Well, I did just hold a presentation on the development of certain vowel-length alternations in Inari Saami a couple of weeks back... Unfortunately, my materials are all in Finnish and Inari Saami (plus, counting sources, some long-dead trees in German), so I guess uploading them here wouldn't really help.
But to very briefly summarize: Proto-Saamic developed consonant gradation, meaning that consonants at the onset of the final syllable of the word (and/or the coda of the penult) were pronounced longer if the final syllable was open than if it was closed, resulting in allophonic alternations like this:
*[tolˑɘ] : *[tolɘn] ('fire', nominative and genitive singular, respectively)
*[lɘsˑtɘ] : *[lɘstɘn] ('leaf', ditto)
Later, many case endings were shortened or lost in most Saami languages, so consonant gradation itself took on the function of distinguishing forms, eg. North Saami dolla vs. dola (from the aforementioned *[tolˑɘ] : *[tolɘn]).
In addition, East Saamic (and eastern dialects of North Saami) have developed vowel length alternations by, as a rule of thumb, lengthening short stressed vowels before short consonants and shortening long vowels before long consonants. For example, the same forms in Inari would be tullâ : tuulâ for 'fire' and lostâ : loostâ for 'leaf'.
The system is quite similar in Skolt Saami, for which a fairly extensive grammar in English exists - and is freely available online. Not that much on the history of the system, but at least a fairly good source on how it works synchronically, I think.