1915-16.] The Size of the Particles in Deep-sea Deposits. 225 
avoid subjective errors of observation, I shall reserve the detailed descrip- 
tion of the instrument and its technique for further publication, and give 
here only the main principles on which it is based. 
Suppose an aqueous suspension of the sample to be contained within a 
cylindrical vessel, in which, close to the bottom, there is a thin circular 
disc of slightly smaller diameter suspended from one arm of a sensitive 
balance. On that disc will then be accumulated all the deposit which 
settles from a liquid column having the upper surface area of the disc for 
its base, and a height equal to the vertical distance from that surface to the 
free surface of the liquid. After a sufficiently long time all the particles 
originally in the liquid column will have accumulated on the disc, and their 
joint weight, P M , will be slightly less than the total weight, P, of the sample 
used for the suspension (owing to the fact that the particles outside the 
circumference of the disc, as well as those contained within the space below 
it, will not fall on the disc but on the bottom of the vessel). By means of 
the balance we can measure from time to time how the weight of the 
deposit accumulated on the disc increases with time, and plot an “ accumula- 
tion-curve ” of the type mentioned above. 
Taking a mean value of the density of the sample, cr, and always express- 
ing the accumulated weights in percentages of the total weight (the latter 
measured after all particles have settled, if necessary after coagulation), we 
need not consider the weight of the deposit in air. 
Fig. 1 gives a schematic idea of the experimental arrangements. The 
disc A was of thin copper foil, heavily gilt, with a surface area of 100 cm. 2 
It was suspended from the arm of a sensitive balance by a very fine wire 
of gilt silver, so as to be quite close to the bottom of the cylindrical vessel L. 
As it is indispensable for an undisturbed accumulation of the sediment that 
the distance h between the disc and the surface of the liquid remain 
practically unaltered throughout the experiment, and also because move- 
ments of the disc will give rise to errors in the weighings due to convection 
currents, it was necessary to use a compensating arrangement (K, to the 
right on the figure) limiting the displacements of the disc to a minimum. 
When the particles accumulated on the disc outweigh the counterweight G, 
so that the disc begins to sink from its zero position, an electric current 
becomes automatically closed by the mercury contacts at E. A small 
Osram lamp, 0, in the circuit then lights up for a moment, at which the 
time is read off a chronograph, and simultaneously a small counterweight 
is added to G. Both these operations may be executed either by the 
observer or automatically by means of an electromagnetic arrangement. 
Before the beginning of each experiment the balance is brought to its zero 
VOL. xxxvi. 15 
