28 Sigurd Johnsen. 
sider Luvarus as belonging to this family. It should, however, be 
noticed that this type of caudal fin is also found in Lampris 
pelagicus (Gunn.), a form whose position in the system has been 
greatly disputed; it generally used to be placed among the Scoin- 
broid forms, but, among other things, the position of the ventrals 
and their number of rays (V. 15—17, abdominal) proved it not to 
belong to this group (V 0O—I + 0—5, thoracic) and it is, therefore 
now placed outside.t) This feature can thus only be used very 
carefully, as a help in defining the systematic position, for the 
arrangement of the skeleton parts of the caudal fin, which pro- 
duces a powerfully built caudal fin, but with only a small area of 
stroke, occurs in two groups, which in other respects are divergent. 
This is evidently a tail structure, which has a certain connection 
with pelagic habits, and causes a particularly powerful propulsion 
in those species which have a fusiform body (rapacious fish, Scom- 
. bridae) or the propulsion of a larger body (Luvaridae, Lampridae; 
plankton feeders). As previously mentioned, Regan has pointed 
out that Luvaridae has points of resemblance both to Acanthuri- 
dae & Scombriformes, more especially to Scombridae; he :con- 
siders Luvarus to be a particulary abnormal and specialized Scom- 
broid. On comparing Regan’s results with what | have written, 
it seems more reasonable to conclude that Luvarus is an early 
branch on the common stem of Scombriformes and Acanthuridae, 
a form which has retained some of its primary characteristics, 
and otherwise taken its own course. 
Boulenger’) places Luvaridae in his division Scombri- 
formes, but says of this group that “it is almost impossible to 
draw up a diagnosis differentiating every one of its members from 
the Perciformes, with which they are closely connected, and from 
which they hardly deserve to be separated.”. As belonging to the 
division Perciformes, he includes Acanthuridae, as well as the other 
families which have generally been placed together as Squami- 
pinnes. In this statement of Boulenger, I find a further con- 
firmation of my assumption that the resemblances which Luvarus — 
shows both to Acanthuridae and Scombrijormes must certainly 
point to actual relationship to both. j 
1) Boulenger, G. A.: On the Systematic Position of the Genus 
Lampris. Ann. Mag. Nat. Hist. X 1902, p. 147. 
2) The Cambridge Nat. History. Vol. VII. Boulenger: Fishes, 
p. 675. 
