Rer. 0. Finher — Convection Currents in Earth's Inferior. 9 



to overlap the edge of a fixed sheet, the stream pushes the former 

 onwards and caixses an overthrust. But if an overthrust in the rocks 

 is produced in a simihxr manner, we must postulate a liquid 

 substratum, whose surface is locally moving horizontally ; and if the 

 earth is a cooling body, and if there is liquid matter in the interior, 

 this kind of movement must needs occiir. The reason of this is 

 that the mode of cooling of a mass of liquid is by convection, the 

 currents ascending in some places and descending in others, while 

 the liquid flows horizontally at its surface from the ascending to the 

 descending portions. 



1 am aware that at the present day (at least in Britain) a belief 

 in the liquidity of the earth's interior is heresy. The orthodox 

 opinion appears to be that it is neither liquid nor solid, but both 

 at once, and so we run with the hare and hunt with the hounds. 

 But such a state of affairs would not admit of convection currents. 



I will briefly mention some phenomena which appear to be 

 explicable on the hypothesis of a liquid substratum holding water- 

 gas in solution and cooling by convection — 



(1) Obviously volcanic action stands first, where Ave see frequent 

 ebullitions of fresh lava, varying in character from one eruption to 

 another, and building up great mountains as if supplied from an 

 inexhaustible reservoir below. 



(2) The same agency, viz. horizontal flow, which, as above 

 suggested, might cause compression to be satisfied with an overthrust, 

 would also account for other forms of compression, such as are more 

 usually characteristic of mountain ranges. 



As to whether the friction of a flow of the substratum would be 

 adequate to supply the force which has compressed the crust in 

 mountain ranges, we may again refer to the analogy of ice. Kansen 

 found that in the Polar regions the thickness of the ice under direct 

 freezing does not usually exceed 4 metres. But it becomes very 

 much thicker by the piling up of broken ice-sheets, driven together 

 and mounting one above another, and he says that the pressure is 

 largely dependent upon the tidal current, especially at the margin 

 of the ice-field.' Now the ratio of the densities of the crust and 

 substratum is probably very close to that of ice and water, so that 

 a similar action of currents in the two cases would be likely to 

 produce similar effects of compression. 



(3) Variations of level, especially such as have been described by 

 Professor Hull and Dr. Spencer, appear to be most readily accounted 

 for hy convection currents, the surface rising over upward and 

 sinking over downward currents, as I have endeavoured to explain 

 elsewhere. - 



(4) Another phenomenon which may possibly be due to convection 

 currents is the irregularity in the small changes of latitude which 

 arise from slight variations in the position of the axis of rotation 

 within the earth. Mathematicians tell us that the anomalies of these 

 movements can be explained by elasticity. But it seems that, although 

 elasticity might modify the period of a regular movement of the pole of 



' Kafiirc, I'^elirunry 11, 1897. - Ainf-ricnn Journal of Science, lOOG. 



