288 



Part III. — Fifteenth Annual Report 



100 fathoms in 1893 and 1896, we find Station II. 0°'4 C. colder in the 

 latter year, Stations III. and IV. practically the same, and Station VI. 



0 ^ warmer. Notwithstanding the widely different characteristics of the 

 two summers, it appears that near the north-western edge of the plateau 

 the mean temperature of the water within 100 fathoms of the surface was 

 nearly the same, and in the southern part of the Faeroe Channel it was 

 actually higher in the colder season. It would be premature to discuss 

 the meteorological significance of these facts before the data from all 

 sources have been compared with them ; but it seems only fair, in order 

 to place the work of the 1 Research ' in the position it deserves, to point 

 out that a low surface temperature over the Norwegian Sea has been fol- 

 lowed by persistent northerly winds over western Europe throughout the 

 last quarter of 1896, and therefore presumably by abnormally high pres- 

 sures over the former region. These northerly winds have almost 

 certainly kept the North Sea full of oceanic water, water which we know 

 to have had a fairly high temperature, and the autumn and early winter 

 has been unusually open and rainy in the British Isles. 



The extensive system of observations organised during 1896 will, it is 

 expected, continue in operation till the end of 1897, and it is to be hoped 

 that the connecting link supplied by H.M.S. ' Research ' during the first 

 year will not be wanting in the second. 



Note. — The following analysis of samples collected by H.M.S. 



1 Research,' made by Mr W. S. Anderson at the Marine Station, Granton, 

 have been sent to me by Mr Irvine. The C0 2 determinations were made 

 with Pettersson's apparatus, but the samples were not collected in 

 exhausted tubes. Numbers of samples correspond to those in Table III. 





Depth 

 in 

 fms. 



X in grms. per 

 Kilogrm. 



Total 

 C0 2 

 per litre 

 in milli- 



grms. 



Alkalimity 

 C0 2 per 

 litre in 



millligrms. 



Probable 

 Density as 

 4 Si5 calculated 

 from x- 



Surf. Faeroe Channel, 



0 



19-567 (?) 



104-8 





1026-30 (?) 



No. 127 



360 



19-287 



102-6 



52-6 



1025-92 



„ 117 



380 



19-295 



1047 



52'5 



1025-93 



„ 105 



500 



19-190 



101-1 



525 



1025-78 



„ 172 



550 



19-521 



105-7 



52-6 



1026-24 



170 



595 



19-525 



104-2 



52-6 



1026-25 



,-, 145 



600 



19-335 



101-5 



52-2 



1025-98 



„ 146 



640 



19-306 



97*2 



52-5 



1025 94 



