There does seem to be some confusion over the names.
There are three commoner small-medium sized dark petrels in Japanese waters.
The first is Swinhoes Storm Petrel - Oceanodroma monorhis
The second is Tristram's Storm Petrel - Oceanodroma tristramii
The third is Matsuidaira's Storm Petrel - Oceanodroma matsudairae
The third species breeds only off the Ogasawaras islands and rarely wanders into closer inshore waters. It can be told from Tristrams's by the white flashes on the primaries and lack of a dark rump. It is around 25cm long and has a wingspan of 56cm.
The second species Tristram's breeds largely off the Izu islands, Ogaswaras and Haiwaiian islands. It iis the smae size as matsudaira's, but in contrast has a pale rump fawn rump and larger pale areas on the secondary coverts reachng the lesser primary coverts/alula and the tertials (the pale area just reahes the tertials on Matsudair's). It has similar flight patterns to Matsudaira's, being rather stiff-winged and gliding for long distances similar to shearwaters and Bulwer's Petrel, rarely if ever hovering over the water.
The first species Swinhoe's is a smaller darker brown bird, with no discernable pale rump, and much smaller paler areas on the secondary coverts only. It also has a less forked and shorter tail than the other two, Matsudaira's being the most forked, and Tristram's somewhere in between. Swinhoe's also tends to be more fluttery and glide for shorter distances than the other two longer-winged species, being only 19cm in length and having a winspan of only 46cm. This species is found mainly in the Sea of Japan and breeds as far north as Vladivostock, but there are other records from the Pacific coast, maily from islands off the north of Honshu, with breeding also recorded on the Izu islands, and I have also seen two birds in the outskirts of Tokyo Bay in April.
Swinoe's, at least in the Sea of Japan is meant to be a later migrant, not arriving until the latter half of May, and departing in September, though (perhaps lingering longer in the Pacific and arriving earlier too..though this is unproven at present), while Tristram's is a winter breeder, being most common from November to June before dispersing more widely and Matsudaira's breeds from December to June before dispersing.
Of course any dark-rumped species of storm-petrel has to be seen well to be totally sure of ID.
Least Storm Petrel has also been recorded in Japanese waters at least once and several unknown birds have been seen in Korean and Japanese waters of a potentially new species perhaps?
For more discussion on this difficult group see the following discussion on Kantorilode:
http://ca.geocities.com/kantorilode/Species/Dark-rumpedStorm-Petrels.html#Dark-rumped Storm-Petrels
Sean
Hiroshima
I forgot to mention that some Leache's Petel can also be dark-rumped and could perhaps be misidentified as Swinhoe's, though they have a deeper forked tail and are slightly bulkier.