Presented at the 2020 Ocean Sciences MeetingWe have made the first comparisons between salinity data from the 1870s round-the-world voyages of HMS Challenger (1872-6) and SMS Gazelle (1874-6) and present-day values in the UK Hadley Centre EN4 data set (Good, Martin and Rayner, 2013). After the elimination of obviously erroneous values, the Challenger analysis contains 1000 samples at 350 stations. Many stations have only near surface values, but a significant number cover the entire water column. The Gazelle data come from 164 stations (approx. 350 samples, the majority with duplicates). Most of the Gazelle data are from the top 200m but with some deep and near bottom samples too. Since salinity was a novel and, as yet, poorly defined property in the 1870s, we derived the salinity values used in this analysis from the surface and subsurface specific gravity measurements made at sea on the two voyages. These were primarily analysed on board ship using hydrometers with the values published in the voyage reports having been reduced using a standard temperature. Despite these limitations we find that the 1870s observations reproduce the present-day surface salinity distribution surprisingly well and that they also capture large scale subsurface structures. We examine the possible causes of large scale differences between the modern and historical data. Comparisons between the HMS Challenger temperatures and modern day values were published by Roemmich, Gould and Gilson (2012) and allowed an upper bound to be placed on ocean warming over 135 years. We examine whether the uncertainties in the 1870s specific gravity/salinity data preclude their use in detecting changes in the global hydrological cycle c.f. (Durack and Wijffels, 2010).