We imaged circumstellar disks around 22 Herbig Ae/Be stars at 25 mu m using Subaru/COMICS and Gemini/T-ReCS. Our sample consists of equal numbers of objects belonging to the two categories defined by Meeus et al. (2001); 11 group I (flaring disk) and II (at disk) sources. We find that group I sources tend to show more extended emission than group II sources. Previous studies have shown that the continuous disk is hard to be resolved with 8 meter class telescopes in Q-band due to the strong emission from the unresolved innermost region of the disk. It indicates that the resolved Q-band sources require a hole or gap in the disk material distribution to suppress the contribution from the innermost region of the disk. As many group I sources are resolved at 25 mu m, we suggest that many, not all, group I Herbig Ae/Be disks have a hole or gap and are (pre-)transitional disks. On the other hand, the unresolved nature of many group II sources at 25 mu m supports that group II disks have continuous at disk geometry. It has been inferred that group I disks may evolve into group II through settling of dust grains to the mid-plane of the proto-planetary disk. However, considering growing evidence for the presence of a hole or gaps in the disk of group I sources, such an evolutionary scenario is unlikely. The difference between groups I and II may reflect different evolutionary pathways of protoplanetary disks.