Just look at the fluoresccence spectrum for a mercury lamp in the link. Anything above 600nm is quite weak, but you might be able to get a signal if you have a lot of fluorescent molecules. But for weak samples, you might not get enough excitation... 

The excitation power you get for A647 is about 8 times lower than for A488 when using a mercury lamp. But in absolute terms you're just 2x weaker than a Xenon lamp.

My advice would be to try, as you'll need to buy the filters anyway, but you probably have higher chances with an A610.

Best

EDIT: Looking at the product number for your camera, I see that it is a color CCD camera. you are going to have a hard time getting proper fluorescence images with that kind of camera... 

On top of that Nikon claims the following

"High Red Sensitivity

A high performance IR cut filter eliminates unwanted infrared contributions from halogen light sources, while maintaining excellent red color rendition for outstanding color reproduction."

Which is exactly what you do not want. 

If you want to do fluorescence with this system, the first thing would be to get a black and white camera, before thinking of the lamp.

http://zeiss-campus.magnet.fsu.edu/articles/lightsources/mercuryarc.html

http://zeiss-campus.magnet.fsu.edu/articles/lightsources/xenonarc.html

Thank you very much for the comments, Olivier Burri. We cannot afford the black and white camera so we end up with DS-Ri2 which at least it is better than DS-Fi2.