This page is very much a "Work In Progress". This is the only information I have on what might be causing white undercoats. I will edit it later as these are just copy/pasted messages I sent to friends.
B Locus Smoke
B-w ~ B locus: white based brown (Semi-dom)
B/B-w = Barely smoke, Hair is almost black "except for the extreme base", eyes probably black or the darkest ruby you’ll never see
B-w/b = chocolate smoke, RE
B-w/B-w = chocolate smoke, but with about double the amount of white than B-w/b, RE
B-lt ~ B locus: light (Semi-dom)
B/B-lt = Off black smoke, less white (Except you can’t tell on A^y/a ???)
B-lt/B-lt = Chocolate smoke, more white
Agouti makes it harder to ID (not shocking, a ticked allele is hard to ID on another ticked allele, wow)
BUT regardless of whether B-lt is homo- or heterozygous, (B-lt/B-lt OR B/B-lt)
When it is also homo OR heterozygous with PED (p/p OR P/p)
they “do not display an absence of pigment in the lower section of their hair shafts”
So even just carrying PED makes B dominant over B-lt
There’s B-lt (light) and B-w (white).
They’re semi-dominant. They show best on black, hard to see on agouti, and don’t show on red (A^y, not sure about recessive yellow; e/e).
When there’s only 1 copy of B-lt or B-w and 1 copy of wild type (B/B-lt or B/B-w), then it creates an offblack mouse with a very very small amount of white at the base of the fur.
When there’s 2 copies, they’re chocolate with double the amount of white base as previously mentioned. It results in about 1/3 to 1/4 of the coat being white next to the skin.
When there’s only 1 copy of B-lt or B-w and 1 copy of regular choclate (B-lt/b or B-w/b), it looks the same as it does when it’s homozygous, except with less white.
However, when B-lt is on the same mouse as Pink Eye Dilute/PED, then the white undercoat will not show up. If the mouse is homozygous PED (p/p) or even if it just carries PED (P/p) then there’s no smoke. WILD.
B locus & Silvered
B/b si/si is also known to cause a white undercoat
Siamese smoke may or may not be caused by B locus alleles.
Due to C-P linkage, it is uncommon for homozygous C dilute mice to carry or express PED (p). When siamese smoke is outcrossed, it is possible that carried/heterozygous PED is altering the expression of the B locus smoke alleles, causing it to disappear. The other explanation is that siamese smoke has nothing to do with the B locus, and is instead caused by a siamese modifier; however, this may yet be explained by different C dilutes having different interactions with B locus smoke alleles. Siamese seems to be strangely common to find smoke in, but not any other C dilutes.