Chaenogobius urotaenia and its two kins are so closely allied each other that their classification is left in confusion yet. A minute scrutiny has been done by the writer dealing with a lot of specimens, on which all available data are based, gathered from various localities of Japan. It is found that there is no trait to assert them simultaeneously from each other. The characteristics given below, however, are useful for taxonomy of them: diverging angle of glossohyal (Fig. 2, J-L; cf. the anthor 1950), upper aspect of tongue (Fig. 2, D-F), presence or absence of scale on nuchal region, and ratio of depth of caudal peduncle against total length (Fig. 1, J). Besides these, length of upper jaw, according to OKADA and NAKAMURA (1948), is available for the purpose. By these traits, the 3 forms may be splited as shown in the next key:
a
1 Forking angle is ca. 65 degrees; the ratio of caudal peduncle, 0.081-0.102 (0.094 on average) …C. urotaenia (HILGENDORF).
a
2 Forking angle is ca. 45 degrees; the ratio of caudal peduncle, 0.061-0.081 (0.070 on average)
b
1 Upper surface of tongue smooth; scale present on nuchal region; upper jaw does not reach to posterior rim of pupil…C. castanea (O'SHAUGHNESSY). b
2 Upper surface of tongue furnished with tubercles; scale absent on nuchal region; upper jaw stretches beyond pupil…C. isaza TANAKA.
The three not only are classified morphologically, but also breed independently in natu-re. Therefore it goes without saying that each of them is independent species.
View full abstract