220 Proceedings of the Royal Society of Edinburgh. [Sess. 
text-book on oceanography Krtimmel also remarks : “ The present state of 
mechanical analysis of deep-sea deposits is therefore anything but satis- 
factory. However, an improvement appears to be in progress, and we 
especially hope for a thorough revision of the whole method.” In the 
sequence he recommends an application of the method of E. A. Mitscherlich,* 
viz. to determine the total surface area of one gramme of the deposit by 
measuring its hygroscopic power instead of by weighing the different 
fractions as in earlier investigations. 
However, apart from the serious objections which can be raised against 
the theoretical side of Mitscherlich’s method, it is obvious that a determina- 
tion of the surface is by no means sufficient if we want to define a deposit. 
For example, a sample consisting of coarse sand intermingled with some 
high-colloidal clay may have the same total surface area as a loamy deposit 
consisting of more uniform particles of intermediate size, yet it will in 
most other respects differ profoundly from the latter, so that there are no 
reasons for classing them together. 
An ideal characterisation can evidently not be attained by dividing the 
sample into a smaller or greater number of groups of various sizes or by 
determining their hygroscopic power or total surface, but only by obtaining 
a continuous distribution- curve of the well-known Maxwellian type, viz. 
by plotting a variable representing in some way the weight or number of 
the individual particles against another variable representing their linear 
dimensions. 
The following investigation has been undertaken by me in order to solve 
this problem, especially with the intention of finding a successful mechanical 
analysis of the medium-sized and the finest particles contained in some 
deep-sea deposits. 
These samples, which belong to the collection of the late Sir John 
Murray, were put at my disposal by the courtesy of Mr J. Chumley of the 
Challenger Office, to whom I here wish to express my indebtedness. 
Although the new method has already been applied by me to the study 
of certain continental soils, it is advisable to give here a brief description of 
its main features. j- 
II. Theoretical Discussion. 
It has already been pointed out that an immense amount of labour 
would be required to split each sample up into a series of fractions (each 
consisting of particles of approximately the same size), sufficiently numerous 
* Bodenkunde , Berlin, 1905, S. 56. 
t For further particulars see Internat. Reports on Pedology , vol. v, 257-312 (1915). 
