124 Proceedings of the Royal Society of Edinburgh. [Sess. 
XI. — On the Application of Manley’s Differential Densimeter 
to the Study of Sea- waters on board Ship. By Norman 
Phillips Campbell, B.A. Oxon. Communicated by Sir John 
Murray, K.C.B. 
(MS. received January 9, 1913. Read February 3, 1913.) 
In a paper * communicated to the Royal Society of Edinburgh, Mr Manley 
describes a differential densimeter and its successful application in the 
laboratory to the rapid determination of the relative densities of sea- waters. 
The paper also contains the suggestion that the instrument would, if 
suspended in gimbals, be rendered available and convenient for use at sea : 
thus making it possible for the oceanographer to carry out observations 
upon the density of sea-water with greater frequency than the usual 
methods conveniently permit. 
During a voyage of the Oruba from London to Colombo in October 
1908, the Orient Steamship Company courteously afforded the present 
author the opportunity of testing the efficacy of such a differential 
densimeter under the varying conditions which obtain on board ship. 
The instrument, which was built in Oxford under Mr Manley’s 
supervision, is shown in fig. 1. 
The densimeter tubes AA are mounted on a mahogany board 7 feet 
long and 6 inches wide ; this board is suspended on gimbals GG. The 
gimbals are hung from the ceiling of the cabin by three copper wires. The 
lead weights WW, attached to the lower end of the board, greatly reduce 
the swinging. The lower ends of the tubes dip into two exactly similar 
flasks : these stand on a shelf C, which is covered with a piece of plate- 
glass. The average length of each column of liquid is 1*52 metre. 
Fig. 2 shows the travelling microscope M (by Elliott Bros.) for com- 
paring the heights of the columns. The drum-head D reads to *001 mm. ? 
but *01 mm. affords sufficient accuracy for the purpose. The microscope is 
mounted on a plate-glass base E, which slides on another piece of plate-glass 
F ; the microscope can thus be brought opposite either tube and adjusted 
by the drum-head and screw S. If the plate-glass shelves C (fig. 1), F (fig. 2), 
were strictly horizontal, the difference in the microscope readings would at 
once give the difference between the heights of the two columns ; it was 
found that C, F deviated from strict horizontal parallelism by 0*66 mm. 
* Proc. Roy. Soc. Edin ., vol. xxvii., Part III., pp. 210-232. 
