CHAi'. v.] GENERAL CONCLUSIONS. 309 



tables of the year 1873, we come, then, to the broad conclusion 

 that the density of the Tipper layers of the North Atlantic is 

 considerably higher than that of any other part of the ocean, 

 and the specific -gravity tables for the spring months of 1876 

 give the same result. I need scarcely say that this is exactly 

 what must have been anticipated, if my view be correct, that 

 the movement of deep water in the Atlantic is mainly due to 

 excess of evaporation over precipitation in its northern portion. 



An element of great nncertainty is undoubtedly introduced 

 into the determination of the specific gravity of surface-wa- 

 ter by the weather. These determinations were made, as usu- 

 al, daily for the section between Stations CCCXXIII. and 

 CCCXXXY., from Montevideo to Tristan d'Acunha, in the 

 beginning of March, 1876; and the mean of these, the tem- 

 perature reduced to 15°-56 C, was 1-02620. Of the eighteen 

 days occupied in running the section, nine were dry and fine, 

 and on nine rain fell either continuously or in showers. The 

 mean for the nine dry days was 1*02639, and for the nine w^et 

 days 1-02591. The maximum surface specific gravity for the 

 section (1-02680) was at Station CCCXXIII. at the point where 

 probably the Brazil Current has most effect on the surface; 

 and the minimum (1-02494) was at Station CCCXXVI. after a 

 heavy fall of rain. The mean specific gravity of the surface- 

 water at the temperature at which it was procured was 1-02502. 



The specific gravity of the bottom-water was determined at 

 ten stations on the section. Reduced to a temperature of 15°-56 

 C, the mean was 1-02601 ; the maximum, 1-02650, was at Sta- 

 tion CCCXXIII. at a depth of 1900 fathoms ; and the mini- 

 mum, 1-02580, was at Station CCCXXVI. at 2775 fathoms. 

 The mean specific gravit}^ of the bottom-water at the depth at 

 which it was procured was 1-02811, showing a difference be- 

 tween the two means of 0-00210, due to difference of tempera- 

 ture alone. 



It seems from these observations that the differences of sur- 



