52 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



that it would not be possible to predict either one of these ele- 

 ments by means of observations made only at one spot. In 

 the Caucasus he finds that magnets have no loss of power be- 

 fore an earthquake, and hence can not be used as a means of 

 warning. Shamaka and its Earthquakes, Tiflis, 1872. 



TERRESTRIAL MAGNETISM AND THE SHRINKING OF THE EARTH. 



Mr. Howorth, in a short note, very suggestive in many 

 other respects, speaking of the foci of upheaval and depres- 

 sion on the surface of the globe, states that we can not resist 

 the conclusion that the earth is stretching itself in the di- 

 rection of its shortest axis ; that its periphery is being thrust 

 out in the direction of the poles. Now as, on the whole, the 

 earth is absolutely shrinking, and local uprisings are due to 

 lateral pressure caused by a subsiding area, it becomes inter- 

 esting to inquire what kind of a strain upon the earth would 

 result in squeezing it out in the direction of the poles. The 

 answer to this, in his opinion, must be that a strain is being 

 applied in the way of a stricture about the world's equato- 

 rial region. This would, he thinks, also have another effect : 

 it would induce magnetism in the earth, and that magnetism 

 would have its poles in the regions of upheaval, as in fact 

 they are. The magnetic poles are strictly, so far as our evi- 

 dence goes, in the very foci of upheaval of the circumpolar 

 regions. 12 A, 1874, IX., 202. 



THE RATE OF CONTRACTION OF THE EARTIl's DIAMETER. 



Mallet has recently made an interesting addition to his 

 paper on volcanic energy, in attempting to calculate, on the 

 basis of certain allowable suppositions, the amount, in vol- 

 ume, of the solid shell of our earth which must be crushed 

 annually, in order to allow the shell to follow down after 

 the more rapidly contracting nucleus. He shows that the 

 amount of crushed and extruded rock necessary for the 

 supply of heat for the support of existing volcanic action 

 can be supplied by that extruded from a shell between six 

 hundred and eight hundred miles thick, and that the volume 

 of material, heated or molten, annually blown out from all 

 existing volcanic cones, could be supplied by the extruded 

 matter from a shell of between two hundred and four hun- 

 dred miles in thickness. On data which seem tolerably re- 



