NO. 2133. FISHES TAKEN BY "ALBATROSS," 18S8— THOMPSON. 427 



by Dollo ^ may be consulted, and also the synonymy given here for 

 each of the species treated. 



For each species a typical example has been selected and used 

 as the type of the description to avoid confusion from the possible 

 inclusion of more than one species under the one name. The range 

 of variation found for as many specimens as possible is then in- 

 cluded in parentheses after each character. No one set of characters 

 is relied on as diagnostic. The scaling of the head is variable in 

 some species and constant in others, but the size of the scales on the 

 head is usually a good character. For the average of the counts and 

 measurements of the species examined the table following the key 

 may be consulted, but for the extremes the individual descriptions 

 must be examined. 



The South American forms may be loosely divided mto two groups, 

 typified (1) by Nototlienia macrocepliala and N. latifrons, and (2) by 

 Notothenia longipes with the remaining species. N. squamiceps is 

 most nearly allied to the first, but differs in its deep body and scaled 

 interorbital region. The first group is characterized by the broad 

 naked interorbital space, lateral eyes, compressed bodies, and few 

 anal rays. They seem to be of uniform coloration and perhaps are 

 not strictly littoral, as most of the other species seem to be. The 

 second group has a more cottoid appearance, a slightly compressed 

 body, narrow interorbital space, and usually a variegated color pat- 

 tern. Among this group N. gilberti is remarkable for its very short 

 snout, and N. elegans for its greatly elongated, slender body and 

 extremely narrow interorbital. 



The measurements made in this review are in hundredths of the 

 body length to the base of the caudal fin and were computed by 

 means of the United States Bureau of Fisheries scale for proportional 

 measurements. The scales were counted in series along the lateral 

 line, not in the line itself, from the angle of the opercle to the last 

 of the muscular part of the tail, and the transverse counts were made 

 from the anal insertion obliquely upward and forward. The count 

 spoken of as from the occiput to the dorsal was made from the 

 posterior line of the head to the first dorsal spine. 



The specimens were each labeled with a number (such as 01-12 

 or 01-11, etc.) to prevent confusion while being examined. These 

 have been left attached and the numbers are given wherever detailed 

 measurements are presented, in order that investigators subsequently 

 dealing with the same material may identify them. Some of the 

 specimens are placed in the collections of Stanford University, and 

 the types, with the remainder of the species, have been sent to the 

 United States National Museum. 



I Trans. Royal Soc. Edinb., vol. 49, 1913, p. 209. 



