8 BULLETIN 189, UNITED STATES NATIONAL MUSEUM 



that the head is contained 3.1 to 3.5 times and the depth 2.8 to 3.2 

 times in the length (that is, in the standard length). "Eye 5.6 to 

 5.9; snout 3.3 to 3.7" means that the diameter of the eye is contained 

 5.6 to 5.9, and the length of the snout 3.3 to 3.7 times, in the head. 

 Other proportions are similarly stated. If there is any departure 

 from this common practice, it is explained in the text. 



The names of the fins, when enumerations of rays are given, have 

 been abbreviated by using their first letter. The formula D. VIII, 

 12, for example, signifies that the dorsal fin is single and that it 

 consists of 8 spines and 12 soft or articulated rays. If the formula 

 is written D. VII-I, 12, the fin is composed of 2 separate parts, or 

 of 2 fins, the first consisting of 7 spines, and the second of 1 spine 

 and 12 soft rays. Similarly, V. I, 5 means that the ventral fin has 1 

 spine and 5 soft rays. As shown in the examples of the fin-ray 

 formulae offered, the spines are indicated by Roman and the soft 

 or articulated rays by Arabic numerals. In some species the last 

 ray of the dorsal and anal is deeply divided. However, it was counted 

 as single, unless the division was complete, in the numbers given 

 in this work. 



The enumerations of gill rakers sometimes are stated with 2 

 numerals, with a plus sign between them, thus "7 + 16." This 

 formula signifies that in this instance the specimen had 7 rakers on 

 the upper and 16 on the lower limb of the first gill arch. The number 

 of scales occasionally is stated thus, "8—76", which signifies that 

 the specimen examined had 8 complete rows of scales between the 

 lateral fine and the first ray of the dorsal, and 76 vertical or obUque 

 series, running upward and backward from the lateral line, counted 

 just above the lateral line. Sometimes it was necessary or desirable 

 to depart from this usual procedure, and of course some scaly fish 

 have no lateral line. In such cases the method of counting is explained 

 in the descriptions. 



The use of keys. — The keys offered are intended to provide ready 

 means for identification. No attempt has been made to indicate 

 natural characters or relationsliips, and only the families, genera, 

 and species occurring in Peru, with few exceptions, have been con- 

 sidered. It is always advisable for the student using the keys to 

 read the description to which the keys direct him before coming to 

 a conclusion as to the correct identification of the material in hand. 



When using the keys first determine to which major group, desig- 

 nated as "a" and ^'aa^\ the specimen in hand belongs. Then take 

 up in the regular order the letters representing the smaller groups 

 under the major group. If the specimen does not agree with the 

 characters mentioned after the single letter, try the double letter 

 and rarely a triple letter, ignoring all the intervening subdivisions. 

 After a little practice the beginner very probably will discover that 

 the keys are not difficult to use. 



