8 BULLETIN OF THE UNITED STATES FISH COMMISSION. 
Gilbert in their Synopsis of 1883, the genus Oligocottus being made to include all 
three species; but they were finally restored in Jordan & Evermann’s Check-List of 
1896. To the genus Oligocottus there have been since added the species acuticeps 
(Gilbert, 1893), embryum (Jordan & Starks, 1895), and borealis (Jordan & Snyder, 
1896); to Blennicottus the variety B. globiceps bryosus (Jordan & Starks, 1896). The 
status of these species has remained unchanged, except that Jordan & Evermann, 
in The Fishes of North and Middle America, have considered Oligocottus acuticeps 
the type of a new genus, Oxy coitus, to which they have transferred also Oligocottus 
embryum. This nomenclature is here adopted, except in the genera Oligocottus and 
Blennicottus , where confusion in specific identification has taken place and a reassign 
merit of specific names is necessary. Girard’s original species, Oligocottus maculosus 
and Blennicottus globiceps, were described respectively from Tomales Bay, north of San 
Francisco, and from the Farallon Islands, off San Francisco. Both are now shown 
to be forms of northerly distribution, the type locality being, in each case, near the 
southernmost limit of the range. Southward along the coast each is replaced by a 
distinct species, both of which are abundant at Monterey Bay. Recent authors have, 
unfortunately, identified Girard’s names with specimens from Monterey Bay, while 
the northern species to which his names should apply have been rechristened Oligo- 
cottus borealis and Blennicottus globiceps bryosus. Therefore these two names are now 
abolished, the species becoming Oligocottus maculosus and Blennicottus globiceps, and 
the southern forms are described as new species, Dialarchus snyderi and Blennicottus 
recalvus, the former being also made the type of a new genus. 
The different species of this group resemble each other to a remarkable degree in 
external appearance, yet most of them are separated by characters which we now 
consider of generic importance. The characters are all remarkably constant except 
that of color, which varies greatly with the surroundings and can not be described 
with great exactness. The color descriptions given in this paper are all from life, 
and have been made broad enough to cover all the specimens examined. The character 
and arrangement of the cirri afford perhaps the best specific distinguishing features, 
and not the slightest variation from the adult plan has ever been discovered in these 
species. The features of generic importance in this group are the character of the 
preopercular spines, the presence or absence of scales, the shape and size of the head 
and mouth, and the nature of the first three or four anal rays of the male. There may 
be one, two, or several rays enlarged, or they may be all of normal size, the number 
of modified rays and the amount of enlargement always remaining the same in any one 
species; furthermore, these enlarged rays may or may not be separated from the rest 
of the fin. I have used the form and size of the preopercular spines for the primary 
divisions of the key, and this arrangement brings together the species nearest alike in 
geographical range and external appearance. The presence or absence of scales can 
not be considered a mark of less importance, however, and the two together stand out 
distinctly as dividing the species into natural groups. All of these species have a slit 
behind the last gill except Blennicottus embryum and Rusciculus rimensis. This seems 
to be an important character in determining the relationships of the species. 
The tables accompanying the descriptions give the various dimensions of the body 
in hundredths of the total length to base of caudal. 
The author is under deep obligations to Dr. Charles H. Gilbert, in whose laboratory 
and under whose direction the work was carried on, and to President David Starr 
Jordan, whose suggestions and advice have been of great help. 
