192 Messrs. Carpenter and Jeffreys on [Dec. 8, 



Surface-water. Intermediate water. Bottom- water. 



Average 1994 19*85 1975 



Maximum.... 20-19 19*94 19*98 



Minimum 19*81 19*70 19*46 



It appears from these Analyses that there is a slight excess of Salinity 

 in the surface- water of the Atlantic, as had been previously observed by 

 Forchhammer * ; the excess, however, being so small as not to neutralize 

 the excess of Density which the deeper water derives from its lower Tempe- 

 rature and from the Pressure of the superincumbent column. Five de- 

 terminations of the Chlorine contained in samples taken at the same spot, 

 from the Surface, and from 10, 25, 50, and 100 Fathoms, gave the follow- 

 ing results : — 



Surface 20*013 



10 fathoms 19*909 



25 „ 19*909 



50 „ 19*909 



100 „ 19*805 



A comparison of these seems to indicate that the excess of density, being 

 limited to a mere superficial film, is entirely due to evaporation ; and the 

 reason why this more concentrated film does not sink, as it does in the 

 Mediterranean (§90), is that its excess of Salinity is so small, that even at 

 the depth of 10 fathoms its effect on Specific Gravity is neutralized by the 

 greater density arising from reduction of Temperature. — The differ- 

 ence between the maximum and minimum, which in Surface-water is only 

 l-52nd part of the whole, and in Intermediate water only l-83rd part 

 of the whole, is in Bottom- water l-38th of the whole; and the minimum 

 of the whole series, namely, 19*46, presented itself in a specimen of 

 Bottom-water taken from a depth of 994 fathoms. This proportion might 

 have been presumed to represent the inferior Salinity of the Polar stream, 

 from which the Temperature of the sample indicated its derivation, were 

 it not that another sample from a depth of 1095 fathoms was found to 

 yield 19*73, or nearly the average proportion, of Chlorine. Two samples, 

 taken respectively from 700 and 322 fathoms, gave 19*63 ; and the maxi- 

 mum of 19*98 occurred in a sample from 717 fathoms. These anomalies are 

 somewhat perplexing, yet the whole range of variation is really very small. 

 Similar anomalies presented themselves in the results of Dr. Frankland's 

 Analyses of samples of Bottom -water collected in the Cold area (Report 

 for 1869, p. 489) ; for whilst the proportion of Chlorine in a sample 

 (No. 64) taken from a depth of 640 fathoms, at a Temperature of 29°*6, was 

 19*88, that of the Surface at 49°*7 being 19*96, it was 20*14 in another 

 sample (No. 54), which, though taken at little more than half the depth 

 (363 fathoms), was shown by its Temperature to have been brought up 



* See bis Memoir on the " Composition of Sea- water " in the Philosophical Trans- 

 actions for 1865, p. 247. 



