Flower number in peas

The John Innes pisum genetic database indicates there are two recessive genes that regulate flower number, fn and fna - "Fn Fna is one-flowered; Fn fna and fn Fna two flowered; fn fna three-to several flowered". Other information I've read (but I can't remember where) suggests this is mediated by environmental conditions. [Note: when JI write Fnfna, they are talking about homozygous genetics - such a pea would actually be FnFn fnafna, bt they shorthand it, as I will, mostly, throughout this post.]

My growouts indicate this is a bit complex. My original varietal growouts were tightly planted - so they were subjected to some environmental stress, but each variety was replicated, and results were consistent. Some varieties had alternate double flowering - the first internode was single flowered, then double flowered, then single flowered up the stem. All of the purple podded plants were exclusively single flowered, as were most of the tall snow peas. Some varieties were consistently double flowered, and none had more than two flowers per node.

This would suggest that all the purple poddeds are FnFna, that the alternate flowering varieties are either Fnfna or fnFna. But what about the consistently double flowering varieties - are they multi flowered, and environmental stress has forced them back to double flowering, or is there something else going on here? Perhaps  either fn or fna is responsible for alternate doubling, and the other for the consistent double flowering?

For the record, one of the most consistent double flowerers was a Nepalese accession, simply labelled 'Salahi' (which I think is a transcription error for 'Sarlahi' but I'm letting it stand). The others were Swiss Giant, and Golden Podded. All my lines I'm assuming are homozygous - that is I'm assuming that these lines have been inbred for a long number of generations, which should result in homozygosity. So, my crosses between multiflower snowpeas (which should be fnFna or Fnfna) and purple poddeds (which should be FnFna) should yield heterozygous F1s in the recessive genes - they shouldn't be multiflowered.

[To spell it out in longhand, a multiflower homozygous pea should be fnfnFnaFna for example - or FnFnfnafna, but let's just go with the first example. The single flowered purples should be FnFnFnaFna. An F1 cross should yield FnfnFnaFna - no homozygosity at the important recessive fn locus. The same would hold for the othermultiflower gene or even in the case of double recessive fnfnfnafna - if I had it.]

So how come some of my F1s are multiflowered? One, Chamber of Death X PurplePodded (Lost Seed Co) has some alternate double flowers in the F1 - but my records suggest they are both single flowerers. This is a curious phenomenon. Stay tuned.