1872.] 



( Shearwater ' Scientific Researches. 



555 



films, and by their successional renewal, the temperature of the whole 

 mass of the water will be progressively reduced, until its density reaches its 

 maximum at 39°*2 ; after which the further application of cold no longer 

 causes the descent of the surface-film. And thus it comes to pass that the 

 temperature of all the water beneath that superficial stratum which under- 

 goes further reduction of Temperature with increase of Specific Gravity until 

 it freezes, is uniform throughout, clown even to the very bottom of such deep 

 basins as some of the Italian and Swiss lakes ; and that, as there is no 

 downward convection of Heat when the summer sun acts on their surface, 

 and the only warming influence will be that of the crust of the Earth 

 beneath, the temperature of 39°' 2 is found to prevail throughout the year 

 from a small distance beneath the surface of the deeper lakes down to their 

 very bottom. 



27. But the case is very different with £e0-water, which does not attain 

 its maximum of density until it freezes (at or below 27°) ; for the downward 

 convection of Cold will continue until any further reduction of temperature 

 produces the formation of ice ; and thus we should expect to find the deepest 

 stratum of /Sea-water that has been acted on by surface-cold showing the 

 lowest temperature, which is precisely what we have ourselves met with in 

 the "Lightning Channel" (§ 14), where we found 29°'G at 640 fathoms, 

 and what Messrs. Payer and Weyprecht met with in Lat. 78° very near the 

 surface (§ 15)*. 



28. Let us next suppose that only a portion of the area of the basin is 

 exposed to the action of surface -cold ; the movement of the liquid takes 

 place somewhat differently. For the surface-film which descends is then 

 replaced, not from beneath, but by an inflow from the neighbouring area, 

 as may be proved experimentally by the means already indicatedf : and 



* Mr. Croll (Phil. Mag., Oct. 1871, p. 248, note) speaks of it as " a well-established 

 <£ fact that in Polar regions the temperature of the sea decreases from the surface down- 

 " wards ; and the German Polar Expedition found that the water in very high latitudes 

 11 is actually less dense at the surface than at considerable depths, thus proving that the 

 " surface-water could not sink in consequence of its greater density." — Now if, as all 

 recent observations concur to prove (§ 15), the temperature of Polar water shows a pro- 

 gressive reduction, with increase of density, from the surface downwards, — save where 

 the melting of ice has reduced the salinity as well as the temperature of surface-water, 

 thus making it lighter as well as colder, — surely the decrease of temperature indicates 

 that as fast as the surface-water is chilled by the colder air above, it sinks and carries 

 its cold downwards. In what other way the deepest water can come to be the coldest, 

 I am at a loss to comprehend. 



t Mr. Croll objects (loc. cit. p. 244) to the Experiment which I described in my last 

 Report (§ 129), on the ground that the mode in which the heat was applied at the op- 

 posite end of the trough would establish a circulation by " a horizontal propulsion of 

 " water caused by the expansive force of vapour, and that the movement was not in any 

 " way due to difference of specific gravity." — Now if Mr. Croll will try this experiment 

 for himself, he will find that the circulation is clearly initiated at the cold end of the 

 trough : the coloured water, as it comes in contact with the ice, tumbling dozen (as it 

 were) to the bottom of it, and then flowing horizontally without any application of heat 

 to the other eud of the trough. And this movement goes on, though at a progressively 



