Since the evidence is leaning toward homozygous lethality in spiders, we may assume that the spider mutation is co-dominant after all rather than dominant. But that has nothing to do with spider to normal breedings.
Breeding a spider to a normal will theoretically give you 50% spider and 50% normal. This is because the spider parent has only one spider mutant allele and can donate it or can donate a normal homologous allele. So, all of the offspring will be one of two types: spider or normal--in equal proportions, theoretically. 50/50
Breeding a spider to a spider should give you 1/3 normals to 2/3 spider because we do not expect to see any supers.
Do the punnett squares, and you'll see the distribution.