102 
THE VOYAGE OF H.M.S. CHALLENGER. 
Although tin deviations are less than for the total quantities of bromine, the net result 
i~ th. same. The ratio of the bromine to the chlorine iu ocean water is, it appears, inde- 
j. nd- nt of the depth, unless the ratio is a function of geographical position and depth, 
and the influence of the former, in my mixtures, has happened to just compensate for 
that »>f the latter independent variable. But this is not probable, the less so, as the Arran 
water gives the same result as the mixtures of Challenger samples. 
i onstaney of the bromine ratio is important as tending to prove that the com- 
position of ocean-water salts is the same everywhere ; because if" it is not, the percentage 
of bromine, being so very small and necessarily affected by the plant life on the surface, 
. d be more liable to fluctuation than that of any of the major components. And yet, 
according to my 77 complete analyses, the percentage of the lime at all events is greater 
in deep-sea than it is in shallow water. Is this result safe and sound, and not perhaps 
after all brought about by analytical errors? This question naturally forced itself upon 
me, and caused me to utilise what was left of our mixtures of water for those 
supplementary determinations of the ratio of the lime to the chlorine which, although 
later in being done, are reported in an earlier part of this Memoir (see pp. 32 et seq.). 
All the quantitative determinations recorded in this chapter (except the 14 abor- 
tive analyses referred to in the introduction) were made under my eyes by Mr. 
John M ‘Arthur, and I feel greatly indebted to him for the zeal and self-denial with which 
of months, devoted himself to this I . <1 ions and troublesome work. 
