148 THE FRESH-WATER LOCHS OF SCOTLAND 
ingredients may be enhanced, as follows : — Sedimentary formations 
impregnated with sodium chloride (e.g. those of marine origin) 
increase the chlorine ; beds of gypsum or pyrites the sulphates (sul- 
phatic waters) ; peaty lands, owing to the solvent action of humic 
acids, the iron and silica ; igneous rocks the sodium and potassium 
(alkaline waters) ; chalk or limestone the calcium carbonate (hard 
waters). High chlorine contents may also be traced to the near 
neighbourhood of a sea-board, and to the influence of human life and 
activity on the drainage area. The effect of local peculiarities is of 
course much greater on small than on large rivers, which latter 
represent a summation of many tributary waters. 
Rivers flowing through arid countries tend to develop quite 
another type of water. In the first place, the rainfall is scanty, 
whence a high percentage of total solids in the river water. Secondly, 
the scarcity of vegetation results in a dearth of carbonic acid, which 
is chiefly derived from decaying vegetable matter. Now, calcium 
carbonate can only exist in solution in presence of free carbonic acid ; 
hence this ingredient dwindles to a minimum. Further, the alkali 
metals leached out of igneous (felspathic, basaltic, micaceous) rocks 
attach preferentially carbonic acid to themselves, and form sodium 
and potassium carbonates, which in turn throw out of solution an 
equivalent of any other calcium salt, e.g. sulphate, which may be 
present ; and the geology of arid regions is frequently of a nature to 
bring the alkalies into special prominence in the waters. In such 
waters, therefore, when they are not rendered predominantly chloridic 
or sulphatic by special local conditions, we commonly find as principal 
solutes sodium and potassium carbonates, calcium in any form being 
in a vanishing minority. The waters, in fact, are alkaline, and can 
be recognised as such by the ordinary litmus test. 
It need scarcely be pointed out that the abnormal waters referred 
to above, whether they be unusually chloridic, sulphatic, or alkaline, 
are, with regard to the inland waters of the world at large, exceptional. 
The majority of discharging lakes, then, may be supposed to be 
filled with a water v^hich holds exceedingly little matter in solution, 
and that chiefly calcium carbonate. Details may be gathered from 
the subjoined analyses of lake waters.^ No apology is offered, or 
needed, for stating these and subsequent analyses in the modern 
rational form, according to which constituents present mainly in the 
ionised state are reported as ions, whilst any carbonic acid present 
over and above what is required to form normal carbonates is sup- 
pressed. The item CO3 in the analyses is not a result of direct deter- 
1 Nearly three hundred analyses of spring, river, and lake waters, mostly 
American, are quoted, with references, by F. W. Clarke, Bull. U.S. Geol. Survey, No. 
330, 1908, from which publication several of the analyses here given are extracted. 
