I hate to say it, but it looks like your laser peak at 499 nm looks like it was truncated, meaning that your fluorescence intensity is 25% of what your detector will measure, not 25% of the laser line. Perhaps it would be better to compare the fluorescence to the second order image that shows up at 200 nm? Why the gel cap? The round shape is going to make it nearly impossible to get reproducible light path. Why not use standard fluorescence cuvettes? You can get 100 of them for $20 here: http://www.amazon.com/Disposable-Vis-Fluorometer-Cuvette-100pcs/dp/B005YOX53Q/ref=pd_sim_sbs_indust_6?ie=UTF8&refRID=02G4HYZN7YVDMEWCK9VK

the fluorecence looks the same even when the laser heads into the sample at a slightly different angle-- is the light path critical? I figured it only mattered if fluorescence decay was what we're after.

The gel cap is a standard object that is more easily available than a cuvette, and seals the oil in quite well, even for storage. Most of those cuvettes seam to leak. But you're right, the cuvettes do look attractive. Jeff sent me a few too.

whelp. forget gelcaps. The article I posted here is right- gelatin fluoresces. the whole blue-green section of this spectra was the cap itself.

I wish we didn't have the fluorescence issues, the gelcaps do allow the capture of a brighter fluorescence than the cuvette holders; the chlorophyll peak is very weak here:

How about using vegetarian capsules? This kind is made out of hydroxypropyl methylcellulose (or HPMC). They also make vegetarian capsules out of pullulan, a polysaccharide derived from tapioca. I'm not sure if either of these two options would fluoresce or not.

that is a great idea! HPMC, or hypromellose looks promising.

It looks like pullulan is used as a Near IR tracer, so it would fluoresce, but maybe not in a range we're looking at, since our spectra skew towards the blue end: http://www.ncbi.nlm.nih.gov/pubmed/24481855