NA TURE 



{^Nov. 4, 1886 



change is less. Assiiminf; this result to be correct, it would seem 

 justifiable to look to Gaujusan as connected with these local 

 changes. Some of the volcanoes in the Kuriles are said to exert 

 a marked influence upon the compasses of ships. When a vessel 

 is lying near certain mountains, as, for instance, in Bear Bay at 

 the north end of Iturup, a dist.ant mountain will have a very 

 different bearing to that which is indicated by the same compass 

 when the vessel is a short distance outside 15ear Bay. In both 

 cases the ship may be lying in the same direction, and the direc- 

 tion of observation is practically along the same line. This 

 leads Prof. Milne to urge, as he has already done, that a mag- 

 netic observatory should be placed on or near one of the nine 

 active volcanoes of Japan. Changes in volcanic activity are 

 probably accompanied by local changes in the magnetic effects 

 produced by subterranean volcanic magmas. These changes 

 may be due to alterations in position, alterations in chemical 

 constitution, and changes due to the acquisition or loss of heat. 

 If such is the case, he argues, the records of a magnetic obser- 

 vatory would lead up to a knowledge of the changes taking 

 place beneath the ground. When it is remembered that vol- 

 canoes like Oshima (Vries Island), where it seems probable 

 that there may be local and rapid changes in magnetic variation 

 taking place, lie in the track of so many vessels, the proposed 

 investigation has a practical as well as a scientific aspect. An 

 investigation of earth-currents at and near volcanoes might be 

 added to the magnetic investigations. 



6. Inti'iisily of Eruptions. — It appears from the accounts of 

 eruptions which are given in the paper that the intensity of 

 volcanic action in Japan has been as great as in any other part 

 of the world. One period of unusu.al activity was between 

 the years 1780 and i8cx3, a time when there was great activity 

 elsewhere in the globe. It was during this period that part of 

 Mount Unsen was blown up, and from 27,000 to 53,000 persons 

 (according to different accounts) perished, that many islands 

 were formed in the Satsuma sea, that Sakurajima threw out so 

 much pumice material that it was possible to walk a distance of 

 23 miles upon the floating debris in the sea, and that Asama 

 ejected so many blocks of stone — one of which is said to have 

 been 42 feet in diameter — and a lava-stream 68 kilometres in 

 length. 



7. TIte Form of Volcanoes. — The regular so-called conical form 

 is very noticeable in many of the Japanese mountains, especially 

 perhaps in those of recent origin. Outlines of these volcanoes, 

 as exhibited either by sketches or photographs, show curvatures 

 which are similar to each other. From a collection of photo- 

 graphs Prof. Milne traced the profiles of a number of important 

 mountains in Japan. These are reproduced in the paper (see 

 Fig. I). From an examination of these figures he found that the 



Fig. I. — Outl.ne of Fujiyama, from a photograph. This may be taken as 

 typical of many Japanese volcanDes. 



curvature of a typical volcano was logarithmic, or, in other 

 words, the form of such a mountain was such as might be pro- 

 duced by the revolution of a logarithmic curve round its 

 asymptote. In his original paper on the subject he said that the 

 form agreed with that which would be produced by the piling up 

 of loo^e material. He ought to have said it was the form assumed 

 by a self-supporting mass of coherent material. Mr. George F. 

 Becker {^American Journal of Science, October 1SS5) continues 

 these observations by an analytical investigation of the conditions 

 of such equilibrium. If the height of a column is a, its radius y, 

 the distance of any horizontal plane from the base .r, the specific 

 gravity of the material p, and the co-efficient of resistance to 

 crushing at the elastic limit k, then the equation of the curve, 

 which by its revolution about the x axis will generate the finite 

 unloaded column of the " least variable resistance " is — 



This latter quantity is of course different for different materials. 

 It can be expressed in terms of -v .ind_>' — 



2k _ _ .)' ^ 



P ~ (tan-' d - l)~* 



d being the angle which the tangent at any point makes with 

 the .X axis. The value of c can be obtained from photographs or 

 drawings of a mountain, while p may be obtained from pendulum 

 experiments or from specimens of volcanic material. With these 

 data we can determine the modulus of resistance at the elastic 

 limit of the materials which compose a mountain on a large scale 

 for many constituents of the earth's crust. Mr. Becker concludes 

 his observations by remarking that a study of the form and 



dimensions of lunar volcanoes would lead to values of-, from 



P 

 whence we might approximately determine whether the lunar 

 lava is similar to that of terrestrial origin. In the table which 

 follows, Prof. Milne has followed out Mr. Becker's suggestion, 

 and calculated the modulus of resistance to crushing at the 

 elastic limit in pounds per square foot for a number of Japanese 



ik 

 mountains. The different values for — for the same mountam 



P 

 is in great measure due to the absence of an accurate scale for 

 the various photographs which had to be investigated. Another 

 difficulty was obtaining a value for r, or the density of the 

 mountain. Prof. Mendenhall, who made a number of experi- 

 ments with pendulums on the summit of Fujisan, says the rocks 

 of that mountain have a density of 1 75. This is when they 

 have air in their pores. As powder the density becomes 2'5. 

 Wada gives the specific gravity of the rock on Fitjisan as 2'6. 



Fig. 2. — Theoretical Mo 



Assuming the density of the earth at 5 '67, then tlie density of 

 Fujisan, as determined by Prof. Mendenhall's experiments, is 

 2 ■08. In the following table the density of the materials of all 

 the mountains mentioned is taken at 2"5. 



Height zk k lbs. per 



'"'"' p p ToT 



Fujisan 12,441 4200 ... ... Photograph 



5000 ... 



4240 ... ... ,, 



■■■ 3500 



[ 5420 \ 



■■• 15450/ 



■■■ 5440 



i 3945 I 



■'■ I4133/ 



4430 ... ... Surveyed section 



... 3640 ... 

 Average for 



Fujisan 4490 2245 350,220 ,, ,, 



Iwakisan 5260 2360 iiSo 174,080 Photograph 



Nantaisan 3800' 2000 1000 156,000 ,, 



Alaid 7773{2;2o}'°7S '63,168 



'^"''^'•''"'° 2745 '310 1655 ,02, iSo Surveyed section 



I3IO J •^■^ ■' 



Comparing the results given in the above table with the 



numbers given in the next section, which are based on experi- 



' This is the height above Lake Chuzenji. 



