The identification of larval forms of marine fishes, of fish fragments found in stomach contents and of fossil fish is often difficult. As an aid to such work a study is being made of the vertebral characteristics of adult marine fishes found off the coasts of Mexico, California, Oregon and Washington and a key based on these characteristics is being constructed.
The possibility of such a key was suggested by Ford (1937), who described at considerable length and with the aid of some excellent photographs the differences in the vertebral columns of several teleostean fishes, approaching the subject from the "functional" angle without any attempt to form a key.
Work was begun on the present study in 1938 and has been continued as material and time were available. Because of the large number of species involved it has been found necessary to divide the study into geographical units. The first unit, covered in this paper, comprises 163 species of fish, omitting the Elasmobranchii, found between Point Conception and San Diego. Work will be continued on fishes from the areas to the north and south of this region and published as completed.
The material for this study consisted mainly of fresh adult specimens which were lightly boiled to loosen the flesh from the bones. To secure less common forms it became necessary to use material which had been preserved in alcohol or formaldehyde. Since no amount of boiling will loosen the flesh from the bones of preserved specimens, the clearing and staining technique described by Hollister (1934) was adopted with some modifications (see appendix). This technique was also used on the smaller fresh individuals to avoid loss of vertebrae in the boiling process.
In naming the various parts of the skeleton, Starks (1901) was followed, with one exception. Starks calls the last fanlike tail segment articulated with the last true vertebral centrum, the "hypural." Since the term "hypural" is used by many authors to designate the spines which are loosely attached to this fanlike plate and the word "urostyle" is used for the plate itself, this latter terminology was adopted. The basis for separating the abdominal and caudal vertebrae is that of Hubbs and Lagler (1947).
In the fin ray counts, the last two rays were counted separately, unless they originated from the same base, in which case they were counted as one ray. Following usual procedure the spine counts are given in Roman numerals and the soft ray counts in Arabic. A comma separates the spine number from the ray number in the same fin. In the case of two dorsal fins, a hyphen separates the individual counts of the two separate fins. Finlets are given in Roman numerals following, and separated from, the ray counts by a hyphen.
It is recognized that fin counts probably will not be of much use in identification of stomach contents, so every effort was made to key the fish without resort to fin counts. In a number of cases when it was impossible to do this the fin counts have been included in the key. They are also given in the descriptive text as corroborative data. Other sources have been consulted to obtain additional fin ray counts. The numbers here given include the greatest known range for each species.
The form and shape of the vertebrae may undergo changes and modifications between the young and the adult forms which may create some confusion where the key is based on adults only. Such a variation has already been described for Clevelandia ios and related forms (Clothier, 1946). Furthermore, examination of additional specimens may indicate that other skeletal features change with growth and, consequently, certain characters here used may later prove of doubtful diagnostic value.
