| Length of largest specimen, 4 in. | 100. |
| Length of head | .20 |
| Height of pectoral | .15 |
| Length of dorsal | .72 |
| Length of caudal | .12 |
| Length of anal | .44 |
| Height of ventral | .09 |
| Height of body | .22 |
| Width of body | .08 |
Colors.—Exceedingly variable, but the general pattern, as preserved in alcohol, consists of a series of vertical bands, alternating with spots of various shapes and sizes, and often densely mottled with dark and light blotches distributed regularly, but not describable. The fins have alternating bands, and in all the specimens the membrane between the third and fourth dorsal spine is as clear as glass, as if intended to be seen through, but probably shines in the water as a sort of signal. When fresh the colors of those from San Diego were as follows: 1st, dark brown, a purple lateral stripe, sides with dark and light brown bars, having silvery blotches between them; below yellowish, top and sides of head blotched with yellow, a bright red ring with a green centre near pectorals, and another near caudal. Dorsal with alternating bars of olive and yellow; pectoral yellow at base, its rays reddish, barred with purple, ventrals and anal smoky.
Another was striped and cross-barred with brown, and mingled with this pattern were blotches of olive-brown, yellowish and purple, but no rings. Fins marked like the body, but paler.
It is possible that the rings observed in the first one, and not seen in any other, were caused by the growth of vegetable parasites, which are often found on fish of similar habits.
Remarks.—This is the first instance of a Myxodes being found on our coast north of the equator, though a nearly allied genus, the Heterostichus, has been long known. I was in some doubt whether to refer the fish to Myxodes on account of the meagre descriptions of the genus accessible, but having sent a copy of the outline to Mr. Gill, I have been confirmed in the correctness of the diagnosis. The following are some of the most important generic characters not shown in the outline of our species:
Branchiæ VI-VI, the apertures freely connected below. Teeth uniserial in each jaw, those of lower jaw largest, some of those along sides larger than the rest. No teeth on vomer. Scales minute, entire, cycloid, closely adherent—none on head or fins.
The two-lobed form of the spinous dorsal does not apparently exist in some of the species. Suspecting that some of the other characters will be found sufficient to distinguish it, I propose for it provisionally the name Gibbonsia, in honor of Dr. W. P. Gibbons, of Alameda County, whose descriptions of our viviparous fishes, published in 1854, by the Academy, have only of late been awarded the credit they deserve.
These beautiful little fish are found at low water in holes among rocks along our coast south of Point Conception, and at the adjoining islands. Their varied and elegant coloration would make them beautiful objects for a marine aquarium, but I had nothing suitable for the purpose, in which I could keep them alive long enough to study their habits.
I obtained three at Pt. Loma, San Diego, three at Catalina I., and two at Santa Barbara I., all manifestly of the same species, though differing individually in color, as above mentioned. They have no popular name.