The existence of an intergalactic globular cluster population in the Coma cluster of galaxies has been tested using surface-brightness fluctuations .
The main result is that the intergalactic globular cluster surface density ( N _ { IGC } ) does not correlate with the distance to the center of Coma and hence with the environment .
Furthermore , comparing these results with different Coma mass-distribution model predictions , it is suggested that N _ { IGC } must in fact be zero all over Coma .
On the other hand , the results for N _ { IGC } and the faint end of the galaxy number counts ( beyond m _ { R } = 23.5 ) are connected .
So N _ { IGC } = 0 settles the slope of this function , which turns out to be \gamma = 0.36 \pm 0.01 down to m _ { R } = 26.5 .

The fact that N _ { IGC } = 0 all over Coma suggests that globular clusters were formed only , or almost only , from protogalactic clouds .
None , or perhaps very few , could have formed in isolated regions .
It also seems inappropriate to advocate a relationship between intergalactic globular clusters and dark matter distributions , although it is true that the relationship could still exist but not be strong enough to have been detected .
Finally , since our conclusion is that intergalactic globular clusters do not exist in Coma , accretion of intergalactic globular clusters might not be significant in galaxy formation and evolutionary processes in the Coma galaxies .