Nov. 4, 1880] 



NA TURE 



19 



on the 4tli, and this is the highest ever ob->erved in ihe month 

 of October ; on the 51I1, 76° was registered, which corresponds 

 with the temperature observed on October 4, 1S59. The returns 

 of the Meteorological Office show that So° was observed on the 

 4th in London and at Cambridge, whilst IT was registered at 

 several stations in the east of England and in the Midland 

 Counties. 



It is difficult to make any satisfactory comparison with pre- 

 vious records, except at one or two places, but these tend to 

 show that so high a temperature at this season does not occur 

 mor.- than about twice in a century. Chas. Harding 



T 



VOLCANOES OF JAPAN 



HE last number (vol. ix. part 2) of the Transactions ol the 

 Seismological Society of Japan is wholly occupied by a 

 IJaper of Prof. Mdne's, on Japanese volcanoes, which is the 

 longest contribution that has yet appeared in the Society's 

 Transactions. The paper is partly historical and partly scientific, 

 and contains, so far as the writer has been able to collect, refer- 

 ences to everything that is known on the subject. Very much 

 comes from his own observations, for he has travelled over the 

 greater part of Japan, and has ascended many of the volcanoes. 

 The paper also contains an epitome of some thirty or forty 

 works in Japanese. On the whole, it is a systematic account 

 of material which has been accumulating for the last eleven 

 years. 



The following are the more important conclusions which Prof. 

 Milne has formulated in the paper : — 



I. Number of Volcanoes. — As Japan has not yet been com- 

 pletely explored, and, moreover, as there is considerable diffi- 

 culty in defining the kind of mountain to be regarded as a 

 volcano, it is impossible to give an absolute statement as to the 

 number of volcanoes in the country. If under the term volcano 

 be included all mountains which have been in a state of eruption 

 within the hist irioal period, those which have a true volcanic 

 form, together with those which still exhibit on their flanks 

 matter ejected from a crater, we may conclude tliat there are at 

 least ICO such mountains in the Japanese Empire. If to this 

 list be added the ruins and basalt wrecks of volcanic cones, the 

 number would be considerably increased. These mountains 

 are distributed as follows : — 



Northern Region.. 



Central Region 



Southern Region.. 



f Kuriles 23 



\ Vezo 2S 



( Northern main island | 

 - Central „ „ \ 35 

 \ Oshima group j 



Southern main island .,. I 



Kiushiu "I 



Southern islands J "^ 



Total 



100 



Of this number about 48 are still active, or have been so during 



tile historical period. These active volcanoes are distributed as 



follows :— 



,T .1 u • ( Kuriles 16 ) 



Nortliern Region., i ,. . 27 



° ( \ezo II ( ' 



Central Region 12 



Southern Region 9 



Total 



48 



From this it will be seen tha' volcanic activity in Japan decreases 

 from the north towards the south. 



2. Niimlier 0/ Eruptions. — Altogether about 232 eruptions 

 have been recorded, and of these tlie greater number took place 

 in the southern districts. This may perhaps be accounted for 

 by the fact that Japanese civilisation advanced frjm the south. 

 In consequence of this, records were made of various phenomena 

 in the south when the northern districts were still unknown and 

 unexplored regions. The greater number of eruptions took 

 place in February and April. Comparing the frequency of 

 eruptions in the different seasons, the volcanoes of Japan appear 

 to have followed the same law as the earthquakes, a greater 

 number having taken place during the cold months. This winter 

 frequency of volcanic eruptions may possibly be accounted for 

 in the same manner that Dr. Knott accounted for the winter 

 frequency of earthquakes. During the winter months the 

 average barometric gradient across Jajian is steeper than in 



summer. This, coupled with the piling up of snow in Ihe 

 northern regions, gives rise to long-continued stresses, in conse- 

 quence of which certain portions of the earth's crust are more 

 prepared to give way during the winter months than they are in 

 summer. 



3. Position and Relative Age of Japanese Volcanoes. — The 

 youngest of the Japanese volcanoes appear to be those which 

 exist as, or on, small islands. On the islands in the Kuriles, in 

 the Oshima group, and in the Satsuma sea, many of the vol- 

 canoes are yet young and vigorous. Moreover, many of these 

 islands have been formed during the historical period. The 

 island-f.jrming period in the Satsuma sea, for example, was 

 about the year 1 780. 



The volcanoes of Japan form a long chain running from N.E. 

 towards S.W. ; but a closer examination of the distribution 

 of the volcanic vents shows that there are probably four 

 lines : — 



(a) The N.E. -S.W. line running from Kamchatka through 

 the Kuriles and Northern Yezo. 



{b) 'i"he curved line following the backbone of the main 

 island, and terminating on the western side of the Yezo anti- 

 clinal. 



{c) The N.N.W.-S.S.E. line of the Oshima group. This 

 line, coming from the Ladrones, passes through Oshima and 

 Fujisan parallel to and near to the line of a supposed fault. 

 Here it intersects the main line running through the main 

 island. Volcanic vents are here veiy numerous. As the main 

 inland line is intersected, while the Oshima line is the inter- 

 sector, it may be argued that the Oshima-Fujisan line of 

 volcanoes are younger than many of those on the main island 

 line. 



[d) The Satsuma line, coming from the Philippines through 

 Sakurajima and culminating in the famous Mount Aso, which is 

 the nucleus of Kiushiu. 



4. Lithological and Chemical Character of Lavas. — Although 

 Prof. Milne has made an extensive collection of the volcanic 

 rocks of Japan, the opportunity for examining them has not yet 

 presented itself, and therefore he can only speak of them in 

 general terms. They are at present being carefully studied by 

 the officers of the Geological Survey. The rocks in his posses- 

 sion are chiefly andesites. Those containing augite, like the 

 rocks of Fujisan, closely approximate to basalts. True basalt is, 

 however, rare. Another common rock is hornblende andesite, 

 some of which contains free quartz. Quartz trachytes occur in 

 the north of Japan, The following table shows the percentages 

 of silica, and ferrous and ferric oxide, contained in the rocks of 

 ten volcanoes : — ■ 



Locality. SiO, FeO Fe.Oj 



1. Norokura 61 72 ,, I '35 ,,. 3'50 



2. Misake 59"97 •■• I'^l ■■■ 3-86 



3. Kusatsu 6i'49 •■ 3'3o ••• 4'35 



4. Amagi (Hakone) 65'34 ••• 2'45 ... S'og 



5. Kom'agadake 56-27 ., 2-19 ,,. 6-69 



6. Moriyoshi 59-17 .., 2-65 .,. 4-15 



_ „, , , ( 60-64. .,, 3-81 .,, 3-14 



S. Hakone (Tonosawa) 48-97 ,,, 4-02 ,,, 4-81 



9 Fujisan 49-00 .,, 5-1 ,..6-06 



10. Oshima 52-00 ,,,i3-7o(?) 



One feature exhibited by the table is that the rocks of Oshima, 

 Fujisan, and Tonosawa are basic, while those like Chokaisan 

 and Moriyoshiyama belonging to the line of volcanoes of the 

 main island, are relatively acidic. More extended observations of 

 this description may show that different lines of volcanoes have 

 llirown out different lavas, or that the lava; of different constitu- 

 tion are of different ages. 



5. Magnetic Character of Rocks. — In a study of the soils in the 

 neighbourhood of Tokio, Mr. E. Kinch refers specially to the 

 magnetite they contain. A great portion of this comes from the 

 disintegration of volcanic rocks. Many of the Japanese lavas 

 have a distinct effect upon a compass needle, and many of the 

 black lavas from the crater of Fujisan will easily turn the needle 

 of an ordinary compass through 360°. Many of the pieces of 

 lava are not only magnetic but polar. Dr. Naumann found a 

 block of augite trachyte on the top of Moriyoshiyama which 

 woulil deflect the needle of a compass through 155°. The 

 most curious observation made by this investigator was that the 

 magnetic declination near (iaujusan has during the last eighty 

 years (when it was about 14'30'E.) decreased 19°, being now 

 about 5° W. As we recede from this mountain the amount of 



