Botany, Imperial University 14B Some New Cases of the Occurrence of Bothriocephalus liguloides Lt By Isao Ijima, l!i;/al.iisl,i, Pli. D. anil Ki:ntai:o Muiîata, LjaLiishi. (Plate V his.) ^'^^ A Magnetic Survey of all Japan, carried out, by Order of the President of the Imperial University, by Cakgill G. Knott, D. Sc. (Edm.), F. Ft. S.E., Professor, and Aikitu Taxakadate, EiaaknsJii, Assistant Professor of Physics, Imperial University, Japan. ( Plates VI - XV.) ^163 Determination of the Thermal Conductivity of Marble. B-.v Ki^njiko Yw^G^v.-A, lUnahiihahushi, Pli. P.., Professor of Physics, Imperial Uni- ■ ., 2G3 vcrsitv Combined Effects of Torsion and Longitudinal Stress on the Magnet- ization of Nickel. By H. Kagaoka, Hi^ak„.hi, of the Imperial I mversity. (Plates XV I-XIX.) ^^^ On the Magnetization and Retentiveness of Nickel Wire under com- bined Torsional and Longitudinal Stresses J "-y n. Nagaoka, iu lieber die Darstellbarkeit willkürlicher Functionen durch Reihen die nach den Wurzeln einer transcendenten Gleichung fortschreiten. von Dr. Phil. R. Fujisawa. Neiden den trio-onometri.sclien Heiben werden, in der mathema- tischen Physik, die ihnen naheverwandten, naeli den AYurzehi einer o-ewissen transcendenten nieichnng f(^rtsch reitenden Reihen vielfach ano-ewandt : daher wird es wiinschenswerth sein, wenigstens für alle Fälle der Natnr, d. h., unter beschränkenden, jedoch für die phy- sikalische Anwendung' hinreichend allgemeinen Voraussetzungen ül)er die Natur der Function, zu l)eweisen, dass die Reibe wirklich gegen den Werth der gegebenen Function convergire. Dass der \'on Sturm und Liouville herrührende, von Heine vervollständigte Beweis un- zureichend ist, hahe ich schon in einer früheren ar1)eit discutirt.* Es handelt sich hier darum, die Convergenz jener in IJede stehenden Reihe darzuthun, wie fin- die trigonometrisfheu Reihen durch die berühmte Arbeit Dirichlet's geschehen ist. Die Reihe in ihrer allgemeinsten Form ist wie folgt beschaffen ; sie schreitet nach den gegebenen Functionen 6 {x, X), welche einen * Ueber eine in der Wärmeleituusstheoiie anftretende, nach den Wurzeln einer transeeu- deuten Gleichung fortschreitende unendliche Eeihe. Inaugural-Dissertation. Strassburg 188ß, Z R. FUJI S AW A Parameter X enthalt, für den ninn alle Wurzeln einer gegelDcnen trans- cendenten Gleichung d (A.) = 0 zu setzen hat. In dieser Form wird aber der in lîede stehende lîeweis wohl schwerlidi durch zuführen sein, ohne dnss man sehr beschränkende ^Voraussetzungen über 9 und 0 zu machen genöthigt ist ; es empfiehlt sich daher, von vornherein den ]3eweis an einem bestimmten Beispiele durchzuführen und dadurch den Weg zu zeigen, wie man auch in andern Fällen zu verfahren hat. Ein so directer AVeg, wie der Dirichlet's ist hier der Natur der Sache nach wohl niclit möglich ; es lässt sich aber dieser Fall auf einen durch das Theorem Diriclilet's erledigten Fall zurüchführen Avie icli in meiner eben citirten Arbeit flu* eine Reihe aus der AA^ärme- ] ei t IUI £i's théorie claroeLhan habe. Es war die Keihe : -.1 n 1 wo X^,?^^,---- die, der Grösse nach geordneten positiven Wurzeln der Gleichung 0 (?J z= cos ;^, + f a - 1 ) sin X -- n f a >• 0 ) bedeuten. Für die Wärmeleitungsnutgnbo genügt es nachzuweisen, dass u mit positis ;ibnehmeiid<-m t gegen /'(r) convergirt. Stellt die für t = 0 formirte Reihe 1 ^. . /, ,NX'''''"^'"(^"f)'''' /(/•) dar, so ist dies n:ich (ùncm bekannten Sntze über iV)tenzreihen hierfür ausreichend. al)er nicht um hwendiü-. Dieser, wie ich glaube, ÜBEK DIE DAKSTELLBAKKEIÏ WILLKÜRLICHER FUNCTIONEN. 3 bisher unbenclitete Umstand .spielt eine wesentliche Kolle bei meinem Beweise ; die Methode des Beweises selbst lässt sieh auch auf die für t = 0 formirte Reihe v anwenden, wie ich ^^einer Zeit Ijemerkt hal)e. Dies zu zeigen ist der Zweck der vin'liei'eiiden Arbeit. S'. 1. ^ A\'ir bes'-hiifti J;. FÜJISAWA und wenn wir die Reihe v für den speciellen werth von a -- 1 mit v' bezeichnen, 80 lautet die.sellje : Von dieser Reihe ist bekannt, dass sie sich aus der Entwickehing von r/(r) cos ^ nach den Sinus der ganzen A'ielfachen von — ergibt, also dass, mit Ausschluss der Grenzen r =- 0 und /• = /, zwis- chen denselben /(?•) zur Summe hat. Es handelt sich darum, zu zeigen dass die beiden Summeu v und v' ireo-en die nämliche Grenze convergiren. §. 2. Wofern 0 ( In) = K L'o« K + (oc-i) hIu ;i„ = ü ist, haben wir sin (;i„ Çj ■ sin (x,, ^^ ) „ • sm (?.,, J ) • «in (^K f ) T sin A.^ cos X„ sin X„ • 0'(^„) also auch sin(;i„ f) • sin(;,, -;-) (a-1) sin(\. -^). sin^A,. ^-) T^ ^iu_A^^^_cos_^ cos \, ■ — o unstetig wird. In den beiden Punkten ;; ^ ± \^ hat sie das Resi- duum Ji 1 sm A.,t cos /l„ ^1, und fiu' :. = + -^^-7; tt das folo-ejide ,,/ 2«-l \ . /■2«-l T/\ . / 2)1-1 ir/)\ Dies vüraugescliiekt, sei nun in der 2 -ebene eine Flache E* vorgelegt, welche Avie folgt entsteht : in den Punkten mir und —mir, unter tn eine positive ganze Zahl verstanden, errichte man auf der ;î;-axe Perpendikel AA' und BB', wo m TT + i ■s/mir. B — — mir + i -y/ifiTC A' ^= mir — i ^WK, B'— — mir — i ^/mir; * Diese Placke E kommt bei Gelegenheit einer ähnlichen Untersuchung bei Heine Crelle's Journal, Bd. 89. vor. 6 R. I'UJISAWA vom Anfang.spunkt 0 aus .schlage man mit dem Hadius \/(/«7r)'^ + w7r Kreisbogen ACE und A'C'B'. Diese dm-eli die in sich zurück- kehrende Linie Ä'At'BB'C'A' eini>eschh3ssene Fliiclie moue E heissen. Alsdann hahen wir nach dem Kesiduensatze von Cauchv wo die Inteiiration über Ä A C B B'C'A zu erstrecken ist. Da die Funktion w eine ungerade Function von z = x ■{■ iij ist, so ist (»tfenljar ('\ ^, w ch = J j^ IV d^^ ^ )[^-f^']^^/' und lassen m, die Reilie der o-anzen Zahlen dnrchlanfend, über alle Grenze hinanswachsen, so haben wir, die Convergenz beiderseits vorausgesetzt, Nun lässt sich aber zeio;en, dass die rechte Seite (heser (ileichnno- in der That verschwindt^. §. 3. Wir haben g-esetzt rt ^=^nrfS'pfip)[Q+J'']dp; mithin :\[oa j ^ -J — f''p Mo^fip) [Mod Q + ^fod 7^] dp. I . l. TT' y Znnaclist bescliiiftigen wir nns mit Q und P. Es ist Q= j I j/ "■ dr.-i j' ir {mir + f?l) dif, R. FÜJI8AWA worin w {z) (a-l)sm(.i:).sin(. -f) cos z .

%^+ sinrr^, Sy ^ i\ro(l sin ( .r + ///) ^ C;;y, nnd, inshesondere, Mod sin -T- ( .r + /?/ ) Mod sin ^^' ( X + /// ) ^ rY ^1 y ^ , alnio :\rod sin Y i ■'*' + ^'/ ) • •'^in y- (a- + i.i/) ^(T")-'(f-0- Entwickelt man «je (.r + /i/), ^ ÜBER DIE DAliSTELLBARKEIT WILLKÜRLICHER FUNCTIONEN. 9 (p {x + iy) X cos X Cy ■}■ y sin X Sy + (a — 1) sin X Cy + i y cos X Cy — X sin X Sy + (a — 1) cos X Sy und liildet alsdann den Modul, so erhält man durch einfache Tv eduction Mod 9 {x + iy) — [x cos X + ( a— 1 ) sin x^ + [yCy + {^-\) Syf . O CT O • O 4- ar Sy- smar Plierin setze man x = inir, sin mir = 0, cos wtt = ± 1 ein, so erhält man Mod 0 {imr + iy) = ^ W7r'+ {y Cy + (a — 1) Syy+ mir' Sif ^^vrir'Cf + (yCy + {,,^1) SyJ , also Mod 0 {in TT + iy ) ^ mir Cy. Es folo-t hieraus Mod 7V {mir + iy) ^ Mod(a — 1) Es ist a her, M'ofern < > <:r <: l, 0 ^ p'^l, für jeden Werth von y ^■(fO-^(f") mir cy ^(tO=^'^' ^'(f'0=^'^' folcflich <^(tO {x'-\- f)i. Sy. -^ , wo Q eine stets von Null verscheidene für erhebliche Werth von y nahezu gleich der Einheit und mit wachs- endem y schliesslich gegen die Einheit con vergirende Zahl bedeutet. Unter Ecslhaltung dieser Ikdcutung von Q findet man weiter für die WerLhe von y, die von Null verschieden sind, Mod {z.io) < Mod (a-1) • Q ^-^ -g^ — Auf dem Kreisbogen ACE ist nun y positiv. Mit Rück- sicht hierauf bringen wir Û sf in die Form : Für erhebliche positive AVerthc von y ist der Faktor nur sehr wenisf von der Einheit verschieden und conver^^irt mit Avachsendem y sehr rasch gegen dieselbe ; dasselbe gilt auch von ß, also auch von ihrem Producte, welches wir der Kürze halber mit (o bezeichnen wollen, so dass 12 îi. FÜJISAWA W l"i ■ / 1 - -2,i/\2 (I_e-'2V)2 Der Faktor ,-2(2--'^). hut, wofern ?/ > 0 j 0 <: r <: / und 0 ^ p <: l , nh Funktion von fj betrachtet, .schien grö.ssten Werth fih* p — l ^ und dieser Werth ist e-sci--- )î, Wir haben demnach Mod {%. lü) ^ Mod ( a -1 ) iv. c'-^^-T-»^ , gültig für jeden der Ungleichheit 0 W e r t h von p . Auf dem KreisboQ:en A C B ist p <: l genügenden y = \^ ni^ir'^ + ?//7r aiuO ; mithin Mod P ^ Mod ( .-!)/ B C A CO e 2(1- V'w'^jr^ + wTT sinö (iö Die beiden Faktoren unter dem Integralzeichen sind positiv ; es ist desshalb nach dem Mittelsatze von Cauchy C e i dO , wo «„einen gewissen Mittel werth von o) auf dem Kreis- bogen A C B bedeutet, welcher für erhebliche Werthe von m nahezu gleich Eins wird, und mit wachsendem m gegen die Einheit convergirt. ÜI3EK DIE DAKSTELLBAKKEIT WILLKÜJtiLICHEK FUNCTIONEN. 13 Es ist nun n'^ TT und -J A , -C sin e e -^ ='" " dû <: je ^ «'" " dO . Ea foliT^t hieraus TT Mod P ^ Mod (a-1) • 0),' ,yj-. ,■ . -7— ^-Tr— - gültig für jeden der Ungleichheit ^ ^ p "^l genügenden Werth von p. Somit erhalten wir für jeden der Ungleichheit 0 :^ /> < Z genügenden Werth von p die Ungleichheit Mod Q + Mod V ^ Mod (a-1) A^miç TT mit den Zusätze, dass Oo eine für erhebliche Wer the von m nahezu der Einheit ü'leiche und mit wachsendem t\i gegen E ins c o n v e r g i r e n d e Zahl bedeutet. §. 4. Wir verstärken die Uni^leichheit Mod à < — ?— fp Mod/(/?) [Mod Q + Mod V^dp, v.l. TT '^Q dadurch dass wir [Mod Q + Mod P] durch die in dem vorangehenden 14 U. FUJISAWA Paragra])li ijachgewiesene Zahl ersetzen, welche die Eigenschaft hat, für jeden 0 ^ /? < Z genügenden Werth von p stets grösser als [ Mod Q + Mod P ] zu sein : Mod J ^^l — • Mod (a-I) T—/^ - — r.l.ir L^/ in-rr + <^o -971 rsr/^ 2^77-^^^" 1 Sp Mod /(/.) dp. Den grössten Werth von Mod/(/j) zwischen den Grenzen 0 und /, welcher der Voraussetzung nach endlich ist, hezeichnen wir mit A ; so ist Sp Mod/(/^) dp ^ fp A dp 0 0 ^ AI' wo die Gleichheit stattfindet, wenn f(p) constant und gleich A ist. Mithin haben wir endlich Mod J < — . Mod (a — 1 ) • ^ ^ 1 . - - + «0 "öTl r'\ — /% , 1 Lässt man nun hierin m, die Reihe der ganzen Zahlen durch- laufend, über alle Grenze hinauswachsen, wofern 0 -< ?• < / ist ; es foh>'t, dass Mod à, also auch J selbst verschwindet zwischen /• = 0 und r — l mit Ausschluss der Grenzen, wenn die Gliederzahl m ins Unendliche wachst. Es ist somit bewiesen (Vergl §. 2. zu Ende) dass die Differenz der beiden Summen v - v' gleich Null ist ; d. h. dass, da v' für sich allein conver2i;irt und mit Ausschluss der Grenzen r=0 und r — l zwischen denselben /(r) zur Summe hat, dasselbe also auch von v ÜBER DIE DARSTELLBARKETT WILLKÜRLICHER FUXCTIO^^EX. 15 Sfilt. Wir o-elano^cn also zu dem Satzo Die unendliche Reihe hat /(r) zur Summe für 0 < r <: / mit Ausschluss der Grenzen r = 0 und r = l. On the Composition of Bird-lime. by Edward Divers, M.D., F.R.S. Prof., and Michitada Kawakita, M.E., F.C.S. Assl. Prof., Imperial University. Bird-lime seems never to have been examined to the extent to yield results deemed worthy of publication, until the year 1884, when J. l*ersonne made known those of his father's and his own examination of it in the Comptes rendus, 98, 1585. When that paper appeared we ourselves had been for some time occupied with the investigation of Japanese bird-lime, and had already obtained results, which proved to be in o;eneral ao;reement with those of Personne's examination, and yet sufficiently unlike them, and in some respects in advance of them, to lead us to continue our work, although he promised farther atten- tion to the subject. I^p to the present date, however, nothing more from him has appeared, and we now otfer this paper as an extension and partial confirmatioîi of liis observations. Bird-lime, or Tori-mocJii, is prepared in Japan, just as it is in Northern Europe, from a species of holly, by macerating and pound- ing its inner bark in water, afterwards picking out the fragments of crushed tissues from the viscid mass. Bird-lime exists, ready-formed in the bark, in great abundance, and is not apparently modified in any way by fermentative action during its preparation. In Europe 18 E. DIVERS AND M. KAWAKITA it is prepared from the Common or Prickly -lecaved Holly, (Ilex Aqiii- foliani), bat in Japan it is obtained from Mochi no hi, the 7. intégra of Thunberg, (Prinns intégra, H. & A.). We are not familiar with bird- lime as prepared in Europe, but judging from descriptions, Japanese bird-lime is like it, except perhaps in not having a greenish hue, though of that even we are not certain, since the Japanese product may well have it sometimes, when quite freshly prepared. Bird-lime is extensively used in Japan, as in Europe, for catching birds and insects, and with the usually attendant cruelty. In manuals of economic botany we find enumerated as peculiar constituents of the holly, a bitter principle named ilicine, an aromatic resin, and bird-lime itself. In the account of bird-lime given in Ure's Dictionarg, the true substance is not well distinguished from the viscid matter of Mistletoe, (Viscum album), examined by Reinsch, from which it appears to be entirely different. Some properties of (Japanese) bird-lime. Bird-lime is pale greyish, nearly opacpie, of faint, peculiar odour, almost tasteless, soft, elastic, tenacious and very adhesive to dry sur- faces, and slightly lighter than water. It can be preserved in water f)r an indefinite time without change, except on its upper surface. Exposed to air it very slowly turns brown outside, and becomes coated with a thin brittle skin. Heated moderately, it gives off water, and above 100° froths, through disengagement of steam. By the loss of its moisture it becomes transparent, brown, and wdiile hot, of the consistency of cold oil. If now allowed to cool, it retains its transparency, and forms a soft solid mass, elastic, tenacious, and sticky, as before, somewhat resembling Canada balsam in appearance. Ether, carbon l^isulphide, chloroform, light petroleum, and ben- zene dissolve bird-lime, leaving a residue which, though of not in- ON THE COMPOSITION OF BIED-LIME. 19 considerable volimie, is of little weight. Cold alcohol scarcely dis- solves it at all, and even hot alcohol, which has some solvent action at first, attacks merely the surface-portion of the mass. The alcohol solution deposits, as it cools, a nearly colourless, transparent, adhesive matter, little ditferent from purified bird-lime itself. Ether is much to be preferred to other solvents, because it yields a clear solution, whereas carbon bisulphide and the rest give milky liquids, owing to the presence of water. The ether solution mixed with alcohol be- comes turbid and deposits a tenacious mass. Freed from water and particles of woody fibre, bird-lime under- goes, when heated, scarcely any change up to about 350°, only becom- ing slightly fluorescent and a little darker in colour, and acquiring a feeble waxy odour. But about the melting point of zinc, it suffers destructive distillation, in which most of it comes over as fatty acids and fluorescent hydrocarbons of waxy and mild empyreumatic odour and buttery consistence, very little permanent gas being formed, and only a small carbonaceous residue being left. Bird-lime burns in the air with a bright smoky flame. It is not very sensitive to reagents. Sulphuric acid dissolves it slowly, forming a red liquid, which blackens only when heated, and which gives, when poured into water, a viscid precipitate like bird- lime, but dark coloured. Boilino- nitric acid of moderate strenîî^th slowly dissolves it, with partial oxidation, this solution also precipitat- ing with water. The sulphuric-acid solution poured into concentrated nitric acid yields on addition of water a precipitate of a feebly nitrated mixture of bodies. Aqueous solutions of potassium hydroxide only slowly and slightly emulsify bird-lime. Fusion with the hydroxide is attended with much darkening in colour and leaves a mass which emulsifies in water. Potassium hydroxide in hot spirit slowly dissolves the greater part of purified bird-lime, producing a dark-. 20 E. DIVERS AND M. KAWAKITA coloured solution. In this way, that is, by continued boiling with strong alcoholic potash, bird-lime has been attacked by both Personne and ourselves, in order to determine its composition. The Constituents of bird-lime. Personne has found bird-lime, prepared from I. Aqnifolinm, to contain water 27 , and vegetable debris and calcareous salts 23 parts per cent., the remaining and essential part being some caoutchouc^ the Compound ether, or ethers, of a new alcohol., and other matters undeter- mined. The acids or acid forming the ethers were also not deter- mined by him. He isolated the caoutchouc by saponifying the ethers with alcoholic potash, which left it undissolved. Japanese bird-lime is much cleaner than that described by Personne, containing only 2 per cent, of dry bark fragments, and no separate lime salts. But its water-content is larger, (probably because it is kept in stock underwater), the percentage lost at 110°- 120° being 38. Caoutchouc forms about 6 per cent., leaving 54 per cent, as the proportion of compound ethers and allied matters. The bark, etc. — Of the 23 parts per cent, found in French bird- lime by Personne, some 13 parts consisted of calcium oxalate. On boiling out the bark fragments from Japanese bird-lime with sodium carbonate, some oxalate was dissolved out but only in small quantity. The bark burnt to ashes gave as much as 6.3 per cent, of ash, princi- pally calcareous and largely phosphate, but with of course some carbonate. But as the whole ash was only one-eighth per cent, of the entire bird-lime, and as only a little of the calcium salts was oxalate, Japanese differs, in this respect, remarkably from French bird-lime. llie caoutchouc. — As we have stated, the caoutchouc can be sepa- rated by boiling out the purified bird-lime with alcoholic potash, and ON THE COMPOSITION OF BIED-LIME. 21 this is the best way of proceeding. It is, however, difficult to get quite free from potash, and needs, to this end, to be repeatedly dis- solved in ether and reprecipitated by alcohol. The caoutchouc can also be separated by dissolving the bird-lime in ether and precipitat- ing the solution with 95 ^/o spirit, but then only very imperfectly, because the main constituent of the bird-lime also precipitates partly. The caoutchouc of bird-lime is pale yellow and transparent, highly elastic, and when heated evolves the well-known penetrating odour. A combustion analysis of it gave us carbon 86.56, and hydrogen 11.31 per cent., so that oxygen to the extent of 2 per cent, was present. Before weighing it out, it had been kept for some time at 120''- 130°. It left when burnt a trace of ash. Other and principal constituents ofhird-lime. — AVe have not fully isolated these by proximate analytical methods, l)ut their general properties appear to be those of the partially purified bird-lime. For when a boiUng spirit-solution of bird-lime is evaporated and cooled, and aii'ain when an ether-solution of bird-lime is mixed with a little alcohol, to separate the caoutchouc, and then evaporated, in both cases the solid matter obtained is quite like the partially purified bird-lime, except in being without colour when deposited from the cooling spirit solution. Products of the saponification of hird-Uine, and their isolation. — Saponification with alcoholic potash yields, besides the residual caout- chouc, firstly, the potassium salt of palmitic acid and a very little of that of a semi-solid acid which we have been unable to purify or identify ; secondly, tiro cnistalline alcohols ; and thirdly, a small quantity of a resinoïd body. The separation of these bodies may be carried out in somewhat different ways, and is unavoidably tedious. The purified bird-lime is boiled for two hours with potash and 95 ^o spirit, in a fiask fitted with a condenser ; the alkaline solution. 22 E. DIVERS A^'D M. KAWAKITA decanted from the cuoutchouc, is poured into dilute spirit, by wliich a voluminous, gelatinous precipitate is produced, consisting of the alcohols with some of the resinoïd body and potassium palmitate. The precipitate is well broken up by stirring, collected on a cloth filter, pressed, and washed with dilute spirit. The washing can only ])e very imperfectly etfected. Three ways of proceeding from this point have been practised by us. In one the precipitate is diifused through dilute spirit, and stirred well and warmed witli calcium-chloride solution. The now much less voluminous precipitate is repeatedly washed with water, dried, and extracted with ether, which leaves the calcium palmitate undissolved. Spirit of 95% may be used in place of ether, but as it dissolves out a little calcium salt its use is less satisfactory. On evaporating the ether (or spirit) the alcohols and resinoïd body are obtained. A second way of proceeding is to warm the precipitate with water and hydrochloric acid, until it has shrunken to a small volume, wash repeatedly with water, press, dry, and extract with light petroleum, which dissolves out the palmitic acid and some of the resinoïd body, and leaves behind all the alcohols and the rest of the resinoïd body. After proceeding in either way, the resinoïd body is separated by repeated extractions with warm 80 % spirit. A small quantity of the alcohols at the same time dissolves, and may be partly recovered by precipitation with a very little water and extracting the precipitate with 80% spirit. The third way of proceeding, which is simpler in execution than the others, but much less effective, is to use 70-80 % spirit in place of the light petroleum, in the second way of work. This dissolves out the resinoïd body as well as the palmitic acid. Personne's method of procedure is to pour the product of saponi- fication into water, to wash the precipitate with much water, treat it ON THE COMPOSITION OP BIRD-LIME. 23 with acetic acid to neutral reaction, again \Yash, dry, dissolve in hot 90% spirit, cool, and crystallise out from the solution the bird-lime alcoliol. This method we have not found to work well, because of the great difficulty in washing properly the voluminous gelatinous precipitate, and in just neutralising it with acetic acid. This precipi- tate contains, besides the alcohols and resinoïd body, much acid potassium palmitate, to which indeed its bulky state is partly due, and we have found it far preferable to convert the potassium palmitate either into the calcium salt or into free acid, as jibove described. Personne seems not to have recognised the presence of any fatty salt in the precipitate of the alcohols. Separation of the alcohols from each other, and their petrification. — The crude solid alcohols can only be fully separated from each other by fractionated extraction with strong spirit, repeated until products are obtained of constant melting-point. The alcohols, already treated, as described, with SO "/<> spirit to remove the resinoïd body, are warmed with successive portions of spirit increasing in strength from about 85 ^o to over 90 ''/o, each portion of the solvent deposit- ing crystals of the alcohols as it cools, and each mother-liquor, by successive evaporations, yielding a series of other crystalline deposits, all similar in appearance. When the last mother-liquors are too small in quantity and too impure to yield a satisfactory product by further evaporation, they are rejected or worked up for the little resinoïd body they contain. The portions of the alcohols least soluble in spirit consist principally of the one alcohol, and those most soluble, of the other alcohol. The intermediate portions yield by re- newals of the treatment with spirit other series of deposits of higher and lower degrees of solubility, the extremes of which are the two alcohols nearly free of each other. The portions of the less soluble alcohol are submitted to further fractionation, until the part undis- 24 E. DIVERS AND M. KAWAKITA solved by hot 90 ^jo spirit, and that dissolved and deposited by it on cooling have the same melting point. It is tlien, finally, dissolved in hot 95 ^o spirit, crystallised out, and again tested as to its melting point. The crystalline deposits most soluble, and consisting principally of the more soluble alcohol, require much further fractionation, in order to sejDarate the less soluble alcohol on the one side, and the resin Old body on the other, and the ultimate yield of the pure alcohol becomes very small. In fractionating it out, spirit of 85 % is used, but finally this alcohol, like the other, is to be crystallised out from 95 °/o spirit, in order to get good crystals. Personne observed the comparative insolubility of the solid alco- hol in 80 Yo spirit, but making no use of this fact, he purified the cake of crude solid alcohol by repeated crystallisations from boiling 90 "/o spirit. During the purification he met with a body of peculiar form, visible under the microscope and less soluble than the solid alcohol in spirit, and this he found to be gradually removed by repeated crystallisations. We have met with no such substance in Japanese bird-lime. Piivification of the rcsinoid hoihj. — This is found mainly in the 80 Yo spirit used to wash the crude alcohols after they have been separated from palmitic acid. AVhen this separation has been efiected in the second way, the spirit contains also some fatty acids. The light petroleum used to dissolve out palmitic acid also contains some of the resinoïd body. In order, therefore, to separate palmitic acid, the residue, after evaporating the spirit or the petroleum, is dissolved in alcoholic potash, the palmitic acid precipitated with calcium chloride, water added, and the precipitate washed, dried, and extracted with ether. Evaporation of the ether leaves the resinoïd body still mixed with some of the alc(jhols, but free from any fatty acid. The impure product is dissolved in strong spirit, and left to evaporate slowly. 0^' THE COMPOSITION OF BIRD-LIME. 25 The alcohols separate as indistinctly crystalline, opaque matter, while the resin forms a translucent, gammy deposit, still containing spirit, on the bottom and sides of the vessel. The resin is redissolved in spirit and tlie solution left to evaporate. Repeating these operations several times yields it in a condition in which it shows no longer any tendency to deposit crystalline matter. Separation and purißcatiou of the fattij acids. — By far the greater part of the fatty salts remains dissolved when the saponified bird-lime solution is poured into dilute spirit. The filtrate and washings from the gelatinous precipitate of alcohols are diluted with water, mixed with liydrochl(3ric acid, and warmed, in order to separate the fatty acids. By similar and well-known methods the portions of these acids precipitating with the bird-lime alcohols can Ije recovered, after separating them as calcium salts from the alcohols and resinoïd body, and added to the main quantity. The crude fatty acids which, when cold, form a soft, brown, solid msss, are dissolved in alcoholic potash and precipitated again with calcium chloride ; the calcium precipitate is washed with spirit wliich removes chlorides and some colouring matter, as well as some of the calcium salt of the soft fatty acid; the precipitate is then washed with ether, which dissolves out, more easily than the spirit, most of the remaining colouring matter and calcium salt of the soft fatty acid ; lastly, it is heated with hydrochloric acid and water, in order to get the crude palmitic acid. Repeating these operations once or twice, and finally crystallising it from its spirit-solution, gives the palmitic acid pure. The s[)irit and ether washings of the calcium precipitate yield by appropriate treatment the semiliquid acid, still in an impure condition. I'almitic acid can als(3 be prepared from bird-lime by destructive distillation. Its purification from hydrocarbons by w\ay of saponifi- cation, presents no great difficulty, and need not be described. 2G r. DIVEKS AND M. KAWAKITA TliG alcohols of bird-limo. To one of the two nicohols of bird-lime we give the name mochylic a!coli.ol^ formed from the Japanese word mochi for (bird) lime or gluti- nous matter, and to the other we attach the name iUctjUc alcohol, cs:ientially the same as ilicic alcohol given by Personne to the single alcohol described by him, but framed more in accordance with the accepted nomenclature for alcohols. Our ilicylic alcohol differs but little from Personne's ilicic alcohol. Both the alcohols of bird-lime are obtained in tufts of small, slender, lustrous prisms, and are distinguishable from each other only in solubility, in melting point, and in composition. Mochiilic alcohol occurs much more abundantly than ilicylic alcohol. It dissolves well in 95-98 % spirit, but is almost insoluble in 80% spirit. It is very little soluble in petroleum spirit in the cold, is readily soluble in ether, and dissolves also in concentrated sulphuric acid, to which, like bird-lime itself, it imparts a red colour. It melts at 234° C and decomposes under atmospheric pressure at a little below the melting point of zinc, the principal product being a viscid matter, apparently the hydrocarbon to be described among the products of the destructive distillation of bird-lime. In a vacuum it sublimes slightly at a little above 160°, and freely and entirely near and above its melting point, without decomposing, or changing in melting point. Heated with palmitic acid in a sealed tube to 150- 160° it yields a body indistinguishable in essential properties from bird-lime, a sticky transparent matter, readily soluljle in ether, but nearly insoluble in the strongest spirit. Our atteinpts to form, by acetic oxide or chloride, mochyl acetate have been unsuccessful. llicijUc alcohol differs from mochylic alcohol in melting at 172°, and in being moderately soluble in 85 - 90^0 spirit, though almost ON THE COMl'OSITION OF BIED-LIME. 27 insoluble in 80 "/o spirit. It l)egins to volatilise in a vacuum below 150°, and sublimes freely near its melting point in beautiful tiifts of needles, still melting at 172°. Heated witb palmitic acid it also forms a body like bird-lime. It fails apparently to yield an acetate, even after long heating at 150-170° with acetic oxide, in which when hot it, as also mochylic alcohol, readily dissolves, partly crystallising out again unchanged on cooling, and partly becoming a dark viscid matter not acetate. Personne found his ilicic alcohol to yield a crystalline acetate with acetic oxide, melting at 204°-6°. The melt- ing point of Personne's ilicic alcohol was 175°, and its boiling point above 350°, but under the reduced pressure of 100 mms. it began to sublime at 115°. In appearance and in behaviour to spirit of different strengths, it was like our ilicylic alcohol. ]5oth mochylic and ilicylic alcohols dissolve in a mixture of sulphuric and nitric acids, and from the solution water separates a gelatinous matter, readily soluble in spirit, and puffing only slightly, when dried and heated. Chemical composition of the two alcohols, — Combustion of the two alcohols has given us the following results: — Mochylic alcohol, m. p., 234° I. II. III. C^eH^eO. Carbon 83.37 83.39 83.28 83.42 Hydrogen 12.29 12.16 12.38 12.30 Oxygen 4.28 100.00 llicijlic alcohol^ m. p., 172° I. II. a,H3,o Carbon 83.09 82.98 83.02 Hydrogen 11.93 11.92 11.95 Oxygen 5.03 100.00 28 E. DIVERS AND M. KAWAKITA Illcic alcohol, m. ])., 175° (Personne'« analyses). I. II. III. IV. V. Mean G„H.hO Carbon 83.25 83.61 83.48 83.07 83.40 83.3G 83.33 Hydrogen 12.18 12.44 12.17 12.24 11.98 12.20 12,22 Oxygen 4.45 100.00 It will bs seen that Personne's numbers vary rather widely, but fall for the most part between those obtained by us for our two alcohols. It will also be seen that the formula he has proposed, as agreeing best with the mean of his analyses, is that of a homologue of our alcohols, the general expression being 0^ Hgu.ß 0. As he worked upon bird-lime from a species of Ilex different from that which yields Japanese bird-lime, it cannot for the present be decided whether ilicic alcohol is distinct from the alcohols here described. The resin Old component of bird-lime. The resinoïd l)ody is obtained in pale-yellow fragments which are brittle, and not sticky like bird-lime. It melts at 110°, and does not volatilise when heated to 220° in a vacuum. Above 360° it darkens, boils, and distils without much apparent change. It is very soluble in spirit, even of only 80 ^o strengtli, also in ether. When its spirit solution is evaporated by heat sufficiently, it separates from its solvent as a viscid liquid still containing spirit, which evaporates by further heating below 100°. Its solubility in spirit is not in- creased by the presence of potassium hydroxide. Heated with the solid hydroxide barely to the melting point, it slowly combines with it, probably at the same time absorbing oxygen. The cooled mass wholly dissolves in water from which hydrochloric acid precipitates a gelatinous body very brittle when dried. We have not farther examined it, ïov want of material. ON THE COMPOSITION OF BIRD-LIME. 29 When bird-lime is kept for a long time, a thin brittle skin forms on its surface, which is readily soluble in spirit. This skin consists probably of the resinoïd body. If it docs not, then Ave have no evidence as to whether the resinoïd body is produced during the saponification of the bird-lime, or exists in it ready-formed, as the result of slow atmospheric oxidation. In composition the resinoïd body differs from mochylic alcohol only in having two atoms less of hydrogen, as the following analyses and calculation show: — I. II. C,oH,,0 Carbon 83.79 83.66 83.87 Hydrogen 11.80 11.92 11.83 Oxygen 4.30 100.00 The fatty acids of bird-lime. The fatty acids of bird-lime are two, as already stated, palmitic acid, and, in small quantity only, a semi-liquid acid, the calcium salt of which is soluble in spirit and in ether. This acid has not been further examined. The other shows all the characters of palmitic acid. Melting point, 61.5°. Analysis (I.) of acid prepared by saponification, and ( II.) of acid obtained by destructive distillation of purified bird-lime: — • I. II. Palmitic acid. Carbon 74.98 74.86 75.00 Hydrogen 12.67 12.55 12.50 Oxygen 12.50 100.00 The potassium salt yielded 13.3 per cent, of potassium. 30 E. DIVERS AND M. KAWAKITA Products of destructive distillation of bird-lime. These have been already enumerated, so far as their nature iä known to us, and the result of analysis of the pahnitic acid hns just been tabulated. The principal hydrocarbon, distilling next after the palmitic acid, was prepared from the middle portion of the distillate, by treating it with hot spirit so as to leave about half undis- solved. This was then washed with cold spirit. The hydrocarbon thus left Avas a thick oil, slightly yellow, but free from fluorescence. On analysis, it gave numbers agreeing with the formula, C2JI41: — Found Calculated Carbon 87.59 87.64 Hydrogen 12.49 12.36 100.08 100.00 Apparently the same body is obtained by distilling mochylic alcohol at the ordinary atmospheric pressure. The decomposition of the main constituent of bird-lime by heat may therefore be thus re- presented:— Mocbyl palmitate. C42 H76 O2 = CjG H^j + C16 H32 O2 and the decomposition of mochylic alcohol by — CgG H^Q 0 = Cog H^i -f H2 0 . The last fractions of the distillate consisted of hydrocarbons yielding nearly 91 per cent, of carbon. No attempt was made to isolate the caoutchouc hydrocarbons present, no doubt, in the mixture. Constitution of bird-lime. Bird-lime is closely allied to the waxes, and consists principally of mochjl and ilicijl pahnitatcs, C^o H/g Oo and C33 Hya 0, . On Anorthite from Miyakejima, hV Yasushi Kikuchi. Assistant Profossni- of Geology, Imperial University. With Plate I. The AnorHiite-crv.stuls which form tlie subject of the ]»re8eijt cominnniciition were collected l)y Messrs, S. Okiil)o and X. ut of the crater as such.* A. Penckt has indicated the pi)ssibility of such * During a recent visit to Siilplmr Islau^l, I have obssrvetl the same phenomenon. Felspar crystals, the nature of which awaits further investigation, are fauml, covered with a black hxva- crast, and forming a loose superficial layer in the vicinity of a volcano on that Isliud. t Studien über lockere vulkanische Auswürfiiuge— Zeitschft. d. deutsch, geol, üesellft. Bd. 30 1874. p. 124. 32 Y. KIKUCHI crystal-ejections near tlie crater of a volcano. A case analogous to that here described is met with in the eruptions of Vesiivins; crystals of Leucite wliicli there play tlie role of Felspar, h.-ing sometimes esjected. Tims Leucite eruptions continued from April 1845 to January 1849, according' to Scacchi,* who considers these well- defined Leucite crystals as beino; derived from the refusion of the crystals already formed in tlie older lava within the volcano. From the basic character of the Anorthite it is to be inferred that it had early crystallized out of the magmn, and that the crystals thus formed were ejected while the latter was still in a liquid condition ; the heavier crystals falling near the crater, while the lighter portions of the ejected matter were carried further off. On this account, they have always a thin vesicular coating of lava, the interi(»r however remaininnf uniniured. Similar kinds of crystals have also been broujjht from the other islands of the Shichitö Chain, viz. Oshima (Yries Lsland) and Hachijö. Most of the specimens of lava, which I have seen are of a black colour, and have a porphyritic structure, mostly porous but some- times very compact. The porphyritic ingredients are Anorthite- crystals of the Microtine type, and Olivine, usually in the form of rounded grains, and very rarely showino- crystal-f)rm, as S. v. CD / •' * O « ■ Waltershausen t observed on the Olivine in the lava of Etna. A coat- ing of red iron oxide is always f nmd on these Olivine grains, often assuming a brilliant metallic lustre. Microscopically examined, the ground-mass is seen to consist of a microcrystalline airsrreofate of Felspar and Magnetite, with sometimes microcrystals of Apatite. * Ueber den Ursprung der vulkanisclien Asche. — Anszu» by Rammelsberg in Zeitschft. d. dentscb. geol. Gesellschft. Bd. 24. 187'2. p. 5J8. t Vulkauische Gesteine in Sicilien ixnd Island, p. 161, ON ANORTHITE FROM MIYAKEJIMA. 33 Generally a brown coloured glass basis is found between these crystals. Anorthite, Olivine, and Augite CiMistitiite the essential microscopic porphyritic components. The order of crystalliziition of these com- ponents can pretty certainly be described as follows: — Aj^atite, Magnetite, Olivine, Anorthite, and Augite. A specimen of Basalt- glass or Tachylite is also known. Under the microscope, it is found to consist of a brown glass, in which well-defined crystals of Anor- thite, Augite-microliths, Olivine and Magnetite are developed. The recent lava of Miyake may therefore be called Anorthite-basalt. The volcanic rock composing the upper part of Fujiyama is, according to Lüdecke and AYada,* Anorthite-basalt. It is probable that this rock constitutes most of the lavas of comparatively recent eruptions in the volcanoes of the Shicliitö Group. The chemical analysis of the Anorthite forming the porphyritic component of the rock f(jund near Tönosawa, situated within the extinct volcano of Hakone, was first published by Wada.* But unfortunately the crystaîlographic and optical characters could not be well investigated because of the nature of these crystals. Certain glassy Felspar crystals which had been brought from the Island of Oshima and known as Saiiidine, were found on examination to be completely identical with the specimens from Miyake. Dr. E. Xaumannt states in his account of the Island of Oshima, that Sanidine crystals occur very abundantly in the lava of that Island. So far, however, I have been unable to find anv Sanidine crystal in anv of the collections brouofht from Oshima, which have been to mv hand. The Anorthite-crystals from Miyakejima are 1 - 4om. in their longest direction, and are always covered with a black or sometimes * Notas ou FujiyaniA — Transactions of the Seism. Soc. of Japan. Vol. IV 1882. t Die Vulkauinssl Ooshima u. ihre jüngste Eruption — Zeitschft. d. deutsch, geol. Geseiht. Bd. 29. 1877. p. 378. 34 y. KIKUCHI reddiyh coloured thin spougy crust of ];iva. This thin coating examined under tlie microscope, is found to consist of a brown amor- phous glass, sometimes with microcrystals of PIagiocla«!3, a few Au"ite u'rains and Mno-netite. The interior of the crystal in fresh spechnens is transparent, and has a hyaline lustre. The cleavage is perfect parallel to V. and M. The hiHer is less perfect than P, and the cleavage-face always exhibits twin-striœ. A slightl}^ pearly lustre is observed on both of these cleavage-faces; on one portion of the brachypinacoidal cleavage-plane, there has been observed in one specimen, a play of colour somewhat like that of Labradorite. Grains of Olivine are always found as enclosures, sometimes arranged in dis- tinct zones. In some portions glass-enclosures are found often arranged parallel to a crystallographic direction. »Some specimens show the beginning of decom[)osition, becoming hazy in appearance. Sino-le individuals are rather rare ; they are mostlv found as twins, or several individuals are grouped together in a most irregular man- ner, assumini!' a ulobular mass. C^rvstal-f jrm.^ — The thin coiilinn' of lava^ which always covered tlje crystals, rendered the determination of the crystal-faces peculiarly difhcult. The reflecting goniometer could Jiot be used, except in the rleavaufe-faces. 1)V nieasurinu' tlie auQ-le between cleavag-e-faces V : i)/, and by iinding the position of the face ij. Avhich is always developed, and the position of tlie Pericline twin-lamella' on i)/, I was able generally to iix the position of the crystal, so tliat the other faces could be determined by approximate measurements with a contact goniometer and by zonal relations. Ir was not easv to get a perfectly clear cleavage-piece for goniometric and optical purjioses, ;is the twin- stria^, Olivine-enclosures, and numerous fissures which make the crystals very brittle, were always present. The cleavage-face parallel to P usually gave a good reflection, but that parallel to J/, a dis- ON ANORTHITE FBOM MIYAKEJIMA. 35 turbsd or diffused reflection. Hence the measurement of this facial anofle did not g-iye always a satisfactory result. With the clearest piece Avhich I could get, the angle of i^ : M (001 : 010) was found to be 94° 8' and the angle P : il/' (001 : 0Ï0) 85° 51', while the corre- sponding angles of the Anorthite of Vesuvius are 94:° 10' and 85° 50'. The foJlowino" faces have been identified : — r = oV (001) M = crA>db (010) T = or.;P (1Ï0) / = cc7V (110) z = cc;pS(130) /• = ccP;3(130) / = :2'P'cÄ(20l) // = ^,V,^ (201) c = 21''àb (021) // = 2'/'d6(021) m= F' (HI) 0 = 1\ (HI) i>= p (111) h = 4'P2 (241) V = 47^,2 (24Î) ir == 4/^2 (241) u = 2L\ (221) X = ,Vr^ (101) Some of the ap[)roximate measurements of the important angles are given below, compared with the angles calculated by v. Kohscharow* for the Vesuvian crystal : — Kokscharow. 7" =. 98Vo° 98° 46' 8" 89 27 2(3 90 32 34 142 13 6 139 48 34 l) •> o >> /» OO O OO 133 14 12 II y y y y y p M' = 89 M = [)()' u = 142 V, p = 139 7, n = 83 V, n' = 133^ The faL-es t, m, h, v, iv, u, are very rare, x lias been observed in one specimen only as a yery narrow band. • Materialien zur Minéralogie Kiisslancls. Bd. IV. 36 Y. KIKUGHI The most prödominating faces are P, M", ?/, 6, m, to which T, I, are somatimas combined. The most primitive form i.s a parallelopiped formed by the faces P, M, y. Usually the hemidom^ e or n trun- cates the edge of P : J/, or P : il/', to which prisms 7', /, and quarter- pyramids 0, p, are added, (fig. 1, 2). The différent modes of facial development give rise to different forms, which may conveniently be distinguished into two types. First Type — The crystal assumes a tabular shapj by the [»re- dominance of the Base 1\ while the Brachypinacoid M appears as a narrow band extending in the direction of the Brachyaxis a (fig. 1, 2). Fig. 3. represents in projection this type of crystal, in which the development of tlie faces is most perfect. Some of these crystals show an irregularity arising from parallel intergrowtli. (tig. -A). Second Type — In this type, the form is also tabular, but the pre- dominating lace is Jl/, while P appears as a narrow band extending along the Brachyaxis. The faces forming this type are the same as those of the first type (tig. 5). Simple crystals oî the first type are more numerous than those of the second ; the latter however appear more frequently as twins. In both types, the form becomes sometimes irregular by the unequal development of the faces on the back and front sides, (renerally o, p, appear on one side as narrow bands, while their opposite faces o', //, become much broader and intersect in a line, which would be parallel to the face x, did this face occur, as it generally does not. But owing to the imperfection of many specimens it is often difficult to say whether the faces o,]>, have greater extension on the back or on the front side. The predominance of the faces 1\ M, y, o, p, and tlie absence or rarity of prisms T, /, resemble to some extent the case of the Anorthite-crystals first described by G. vom Rath,* found as a second- • Beitraege zur Pétrographie— Zeitschft. d. deutsch, geol. Gesellft. 13d. 27. 1875 p. 302. ON ANORTHITE FROM MIYAKE JIMA. 37 ary minernl in the contact-rock of Pesmeda-Alps, in Monzoni, Tyrol. Bat the difference between the two lie3 in the fact tliat the Pesmeda- Anorthite is elongated along the Mncroaxis h, Avliile the Miyake- crystal has o-reater elong^ation alono- the Brachyaxis ci. T^vins — Simple crystals are rare ; most of the crystals being found as penetration-twins. The twin law is analogous to that of the Carlsbad type in Orthoclase. The twinning-axis is the A^ertical axis c, the plane of composition the Brachypinacoid M ( r^P^). Bnt as the crystals have often no prismatic edge, the two individuals have apparently no common direction. An exact measurement of the angles made l)y the faces of the two individuals is impossible. The angular divergence of the Brachyaxis, as measured along the edo-es of P : /)/ and V : M is approximately 128°. Fig. 6 rejiresents a typical form of the twin, the faces consisting of P, i)/, ?/, f the "Rhombische Schnitt" <^n the cleavage-face of J/, is indicated by very fine stria^ making an angle of -15° to -17° with the edo-e of P: M (fig. 1). These twin-lamellae are extremely fine, but they become comparatively much ))roader in the crystals im- bedded in the lava, sometimes attaining a breadth of 7, mm. Fig. 11, 12, show these broader twin-lamella^ as observed in sections of por- phyritic crystals, cut nearly perpendicular to the edge 1' : 3/. Optical characters — The direction of extinction has been ex- amined on the cleavage-faces of 7', and of 3/ under parallel polarised lio-lit. As the mean of several observations, tlie direction of maxi- mum extinction on the fa'C P was found to be-38° to— 10°. Avhile that of the face 31 was -40° to-4r; tlie latter does not completely extinçruish the lioht, owini»- to the interposition i-f the Pericline twin- lamellai. • Sartorius-Lasaulx, Der Aetua. Bd. II. p. 505. also Zaitsclift, f. Krystallographie, Bd. 5. 1881 p. 32G. ox ANOßTHITE FEOil MIYAKEJIMA. oU In convergent polarised light, the cleavage piece parallel to P shoAVs one of the axial rings at the margin of the field, while that parallel to J/ shows another, but in the latter the centre of the rino- is much nearer to the maro-in of the field. These observations ao-ree completely with Max Schuster's* on the Anorthite of Vesuvius and Pesmeda. The position of the positive Bisectrix Avould hence be nearly perpendicular to the face e ('2P'oc). The section cut as nearly as possible parallel to this face, shows in a convergent polarised ray, a biaxial interference-figure of a wide angle. Further optical charac- ters can not at present be siven. The porphyritic crystals which are fou.nd within the basaltic lava, show somewhat ditferent habitus, as compared with the crystals already described. The f jlloAving observations ma}^ serve to indicate the characters of the porphyritic crystals as seen under microscopic slides of the lava. The section cut nearly j-arallel to P (oP) presents an almost rectangular outline formed by the traces of the faces y and M. Fig. 10 represents a typical form (jf the section ; sometimes the corners being differently cut oif as in fig. 9. Generally a few lamella3 of the Albite type are found parallel to the cleavage-trace of i)7, the direction of ex- tinction making 40° on both sides of the boundary. A noteworthy feature which is often observed in this section, is the presence of fine diagonal fissures (fig. I), 10), intersecting at angles of nearly 85° and 95°, the acuter angle being nearly bisected by the cleavage-trace of i)/. Such fissures are also to be seen on the cleavage-face of P of separate crystals with the effect of making them Aery brittle, but the larger porphyritic crystals imbedded Avithin the lava, exhibit the structure much more conspicuously. Such crystals Avhen taken from the matrix fall into fine splinters along the fissures. Usually one of * Die optische Orientirung der Plagioklase.-Tschermak's. Min. ii. Petro. Mittli. Bel. 3. 1881. p. 213. 40 Y. KIKUCHI these sets of fissures is more perfect and distinct than the other, as in fig. 10, where the crystal has been ])roken along one of the directions of the fissures. G. Tschermak* has observed and photographed an analogous series of fissures found in an Anorthite-crystal within the tufaceous Meteorite (Eucrite) of Stannern. ]Jut in the crystals from Miyake, they are more numerous and distinct, A third series of fissures parallel to ooPoo , shown in Tschermak's figure has not been observed in the Miyake Anorthite. An anomalous optical phenome- non is observed in the crystals showing this fissured structure. When a slide of such a crystal is examined under crossed. Niçois, the extinction of light takes place in some parts, as usual at a direction making -38° to -40° with the trace of M, but the entire field never becomes dark ; the dark portion being distinctly marked off from the luminous part. The boundary of the two parts is formed by zig- zag lines which correspond to the directions of the fissures and to the cleavage-trace of M. Fig. 1) represents the section of a crystal formed by the parallel intergrowth of two individuals. The upper part, almost devoid of fissures, is separated by an irregular boundary from the lower half, in which very characterestic series of fissures are to be seen. The crystal is twinned according to the Albite-type, consist- ing of two principal twin-individuals aa and hh, which extinguish the light symmetrically on both sides of the trace of M at an angle of 38°. The part marked cc never becomes dark in any position, only showing a light bluish-violet tint when the portion aa becomes entirely dark. Sometimes patches of such luminous parts are found scattered M'ithin the dark field, assuming a damasked appearance. The phe- nomena just described may probably be due to a strain resulting from the sudden contraction experienced at the moment of eruption. In fig. y, the mode of attachment of the upper and the lower halves shows that the upper went on growing in the direction of the lower * Mikroskopische Beschaffenheit der Meteoriten. Taf. II, fi<^. 1. ON ANORTHITE FKOM MIYAKEJIMA. 41 lialf, after the lattei* had experienced a sudden contraction, the magma still retaining a certain degree of fluidity. The section parallel to Jl/(ooPco), is generally a parallelo- g-ram with anHes of 98° and 82°, but sometimes becomes rhombic in outline by the development of the faces o, p, combined with P, the angles of the rhomb measuring 52° and 128° (compare fig. 7). Peric- line-lamellas making an angle approaching to -16 with the trace of I\ are often found. The section perpendicular to the zone of P : BI, presents a rec- tangular form formed by the traces of P and M, the corners being usually cut off by the faces e, u. Fig. 11. 12, show the usual forms of this section ; fig. 11 representing the first type, fig. 12 the second type, according to the classification of separate crystals already given. In this section, broader l^erichne-lamellaî are inserted, which would be parallel to the trace of P, if the section falls just parallel to the jMacroaxis. Some of these lamellae thin out at the extremities. The broad band arranged parallel to M is probably twinned according to the Albite type (fig. 12). This band itself is divided into two halves a, h in the manner of a penetrating twin, with different optical ori- entation, (a/l) = 40°/25°), so that these halves may be twinned according to the Carlsbad law. The sections parallel to M and perpendicular to the Brachyaxis are traversed by irregular fissures, some of whicli riui nearly in the direction of the Vertical axis. The substance of these felspars is always fresh, and nearly homogeneous; some variation in chemical composition being sometimes indicated by zonal structures as in fig. 12, wdien examined under polarised light. The enclosures are Olivine-grains, glass, and gas pores which are sometimes arranged parallel to a crystallographic direction. Olivine-grains are generally found as enclosures within the Anorthite, but occasionally we observe the inclusion of Anorthite within the Olivine. 42 Y. KIKUCHI C(UT08ion-fignrps. — Souie experiments liave heen tried in regard to tlie clinracter of corrosion-figures (Aetzfiguren) on cleavage-faces. Clear cleavage-pieces were selected, and acted npon hy dilute fluoric acid, and by hydrochloric acid. In some natural cleavage-faces which have been long exposed to the action of atmospheric agencies, corro- sion-fignres are found similar in character to those produced artifi- cially. These figures are most distinct under a magnifying power of 200-300 diameters. On the face 7^=001 triangular impressions are formed (fig. 13). These triangles are asymmetric, rounded on one side, while the other sides are nearly straight; the angle formed by the latter being directed backward, as observed by Wiik.* On the face 7^'= 001, the figures are reversed as shown in fig. 14, the angle formed by the straight sides being pointed forward. On observing the difterent figures, we find some of them formed of dif- ferent faces apparently of îi hemidome and quarter-pyramids (fig. 13 a, /), c), different stages of such development being seen. The figures produced by hydrochloric acid on these faces are smaller and the angle formed by the straight sides more acute than those produced by fluoric acid. Wiik considers the two straight sides as being formed of /(Go.P;3)and2;(oo;P3). On Ihe face ilf=010 are produced the elliptical figures observed by AViik. The one end of the longer axis of the ellipse is pointed, and directed upward and backward (fig. o). The direction of the hmger axis is inclined in the same sense as that of the Pericline-lamella?, makino- an ano-je of nearly ^^° with the edge of P : il/. 4'he figures which are produced on the face M' = 010 are of the same form, but reversed iu position, as shown in fig. 16, the pointed end of the figure ])eing directed dc^vnward and forward, and arranged in the same direction as those on M. * R'?ferat in Zeitsclift-. f. Krystallographie. Bd. 7. p. 188. ON ANOIITHITE FEOM MIYAKE.JIMA. 43 The figures produced by fluoric acid are nearly all elliptical (fio-. 15, 16, h) while those produced by hydrochloric acid are round-sided asymmetric triangles (fig. 15, K),^^), one of the sides being apparently' parallel to the direction of the Pericline-lamella^. Fig. 17 is intended to show the relative position of these figures with reference to a simple primitive form consisting of the faces P, M, and y ; the figures on P' and M' being supposed to be looked at through the face y. For the chemical analysis of the mineral, I am indebted to Mr. Y. Kitamura, of the Geological Survey. The sample taken for the analysis was obtained from fresh specimens of separate crystals. The lava-coating was removed, and transparent portions of the interior carefully selected. The small grains of Olivine attached to the powdered sample were removed by means of Thoulet's solution, until the microscope showed the ideal purity of the sample. The result of the analysis is shown below in Column I. For the sake of com- parison, we quote here in Column II, the analysis of a similar kind of Anorthite as that of Miyake, from a porphyritic volcanic rock of Tönosawa in Hakone, given by Wada.* I. II. Miyake Tönosawa (Anal. Y. Kifcainura) (Anal. M. Hida) Si 0, 44-03 Vo.... 44-16 Vo AhO, 36-80 .... 31-87 Fe,0, 1 OO Ca 0 19-29 .... 20-90 MgO -20 -53 Na,0 .... -23 .... -32 IC 0 .... , -55 H, 0 .... -12 -60 100-67 100-26 1. c. p. 33. 44 Y. KIKUCHI Speclftc gravity = 2*761 as determined by Thoulet's solution. Before the blowpipe, somewhat difficult to fuse. Very fine splinters which can sometimes be obtained from the fissured variety within the lava, fuse comparatively ensily to a clear glass, glowing intensely, and im- parting a reddish -yellow tint to the flame. In hydrochloric acid, the powder is entirely decomposed, leaving gelatinous silica. ON ANORTHITE FROM MIYAKE JIMA. 45 Explanation of the Plate I. Fig. 1. — A simple crystal of the first type in the position adopted by Des Cloizeaux and vom Kath. On the face il/=010, the direction of the Pericline-lamella?, or the trace of the " Rhombische Schnitt" is indicated by fine lines, which make an angle of about -16 with the ed"e P : M. Fig. 2. — Same as fig. 1. A very common form of the first type. Fig. 3. — A horizontal projection of a crystal of the first type ; the plane of projection perpendicular to the Brachyaxis. Fig. à. — Same as fig. 3, showing parallel intergrowth ; the crystal is imperfect at the middle portion and an irregular attachment of smaller individuals is observed in the specimen a. Fig. Ö. — A horizontal projection as in fig. 3, 4. Some imperfec- tion is observed in the lower portion. Fig. 6. — A common form of the penetration-twin analogous to the Carlsbad type. In this form it often happens that the faces o, ]), are unequally developed on ditterent sides as in the figure. The intersection-edges of o\ p, M' become very nearly parallel to those of jP, e, M. Fig. 7 . — A rhomb-shaped modification of the twin, the same as in fig. 6. The two individuals which are distinguished by the thicker and the lighter lines, are composed of several subindividuals. The faces 0, p, are developed on both sides very extensively, the edge P, d, il/ becoming very nearly parallel to the edge o,p',j\l'. The projection is made on a ])lane perpendicular to tlie edge P, x, //, of one individual, so that the edge P, .t, // does not coincide exactly with the first. Tlie face t becomes here relatively large, and ^ ery nearly parallel to y'. Fig. 8. — Horizontal projection of tlie twin, similar to those of 46 Y. KIKUCHI üf. (3, 7, Ijut only the halves of the two iiidividiials attached by 3/ without penetratuig. Fig. 9. — Section of a porphyritic crystal within the hiva, niagni- iied about 20 times, cut nearly parallel to the face F, consisting of the nonfissured upper, and the fissured lower, halves in a parallel position. These hidividuals consist of two principal twin-individuals a, (y, and h. h, of the Albite type, the part c, c, showing an optical anomaly. Fi(j. 10. — Same as fig. 9, magnified 10 times, the rectangular outline being f jrmed l)y J/, //. ^ t'l'y characteristic diagonal fissures are here developed, similar to those of the lower half of fig. 9, the upper right and the lower left corners being removed by the one set of fissures which is more perfect than the other. A single lamella of the Albite type traverses the middle portion. F'kj. 11. — Section of a porphyritic crystal, cut nearly perpen- dicular to the zone F : il/, magnified 5 times, showing the broader l-*ericline-lamellii?. Fig. 12. — Same as fig. 11, magnified 10 times, showing the fine transversal Pericline-lamelke, and the broader vertical lamelhe rt, ?>, probably of Albite and of Carlsbad types. Zonal structure is well exhibited in this section under polarised light. Fig. 13. — Corrosion-figures on the cleavage-face 7' ; the acute angles of the triangular impressions are pointed backward. 1'hey are shown magnified in a, h^ c. Fig. 11. — The same as seen on the cleavage of F\ of the same form as those on F, but the acute angles are pointed forward. Magni- fied in a, h. Fig. Id. — The same of elliptical forms, on the cleavage-face of i)/. The fine lines represent the direction of the trace of the "Rhombische schnitt '' a, figure produced by hydrochloric acid ; h, by fiuoric acid, magnified. Fig. IG. — The same as seen on the cleavage- face 37'; a, h as in fio^. 15. ox AXOKTHITE FROM MIYAKE JIMA. 47 Fig. 17. — A diagrammatic representation of the relative positions of corrosion-figures on the faces P, P', il/, 3/', in a primitive form consisting of P, M, y ; the figures on P' and J/' being supposed tu be looked at through the face y. The Source of Bothriocephalus latus in Japan, Isao Ijima, Ph.D. Bothriocephalns latus, which was formerly thought to he restricted to Europe .in its distribution, is the commonest tape-worm found in Japan. It is met with everywhere in the country, of course with local differences as to its frequency. The identity of the species with the European f )rm admits in my o})inion of no doubt, I would not have made this remark, had not Küchenmeister of late errone- ously assumed that in Europe more than one species is included under the name of B. latus. As far as my experiences go. Taenia mediocanellata occurs in Japan much more rarely than B. latus ; among dozens of tape- worms that I have examined, only a single specimen of the former species M'as found. That fish, the recognized source of Bothriocephalus, is used as food much more generally in Japan than beef (the source of Tœnia mediocanellata), sufficiently explains the above-stated fact. Tcenia solium, if ever it occurs in Japan, must be exceedingly rare. I have indeed, never as yet met with it in Tokyo. This is un- doubtedly due to the fact that pork as an article of food is even less used than beef. Unfortunately means are not yet at my disposal for determining the occurrence of human tape-worms at any locality on statistical grounds. 50 T. IJIMA As is well known, Prof. M. Braun ^^ was the first to show that in the Baltic region (Dorpat) the plerocercoid larvœ of B. latus infest the viscera as well as the flesh of the pike (Esox lucius) and of the burbot (Lota vulgaris). He proved that the above-mentioned tape-worm, so common in Dorpat and vicinity, is derived from using these infected fishes, and the pike especially, as food. Küchen- meister ^^ alone stood out against considering the pike as the inter- mediate host of B. latus, substituting certain Salmonida3 (Salmo salar) in its stead. His opinion has however sutfered a well-grounded refutation by Braun and Leuckart,^^ so that the former's discovery stands as an established fact. According to Leuckart, Grassi in Catania also succeeded in raising B. latus in himself from larvœ taken from the pike. And Parona*) in Lombardy found larvœ, which were experimentally proved to belong to B. latus, not only in the pike but also in the perch (Perca fluviatilis). Further it turned out from Zschokke's and my observations that some Salmonidœ are also to be counted amonsf the intermediate hosts of B. latus. Zschokke's researches 5) at Geneva, one of the places noted for the occurrence of B. latus, have proved that there its principal source is Lota vulgaris and after it Perca fluviatilis, while infection from pike and salmon is considered as occurring only in exceptional cases. Of the latter, Salmo umbla was found almost regularly infested by Bothriocephalus-larvai. 1) Zur Entwicklungsgeschichte des breiten Bandwurmes. Würzburg 1883. 2) "Wie steckt sich der Mensch mit Bothriocephalus an?" Berlin, klin. Wochenschrift. Nr. 32, 33. 1885.— "Die Finne des Bothriocephahis und ihre Uebertragung auf den Menschen." Leipzig 1886. — " Weitere Bestätigung meiner Behauptung, dass die Finne des Hechtes nichts mit Both, latus zu thun hat." Deutsche medic. Wochenschrift. Nr. 32. 188S. 3) See Leuckart " Zur Bothriocephalus-Frage." Centralbl. f. Bactériologie u. Parasiten- kunde. Nr. 1, 2. 1887. 4) In " Estratto dei Rendiconti del R. Istituto Lombardo. Ser. II. Vol. 19. 188G. 5) " Der Both, latus in Genf." Centralb. f. Bact. u. Parasstenk. Nr. 13, 14. 1887. THE SOUECE OF BOTHRIOCEPHALUS LATUS IN JAPAN, 51 From what ha.s been said it is appai*ent that the (last) inter- mediate host of B. latus includes several species of fish which may belong to different families and that the principal source of that tape- worm may vary according to localities. In Japan, where as already said, B. latus is abundant, none of the above-mentioned fishes are known to occur, although the pike is said to exist in Sasrhalien. What fish is then the source of B. latus in Japan? There has Ijeen among the Japanese a popular belief that tape- worms develope from eating certain fishes. Onchorhynchus Perryi Hilgd. (Mas2i) and Onchorhynchus Haberi Hilgd. (^Sake) were the most suspected, a belief which was certainly without any scientific basis. Guided by this suspicion, however, I examined in May 1886 a specimen of Onchorhynchus Perryi, and was not disappointed. Seven Bothriocephalus-larva', unmistakeable by the configuration of their head, were found imbedded in the trunk-muscles. In form, size and motions, they corresponded exactly with Braun's description and figures of the larva of B. latus. Concerning the appearance of the larvœ I have nothing of impor- tance to add to what is already known. Nevertheless a few words about them mi"ht be useful, since zooloiii- cal Hterature is not accessible to many in Japan. The larva of B. latus is a slender elongated worm of white color. The body is properly speaking not flat but rather thick. Its length varies from 8 to 30 mm, its breadth from 1 to 3 mm. When contracted, the leno'th is reduced by more than one-half, but the ])readth rela- tively increases so that it acquires a thick wrinkled ""^ „ , . " A Larv:t of Botunoce- form. The head is then involuted and shows a cleft-like onchorhyLchus Per- ryi. a, iu contracted; depression at the end. If such a larva be put into AbÔut*'2x"maguified." 52 I. IJIMA luke-warm water or salt-solution, it begins to move energetically. The head is alternately thrown in and out, the body at the same time bending and stretching and exhibiting also peristaltic motions which travel from end to end. The head, when extended, is club-shaped, hardly 1 mm in length and bears two shallow longitudinal grooves, thus essentially agreeing in form with the head of the fully grown Bothriocephalus latus. The body is solid as in the mature w^orm but without any trace of strobilization or of sexual organs. Ex- amined under the microscope, it clearly shows some of the excretory vessels and calcareous bodies, of which there are comparatively many. The larva, as it lies in the flesh, is tolerably extended but with the head always drawn in. It rarely lies straight but usually irregularly coiled. No sort of capsular membrane invests it. Within the body of the host, it by no means makes such active movements as described above ; but it is exceedingly probable that it may at any time shift its position within the host. At least, there can be no doubt about this much, that the larvœ found in various parts of fishes have wan- dered in from the intestinal canal. Braun found them not only in the muscles of the pike, but also in the various organs of its body- cavity, some hanging on the intestinal wall, others free in that cavity. Moreover he found on the liver-surface tracks of their wanderings. In Onchorhynchus Perryi, I have as yet never found them in any other place than in the trunk-muscles. To prove that the Bothriocephalus larvae obtained from Oncho- rhynchus Perryi belonged to B. latus, it was necessary to experiment with them. So I swallowed two larvae, one of which was however injured :and my assistant, Mr. M. Kikuchi, kindly offered his services and swallowed three. This was done on May 10*^^ 1886. From the 2^^ or the 3^'^^ day, I began to experience now and then slight pains in the duodenal region. This lasted for some days and THE SOURCE OF BOTHHIOCEPHALUS LATUS IN JAPAN. 53 then entirely stopped, so that I felt myself in my iisaal health until the 28^^ of the same month, when slight diarrhœa. began to incon- venience me. On June 1^^ a piece of B. latus, 22.5 cm long, was discharged. Since it had the characteristic terminal proglottis, it was certain that no segments had been previously lost. On the following day, viz. 22 days after the beginning of the experiment, the remainder of the tape-worm body including the head was obtained by means of anthelmintic. It seemed that only one (probably the uninjured) of the two larvœ swallowed had developed into the tape-worm. In Mr. Kikuchi's case, intestinal complaint commenced a little later than in mine. In both cases, fœces were subjected to micro- scopical examination from time to time, but Bothriocephalus-eggs were never met with. It seems that it requires a considerable length of time before the worm is ripe enough to let some of its eggs escape from the uterine opening. On June 27*^^ anthelmintic was given to Mr. Kikuchi. Unfortunately however no worm was obtained. Apparently it became lost, proper precautions not having been taken to secure it. After this his complaint entirely ceased. The total length of the worm that had grown in me was 315 cm (over ten feet!) and the number of proglottides, as far as could be counted, 1467. Of these the last 617 had their uterus already filled with eggs. Considering that a larva of insignificant size had acquired the above-stated respectable dimension during a period of only '2'1 days, the rate of growth is really wonderful and one might doubt the genetic connection between the larva voluntarily swallowed and the tape-worm produced, were it not for the similar results arrived at by Braun and Zschokke. The former experimented not only on cats and dogs but also on three of his students, who had voluntarily offered themselves for the purpose. They were first ascertained to have no tape- worm in the body and then each swallowed 3 or 4 0^ I. IJIMA larvae. Symptoms appeared in 3 weeks and in a month every one of them hegan to discharge Bothriocephalus eggs. At this, an- thehninthic was given and in all more than 6 specimens of B. latus were ohtained, averaging about 335 cm in length. Zschokke also obtained positive results in several experiments on man. In 4 cases symptoms appeared in lG-22 days after swallowing the larvœ and the tape- worms obtained in 20-26 days measured about 129 cm on an average. Some specimens of Bothriocepha]us-larva3 from Onchorhynchus Perryi were sent to Prof. Leuckart for examination. He did not hesitate to pronounce them as identical with those found in the pike of Europe. Thus, the fact is established that Onchorhynchus Perryi does harbour the larva of B. latus ; and I believe I am right in claiming that fish to be at least one of the chief sources of B. latus in Japan. Subsecjuently I have had occasion to examine 7 examples of that fish at different periods of the year and found Bothriocephalus-larvœ in all but one. The single negative case was a young 0. Perryi from the Hokkaido. It was in a state of putrefaction, making the task of search so unpleasant, that a thorough examination was abandoned. Another Hokkaido specimen, said to have been sent in ice contained at least 4 larvae. The other specimens Avere brought to the Tokyo market from neighboring districts, probaljly from the Tonegawa. The number of larvœ lodged in one fish was not large, 7 having been the maximum. In one case, only a single larva could be found after a laborious search. Even then it would be unsafe to assert that no chance was left for some to escape unnoticed. At all events, the number of the B. larvœ in 0. Perryi is much less than that found by Braun or Zschokke in the pike, burbot, etc. But to judge from my somewhat limited experience, their occurrence in O. Perryi seems to THE SOURCE OF BOTHRIOCEPHALUS LATUS IN JAPAN. 55 be tolerably constant. Another reason for regarding 0. Perryi as at least one of the chief sources of B. latus, is the undeniable fact that where that fish abounds, the tnpe-worm is also abundant. Toyama in the province of Ecchiu is such a place. In Yezo, especially anion o- the Ainos, tape- worms are said to be common. Considering that 0. Perryi occurs there in plenty, these tape- worms presumably belono- to the species Bothriocephalus latus. 0. Perryi is in some parts of Japan often eaten raw (Sashimi) like many other fishes. Under such circumstances, infection with Bothriocephalus is highly probable and all the more so since its larvœ in the flesh might easily be mistaken for a piece of fat, tendon or nerve. Once a friend told me, soon after his return from a short visit to Saikyö, that he had there tried the ^^ sashimi" of the fish in Cj[uestion. I warned him of what change might perhaps occur in his health within 3 weeks' time, little suspecting that he would, as he really did, present me with a fine specimen of B. latus at the end of a month. In Tokyo, where 0. Perryi is altogether scarce, its ^^sashimi" is not generally indulged in, a fxct which probably explains the comparative scarcity of B. latus in this city. The life-tenacity of Bothriocephalus -larvae seems to be of no small degree. They may retain vitality for a week outside their host and Braun even found them alive not only in weakly salted pike but also in such that had been frozen. It is however certain that the heat of thorough cooking or roasting sufiices to render them harmless. At the same time it must be admitted that ordinary methods of cooking do not exclude all chances of infection, as is clear from the way we obtain our Tœnia from beef or pork. A naturalist friend in Tokyo once discharged a B. latus, which he is inclined to regard as having been received from the flesh of 0. Perryi, which he had eaten roasted about one month previously. The fish had been 56 I. IJIMA sent from Yezo preserved in Sake-dveg ÇKasii). A few other enquiries were made of persons who had been tape-worm patients and many of them recollected having eaten 0. Perryi in some form or other. To all appearances then 0. Perryi is the chief source of B. latus. I do not wish however to emphasize this too strongly, since my examination of other river-fishes is still very incomplete. Onchorhynchus Haberi, another species of salmon common in Japan, closely resembling 0. Perryi in habits and other respects, is peculiarly open to suspicion. Nevertheless, three specimens of this fish, obtained in the Tokyo market and carefully searched by Mr. Kikuchi and myself, showed no trace of Bothriocephalus-larvre. Carassius auratus (Funa), Cyprinus carpio (A'oi), Plecoglossus alti- velis (Ayii^ and a species of Salmo (Yamahe^ were also searched but with neofative results. However the number of individuals examined was, as in the case of Onchorhynchus Haberi, too small to allow of much weio-ht being attached to the conclusions. A more complete investigation of our fresh-water fishes is very much to be desired. Earthquake Measurements of recent years especially relating ^to Vertical Motion. S. Sekiya Professor of Seismology, Imperial University. Thi« paper contain.s the records of earthquake observations made during the two years from September 1885 to September 1887. Severe shocks as well as feeble tremors are arranged in the accom- panying table, and in some shocks separate notes are added by way of fuller description. The measurements were made at two places in Tokyo ; one set at Hitotsubashi where the ground is soft and marshy, and the other set at Hongö where the soil is hardened alluvial mud. Vertical motion which forms the principal subject in this paper has not hitherto been so much studied as horizontal motion. This is on account of its comparatively rare occurrence, and when it occurs its smallness makes it of secondary importance to the more prominent horizontal movement. Reference will be made to Plates IX, XI and XXYI, Vol, I. of this Journal, in which the characteristic features of vertical motion occurring in conjunction with horizontal motion can be seen. In this country aljsolute vertical motion ^ was first measured by Mr. E. Knipping between 1878 and 1880. During that period 1 Mittheilungen der Deutschen Gesellschaft, etc. Ostasiens Vol. 17, May 1879 and Transactions of the Seismological Society of Japan Vol. I. p. 72. 58 s. SEKIYA eight observations were made ; in four cases the vertical motion reached 0.02 m.m. The laro'est value 0.56 m.m. was observed in the severe earthquake of February 22nd, 1880. In 1883 Prof. J. Milne, in conjunction with Prof. T. Gray, made experiments on artificial earthquakes. ^ A^ibrations then were caused by letting fall a heavy Aveight from various heights or exploding dynamite in holes made in the ground. His results were principally as follows. (1) In the soft ground vertical motion appears to be a free surface wave which outraces the horizontal components of motion. (2) Vertical motion commences Avith small rapid vibrations and ends with vibrations wdiich are long and slow. (3) High velocities of transit of seismic waves may be ol^tained by the observation of this component of motion. It is possibly an explanation of the pre- liminary tremors of an earthquake and the sound })henomena. In the table are given the following quantities. i. — Maximum Motion (2r) or the largest range of the displacement of the ground in each shock. 2. — Complete Period ( ^ ) of the maximum motion or the time taken to make a complete for-and-back motion of the ground. 3. — Maximum Velocity ( -y ) of the ground, or v = -^ — . i^. — Maximum Acceleration = — . r The last two quantities were calculated Ijy assuming for con- venience sake the motion of the o-found to be harmonic thouo^h it is not exactly so in actual cases. o. — Direction of the maximum horizontal motion of the ground. 6. — Duration of the earthquake, i. c., the interval of time from the commencement to the end of the disturbance. It is almost 1 Transactions of the Seismological Society Vol. VIII. EARTHQUAKE MEASUREMENTS OF RECENT YEARS. 59 impossible to measure the absolute duration of earthquakes as they usually begin with exceedingly feeble tremors and end with very slow undulations. 7. — The distance and direction of the origin of each earthquake from Tokyo, and its area. These w^ere kindly supplied by the Imperial Meteorological Observatory. Existence of vertical motion, the range and the direction of horizontal motion, etc., may be examined in reference to the position of the origin of shocks and their area. By Tremors are meant feeble shocks whose range of motion is less than one-tenth of a millimeter. Local shocks marked hocal are small earthquakes shaking only limited regions of the country usually from five to fifteen miles around. Prof. J. A. Ewing's Horizontal Pendulum and Vertical-motion Seismographs were mainly used in making these measurements. They are automatically started by the earthquake motion when it attains one-fifteenth part of a millimeter. By increasing the sensi- tiveness of the instruments the number of records may be propor- tionally increased. The records, unless otherwise stated, are those obtained at Hitotsubashi. 1885 No. Date Mux. Iloriz. Xfot on iu m.m. Compliiti» Piiiod of Max. Horiz. Motion iu Sec. Max. Velocity, in m.m. per Sec. Accel, iu m.m. per Sec. per Sec. Dirix-lion of Max. Horiz. Motion. Duration of Horiz. Motion. Max. Vertical Motion in m.m. Complete Period of Veitical Motion in Sec. Duratiou of Vertical Motion. Distnnco and direction of origin from the observa. tory iu milef. Radius of propapatiou of seismic waTea iu miles, or area of dis- turbance in sq. miles. 1 Sept. 2, 8.36. 0. pm. 0-3 0-8 1-2 9G NNB min. Bec. 2.30 0-05 0-70 mln. Bee. 0.46 46 miles due North inland. 54 m. 2 Sept. 2G, 0 30. 0. p.m. G-5 2-2 9-3 2G-6 N76°W 3. 35 0-14 0-5G 1.42 110 miles S SOW in the ocean. 146 111. 3 Sept. 28, 5 2S. 0. am. 3-8 1-7 7 25 8 EW 3.00 Trace The same in general features as No. 2. 4 Sept. 29, 8.3G.16. am. 01 1 0 3 2- SN 1.45 Local 5 Oct. 1, 1. 9. 0. pm. 1 0-7 4-5 40-5 SN 2.00 0-01 004 0.09 59 miles N 35 E on the sea shore. 90 m. 6 Oct. 3. 01 1 0-31 2- SN 0.50 Local 7 Oct. 7, 7.3445. a.m. 0-4 0-7 1-8 16-2 N1.5°E 0.18 Local 8 Oct. 9, 7.54. 0. p.m. 0-1 0-8 0-4 3-2 EW 2.00 Local 9 Oct. 11, 5.28.18. a.m. M 1 3-5 22-3 WNW 4.03 0.02 0.06 71 miles N60E in the sea. 83 m. 10 Oct. 15, 9. 2.29. a.m. 0-3 0-8 1'2 9-G \V15°N 1.10 l;s miles WSW 14,-333 sq. in. 11 Oct. 15, 8.18.43. p.m. 1-0 0-7 4-5 40-5 ESE 2.28 0.03 0.07 35 miles SSE in Tokyo Bay. 33 in. 12 Oct. 18, a.m. Tremors EW 0.20 Local 13 14 Oct. 18, 0.15. 0. p.m. Oct. 21, 1.15. 0. a.m. 0-3 0-4 90 0-9 IG 1-4 17-1 9-8 NS NW 1.00 2.02 N Cb'E 69 miles 98 miles N 73'W 61 m. 19,500 sq. m. l.j Oct. 24, 5.12.18. p.m. 0-7 0-9 2-4 IG-0 SGO^W 1.15 81 miles nearly E iu the sea. 93 m. 10 Oct. 2G, 10 41 11. p.m. 2-2 1-4 4-9 21-8 EW 3.20 llii miles S in the sea. 139 m. 17 Oct. 30, 8.31.16. p.m. 0-3 1-0 0-9 5-4 WNW 2. 30 Trace 415 miles N W in the ocean. 34,700 sq. m. 18 Not. 16, 1.53.36. p.m. Tremoi s 0.30 15 miles SSE 22 m. 19 Xov. IS. Tremors NS Local 2IJ 21 Dec. 3, 6. 1.42. a.m. De.-. 7, 1. 2. 0. n.-n. 0-2 2-1 0-G IG 1-0 4-1 10- 160 NS NNW 1.00 5.02 20 miles N N E inland. 9.< miles E13-N in the sea. 29 m. 17,120 sq. ui. 22 Doc. 19, 2.12. 0. a.m. Tremors NS 0.43 Local 23 Dee. 19, 6.28. 0. p.m. 2-8 1-7 5-2 19-4 S55°W and then tn X3U \V 1.4G 0-22 0-05 0.45 37 miles E35*H inland. 160 m. 24 Dec. 25, 1.13.30. p.m. 0-2 0-7 0-9 81 EW Trace 0.10 Local 25 Dec. 28, 10. »; :!0 p. m 3 5 1-4 7-9 35-7 EW 3.30 1 100 0-06 1.62 29 miles N N E inland 98 m. 1886. No. 26 27 28 '29 30 31 32 33 34 35 ■)6 37 38 3!) 40 U 12 13 44 15 16 17 48 49 50 Dato Jan. 4, 8.31.30. p. m. Jan. 5, 4.26 42. p. m. Jan. E>, 6.48. 0. a. m. Jan. 18, 9.15. 0. p. m. Feb. 18, 3. 0. 0. a. m. Feb. 19, 9.51.11. a. m. Feb. 22, Feb. 24, 7.34. 0. a. m. Feb. 24, 3.36.25. p. m. March 2, 5. 3.49. a. m. Marcb 3, 3.36. 0. p. m. Mareli 13, 6.25. 0. p. m. March 26, 6. 6. 0. p. ni. April 1. April 4, 1. 0. 0. a. m. April 13, 4.45. 0. a. m. April 14, April 23, 4.22.22. a. m. May 2, p. m. May 3, 0. 0. 0. p. m. May 5. May 8, 10.14. 0. p. in. May 9, 3. 0. 0. p. m. I May 11, 2.31 58. p. m. May 16, 0. 7 16 a in Max. .Complete Max. Period of . Max. ! Velocity, Horiz. Motion iu Sec. i per Sec. Horiz. Motion in m.m. 0-4 0-8 0-1 0-3 0-2 O-l 0-1 0-5 0-3 0-6 Tremors 0-7 0-1 0-1 0-4 1-2 0-8 0-7 0-] 0-2 0-2 2-1 O-.J 0-5 3-3 in m.m. 0-7 1 1-2 0-8 0-7 1 0-7 0-6 1 0-8 0-9 (l-S 1 I 0-9 0-0 0-8 09 0-7 0-9 0-S 1-2 0-9 1-8 2-5 0-3 1-2 0-9 0-3 0-4 2-6 0-9 2-4 2-4 0-4 0-3 1-3 4-2 2-8 2-7 0-4 0-9 07 8-2 1-3 1-7 8-6 Accel. iu m.m. per Sec. per Sec. 16-2 15-8 1.8 9-6 8-1 1-8 3-2 27-0 o-O 19-2 16-2 3-2 1-8 8-0 29-2 19-6 20-8 3-2 8-1 4-9 64-0 6-8 11-6 14-8 I Direction of Max. Hoiiz. Motion. Duration of Horiz. Motion. EW EW SN NS EW EW EW NW EW NNE EW WNW SSE EW EW S54E EW ENE WE ESE EW N47°E BSO'N EI CS NöO'W ni'n sec. 0. 54 0.75 2.00 0.65 0.22 0.15 0.40 1.45 1.20 2.10 0.15 2.00 0.65 0.65 2.10 4.02 0.40 1.26 0.40 2.00 0.54 2. 24 2. 50 1.25 3. 05 Max. Vertical Motion in m.m. Complete Period of Vertical Motion in Sec. Trace 0-08 Trace Trace 0-05 0-33 0-4 0-3 0-4 0-5 Duration of Vertical Motion mln. sec. Distance and direction of origin from the obseva- tory in miles. Eadiu.s of propat^atini of seismi-' waves ill miles, in- ai-ea of dis- turbauce iit sq. juileb. 0.56 0.04 0.15 0.10 0.44 0-8 1.25 ^9 miles N 6G° E in the sea. 66 miles due E on tJie sea shore. Local Local Local Local Local 29 miles, due N inland. Local 73 miles due E in the sea shore. Local 68 miles nearly E on the sea shore. 17 miles S S E in the bay. Local l;î4 miles NNE away in the sea. 3J2 miles S E iu the ocean. Local 171 miles N W on the other side of the island Local Local 29 miles NNE iuland. 32 miles WNW inland. Local Local 56 miles nenry N W iiiliiud. 61 m. 73 m. 73 m. 78 m. 73 111. 24 111. 146 m. Disturb- ance 33,U0li sq. m. 16,4(0sq.iii. 32 m. 85 Ml. 18,530 sq. m. 1886. No. Date Max. Horiz. Motion in m.m. Complete Period of Max. Horiz. Motion in Sec. Max. Velocity inm.m. per Sec. Accel, inm.m. per Sec. per Sec. Direction of Max. Horiz. Motion. Dui'iitiou of Horiz. Motion. Max. Vertical Motion in m.m. Complete Period of Vertical Motion in Sec. Duration of Vertical Motion. Distance aud direction of origin from the obscrvr- tory in milep. 29 miles NNE inland. Tîndius of propagation of seismic waves iu miles, or area of dis- turbance in eq. miles. 85 111. 51 May 18, 8.12.51. p.m. 0-5 0-9 ]-7 n-6 S50^E mln sec. 2.15 0- 1 0- 3 min. SCO. 0.36 52 53 Mav 30, 9.38 18. p.m. June 3, 3. 6.37. p.m. 0-4 0-4 0-9 M 1-4 1-1 9-8 6-1 EW (mainly) EW 1.30 1.31 Ti-ace 57 miles NNE inland. N E "r. ttiles in sea shore. 57 ni. 76 111. 54 June 6, 6. 0. 0. p m. 0-5 0-7 2-2 19-4 EW 1.06 Trace 25 miles. 55 Juno 11, 1.45.44. am. t 0-1 ||0-4 0-7 0-6 0-4 2.1 3-2 22-1 NW EW 1.02 1.15 E SE iu Tokyo Bay 15 miles. 35 ra. 56 June 12. 0-5 0-7 2-2 19-4 WE 1.15 Local 57 June. 13. 0-3 0-9 1-0 6-7 EW 0.36 Local 58 June 14, 6.25.19. p.m. 0-6 0-8 2-4 19-2 ESAV 1.10 0. 1 0- 5 0.30 In or near Tokyo. 35 m. 59 June 22 0-3 0-8 1-2 9-6 EW 0.52 Local 60 June 28. 0-2 0-8 0-8 6-4 EW 0.31 Local 61 July 2, 0.33. 6. p.m. Il 1-8 1.0 0-9 2-2 6.2 13-8 42-7 ESE WNW ESE 1.27 2.24 0. 3 0. 3 1. 3 0. 8 1.30 1.22 N E in Ihe Pacific ocean. Extensive earthquake sbakins the whole of North Japan. Tokyo on its edge. 62 July 23, 0.57. 0. a.m. ^ 0-6 09 2-1 14.7 NS 1.24 NW un miles on shore line. Estensive shock Tokyo ou its edge. 63 August 3, 2.U.40. a.m. Tremors EW 0.20 N SO miles iulnad. 30 m. 61 .August 9, 11.24. 0, am. to 110-5 0-9 0-8 1-7 2-0 11-6 16-0 NB SW NE 1.44 1.10 The same as No. 62 in general featuren. 65 August 29, 8.34.54. p.m. 0-6 0-8 2-4 19.2 ESE 1.10 C>Q Sept. 6, 0.38.53. p.m. 0-2 0-7 0-9 8-1 EW 0. 35 Local 67 Sept. 12, 8.43 22. p.m. 0-8 1-2 2-] 11-0 830°W 0.50 0. 1 0. 4 0.16 111 or near 'J'ukyo- 65 m. 68 69 Sept. 15, 3. 9 23. n 111. Sept. 16, 1. 2.57. p.m. 0-2 0-1 1-1 0-4 0-6 0-8 3-6 12-8 S3tW sw 1.13 0. 50 Ti-ace Ti-ace Local NNW ■in luiles iiilnud. Tokvo on edge. 50 m. 70 Sept. 21, 8.17. 0. p m. io-2 |lo-t 0.9 0.8 0-7 1-6 4-9 12-8 EW NS 0.53 1.31 71 S.'pt. 30. Ô-8 0.8 3.1 24-0 SN 0. 58 Local 1 8 H 0 - 1 B 8 7. Duration r,"i Max. Complete Max. Accel. Direction Duration Max. Complete Distance and Ra'lius of i Horiz. Period of Max. Velocity in m.m. of of Vertical Period of of direction of propagation j of seismic 3 No. Date Motion Horiz. Motion iu m.m. per Sec. Max. Horiz. Horiz. Motion Vertical Motion Vertical origin from the observa- wnves iu miles, or area of dis- in mm. in Sec. per Sec. per Sec. Motion. Motion. in m.m. in Sec. Motiou. tory iu miles. turbüuce iu sq. mue«. £. iiiiu. sec. miu. Bee. I 0-2 0-6 1-0 10-0 sw 24 0.1 0.4 0.15 72 Oct. 4. 2.35.25. p.m. à ■ NNE 25 miles iulaud. 40 m. ||0-3 0-9 1-0 6-7 WNW 1.2Ö 0.1 0.5 0.2i 7o 0<;t-. 22, 3.49.14. a.m. " 0-5 0-8 2-0 16-0 NS 1.05 ENE 34 miles iulaud. 70 m. 1 "■* Oct. 25, 10.11.18. p.m. 0-4 1-2 M 6-1 SE 1.40 O.l 0.5 0.34 The same in geuoral features as No. 72. 1 0-3 1.9 0-9 5«4 NS 1.15 Trace 75 Nov. 1, 6.13. 5. a.m. 0-8 2-0 16-0 NS 0.59 0.1 0.5 0.16 1 ''•^ Xov. 2, 8 21.46. p.m. 5- 0-2 1-1 0-6 3-6 WNW 1.40 0.1 0.6 0.33 N SO miles inliiud. N E in the ÖS miles j Moderaip | sized sboL-k 1 77 Pacific ocean. Dec. 4, 2. 0.39. p.m. 1 0-3 0-8 1-2 9-6 SW 1.50 0.1 0.3 1.12 N E 120 miles iu the Pacific 140 m. ocean. 78 Dec. 6, 0.40. 0. p.m. Tremors 0.29 Local 79 Dec. 8, 11.58.16. a.m. 0-3 0-9 1-0 6-7 SN 2.02 N SO miles iulaud. 35 m. 80 Dec. 11, 10.16.25. p.m. 0-4 0-9 1-3 8-5 SN 1.16 Trace S E 150 miles in ocean near the shore. 120 111. 81 Dec. 12, 10.11.55. p.m. Lost. 82 Dec. 21, 3. 7. 2. a.m. Tremors 1 ^"3 Ü--5 1-9 24-1 S50°W 0.37 0.1 0-4 0.26 83 Dec. 26, 5.48. 5. p.m. i|0-8 0-6 4-2 44-1 NNW 0.51 0.3 0.6 0.21 In or very near Tokyo. 65 111. 8i Dec. 29, n. 5 43. am. ' 0-5 0-6 2-6 27-0 SB 1.0 0.2 0.5 0.18 N '2-2 miles iulaud. 43 m. J 7-3 2-0 11-5 36-2 SSW 6.24 1.3 1.0 1.12 85 .Jan. 15, 6.52. 0. p.m. if 21 2-5 26 63 S65°W (Principal m 6.35 otion2min.) 1.8 0.9 1.38 35 miles. 2U0 m. 5-= ( ) 86 Jan. IG, 10.16.19. pm. ^ 0-6 0-7 2-7 24-3 variable 1.55 The same iu general features as No. t-S. 87 Jan. 24, 10.40.50. pm. 0-4 0-7 1-8 16-2 EW 2.06 Local Exlended 88 Jan. 2S, 3.54. 8. pm. 0.4 11-8 1'6 12-8 ENE 1.05 N W in the Paciâc oceau. aloutr the coasts Ol' North-Jaimii 89 Feb. 2, 2. 8.14. pm. U-8 00 28 22-1 S30W 1.58 178 WSW 180 m. 00 March 2, 5.33.21. p.m. Tremors Origin inland. Small eartli- qimke. 91 March 20, 11.32.53. p.m. Tremors Local 02 March 23. 0-3 0.8 1-2 96 WE 0.45 Local 1887 No. Date Max. Horiz. Motion in m.m. Complete Period of Max. Horiz. Motiou in Sec. Max. Velocity in m.m. per Sec. Accel, in m.m per Sec. per Sec. direction of Max. Hoiiz. Motiou. Duration of Horiz. Motiou. Max. Vertical Motiou in m.m. Complete Duration Period of „f Vertical ,, . , ,, . Vertical Motiou in Sec. ' Motion. ! Distaucd aud diitctiou of Kiigiu from the obseva- tory iu miles. Eiidius of propn^iitinii of seismic waves iu milci, or area of dis- turbauce iD sq. miles. 9S 1 April i, 8.46. 0. a.m. Tremors m'n sec. min. sec. N. 58 miles iulaiid. Tokyo ou edge. ijo m. '.n April 9, 11.49.54. a.m. Tremors The same as N. !'3. 05 April 16, 3.35. 0. a.m. 5 Ü-2 a l|0-4 0-7 0-8 0-9 1-6 8-1 12-8 variable SW NE 0.40 1.52 Träte 0-1 0-8 0.28 ENE ;î5 mile:>. o7 m. 96 April 20, 2.35. 0. a m. " 0-3 1 0-9 ü-i EW 0.40 Local !»7 April Ü3, G.30. 0. p.m. TrcUKirs Local 08 April 27, 9.3Ü.3.S. p.m. 0-5 2-7 0-6 1-4 SN 2.20 iu Tokyo Bay 17 miles. oo m. 09 April 29, 11.12.10. a.m. 1-4 Lost (occurred) hi;) May 2, 11.25.40. a.m. t u-i 0-6 2-1 22-1 NB 0.57 0-1 0-5 0-17 Iu or uear Tokyo. 101 May 4. 0-8 1 2-5 15-6 ENE 1.12 Local 1(12 May 5, 2.35.10. am. Ü-1 0-9 0-3 1-8 EW 0.35 Local 10:J 1 II 1 lii.j 1 0(3 May 6, 3.49.50. p.m. May 7, 7.13.12. a.m. May 9, 0. 9.14. a.m. May 17, 4 19.44. p.m. Tremors U-2 Ü-8 0-2 0-7 0-7 0-3 0-9 3'G 2-1 8-1 31-9 44.1 NS S65°W NS 1.27 2.11 0.20 Local Tlie sfime iu ^euenil features OS No. hH. but suuiUer iu ex- tent. Local N. 3n miles iniand, Tokyo ou tliR edge of the disturbauce. 40 111. 1U7 May 21, 9.46.20. p.m. 1 (>4 0-6 2-1 22-1 NS 0.35 N 20 miles, Tokyo uear origiu. b) 111. 108 May 29, 0 50.52. a.m. 1 0-6 l-B M 1-0 6-0 3-3 34.3 NS WIO-N 4.02 4.30 0-2 Lost M 2 22 N N E l:0 miles on sbore-lines, Tokyo in the middle of dis- turbed area. ExteubiTe shock. 1 00 June 1. " u-1 0-7 0-4 3-2 EW 0.51 Local ilO Juue 17, 1.41.41. a.m. ] Tremors Local 111 Juue 20, 8.38.30. a.m. if 0-2 0-8 0-9 0-8 1-4 6-4 9-8 EW variable 1.20 1.35 N E 7r) milei oil shure-line. Tokyo ou the cage of d. a. 1(10 miles. Moderately extensive eai'thquake. 112 June 21, 2. 2.35. p.m. Tremors 1.05 Local 113 Juue 22, 7.42.39. a.m. || 0-2 ii>3 0-6 0-7 1-0 1-3 10-0 11-2 SN NW 0.46 1.38 E 40 miles ou dbore-liue, Tokyo ou edge. bo m. 1887. No. 111. 11. J lie. 117 118 119 Date •Tune 30. 8. 0..35. a.m. July 2, 3.16.2'1. p.m. July i. July 11, 3. 7.42. p m. Max. Hoiiz. Motion iu m.m. Tremors o I 0-Ü i|Ü-6 0-2 S 0-2 i ].0 July 22, ;": 8.27. 0. p.m. %ii.,i Auijast 15, ]l-9 0.59.15. a.m. Complete Period of Max. Horiz. Motiou in Sec. 0-G 0-9 0-9 0-5 1-4 1-8 0 Max. Velocity iu m.m. per Sec. 3-1 2-1 0-7 31 2.3 3-3 30 Accel. iu m.m. per Sec. per Sec. 320 14-7 4-9 9-Gl lOG 12-1 9-5 Direction of Max. Horiz. Motion S20°E NNE EW SN EW N30W SN Duiatiou of Horiz. Motion. 1.30 1 . 01 0.32 0.24 1.38 o. o/ 3. O.J Max. Vertical Motiou in m.m. Trace Trace 0. 1 Complete Period of Vertical Motiou iu Sec. 0. () Duration of Vertical Motion. Distance aud dircctiou of orif»iu from the observa- tory in miles. 1.01 W -15 miles, Tokyo ou edjje. Local The same iu general feature as No. 02 aud No 61. Exteusivo I earthquake. N E 105 miles ou sbore-liue, Tokyo ou tlie edge. Gß s. SEKIYÀ Notes. For reference see corresponding numbers in the tables and in notes. No, 1. — This earthquake was moderate in its size being enclosed within the radius of 47 miles. It affords a good example of both horizontal and vertical motions. The maximum horizontal motion occurred at the third second frc^m the commencement of the shock ; at this thne the vertical motion was still exceedingly feeble although it was recoo-nazible from the beç^innino;. It reached its maximum 3 seconds later than the horizontal motion which had been then much reduced in its amplitude. The vertical motion was smaller than the horizontal motion in the ratio of 1 to 6 ; its period was quicker in the ratio of 7 to 8 and its duration of motion was shorter in the ratio of 1 to o'3. The direction of the maximum horizontal motion was NNE and SSW while the origin of the earthquake lay in due N from the observing station. No. 2. — This shock o-ave the second lar^'est motion recorded in the Table. The horizontal motion was comparatively feeble during the first 20 seconds, but gradually augmented and remained active during 80 seconds. The vertical motion appeared from the beginning but was A^ery small notwithstanding the large horizontal movement that accompanied it. The ratio of the former to the latter was 1 to 46 in amplitude, 1 to 4 in period and 1 to o'^ in duration. In this and in the f)llowino- shocks it will be observed that the direction of the local movement of the ground at the observing station and the direction of the origin of the shock from the city did not generally coincide. No. 3. — This earthquake disturbed the same portion of the country as No. 2, but with less force. The ground moved almost EAE,THQUAKE MEASUREMENTS OF RECENT YEARS. 67 equally in all directions. Moue than 120 complete waves whose periods varied from 0*7 seconds to 3 seconds were registered. Not- withstandingf the existence of the considerable horizontal motion no vertical motion appeared. No. 4. — This was one of the local shocks which frequently occnr in this and in other parts of the country. Its area of disturb- ance is often not more than a few square miles. The motion is genernlly feeble in those local shocks. No. 5. — The ratio of the vertical motion to the horizontal motion was 1 to 10 in amplitude and 1 to 13 in duration. No. 13. — More than 50 distinct waves of small amplitudes were counted. No. 14. — This extensive earthquake originated among the moun- tain di.^trict of Shinano which is one of the highest portions of the country 2,000 ft. above the sea level. There are one active and many extinct volcanoes. The seismic waves were not propagated much beyond Tokyo. No. 21. — Tokyo was in the middle of the shaken district. No. 23. — Both horizontal and vertical tremors were visible from tlie beginning ; but at the fifrh second tliere suddenly appeared a large horizontal motion (maximum). Distinct vertical waves came a few se("onds later. See Plate XT, Vol. I of this Journal. No. 25. — 'The motion commenced slowly and was not preceded by Cjuick tremors as is usually the case. The Observatory was com- parativelv near the orio-in of the disturbance. No. 33. — This was a middle sized earthcpiake in which the observing station was near its origin. The maximum horizontal 68 s. SEKIYA motion occurred 6 .seconds from tlie commencement sind the maxi- mum vertical motion 2 seconds later. Several distinct verHcnl waves of the average period of 0'3 seconds were registered. No. 39. — Trace of the vertical motion was visible though the shock was only local and the motion small. No. 41. — The whole of North Japan was disturbed by this sliock, tlie observing station being near the southern extremity (^f the dis- turbed district. No. 47. — The origin which was inland was comparatively near the city. There were hardly any vertical tremors during the first few seconds while there were considerable horizontal tremors. A decided horizontal motion occurred at the beginning of the sixth second ; more pronounced vertical motion began one and half seconds later and its maximum ocurred several seconds after. See l^late IX, Vol. \ <^f this Journal. No. 50. — This was another large earthquake in which the seismic waves were propagated from the origin some 120 miles both north and south, and 61 miles toward the west, where they were stc^pped by the mountain. On the east they reached the Pacific Ocean. The observing station was comparatively near the origin. This shock was preceded by tremors of quick period during the first eight seconds, then there suddenly appeared the maximum horizontal motion ; at this time tlie vertical motion, which was visible from the beo-innino; was vet verv small — -0-0(S m.m. with a period of 0"-l second ; afYer (î seconds it reaching the maximum, and continued for eighty-five seconds with decreasing amplitudes and with lengthening periods. The ratio of the vertical motion to that of horizontal motion was 1 to 8'3 in amplitude, 1 to l'ö in period and 1 to 2*2 in duration. 55 55 55 55 55 55 EARTHQUAKE MEASUREMENTS OF RECENT YEARS. 69 No. Öl. — This earthqnnlve, nltlioiigh it was quite extensive and its origin was comparatively near the observing station, produced small motions. The vertif^al motion was visible from the commenr-e- ment, and exhibited its maximum at the seventh second when the horizontal motion was also largest. No. 61. — This extensive shock disturbed the whole of North Japan, Tokyo being near the edge of the distur1)ed area. The pecu- liarity in this shock was the unusually large vertical motion with its slow period. The ratio of vertical motion to horizontal motion in Hongô 1:2*5 „ „ „ „ „ „ Hitotsubashi 1:6 ,, horizontal ,, in Hongö to that of „ 1:2*5 55 vertical ,, ,, ,j ,, ,, ,, ,, i:i No. 62. — Originating on the shores of the Japan Sea, the shock crossed the whole breadth of the main island. Nearer the origin the motions were very violent and somewhat destructive ; it stopped the flow of springs and shattered houses. No. 64. — In general features this shock resembled that of No. 62. It disturbed the same parts of the country and likewise caused considerable damage thoug-h in less deo-ree. No. 67. — Vertical and horizontal motions began at the same moment, but the maximum of the latter preceded that of the former by several seconds. No. 72. — Tökvö was comi^arativelv near tlie ori^'in. No. 75. — Tokyo was near the edsfe of the disturbed area. The maximum horizontal and vertical motions were simultaneous. No. 76. — Tokyo was near the outskirt of the aitected district. 70 s. SEKIYA No. 77. — It was quite a strong shock nearer its origin which was in the sea not far from the shore. The maximum vertical motion arrived several seconds before the maximum horizontal motion. No. 83. — A^ertical motion was comparatively large considering the smallness of the horizontal motion ; moreover it was clearly pro- nounced exhibiting eight distinct waves. Its maximum appeared a few seconds after the horizontal maximum. No. 8Ô. — This is one of the two largest earthquakes in 1887. The origin of the shock was in SW about 35 miles from the Observa- tory. The seismic waves propagated nearly 200 miles to the west and north-east along the Pacific sea-board. On the north-west they approached to the shores of the Japan Sea. They shook, in all, about 82,000 square miles of land area. At Hitotsubashi, after few seconds from the commencement of the shock the o-round moved suddenly 3 m.jn. toward the west. At the thirtieth second the maximum horizontal motion recorded in the Table was observed, which apparently corresponded with the maximum horizontal motion in Honml. More than sixty distinct shocks were recorded. At Hongö, the earthquake commenced with quick tremors. During the third second there appeared for the first time a vigorous horizontal motion in NW and SE (i. e. at riofht anodes to the line joining the origin of the disturbance and the Observatory) accompanied by a consideral)le vertical displacement. The maximum vertical motion occurred at the ninth second. The maxinuiin horizontal motion occurred at the thirty-third second. Decided vertical and horizontal motions simultaneously occurred at the second second. See Plate XXVI, Vol. I of this Journal. EARTHQUAKE MEASUKEMEN'Tö OF RECENT YEARS. 71 The ratio of the max. h. in. in Iltjngo to tli:ir of liitotaiibasihi 1:8 J) 5) 35 35 5? ^* ^^" 33 33 3» 33 33 33 io:io ,, ,, ,, ,, ,, V. m. to mnx. li. ni. in Hitotsuhashi 1:12 J? 33 33 33 33 33 33 53 33 33 33 33 IlOHgO 1:0 No. 95. — Tokyo was on tlic edo;e of the disturbed area. 'J'here were a ï^^w distinct \ ertieal wa\'e.s at Hitotsubashi. Summary of Results. The results obtained iVom n study of tlic ])reeedin;jf lable in:iy be summarised as folh)ws. In most of the earthquakes vertical motion did not aj^pcar, tliat is, the ground moved entirely in a horizontal plane. This was on account of the origin of the disturbance being far away ïvoxn the observing station. In the above table, if we reject the slight shocks marked trcvior, and the doubtfid cases, we see that out of 100 earth- quakes vertical moti(3n occurred in 28 only, /. 6'., once in 3"() shocks. Taking a^•erages of all the quantities in^•olved in these 28 cases, we find : — Maximum horizontal motion 1*2 m.m. Complete period of maximum horizontal motion .... 1 sec. Duration of horizontal motion 12-1 sec. Maximum vertical motion 0'18 m.ni. Complete period of maximum \ertical motion 0*56 sec. Duration of vertical motion ■12 sec. Treating similarly the horizontal motions in 05 earthquakes recorded at Hitotsubashi (soft soil), we find : — Maximum horizontal motion 0*73 m.m. Complete period of maximum horizontal motion ... 0*76 sec. Duration of horizontal motion 117 sec. • 72 s. SEKIYA Also from alike treatment of the horizontal motions in 18 earth- quakes recorded at Hongö (hard ground), we get : — Maximum horizontal motion 0*37 m.m. Complete period of maximum horizontal motion ... 0*76 «ec. Duration of horizontal motion 74 sec. The values of the horizontal motion given in the first tîibulated set of averao'es are laroer than those in the second and third. The reason is, clearly, that shocks containing vertical motion are generally larger than those witlKnit, so that the sec(jnd and third sets, Avhich include shocks both with and without vertical motion, naturally give smaller averages. Although the second and the third sets, referring as they do to ditferent shocks, are not strictly comparable, they nevertheless show in a general way that in hard ground the motion is smaller, the period quicker, and the duration shorter, than in soft soil. Their ratios are 1 to 2, 1 to 1'3 and 1 to 1'5 respectively. In the above sets of averages the records of the somewhat des- tructive earthquake of January 15th, 1887, were not included as that Avas much larger than the ordinary shocks we are dealing with. For the sake of comparison the characteristics of that shock as registered at Hitotsubashi are now" given. Maximum horizontal motion 21 ni.m. Complete period of maximum horizontal motion ... 2* 5 sec. Duration of horizontal motion G min. o-l sec. (Principal motion 2 min.) Maximum vertical motion 1*8 min. Complete period of maximum vertical motion 0*9 sec. Duration of vertical motion 98 sec. For details see No. So. In the earthquake of October loth, 1884, EARTHQUAKE MEASUREMENTS OF KECEN'T YEARS. 7S Ϋ inaxiinuiii horizontal motion of 42 ni.ni. ^vith n complete period of 2 seconds was recorded at the above named place. When vertical motion occinTed it was invariably smaller than the horizontal motion as is obvious from the first set of averajres. The a\erage ratio of the two components of the motion was 1 to 6, or the former was only one-sixth of the lattei-. The period of the vertical motion was shorter than that of the horizontal motion. Their a^'era2■e ratio bein"- 1 to 1"8. That is ■when the ground made one to-and-fro motion it also perf)rmed during the same time nearly two complete up-and-down oscillations^ Tn all the preceding tables the periods of the maximum vertical motion are given, l)ut the periods in the other parts of the disturbances "were much shorter. i'bey varied from 0*2 seconds to 0*5 seconds. Exceedinolv feeble tremors of a few seconds duration îrenerallv ]>receded the principnl motions as in the case of the horizontal motions. The duration of the vertical motion was much shorter than that of the horizontal motion. It occurred invariablv dnrina* the early stages of the earthquake and generally ended before the horizontal components. The average ratio of the two durations was 1 to 3^ or the horizontal motion continued three times lon^'er than the vertical. Tlie ^ertical motion almost invariably appeared when the horizontal motion had reached 1 m.m. which was more than the average amplitude in ordinary earthcpiakes. Out of 100 shocks there were 18 cases in which the ground moved more than 1 m.m. Out of these 18 earthquakes vertical motion occurred in 11 cases, or 78- per cent, and did not appear in the remaining 4 cases. But on the other hand vertical motion also appeared in certain cases when the horizontal motion was less than 1 m.m. Out of the 74 s. SEKIYA 28 shocks already specified as showing vertical motion, 14 showed a horizontal motion of more than 1 m.m., while in the other 14 cases the horizontal motion was less than 1 m.m. Again when we analyze the 14 cases which had vertical motion with less than 1 m.m., we see that in 9 shocks tlie o])serving station was in, ov comparatively near to the centre of the disturbed districts, and in the remaininsf 5 cases it Avas at a considerable distance from the origin. In other words vertical motion generally appeared when the orisfin of the disturbance was near tlie observin"* station. In such a case the vertical motion might have come directly through the earth-crust from the origin, and not in the form of free surface waves. It must be, however, noted that there Avere 4 cases in which the observing station w-as comparatively near the centre of dis- tiu'bance, Ijut in which vertical motion did not (jccur. Tbey were all small shocks, with maximum motions less than 1 m.m. In earthquakes showing l)oth horizontal and vertical motions, feeble but cpiick-period tremors of both types simultaneously preceded the principal movements. The more decided and pronounced motion usually appeared first in the horizontal component, and then came the large vertical movements. For the diagrams oï motion see Plate IX and l^late XI, Vol. I of this Journal. Tremors, or rpiick-period minor waves generally j)recede earth- quake proper and their probable connection with sound phenomena has been discussed in tliis country by Profs. Ewing^, Milne'^ and Knott.^ The vertical motions Avhich have just been considered possess in great measure the chnractcristics of these minor tremors. It is a well-known fact that the movement of the grcjund at the time of earthquakes is very complex, and that the ground moves in 1 Memoirs on Earthquake Measurements, p. 11. 2 Earthquake Notes — Sound Phenomena, Trans : Seis : Soc : Vol. XII. 3 Earthquakes and Earthquake Sound, etc., to be published in Trans: Seis: Soc: Vol. XIII. EARTHQUAKE MEASUREMENTS OP RECENT YEARS. "5 alj aziinutUs during a prolong-ed .-iliakiiig. Fn the earthquakes discussed ill the present paper tlie direction and the distance of the n.— Slate. «.— Gray- wacke-sandstone. o. — Common liornstoue. p. — Upper schal-itein. p' — Diabase sheets. ?.'— Fusnlina limestone. . Mr. Ötsuka, a graduate of our university, took up, at the writer's request, the district in question for the subject of his thesis, m order to explore thoroughly its geognostic condition. As the result of his study in field and laboratory, he has eventually succeeded in recogniz- ing a certain regular order of sequence of strata younger than the so- called crvstalline schists. From various observations taken as far as possible at différent points of the Chichibu mountains by the Avriter and Mr. Otsuka, and from the data ascertained by Mr. Otsiika,^) with which the writer's scheme is also incorporated, the latter chiefiy referring to the ? pre- Carboniferous and Sambagawan series, we were eventually enabled to recognize the following stratigraphie order of rock series, counting from the very lowest that is ever developed in this region ; it must at the same time be remarked that the names of fossils here given should Ije taken with reserve, and considered as only of an approximate value, as we can not pretend to any thing more than this at the present state of our knowledge : — (Compare Diagrain I. page 78.) ' 1 Xormal sericite-schist together with the piedmontite-schist in its upper horizon. 2 Green, and black spotted schist. 3 Lamellar epidote-sericite-gneiss. •i Amphibole-pyroxene schist, })yroxene-amphibole schist, pyroxene-epidote, and amphibole-epidote schist, together with serpentines, gabbros and the gabbro-diorites. ce C3 03 1) On the Geology of the Mountain-districts in Chichibu and Kanra, 1887. (Manuscript^. 80 B. KOTO o u CM • r-l o ce Q) U Pa 8 9 Pi o u be o O m Ol bx) o Ol 5 Red, and white, platy qnartzite. 6 Adinole-slate overlying the preceding. 7 The lower schalstein, intercalated with the adinole-«late, and also the limestone with crinoidal stems and corals. Graywacke-sandstone and slate, intercalated in their loAver horizon with the adinole slate. Siliceous slate or common hornstone overlaid by another series of the adinole-slate. The upper schalstein and the siliceous Iladlolarian slate. Diabase- sheet sometimes occurs in this horizon. The FusuUna-\m\e,^ione.j etc. The Jurassic subgroup composed of the medium -grained, grayish sandstones and shales. («) Ihe above-mentioned sandstones afford rich remains of Cijrena, Mdania, Mijacitcs ; while (6) in shales we have found the remains of plants, viz., TJnjrsoptcris, Nilsonicif Dichsoiiia, Podozavtitcs, rccoptcris and Zamitcs. 14 The Cretaceous subgroup consists of thick sandstones and shales with Tri