Trying to tell apart protein crystals from non-protein crystals in crystallization drops has been a longstanding problem
in protein crystallography. The ultimate differentiator of course is the X-ray diffractometer. However, this also is
impractical in many cases - the crystals may be too numerous, they may be too small, etc.
Look at this example of crystalline objects that are only a few micro-meters long:

Birefringence gave uncertain results. At this point, one option is to plow ahead and optimize the condition, use the
microcrystals for seeding, etc., but is it worthwhile when time and resources could be spent on cases with more certain
prospects for success?
Other examples that are useful for crystallographers are the case of phase separation - is the protein inside or
outside the "oily" drops?
When there is precipitate in the crystallization, is it protein?
UV imaging, it turns out, can address these questions in situ in a rapid, non-destructive way.
The above example under UV fluorescence imaging:

An example of phase separation of drops ripe with protein:

Protein microcrystals mixed with protein precipitate:

References:
1. Judge, RA, et al. [2005] "An ultraviolet fluorescence-based method for identifying and distinguishing protein crystals",
Acta. Cryst. D61, 60-66.
2. Lunde, S, et al. [2005] "UV microscopy at 280 nm is effective in screening for the growth of protein microcrystals", J. Appl. Crystal.
38, 1031-1034.
3. Gill, HS [2010] "Evaluating the efficacy of tryptophan fluorescence and absorbance as a selection tool for identifying protein crystals",
Acta. Cryst. F66, 364-372.