Why are variables lowercased?

The obvious place to look up something like this is Cajori's book A History of Mathematical Notations (in 2 volumes), volume 1 of which is freely available on the internet (URL below). Both volumes are currently available as Dover reprints.

http://www.archive.org/details/historyofmathema031756mbp

In searching the on-line .pdf file for "small letter" and "capital letter" (these two phrases worked best for me), I found that styles differed in the generation or two before Newton and Leibniz, with uppercase letters used by Francis Vieta (1590s) and lowercase letters used by Thomas Harriot (1631) and Descartes (1637). There may be a convention that began with Leibniz in which numerical variables are lowercase and geometrical variables (e.g. for points) are uppercase, google "Leibnizian procedure", but I'm not very sure about this. In any event, even if Leibniz began such a convention, I'm sure there still would have been a lot of individual variation in the late 1600s to late 1700s. However, regardless of who is responsible for these algebraic and geometric notation conventions (lowercase for algebra, uppercase for geometry), I believe they were widely used in the literature beginning at least by the late 1700s (this being based on my own observations of many hundreds of 19th century journal volumes and books I've looked through in the past 30+ years).