"Lies, damned lies, and statistics" as they say. Serbia and Honduras are anomalies. Within the entire data set there's a significant positive correlation between "Guns per 100 inhabitants" and Homicide rate.
Are they anomalies, too, or what do you mean by anomaly? I do not want to edit the table anymore, but someone also mentioned Switzerland, indeed, extremely low rate of firearm-related homicide (0.13) with high rate of gun ownership (27.6-41.2). There are many more.