32 



NATURE 



\Nov. 14, 1889 



aspects ; though, as I conceive, the geographical, climatological, 

 and geological elements in the problem are not to be arbitrarily 

 distinguished. Now I am far away from all books of reference, 

 and it is of course essential that I make myself acquainted with 

 what has already been done in these subjects, and I venture to 

 ask for any hints as to the bibliography of them. Can you tell me if 

 anyone has done for geology what Hirsch, of Berlin, has done fo-- 

 geography (in his work on the distribution of disease) ? Is there 

 any authority on the chemistry of soils, and what I roughly call 

 their physiology and pathology, their structural and functional 

 changes under influences — climate notably — and their own in- 

 trinsic, and the deeper geological interactions ? 



A. Ernest Roberts. 

 Meywar Bheel Corps, Kherwara, Central India, 

 September 9. 



The Earthquake of Tokio, April 18, 1889. 



Dr. von Rebeur-Paschwitz's letter, which appeared in 

 Nature, vol. xl. p. 294, is of special interest to us in Japan, 

 countenancing as it does the conjecture that the very peculiar 

 earthquake felt and registered here on April 18 was the result of 

 a disturbance of unusual magnitude. It was my good fortune 

 on the day in question to be engaged in conversation with Prof. 

 Sekiya in the Seismological Laboratory at the very instant the 

 earthquake occurred. We at once rushed to the room where 

 the self-recording instruments lay, and there, for the first time in 

 our experience, had the delight of viewing the pointers mark 

 their sinuous curves on the revolving plates and cylinders. At 

 first sight it seemed as if the pointers had gone mad, tracing out 

 sinuosities of amplitudes five or six times greater than the 

 greatest that had ever before been recorded in Tokio. There 

 was not much sensation of an earthquake ; indeed, after the 

 first slight tremor that attracted our attention, we felt nothing 

 at all, although in the irregular oscillations of the seismograph 

 pointers we had evidence enough that an earthquake was 

 passing. Very few in Tokio were aware that there had been 

 an earthquake till they read the report of it in the next day's 

 papers. Thus the motion, though large, was too slow to cause 

 any of the usual sensations that accompany earthquakes, and 

 suggested a distant origin and a large disturbance, with a con- 

 sequent wide extension of seismic effect. Excepting the slight 

 tremors recorded at Potsdam and Wilhelmshaven, there has 

 been, so far, no evidence of any such far-reaching action. 



My object in writing this note, however, is to correct an error 

 of calculation which Dr. von Rebeur-Paschwitz has unwittingly 

 made. He has assumed that Tokio standard time is mean local 

 time. On the contrary, the standard time for all Japan is the 

 mean solar time for longitude 135° E., — that is, nine hours in 

 advance of Greenwich mean time. Hence, instead of the Tokio 

 earthquake having preceded the German disturbance by ih. 

 4'3m. it preceded it by only 4Sm. This correction increases the 

 velocity of transmission to 3060 metres per second. We must 

 assume, then, either that large disturbances in the heart of the 

 earth travel with exceptionally high speeds, or that the origin of 

 the disturbance was a considerable distance from Tokio. The 

 latter assumption seems sufficiently satisfactory, if in other 

 respects Dr. von Rebeur-Paschwitz's views meet with approval. 



Cargill G. Knott. 



Imperial University, Tokio, Japan, September 25. 



A Brilliant Meteor. 



Yesterday evening, November 4, at 7.55 p.m., I was for- 

 tunate enough to observe a very brilliant meteor. It became 

 visible almost exactly at the zenith, or a little west of it, and 

 moved, as nearly as I could judge, due east, magnetic ; it re- 

 mained visible for about from one to two seconds, disappearing, 

 finally, rather low down on the eastern horizon. For the first 

 half of its journey it was of a dazzling white brightness, and then 

 it suddenly became a dull red spark. The light emitted from it 

 when brightest reminded me of the light from an arc lamp, and 

 was very much brighter than any of the fixed stars. 



As it was so short a time in view, and there were no stars 

 visible, I could only approximately estimate its point of appear- 

 ance and path. There were a few clouds about, mostly in the 

 west, and the moon was behind them. Paul A. Cobbold. 



Warwick School, November 5. 



ON THE HARDENING AND TEMPERING OF 

 STEEL} 



II. 



'T^HE following considerations appear to have guided 

 ■■■ Osmond in beginning his investigations (see ante, 

 p. 16). Bearing in mind the fact that molecular change in a 

 body is always accompanied by evolution or absorption of 

 heat, which is, indeed, the surest indication of the occur- 

 rence of molecular change, he studied with the aid of a 

 chronograph what takes place during the slow cooling 

 and the slow heating of masses of iron or steel, using, as 

 a thermometer to measure the temperature of the mass, a 

 thermo-electric couple of platinum and of platinum con- 

 taining 10 per cent, of rhodium, converting the indica- 

 tions of the galvanometer into temperatures by Tait's 

 formulae. 



Fig. 5. 



--© 



Fig. 6. 



Figs. 5and6show the actual mode of conducting the experiments. F(Fig. 5) is 

 apiece of steel into which a platinum and platinum-rhodium couple, t, i\ 

 is fixed. It is inclosed in a glazed porcelain tube and heated to bright 

 redness in the furnace, s (Fig. 6). This tube, t, may be filled with any 

 gaseous atmosphere, c is a bulb filled with chloride of calcium. The 

 metal under examination is slowly cooled down. The wires from the 

 thermo-couple pass to the galvanometer, g. The rate of cooling of the 

 mass is indicated by the movement of a spot of light from the galvano- 

 meter mirror at }n, on the screen, R, and is recorded by a chronograph. 

 The source of light is shown at l ; m is a reflector. 



In the next diagram (Fig. 7) temperatures through which 

 a slowly-cooling mass of iron or steel passes, are arranged 

 along the horizontal line, and the intervals of time during 

 which the mass falls through a definite number (6'6) of 

 degrees of temperature are shown vertically by ordinates. 

 See what happens while a mass of electro-deposited iron 

 (shown by a dotted line), which is as pure as any iron can 

 be, slowly cools down. From 2000° to 870° it falls uni- 

 formly at the rate of about 2 "2° a second, and the intervals 

 of temperature are plotted as dots at the middle of the 

 successive points of the intervals. When the temperature 

 falls down to 858°, there is a sudden arrest in the fall of 

 temperature, the indicating spot of light, instead of falling 

 at a uniform rate of about 2° a second, suddenly takes 26 



' A Lecture delivered on September 13, by Prof. W. C. Roberts- Austen, 

 F.R.S., before the members of the British Association. Continued from 

 p. 16. 



