[ 267 ] 



No. 33. 



SOME EXPERIMENTS ON THE CONVECTION OP HEAT IN 

 VERTICAL WATER COLUMNS. 



By H. H. POOLE, Sc.D. 



[Kead November 27. Priuted December 7, 1923.] 



Introduction. 



In view of the importance of radioactive laeating in modern geological theories, 

 considerable interest attaches to the question as to how far the estimated 

 upward flow of heat through the earth's crust accords with that to be expected 

 from the known radioactivities of the surface materials. The figure obtained 

 by multiplying the thennal conductivitj^ of dry rock by the average temperature 

 gradient just below the surface is surprisingly small. The question has 

 recently been discussed by the writer elsewhere.^ It would appear that there 

 are several factors tending to reduce the gradient near the surface, one of 

 these being the effect of underground water, which may bring heat to the 

 surface either by means of thermal springs rising from considerable depths 

 or by convection currents in "stagnant" water in pores, pockets, or fissures 

 in the rocks, thus adding to the effective thermal conductivity of water-logged 

 strata. 



An account is here given of some experiments on the convection of heat in 

 vertical water columns, made with a view to estimating, if possible, the probable 

 magnitude of the heat flow in stagnant water. As might be expected, the 

 effect appears to be extremely complex, depending greatly on the size, and 

 especially on the diameter, of the column. 



Experimental Details. 

 Fig. 1 represents in section the general outlines of the apparatus used, 

 the scale being 1/5. The column of water under test was contained in a 

 glass tube A, open at both ends, connecting a copper vessel B, about 6 cm. 

 deep by 4 cm. in diameter, with . an open upper vessel C, the cork stoppei's 

 through which A passed being rendered water-tight with paraffin wax. The 

 outer cylindrical surface of B was uniformly covered with a heating 

 winding of silk-covered 28 S.W.G. Nichrome wire, wound directly on 

 the vessel, and covered with molten paraffin wax. This winding extended 

 from near the bottom of the vessel to within 1 cm. of the lower surface 

 of the cork. Inside the vessel were placed 24 copper rods, each 5 cm. by 

 0-4 cm., 16 of them being approximately vertical, reaching from the top to 

 the bottom of the water space, near the wall of the vessel, and the remaining 

 8 crossing it diagonally at an angle of about 45° to the vertical. This 

 arrangement of rods, which is not shown in the figui-e, left a conical region 

 just below A unobstructed, while ensuring that the water in it was at an 

 approximately uniform temperature. G was kept cool by a stream of tap-water 



' PhU. Mag., September, 1923. 



SaiENT. PBGC. R.D.S., VOL. XVII, NO. 33. 3 D 



