186 REPORT— 1876. 



salt water at its temperature of cougelation is denser tliau at any higher tempera- 

 ture, its weight would cause it to sink ; and it would in time, did no other cause 

 intervene, occupy the whole of the space in the ocean not influenced hy the sun's 

 heat. But in considering the effect of the heat imparted to the surfaces, we have 

 also to consider the effect of evaporation and precipitation. In the equatorial 

 regions evaporation is rapid, so that the surface-hlm would become cleared through 

 increased salinity were it not for the increased temperature and large precipitation, 

 as well as to its being transported by the friction of the trade-winds and eartli's 

 motion to the westward. This surface-film, constantly moving westward in the 

 equatorial regions, meets in the Atlantic with an obstructing point of the South- 

 American continent, which directs it to the northward, so that the greater part of 

 the water directly heated by the sun's rays in the tropical regions is forced into the 

 North Atlantic. As the salinity of this water is greater than that of the subjacent 

 layers, and its increased temperature only renders it less dense, directly a portion 

 falls in temperature in the colder regions of the temperate zone, the surface- 

 film sinks and imparts heat to the water beneath. Consequently the isotherms 

 ■will be foimd at greater depths where the heated smface-tilms are constantly 

 descending than when, owing to their being less dense than the subjacent layers, 

 they remain on the surface. 



ECONOMIC SCIENCE AND STATISTICS. 



Address hy Sir Gteorge Campbell, K.C.S.L, M.P., D.C.L., President of the 



Economic Section. 



I FEEL a difiiciilty in imdertaking the Presidency of this special and important 

 Department of the British Association, in this great city, which contains so 

 many men masters of so many branches of economic subjects. But, Scotchman 

 as I am, I have felt that I could not decline the honour proposed to me in the 

 commercial capital of my own country ; and I remember with pride that perhaps 

 in no place in the British Empire could economic subjects be discussed with 

 so great advantage. Other places have special industries. Glasgow has many,- 

 and she excels in them all. 



I tmderstand it to be the object of the Association that in the treatment of the 

 subjects presented to us we should stiidy, in this as in other departments, to follow 

 as far as may be a strictly scientific method of inquiry, not lapsing into the di.s- 

 cussion of political details, but attempting to ascertain the principles on which 

 economic results are founded, and to define the niaiu lines of economic truth. It 

 may not always be possible to draw the boundary between science and practice ; 

 but I am sure that we shall all try as much as possible to avoid matters which 

 involve party or personal questions, and to maintain a calm and scientific attitude 

 in our treatment of the many subjects which come within the range of this 

 Section. 



The Section was originally called that of " Statistics ;" and all economic inquiry 

 must be based on or tested by Statistics. At first sight Statistics expressed in figm-es 

 might seem to constitute the most exact of sciences ; but in practice it is far other- 

 wise. In nothing is so great caution necessary ; there is too gi-eat temptation to 

 reduce to figures facts which are themselves not sufficiently ascertained ; too often 

 an exactness is claimed for these figured results which is altogether fallacious and 

 misleading. In fact there is a use and an abuse of figm-es ; and one is sometimes 

 tempted to sympathize with the cynical philosopher who said that nothing is 

 more misleading than facts, except figures. It is especially necessary to distinguisli 

 between figures which are really ascertained, and those which are nierely drawn by 

 deductions from rough and conjectural facts. A false appearance of exactness 



