It has long been observed that there is a correlation between the physical qualities of a watercourse and the linguistic qualities of its name; for instance, of two river-names, one having the linguistic quality of river as its generic element, and one having burn, one would expect the river to be the longer of the two. Until now, a phenomenon such as this had never been formally quantified. The primary focus of this thesis is to create, within a Scottish context, a methodology for elucidating the relationship between various qualities of hydronyms and the qualities of the watercourses they represent. The area of study includes every catchment area which falls into the sea from the River Forth, round the east coast of Scotland, up to and including the Spey; also included is the east side of the River Leven / Loch Lomond catchment area. The linguistic strata investigated are: Early Celtic, P-Celtic, Gaelic and Scots. In the first half of the introduction scholarly approaches to toponymy are discussed, in a Scottish and hydronymic context, from the inception of toponymy as a discipline up to the present day; the capabilities and limitations of these approaches are taken into consideration. In the second half the approaches taken in this thesis are outlined. The second chapter explains and justifies in more detail the methodology and calculus used in this thesis. The subsequent chapters examine the following linguistic components of a hydronym: generic elements, linguistic strata, semantics and phonological overlay. In each of these chapters the methodology is harnessed as an analytical tool to generate new findings for hydronymic research. The conclusion consists of a summary of the findings and a review of the performance of the calculus. It emerges that these analytical tools are of use to the field of toponymy in two ways. Firstly, they formalise and challenge previously unquantified statements made in the field of toponymy. Secondly, they elucidate hitherto unnoticed phenomena. It is suggested that in the future this methodology be applied to other datasets (particularly hill-names) and to other regions in Scotland and the world at large.